Neurolinguistic Aspects of the Japanese Writing System

By Michel Paradis

Neurolinguistic Aspects of the Japanese Writing System provides an account and an analysis of cases of dissociation between kana (syllabic) and kanji (ideographic) script in reading and/or writing. Organized into five chapters, this book begins by discussing the aspects of the Japanese writing system relevant to neurolinguistic research. Experimental kanji/kana processing studies and clinical case reports are then presented. This book also explains the clinical dissociations in performance between aspects of the writing system. This book will serve as a model for further studies in which a similarly detailed analysis is attempted of the neurolinguistic structure of other non-Western orthographies.

• Language: English
• Publisher: Academic Press
• Release date: Oct 22, 2013
• ISBN: 9781483288949

Michel Paradis

Michel Paradis is a leading scholar and lawyer of international law and human rights. He has won high-profile cases in courts around the globe and worked for over a decade with the US Department of Defense, Military Commissions Defense Organization, where he led many of the landmark court cases to arise out of Guantanamo Bay. He also holds the position of Lecturer at Columbia Law School, where he teaches on the military, the constitution, and the law of war. He has appeared on or written for NPR, MSNBC, The New York Times, The Washington Post, The Wall Street Journal, Foreign Policy, Lawfare, America, The Intercept, and the late Weekly Standard. He lives with his wife, daughters, and yorkie in Manhattan.

Book preview

Neurolinguistic Aspects of the Japanese Writing System

Michel Paradis

Department of Linguistics, McGill University, Montreal, Canada

Hiroko Hagiwara

Department of Linguistics, McGill University, Montreal, Canada

Department of English, Kinjo Gakuin University, Nagoya, Japan

Nancy Hildebrandt

Department of Linguistics, McGill University, Montreal, Canada

Table of Contents

Cover image

Title page

Copyright

Foreword

Preface

Glossary

Chapter 1: ASPECTS OF THE JAPANESE WRITING SYSTEM RELEVANT TO NEUROLINGUISTIC RESEARCH

Publisher Summary

Chapter 2: EXPERIMENTAL KANJI/KANA PROCESSING STUDIES

Publisher Summary

PSYCHOLINGUISTIC STUDIES OF KANJI/KANA PROCESSING

KANJI/KANA LATERALITY STUDIES

SUMMARY OF KANJI/KANA LATERALITY STUDIES

CONCLUSION

Chapter 3: CLINICAL CASE REPORTS

Publisher Summary

Chapter 4: CLINICAL DISSOCIATIONS IN PERFORMANCE BETWEEN ASPECTS OF THE WRITING SYSTEM

Publisher Summary

BETTER PERFORMANCE IN KANJI THAN IN KANA

BETTER PERFORMANCE IN KANA THAN IN KANJI

BETTER PERFORMANCE IN ORAL READING THAN IN READING COMPREHENSION

BETTER PERFORMANCE IN READING COMPREHENSION THAN IN ORAL READING

THE USE OF ON- AND KUN-READINGS

OTHER DYSLEXIC AND DYSGRAPHIC ERRORS PECULIAR TO JAPANESE PATIENTS

HEMISPHERIC INVOLVEMENT IN KANJI/KANA PROCESSING: CLINICAL EVIDENCE?

IMPLICATIONS FOR A PSYCHOLINGUISTIC MODEL OF READING

Chapter 5: CONCLUSION

Publisher Summary

REFERENCES

Author Index

Subject Index

PERSPECTIVES IN NEUROLINGUISTICS, NEUROPSYCHOLOGY, AND PSYCHOLINGUISTICS

Copyright

COPYRIGHT © 1985, BY ACADEMIC PRESS, INC.

ALL RIGHTS RESERVED.

NO PART OF THIS PUBLICATION MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM OR BY ANY MEANS, ELECTRONIC OR MECHANICAL, INCLUDING PHOTOCOPY, RECORDING, OR ANY INFORMATION STORAGE AND RETRIEVAL SYSTEM, WITHOUT PERMISSION IN WRITING FROM THE PUBLISHER.

ACADEMIC PRESS JAPAN, INC.

Hokoku Bldg. 3-11-13, Iidabashi, Chiyoda-ku, Tokyo 102

United States Edition published by ACADEMIC PRESS, INC.

Orlando, Florida 32887

United Kingdom Edition published by ACADEMIC PRESS, INC. (LONDON) LTD.

24/28 Oval Road, London NW1 7DX

Library of Congress Cataloging in Publication Data

Paradis, Michel.

Neurolinguistic aspects of the Japanese writing system.

Bibliography: p.

Includes index.

1. Dyslexia--Japan. 2. Neurolinguistics. 3. Language disorders--Japan. 4. Japanese language--Writing. 5. Japanese language--Orthography and spelling. 6. Japanese language--Syntax. I. Hagiwara, Hiroko.

II. Hildebrandt, Nancy. III. Title.

RC394.W6P37 1985 616.85′53′0089956 84-16860

ISBN 0-12-544965-8 (pbk.: alk. paper)

PRINTED IN THE UNITED STATES OF AMERICA

85 86 87 88 9 8 7 6 5 4 3 2 1

Foreword

Spoken languages, qua formal objects, are not restricted to a single phenotypic type. Rather, their grammars exhibit considerable structural variation that can be characterized, in traditional terms, as primarily isolating, agglutinating, or inflecting. Similarly, when languages are expressed in more or less permanent visual form (in cuneiform, or Cyrillic script, for example), their orthographies differ widely in the level or levels of linguistic form that are represented explicitly and (more or less) systematically.

Just as current syntactic theory attempts to specify parametric variation within the constraints of Universal Grammar, so a theory of orthography should characterize both the universal constraints and allowable variants that define the script-types accessible to human psychobiology. Just as the child can acquire any natural language (upon adequate exposure), so can he or she learn to use a variety of culturally evolved writing systems. Despite this fact, it is often argued that the acquisition of literacy is, in some sense, an unnatural skill. Scholars are prone to remind us that the history of true writing systems can only be traced back to about 3500 B.C.E. (although Schmandt-Besserat, 1980, has argued cogently for the existence of precursors dated some 5000 years earlier). The history of mass literacy (in the Western world) is considerably briefer, and over considerable areas of the globe literacy is even now the prerogative of a small, privileged class; even in societies where the educational system aims to achieve 100% literacy, a small but distressing minority of children experience great difficulty in learning to read and never achieve the fluent command of visual language skills that are necessary to participate fully in a technological society.

Small wonder, then, that linguists should so often have regarded visual language as the disadvantaged relative of spoken language, and that educators should periodically contemplate orthographic reform as a means of fitting the medium to the message. Without wishing to downplay the very real problems involved in achieving the goal of global literacy, there is a more positive and, to my mind, more interesting way of looking at the issue. Given adequate exposure, sufficient motivation, and decent teaching, the vast majority of children are perfectly capable of acquiring good reading and writing skills. Furthermore, they are capable of this feat when exposed to the orthographies of French or English, Arabic or Hebrew, Korean or Navajo, or any of the other script-types that can be found in five continents. Some children achieve an understanding of written language before formal schooling, and many have mastered the basic skills of literacy after only a year or two of instruction. No account of the psychobiology of written language can afford to ignore such basic facts. Written language cannot be dangerously close to the epistemological boundary of what the human brain is incapable of comprehending and manipulating.

Much of what we know about the structural organization of reading and writing skills and their material substrate in the brain has come from detailed studies of the fractionation and partial loss of written-language competence in previously literate adults who have sustained damage to the central nervous system (Geschwind, 1962). More recently, the paradigm of information-processing psychology has allowed for a considerable integration of studies of normal and impaired reading whereby pathological data are interpreted as the selective loss of particular modular components and mechanisms that underlie fluent, mature skill (Coltheart, 1981; Newcombe & Marshall, 1981; Patterson, 1981). Introduction of the technique of split visual-field presentation, whereby a stimulus can be projected initially to either the left or the right half-brain (Franz & Davis, 1933) has likewise served to revive theoretical interest in differential hemispheric specialization for the linguistic and visuospatial aspects of written language (Zaidel, 1983).

For obvious reasons, most of this work has been conducted with people whose native orthographies are the alphabetic scripts of West Europe and North America: English, French, Italian, and German, in particular. Whilst there are significant differences in the ways in which these languages are mapped into written form the structural characteristics of their orthographies fall within a very narrow band of the typological variants of writing systems. Outside of Europe and North America, only one other culture—that of Japan—has so far supported an extensive research tradition in which the patterns of selective loss and preservation of visual-language skills consequent upon brain damage have been evaluated in any detail. Much of the documentation of these deficits, written for the most part in Japanese, has, of course, been unavailable to Western students of the acquired dyslexias and dysgraphias. Much of the information that has percolated into Western neurolinguistics has been grossly oversimplified or downright misleading, as Westerners have typically failed to appreciate the full complexity of the Japanese writing system(s). Most of us know that Japanese is a mixed system, syllabic (kana) and ideographic (kanji), but we have all too often glossed over the richness of structural information that hides behind these seemingly innocent descriptions.

The development of a truly universal neurolinguistics of written language will for long be in debt to Michel Paradis, Hiroko Hagiwara, and Nancy Hildebrandt. For the first time in English, they have provided us with a full and lucid analysis of the fine grain of the Japanese writing system, and have summarized all the clinical reports of dyslexic fractionation in Japanese back to the turn of the century. This major monograph, linguistically informed, clinically astute, and organized with an eye to theory construction within an information-processing framework, is the firm foundation on which future work will build. It will, one hopes, serve as a model for further studies in which a similarly detailed analysis is undertaken of the neurolinguistic structure of other non-Western orthographies.

One particularly valuable aspect of Paradis, Hagiwara, and Hildebrandt’s work is that they have brought together in one monograph clinical studies of acquired dyslexia and experimental investigations of hemispheric specialization for written language in normal subjects. Our journals bear horrendous witness to how easy it is to run ill-thought and badly controlled experiments on laterality effects. The temptation to display a motley collection of kana and kanji in the two visual fields and to interpret the results in terms of some overreaching, oversharp dichotomy has often proved overwhelming. Western (and, to some extent, even Japanese) scholars have succumbed to conclusions that are simply unfounded because the stimulus materials have been confounded. Paradis, Hagiwara, and Hildebrandt provide a masterly summary and critique of this literature that should, at very least, provoke a dramatic increase in the level of sophistication that future experimentalists may reasonably aspire to.

Modern information-processing accounts of normal reading and pathologies of reading have a provenance of only a little over a decade; they were constructed on the basis of a very limited range of data from an even more limited set of writing systems. Despite the narrowness of this empirical base, it is most heartening to see that those models do appear to provide a theoretical framework within which to interpret the pathologies of reading and writing seen in the radically different orthographic system of Japanese. Whilst current models are computationally much underspecified, they do seem to be a step in the right direction. The systematization of Neurolinguistic Aspects of the Japanese Writing System that Paradis, Hagiwara, and Hildebrandt have achieved brings forcibly to our attention the absolute necessity of developing these models within the context of a universal theory of reading and writing. And universal in a linguistic context means, of course, that language- and culture-specific variation must be accounted for. There is every reason to hope and expect that the next decade of neurolinguistic research on writing systems will be even more exciting than the last, and that our theories will be considerably deepened by insisting that they be made responsive to data on a fair sample of the phenotypically distinct script-types that children can and do master naturally.

John C. Marshall,     Oxford, England

REFERENCES

Coltheart, M. Disorders of reading and their implications for models of normal reading. Visible Language. 1981; 15:245–286.

Franz, S.I., Davis, E.F. Simultaneous reading with both cerebral hemispheres. Studies in cerebral function, IV. Publications of the University of California at Los Angeles in Education. Philosophy, and Psychology. 1933; 1:99–106.

Geschwind, N. The anatomy of acquired disorders of reading. In: Money J., ed. Reading disorders. Baltimore: Johns Hopkins University Press, 1962.

Newcombe, F., Marshall, J.C. On psycholinguistic classifications of the acquired dyslexias. Bulletin of the Orton Society. 1981; 31:29–34.

Patterson, K.I. Neuropsychological approaches to the study of reading. British Journal of Psychology. 1981; 72:151–174.

Schmandt-Besserat, D. The envelopes that bear the first writing. Technology and Culture. 1980; 21:357–385.

Zaidel, E. Disconnection syndrome as a model for laterality effects in the normal brain. In: Hellige J.B., ed. Cerebral hemisphere asymmetry: Method, theory and applications. New York: Praeger, 1983.

Preface

The purpose of this book is to provide English readers with an account and an analysis of cases of dissociation between kana and kanji script in reading and/or writing reported in the clinical and experimental literature since the beginning of the century. It includes reports previously available only in Japanese (the language of the vast majority of publications on the subject). To this end a brief description of the Japanese writing system is provided, particularly of those features relevant to a meaningful interpretation of the data obtained from clinical and experimental studies. Indeed, because many of these features have been ignored in most past investigations—ignored by Japanese investigators for lack of linguistic sophistication and by Western investigators for lack of sufficient knowledge of the Japanese writing system—great caution is advised in the interpretation of results published so far. These results are presented along with reasons why they are at best ambiguous and at worst completely uninterpretable. Alternate possible explanations are suggested and constraints on the choice of stimuli to serve in future experimental or clinical research are outlined. The use of these differentiated stimuli may eventually provide grounds for a radical reinterpretation of available data, with far-reaching consequences for the diagnosis of dyslexia and associated deficits in Japanese patients.

The study of dyslexia in Japanese, a language that combines two very different types of graphic symbols in its writing system, may tell us something about reading strategies in general. As such, it is of interest not only to the researcher concerned with Japanese dyslexia but to the psychologist, the neurologist, and the linguist. In particular, data in support of the psychological reality of hypothetical constructs of reading processes such as box diagrams are of benefit to the psychologist, and the detection of anatomical correlates of cognitive skills involved in reading is of relevance to the neurologist. In addition, it is of importance to the linguist to investigate the relationship between the spoken and the written form of a language that contains a relatively small number of possible syllables and a very large number of homophones, and whose writing system integrates a syllabic system (with very few grapheme-to-syllable correspondence exceptions), and an ideographic system (with most characters being multiple homographs).

Thus, in the present volume each previously published case study of dyslexia in Japanese is extensively summarized; all cases are tabulated and thoroughly analyzed in the search for correlations between patterns of impairments and site of lesion, degree of literacy, associated aphasic or agraphic deficits, age, sex, and severity of the alexia. This survey of clinical cases should dispel some prevalent misconceptions about the incidence of differential impairment of kanji subsequent to right hemisphere damage. The survey of the Japanese laterality literature should likewise show that there are no differences with respect to results from English studies, once methodological variables are taken into account.

All three authors worked together to produce this monograph, each bringing to the task different skills and a different way of looking at the Japanese writing system. Hiroko Hagiwara provided the insights of a native speaker into the facts of Japanese, translated the case reports from Japanese into English, and analyzed the clinical literature. Her contribution is based on her M.A. thesis for the Department of Linguistics at McGill University. Nancy Hildebrandt, a nonnative speaker of Japanese, was able to study the system from the outside and analyzed the experimental literature. Michel Paradis approached the system with no prior knowledge of Japanese but rather in the manner of a linguist who elicits information in the field, thereby educing the parameters relevant to psycholinguistic and neurolinguistic investigation, and interpreting the identified patterns of data. Although each author was primarily responsible for writing the initial draft of a given chapter, each was also responsible for reviewing them all.

We are indebted to many colleagues for constructive comments in conversations and at meetings over the past 2 years. In particular, we wish to thank Drs. Derek Besner, Justine Sergent, Atsushi Yamadori, Morihiro Sugishita, Makoto Iwata, and Molly Mack for providing valuable comments and advice. The contents of this book grew out of more general research on aphasia in bilinguals funded by the Quebec Ministry of Education FCAC grant EQ 1660.

Glossary

Ateji In the narrow sense used here, ateji refers to a multi-character kanji whose component characters are pronounced in accordance with a standard pronunciation (albeit not by virtue of a phonemic radical) but whose usual meanings are irrelevant to the meaning of the word (or at least are far from transparent). Example: (sarcasm) pronounced [hiniku] is written with the kanji for skin [hi] and the kanji for meat [niku]. Ateji is the reverse of the semantically compositional jukujikun (q.v.). See Table 1.12 for more examples.

Choo-on Geminate vowels, some of which have an irregular representation. See Table 1.2.

Furigana Small kana written above a kanji (in horizontal writing) or to the right of the kanji (in vertical writing) to specify its unfamiliar pronunciation.

Gogi Word-meaning. Hence, gogi aphasia: word-meaning aphasia, a syndrome similar to transcortical sensory aphasia, generally accompanied by impairment of kanji processing.

Grapheme As used in this work, a written symbol such as a letter, a kana (q.v.) or a kanji (q.v.). In most current literature on dyslexia, grapheme is more restrictively taken to mean any letter or group of letters whose realization is a single phoneme.

Hiragana Syllabic characters normally used to write grammatical morphemes such as postpositions, conjunctions, honorific prefixes, case markers, derivational suffixes and inflections. See Table 1.1. N.B.: Before 1920 katakana (q.v.) were used for this purpose.

Homograph A word identical in written form with another word of the same language but different in sound and meaning. English example: wind, (1) [waιnd] to have a curving course; (2) [waιd] a movement of the air. Japanese example: (1) [see] nature, sex; (2) [∫oo] disposition; (3) [saga] custom.

Homophone A word identical in sound with another word of the same language but different in meaning and often in written form. English example: knead and need. For Japanese examples, see Table 1.6.

Ideogram A written character symbolizing the idea of a thing without systematic representation of its pronunciation. Japanese ideograms (kanji) represent morphemes of the spoken language.

Jukujikun A multi-character kanji whose pronunciation is independent of any pronunciation of each component character. There is no fit between the pronunciation of the word and that of any component character. The characters may or may not be semantically compositional. When they are—that is, when the sum of the meanings of the component characters yields the meaning of the whole—the jukujikun may be considered the reverse of ateji (q.v.) since here the pronunciation, not the meaning, is arbitrary. See Table 1.12, Nos. 3 and 4.

Kana A written character (derived from Chinese writing but extremely simplified) used to represent Japanese (short) syllables. Strictly speaking, a kana represents a mora, or syllable of short duration. It is not clear whether in Japanese there are CVV syllables of long duration or whether, in CVV syllables, vowels represent two successive short vowels, in which case mora and syllable would be coextensive in Japanese. In careful speech, at least, two successive identical morae (e.g. [oo]) are pronounced as geminate vowels, not as a long vowel. Hence, for the sake of simplicity we will say that a kana corresponds to a syllable of spoken Japanese. There are two types of kana, hiragana (q.v.) and katakana (q.v.). Each syllabary has a one-to-one correspondence between grapheme and syllable, with very few exceptions. See Tables 1.1 and 1.2.

Kanji Literally: Chinese [kan] character [d3i]. Ideographic character of Chinese origin used in the Japanese writing system to represent a lexical morpheme of spoken Japanese. Hence the expression kanji character would be pleonastic, and therefore will not be used. Lexical formatives (content words) are represented by one or more kanji in accordance with the number of component morphemes. The word kanji applies to a single character as well as to a multi-character word (a polymorphemic word).

Katakana Syllabic characters normally used to transcribe foreign words and to signal emphasis, as italics in English. See Table 1.1. N.B.: Before 1920 hiragana (q.v.) were used for this purpose.

Kun-reading A pronunciation of a kanji corresponding to a spoken morpheme of native Japanese origin. See Table 1.7.

Logogram A written character representing an entire word of the spoken language without systematic representation of its pronunciation.

On-reading A pronunciation of a kanji corresponding to a spoken morpheme of Chinese origin (by now well integrated into the Japanese system). One may compare morphemes of Chinese origin in Japanese to morphemes of Latin and Greek origin in English (e.g., poly and glot in polyglot or semi in semivowel) as opposed to morphemes of Anglo-Saxon origin (e.g., child and hood in childhood). See Table 1.7 for examples.

Paragraphia An error in writing due to the substitution of a segment or segments.

Paralexia An error in reading due to the substitution of a segment or segments.

Phonemic radical A part of a character that indicates a certain pronunciation. Different characters that contain the same phonemic radical are pronounced the same way (see Table 1.8). However, there are exceptions in that some characters which contain a given phonemic radical are nevertheless not pronounced in accordance with it (see Tables 1.9 and 1.10). Moreover, phonemic radicals are only indicative of an On-reading (q.v.), never of a Kun-reading (q.v.) of the same character.

Semantic radical A part of a kanji which is found in several different characters and is indicative of a given basic meaning (e.g., the radical is found in many words related to water: pond, lake, sea).

Soku-on A geminate consonant marker written with the small kana character [tsu]. (See Table 1.1.)

Spatial frequency The visual equivalent of auditory frequency (see Campbell, 1974). In very simple terms, large patterns, contours, etc. have a low spatial frequency, while figures with higher visual resolution have a high spatial frequency.

Yoo-on A small-sized kana character which changes the high front vowel of the preceding syllable to a semivowel and adds a new vowel, for example, [∫i] + [ja] –→ [∫ja].

CHAPTER 1

ASPECTS OF THE JAPANESE WRITING SYSTEM RELEVANT TO NEUROLINGUISTIC RESEARCH

Publisher Summary

A superficial reading of the literature on dissociations between performance with kanji and kana characters in Japanese alexic patients may lead the non-Japanese reader to assume that there are two writing systems in Japanese, one syllabic and the other ideographic, and that some patients selectively lose access to one or the other. This chapter describes the interpretation of two script types, namely, kana and kanji, which are used in systematic combinations in sentences. A kanji is a graphic symbol representing a lexical morpheme with no systematic relationship to the corresponding spoken sounds, each morpheme being represented by a specifically shaped character. A kana is a character that stands for a syllable. The Japanese speaker uses one writing system that integrates these two types of script, and this property of the system poses large number of serious problems when the performance of two scripts are compared. A major consequence of this system is that when one tries to compare performance on kanji stimuli and kana stimuli, there is an unavoidable part-of-speech effect, because kanji characters in isolation are used to represent nouns whereas kana characters are used to represent grammatical morphemes. To overcome this difficulty, performance on nouns written with a kanji can be compared with performance on nouns written with kana characters. However, this leads to additional difficulties such as length-of-grapheme-representation effect, strangeness effect for stimuli and semantic-category effect. In general, the sole dimension considered as a factor in differential performance with respect to the two Japanese writing systems has been based upon a distinction between the phonemic and visual dimension. Differences in performance have been attributed to differences in visual processing for kanji and phonological processing for kana.

A superficial reading of the literature on dissociations between performance with kanji and kana characters in Japanese alexic patients may lead the non-Japanese reader to assume that there are two writing systems in Japanese, one syllabic, the other ideographic, and that some patients selectively lose access to one or to the other. Yet, kana and kanji are not alternative writing systems, but two scripts, both used in systematic combination in sentences.

A kanji is a graphic symbol representing a lexical morpheme with no systematic relationship to the corresponding spoken sounds, each morpheme being represented by a specifically shaped character (however, see p. 11). It is not strictly speaking a logogram, since it does not stand for an entire word, except in the case of monomorphemic words. Polymorphemic words are represented by more than one kanji. Hence, for lack of a better word (such as morphogram) we refer to it as an ideogram, since in some sense a morpheme represents an idea, more specifically a unit of meaning, an object of mental representation. Both logogram and ideogram are found in the literature in reference to kanji.

A kana is a character that stands for a (short) syllable (sometimes called a mora, but the point can be made that in Japanese there are no long syllables, except for two successive identical vowels pronounced as one long vowel in fast speech but represented by two kana). The number of possible syllables in Japanese is limited to 102, and if we consider not only the shape of the characters but their size as well as diacritical marks, there is a one-to-one correspondence between kana and syllable (see Table 1.1). That is to say, each syllable is represented graphically by a distinctive kana character and each kana character stands for a specific syllable sound, with only five minor exceptions. In the case of double [o]