Japanese Language,

Gender, and Ideology

STUDIES IN LANGUAGE AND GENDER
Mary Bucholtz, General Editor

Advisory Board
Penelope Eckert, Stanford University
Kira Hall, Yale University
Janet Holmes, Victoria University of Wellington, New Zealand
Miyako Inoue, Stanford University
Don Kulick, University of Stockholm
Sally McConnell-Ginet, Cornell University
Marcyliena Morgan, University of California, Los Angeles/Harvard University
Deborah Tannen, Georgetown University
Ana Celia Zentella, Hunter College, City University of New York

Reinventing Identities: The Gendered Self in Discourse

Edited by Mary Bucholtz, A. C. Liang, and Laurel A. Sutton

Pronoun Envy: Literacy Uses of Linguistic Gender

Anna Livia

Japanese Language, Gender, and Ideology: Cultural Models and Real People
Edited by Shigeko Okamoto and Janet S. Shibamoto Smith
Japanese Language,
Gender, and Ideology
Cultural Models and Real People

Edited by
Shigeko Okamoto
Janet S. Shibamoto Smith

1
2004
3
Oxford New York
Auckland Bangkok Buenos Aires Cape Town Chennai
Dar es Salaam Delhi Hong Kong Istanbul Karachi Kolkata
Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi
So Paulo Shanghai Singapore Taipei Tokyo Toronto

Copyright 2004 by Oxford University Press, Inc.

Published by Oxford University Press, Inc.
198 Madison Avenue, New York, New York 10016
www.oup.com
Oxford is a registered trademark of Oxford University Press.
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
electronic, mechanical, photocopying, recording, or otherwise,
without the prior permission of Oxford University Press.

Library of Congress Cataloging-in-Publication Data

Japanese language, gender, and ideology : cultural models and real people /
edited by Shigeko Okamoto, Janet S. Shibamoto Smith.
p. cm.(Studies in language and gender)
Includes bibliographical references and index.
ISBN 0-19-5166I7-5; 0-19-516618-3 (pbk.)
1. Japanese languageSocial aspects. 2. Sociolinguistics. I. Okamoto, Shigeko.
II. Shibamoto Smith, Janet S. III. Series.

PL524.75.J36 2004
306.44'0952dc22 2003066233

9 8 7 6 5 4 3 2 1
Printed in the United States of America
on acid-free paper
 ACKNOWLEDGMENTS

We would like to express our gratitude to all the colleagues who contributed chap-
ters to this volume. We have greatly enjoyed working with them and have learned a
lot in the process. We are deeply grateful to Mary Bucholtz, general editor of the
Studies in Language and Gender Series, for her insightful comments and unflagging
support and encouragement, and the two anonymous reviewers for their valuable
comments on the draft of each chapter. We also thank Peter Ohlin, our editor at Ox-
ford University Press, and the Oxford University Press staff. Special appreciation
must go to artist Tanno Yumiko, who generously provided the perfect visual image
for the cover of this volume. On a more personal front, we want to acknowledge our
husbands, Per Gjerde and David Glenn Smith, for their support and patience through-
out this project. We are happy to have had such cooperative and intellectually en-
gaged partners over the last two years.
The creation of this collection has been a long process but also an intellectually
rewarding journey. There have been numerous occasions when we exchanged our
views by e-mail, over the phone, oron all-too-rare but very treasured occasions
in person. Both of us enjoyed these exchanges and have benefited enormously from
this extended conversation. Our names appear in alphabetical order.
This page intentionally left blank
 MARY BUCHOLTZ , SERIES EDITOR

FOREWORD

The Japanese language has played a foundational role in the development of lan-
guage and gender as an area of linguistic research, and the language use of Japanese
womenand, increasingly, mencontinues to serve as a touchstone for theoretical
and empirical advances in the field. This scholarly interest in Japanese was stimu-
lated in the first instance by the widely discussed linguistic style known as womens
language. For pioneering English-speaking feminist linguists in the 1970s and 1980s,
womens language in the Japanese case provided a powerful illustration of the
intimate relationship between gender and language, one that was seen as both like
and unlike linguistic arrangements of gender in the West. Japanese was seen as dif-
ferent from English in having inscribed gender into its very structure, through the
prescribed use of differential grammatical and lexical forms for women and men.
Yet despite this tendency to treat Japanese womens language as alien and exotic,
at the same time the linguistic situation of Japanese women was often seen as differ-
ent from the circumstances of English-speaking women only in degree, not in kind.
Thus womens language in Japan came to be examined for its potential in inform-
ing the understanding of language and gender in other societies as well.
In addition to its utility for Western feminists seeking to theorize language and
gender, womens language has long been the object of scholarly scrutiny in Japan.
Male Japanese scholars had been discussing this speech style for at least a century.
In the late 1970s female Japanese linguists began to investigate womens language
from a feminist perspective, and their findings often challenged nonfeminist percep-
tions of language and gender. Some researchers identified a number of fissures in
what had been up until then a seamless and largely unquestioned cultural ideology
viii FOREWORD

of how Japanese women ought to speak. The recognition of the ideological dimen-
sion of womens language was a crucial breakthrough that opened up new avenues
of investigation.
At the same time, feminist researchers sought to expand the range of issues ex-
amined within language and gender research in Japan. The fascination with womens
language among both Japanese and non-Japanese had led to the tendency to over-
look other crucial aspects of the relationship between language and gender, and re-
searchers worked to bring these other topics into analytic view: issues such as the
diversity of Japanese womens identities and subject positions and hence of their lin-
guistic practices; the links between language use and linguistic ideologies related to
gender; the interconnection of gender and sexuality and the linguistic aspects of sexual
identity and practice; and the central but unacknowledged importance of men and
masculinity in the study of language and gender. Many of these issues had begun to
be addressed by feminist linguists working on other languages and cultures as well,
but their investigation in the Japanese context has been especially significant in shift-
ing scholars from a near-exclusive focus on womens language to the full range of
issues that connect gender to language. Informed by and contributing to new strands
of feminist thought within language and gender studies, Japanese Language, Gen-
der, and Ideology: Cultural Models and Real People brings together an impressive
group of international scholars to advance this research goal.
This volume forges bold new directions in research on the much-studied and
much-misunderstood topic of womens language while heading out into previously
uncharted territories of linguistic scholarship in Japanese. In the following pages,
womens language receives extensive and critical attention from scholars, who
historicize and contextualize the emergence and development of this style as part of
a wider Japanese gender ideology. Other contributors focus on the numerous women
of various classes, regions, ages, and sexualities who do not always or ever use
womens language, as well as groups of men who draw on its resources to position
themselves outside of hegemonic Japanese masculinity. Still other chapters examine
emergent issues of language, gender, and sexuality in Japan, including questions of
scholarly and popular representations of language and gender, gendered cultural styles
and practices, sexual identity and romance, and the linguistic practices of communi-
ties of speakers in a rich array of social contexts. Japanese Language, Gender, and
Ideology includes researchers based both within and outside of Japan, and it is an
especially significant achievement of the volume that it makes available in English
for the first time the writings of several important contemporary feminist Japanese
linguists.
The purpose of Oxfords series Studies in Language and Gender is to promote
the most innovative, significant, and enduring research in the study of language,
gender, and sexuality around the world. The inclusion of this volume in Studies in
Language and Gender signals the continuing importance of feminist scholarship on
the Japanese language as well as the contributions of this body of work in moving
the field into exciting new arenas of inquiry.
 CONTENTS

Note on Naming xi
Contributors xiii
Introduction 3
Shigeko Okamoto and Janet S. Shibamoto Smith

PART I HISTORICAL AND THEORETICAL FOUNDATIONS

1 Cultural Ideologies in Japanese Language and Gender Studies:
A Theoretical Review 23
Sumiyuki Yukawa and Masami Saito
2 Ideology in Linguistic Practice and Analysis: Gender and Politeness
in Japanese Revisited 38
Shigeko Okamoto

3 Gender, Language, and Modernity: Toward an Effective History

of Japanese Womens Language 57
Miyako Inoue

4 Japanese Female Speech and Language Policy

in the World War II Era 76
Rumi Washi
5 Shifting Speakers: Negotiating Reference in Relation
to Sexuality and Gender 92
Wim Lunsing and Claire Maree

PART II LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

6 Language and Gender in the (Hetero)Romance: Reading
the Ideal Hero/ine through Lovers Dialogue in
Japanese Romance Fiction 113
Janet S. Shibamoto Smith
x CONTENTS

7 Lets Dress a Little Girlishly! or Conquer Short Pants! Constructing

Gendered Communities in Fashion Magazines for Young Peop1e 131
Momoko Nakamura

8 You Are Doing Burikko! Censoring/Scrutinizing Artificers of Cute

Femininity in Japanese 148
Laura Miller

9 Women and Words: The Status of Sexist Language in Japan as Seen

through Contemporary Dictionary Definitions and Media Discourse 166
Orie Endo
Janet S. Shibamoto Smith, Translator

PART III REAL LANGUAGE, REAL PEOPLE

10 Farm Womens Professional Discourse in Ibaraki 187
Yukako Sunaoshi

11 Lesbian Bar Talk in Shinjuku, Tokyo 205

Hideko Abe

12 Prosody and Gender in Workplace Interaction: Exploring Constraints and

Resources in the Use of Japanese 222
Yumiko Ohara

13 Alternative Femininity: Personae of Middle-aged Mothers 240

Yoshiko Matsumoto

14 Japanese Junior High School Girls and Boys First-Person Pronoun Use
and Their Social World 256
Ayumi Miyazaki

15 Japanese Mens Linguistic Stereotypes and Realities: Conversations from

the Kansai and Kanto Regions 275
Cindi Sturtz Sreetharan

Index 291
 NOTE ON NAMING

The names of our chapter contributors and our names are listed according to Western
convention, that is, with given name preceding family name, irrespective of nation-
ality. In the text of the chapters, Japanese names are given according to the Japanese
convention, family name preceding given name, with some minor exceptions where
the individuals themselves have made other choices in professional contexts. Western
names appear in given namefamily name order. We recognize these inconsisten-
cies, as we participate in a global economy of intellectual endeavor that incorporates
numerous disparate cultures of naming.

xi
This page intentionally left blank
 CONTRIBUTORS

Hideko Abe Associate Professor, Department of Foreign Lan-

guages, Western Michigan University
Orie Endo Professor of Humanities, Faculty of Language and
Literature, Bunkyo University
Miyako Inoue Assistant Professor, Department of Cultural and
Social Anthropology, Stanford University
Wim Lunsing Free scholar based in Leiden
Claire Maree Full-time Lecturer, Department of English, Tsuda
College
Yoshiko Matsumoto Associate Professor, Department of Asian Lan-
guages, Stanford University
Laura Miller Associate Professor, Department of Anthropology,
Loyola University
Ayumi Miyazaki Graduate student, Graduate School of Education,
Harvard University
Momoko Nakamura Professor, Department of Economics, Kanto Gakuin
University
Yumiko Ohara Instructor, Liberal Arts Department, Hawaii Tokai
International College
Shigeko Okamoto Professor, Department of Linguistics, California State
University, Fresno
Masami Saito COE Post Doctoral Research Fellow, Ochanomizu
University
Janet S. Shibamoto Smith Professor, Department of Anthropology, University
of California, Davis
Cindi Sturtz Streetharan Assistant Professor, Department of Anthropology
California State University, Sacramento

xiii
xiv CONTRIBUTORS

Yukako Sunaoshi Lecturer, the School of Asian Studies, the University

of Auckland, New Zealand
Rumi Washi Instructor, Institute for Japanese Studies,
Nagoya Gakuin University
Sumiyuki Yukawa Professor, Faculty of Humanities, Toyama University
Japanese Language,
Gender, and Ideology
This page intentionally left blank
 SHIGEKO OKAMOTO
JANET S . SHIBAMOTO SMITH

Introduction

I.1. Toward a new framework for Japanese

language and gender studies

Models of Japanese society and culture, in particular the Nihonjinron theories on

the Japanese literature, commonly stress such attributes as homogeneity, harmony,
consensus, and hierarchy, including gender stratification (Yoshino 1992, Befu 2001,
Sugimoto 2003). In line with this framework, Japanese sociolinguistic practice (e.g.,
the use of honorifics, address terms, and indirect speech styles) as part of Japanese
cultural behaviors has been described primarily based on context-independent, ab-
stractly normative usage in Standard Japanese. Speech produced by real Japanese
people in real situations that differs from that appropriate language use has been
considered erroneous or deviant, or has simply been ignored. Descriptions of
Japanese female and male speech are no exception. Despite decades of intensive
research into Japanese speaking practices and into Japanese womens language
(see, e.g., Ide 1979, 1997; Shibamoto 1985; Ide, Hori, Kawasaki, Ikuta, & Haga 1986;
Reinoruzu[Reynolds]-Akiba 1993), our understanding of Japanese women and men
as speaking subjects still remains largely obscured by essentialized depictions of
the Japanese woman and the Japanese man and their languagesor joseego
womens language and danseego mens languagecharacterized by a number
of stereotypical linguistic features held to be differentially used by average Japa-
nese women or men.
These stereotypical features include broad characteristics of speech styles, such
as polite, gentle, unassertive, and empathetic speech, as well as more specific features

3
4 INTRODUCTION

(particularly in Standard Japanese) such as honorifics, self-reference and address

terminologies, sentence-final particles, indirect speech acts, interjections/exclama-
tory expressions, and voice pitch level (see Shibamoto Smith 2003 for a summary of
these features). However, we argue that womens language and mens language
characterized as such are cultural constructs informed by the hegemonic ideology of
language and genderconstructs that have been widely disseminated as linguistic
norms in Japanese society not only through popular media but also by language policy
makers and linguists like ourselves. In fact, the latter grouppolicy makers and lin-
guistsseems to have had considerable influence on the former. Previous studies of
Japanese language and gender have tended to focus on these normative usages with-
out examining their implications for the real language practices of real speakers. In
particular, real women and men who are notby virtue of their age, class, regionality,
sexual orientation, or other characteristicsidentifiable as average speakers and
their speaking choices, which may or may not conform to the normative usages, have
effectively been erased from the record.
However, the view of Japanese as homogeneous, interdependent, and so forth
has been increasingly criticized in recent years as partial and misleading, represent-
ing hegemonic ideologies as Japanese cultural behaviors (e.g., Yoshino 1992; Befu
1993, 2001; Sugimoto 2003). In order to gain a better understanding of Japanese
society and culture, a growing number of researchers are now investigating diver-
sity and heterogeneity in Japanese society. These researchers pay more attention to
previously understudied sociocultural subgroups (e.g., blue-collar workers, working
women, and ethnic minorities) and social situations (e.g., Weiner 1997, Eades, Gill, &
Befu 2000, Sugimoto 2003). In Japanese sociolinguistics, diversity has been studied
primarily in relation to differences in regional dialects. However, there is now grow-
ing awareness of the need for studying linguistic variation as it relates to social and
contextual diversity that involves real Japanese people interacting in real-world situa-
tions (see, for example, Maher & Yashiro 1995, Cook 1998, Sato & Okamoto 1999,
Saft 2001; see also papers presented in conjunction with Miller 1995 and Okamoto &
[Shibamoto] Smith 1998). In particular, researchers on Japanese language and gender
have been actively engaged in the investigation of linguistic diversity and its meaning.
Since the early 1990s, there has been increasing interest in the study of within-gender
variability in speech practice, in particular, nonnormative uses (Okamoto & Sato 1992;
Smith 1992; Takasaki 1993; Okamoto 1995; Sunaoshi 1995; Inoue 1996; Ogawa &
[Shibamoto] Smith 1996, 1997; Maree 1997; Matsumoto 1999; Abe 2001; Sturtz 2001).
These studies necessitate consideration of the relationship between linguistic practice
and ideologiesa central issue addressed in this volume.
This volume brings together studies that contribute to advancing Japanese lan-
guage and gender studies in this new direction. It is also in line with the orientation
of recent language and gender research in general, which cautions against essential-
ist approaches based on abstractions and binary oppositions of women and men, while
emphasizing the importance of examining local linguistic practices of real speakers
as social agents and their context-dependent and potentially multiple social and prag-
matic meanings in relation to dominant gender and language ideologies (see, among
others, Hall & Bucholtz 1995, Bergvall, Bing, & Freed 1996, Bucholtz, Liang, &
Sutton 1999). This volume builds on and contributes to these trends. It begins with a
 INTRODUCTION 5

close examination of the ideologies and cultural models that underlie our concep-
tions of the idealized Japanese woman and man, as these may have significant regi-
menting effects on speaking practices of real women and men (parts I and II); it then
calls into question these very ideologies and models via a close examination of so-
cially and ideologically diverse Japanese women and men in real verbal interactions
(part III). It presents a significant move forward for Japanese language and gender
studies by attempting to overcome the limitations of previous studies in several ways.
First, this volume highlights the linguistic and cultural ideologies that provide
the backdrop against which Japanese men and women choose how to speak and in-
terpret what is said. Language ideologya rationalization of perceived language
structure and use (Silverstein 1979:193)has not in general been a serious focus of
study in Japanese sociolinguistics, perhaps because it is considered to be somewhat
unscientific, or a distraction from primary and thus real linguistic data, as it has
been in American linguistics (Woolard 1998:11). However, as ideology has come to
be recognized as a powerful shaping force on language production and interpreta-
tion generally (Kroskrity 2000) and as Silverstein (1985) illustrates for language and
gender in English, we attempt to demonstrate in this volume that considering lan-
guage ideology is essential to Japanese sociolinguistic research, as it serves as a
mediating link between social forms and form of talk (Woolard 1998:3). Previous
studies in Japanese language and gender have often confounded native speakers lin-
guistic ideology with their practice by characterizing Japanese female and male speech
patterns in terms of distinct differences based on normative usages for women and
men. Recognizing the importance of distinguishing language ideology and practice,
this book problematizes the ideologies and cultural models that underlie our con-
ceptions of the Japanese woman and the Japanese man as these relate to speak-
ing practices; it considers how these ideologies and models are constructed, both
historically and today, and how they relate to real Japanese women and men in real
verbal interactions.
Second, the volume questions binary distinctions in social and linguistic catego-
ries such as Japanese women and men and Japanese female and male languages. Such
distinctions, often taken for granted in previous studies, assume that each category
is internally homogeneous, consisting of members who share characteristics distinct
from the members of the other category. And the category subjected to the most scru-
tiny has generally been women. In contrast, this volume emphasizes the importance
of examining and accounting for real language practices of both female and male
speakers whose identities in one way or another position them relative to the models
of the idealized average Japanese woman and man. They may fit the normative
patterns, for reasons that are context-dependent and strategic. These uses need ex-
planation. They may fall outside the normative models; the talk of these speakers,
too, requires explanation. The chapters in this book demonstrate that both normative
and nonnormative patterns of language-use-in-context should be identified and ac-
counted for in a principled way without delegitimizing the latter use by labeling it
incorrect or marginal. They show, in other words, that a sociolinguistic account of
Japanese language and gender relations can encompass a much wider group of Japa-
nese speakersfemale and malespeaking in more diverse social situations than
has hitherto been the case.
6 INTRODUCTION

Third, the volume demonstrates that research on Japanese sociolinguistic phe-

nomena cannot remain focused on static and abstract descriptions of the relationship
between language and social categories. In particular, the essays here move the field
away from the common but problematic assumption that one linguistic form, such
as a morphological element or a pitch height, can be directly linked to a single social
category, such as gender. This assumption presupposes that one chooses linguistic
forms simply because one belongs to the category woman (or man). However,
speakers do not automatically choose language because they belong to a particular
social category. Irvine argues that the relationship between the distributions of lin-
guistic and social forms is not straightforward; rather, it is more productively sought
in cultural ideologies of languagethose complex systems of ideas and interests
through which people interpret linguistic behaviors (1998:52). In this volume,
indexicality, or the relationship between linguistic expressions and social mean-
ings, is understood, not as direct and fixed, but rather as multiple, flexible, context-
dependent, and mediated by speakers beliefs and attitudes concerning language
use, which are variously influenced by hegemonic ideologies. That is, the social mean-
ings of a particular linguistic expression may be interpreted differently, depending
on the context as well as on the speakers (or the interpreters) beliefs about language
use. For example, a particular stereotypically gendered form in Japanese (e.g., a sen-
tence-final particle) may not necessarily be used to index the speakers gender, al-
though it may be interpreted as doing so by hearers. Sociolinguistic phenomena must,
thus, be understood through close examination of local contexts of language uses,
including ideational contexts. In addition to providing empirical analyses of diverse
real speaker data, then, the volume addresses how linguistic expressions are linked
to social meanings in specific contexts and attempts to account for the complex and
dynamic nature of the relationship between language and gender.
Fourth, as a corollary of the third point, the volume considers the use of lan-
guage in relation to the construction of identities and interpersonal relations (Wodak
1997, Bucholtz, Liang, & Sutton 1999). It examines the speakers active involve-
ment in language choice as a performative strategy for creating the desired social
context, in particular identities and relations. Bucholtz argues that identities, far from
being given in advance for individuals to step into, emerge over time through dis-
cursive and other social practices (1999:12). Viewing identities as multifaceted and
dependent on the relational context, this volume examines language use as a series
of strategic speaker choices made in the interests of creating a context-appropriate social
persona. With regard to gender, language is thus seen as a resource upon which speak-
ers as social agents draw in order to construct diverse forms of gender identities and
relations. Understanding such creative use of language is possible only by recognizing
the flexible and multiple relationship between language and social meanings.
Finally, this volume presents a methodological challenge to the field of Japa-
nese language and gender as well. Reliance on introspection and survey data has been
common in Japanese sociolinguistics, as noted by Shibamoto (1987), but these meth-
odologies tend to provide information only about normative or prescriptive usages
that reflect dominant linguistic and cultural ideologies. However, it is important to
clearly distinguish between beliefs about how one should talk and the way one actu-
ally talks. The chapters in this volume that examine language practices (part III) uti-
 INTRODUCTION 7

lize alternative methodologies to surveys and introspection. They all analyze actual
speech data gathered in specific social contexts employing sociolinguistic or discourse
analytic methods; furthermore, most analyses are supported by rich ethnographic data.
Such analyses reveal a complex and dynamic relationship between language and
gender that cannot possibly be envisioned by macrosociological ideologies.
These five issues are explored in the ensuing chapters, which are grouped into
three parts: part I, historical and theoretical foundations; part II, linguistic ideologies
and cultural models; part III, real language, real people.

I.2. Historical and theoretical foundations

Part I tackles ideological issues head-on. As mentioned earlier, one of the most no-
table gaps in Japanese language and gender research has been a clear distinction
between ideology and practice. This distinction entails (1) a close examination of
the normative ideology of language and gender and its implications in the Japanese
context, (2) a thorough investigation of the historical development of gendered speech
norms and their associated ideology, and (3) a principled analysis of the relation
between the normative ideology and the range of real speaking practices that consti-
tute Japanese verbal life. The chapters in part I all address one or more of these issues,
starting with a critical overview of the Japanese language and gender research field
and its ideological underpinnings and presuppositions (Yukawa & Saito), then pro-
ceeding to a reexamination of the ideology of womens speech and its implications
for linguistic analysis and practice (Okamoto), to tracing the historical development
of contemporary stereotypical feminine and masculine Japanese speech-as-
ideology (Inoue, Washi), to a look at the hegemonicand often damagingpsy-
chosocial effects of these ideologically and culturally salient styles on oft-ignored
speakers, in particular, lesbians and gays (Lunsing & Maree).
The chapters in part I provide an invaluable look back over the development
of what is often called joseego womens language in the first half of the twen-
tieth century, both as a social narrative of appropriate behavior for Japanese women
and as a story in part constructed by linguists but used more broadly for (largely
exclusionary) social ends. Part I also considers how the ideology of gendered speech
norms that developed over the years affects both linguistic practice and analysis.
Recognizing that we researchers of language and gender ourselves are often suscep-
tible to the hegemonic language and gender ideology, these chapters go a long way
toward providing a coherent framework that takes a distinction between ideology
and practice into account in investigations of Japanese speech. Part I closes with a
powerful statement of the importance of this distinction for a particular group of
nonnormative speakers (lesbians and gay men) struggling to find suitable speak-
ing practices usable to locate themselves socially in ways that do not violate their
inner sense of self.
Sumiyuki Yukawa and Masami Saitos comprehensive overview of Japanese lan-
guage and gender studies in the latter part of the twentieth century and into the present
traces the historical development of the field from the era of a persistent essential-
ism that relegated women and men into oppositional and internally undifferentiated
8 INTRODUCTION

categories toward constructivist frameworks that can incorporate both the dominant
discourses of femininity and the real behavior of speaking women. A most welcome
component of this survey is the rare chance to read in English about the pioneering
work of Jugaku Akiko, whose 1979 volume, Nihongo to onna The Japanese lan-
guage and women, established a framework for language and gender research cen-
tered around the notion of onnarashii kotoba feminine speech as a normative
constraint on womens speech behavior. Yukawa and Saito delineate Jugakus under-
standing of onnarashii kotoba as distinctive from other researchers treatment of this
concept in including not only language used by women but also the language used to
women and the topics deemed suitable for women, all of which reveal the workings
of linguistic and cultural ideologies in the structural oppression of women. Yukawa
and Saito note that while it is sometimes overlooked, Jugakus work remains ahead
of its time even today. Their chapter also reviews the work of Japanese feminists on
issues concerning language and gender, in particular sexism in Japanesevaluable
information that has not been readily available in English. It focuses on two groups
of feminists: activists in the uuman ribu womens lib movement in the early 1970s
and feminist scholars in the 1980s and 1990s. Yukawa and Saito demonstrate that
the former were concerned with oppression through language and were, in fact, en-
gaged in subversive linguistic practices, and call our attention to the efforts made
by the latter group toward eliminating institutionalized manifestations of gender
ideologies, such as sexist representations in dictionaries, newspapers, and other print
media. After documenting the Western-influenced Japanese language and gender
research led by Ide and her colleagues that formed the mainstream of research into
Japanese womens language through the 1980s and into the 1990s, the authors close
their chapter by outlining new research trends aimed at clarifying the relationship
between language, identityhere, gender identityand ideology. It is a matter of
some pride to us as editors that so many of the scholars mentioned by Yukawa and
Saito as contributing to these new lines of research have chapters in this volume.
Shigeko Okamoto turns to a sustained examination of just those language and
gender ideologies that have greatly influenced both linguistic practice and analysis
in the Japanese case. Her analysis centers around the generalization that women speak
more politely than men, arguing that close examination of such generalizations is
necessary in order to understand their ideological underpinnings as well as their rela-
tion to actual speaking practice. Through an extensive interrogation of the linkages
between politeness, honorifics, and gender made by linguists and other researchers,
on the one hand, and purveyed as social norms to the public via education and media,
on the other, Okamoto demonstrates the remarkable consistency of messages to
women that, in order to be onnarashii feminine, they should speak in ways that are
gentle, polite, and refined. In the examination of actual speaking practices that fol-
lows, however, she demonstrates that there is much less consistency in the ways that
women and men can be observed to speak. Okamoto then elucidates the relation
between the ideological and actual by critically examining the three ideologically
nonneutral assumptions that often serve as the foundation for theorizing about the
nature of womens language in Japan: (1) that most or all women share the same
attributes and therefore (should) use language in the same way, (2) that certain lin-
guistic expressions, such as honorifics, are inherently polite, and (3) that speaking
 INTRODUCTION 9

politely, that is, showing deference or refinement through the use of honorifics and
other expressions, indexes femininity or the female gender. Her critique of these three
assumptions and her reframing of them as simply one of several possible orienta-
tions toward social norms opens the door to much new empirical work on not just
the diversity of language use but also diverse interpretations of variant language forms
as individuals adopt, resist, or simply ignore normative statements concerning femi-
ninity and masculinity while engaging in real speaking events. Her point that femi-
ninity and masculinity manifest simultaneously at two levels, the ideological and the
practical, and that the speech forms chosen may not be interpreted in the same way
on both levels is pursued by the chapters in part III.
The next two chapters address the historical development of contemporary con-
structions of womens speech norms. They address the processes of construction and
dissemination of the contemporary complex of forms in Standard Japanese that form
normative joseego today. Miyako Inoues chapter focuses on the modernizing period
(late nineteenthearly twentieth centuries) and investigates the process by which
women and their linguistic practices came to embody the shifting boundary between
tradition and modernity. Her description of language modernization, built on pre-
vious scholarship (e.g., Komori 1992, 2000), is centered around the standardization
of Japanese and the development of the modern Japanese novel. It is, Inoue argues,
in the dialogues and reported speech of these new novelswhere womens voices
were objectified, reified, and re-presentedthat Japanese womens language was
constructed. Inoue describes for the reader the process by which writers selected,
systematized, and disseminated particular feminine forms, forms neither inherently
feminine nor inherently refined, which, through the popular consumption of katee
shoosetsu domestic novels by a female readership, served to ascribe these forms to
that new icon of appropriate or socially desirable femininity, the middle-to-upper-
class Tokyo woman. This newly created language, argues Inoue, is the origin of
todays Japanese womens language. Much work remains to be done both on the
specific processes by which the forms identified in her chapter came to constitute an
ideology of womens language and on their links to earlier articulations of ideas
about women and speaking practices, but Inoues investigation into the historical
process of creating a modern Japanese language and a womens language within
it clearly demonstrates the importance of historicizing womens speech, of incorpo-
rating an understanding of the sociopolitical nature of the indexical process (and an
appreciation of its power) into our analyses of sociolinguistic variation, and of the
political potential of reported or represented speech, a topic that occupies part II of
this volume.
Rumi Washi also argues for understanding Japanese womens language, which
she refers to as female speech, as an artificial construct. In her chapter, based on
an extensive survey of historical documents, she examines the relationship between
language policy, in particular language standardization policy, and the education of
women as it relates to female speech in the period between 1920 and 1945 (that is,
the period before and during World War II). Washi first outlines the various strands
of influence on women and their speaking practices in general during this period,
then turns to a detailed investigation of one semiofficial body, the National Language
Association, and its relationship to female educators. She first traces language policy
10 INTRODUCTION

developments from the 1930 founding of the National Language Association, charged
with rationalizing, improving, and promoting the national language, to the establish-
ment of a womens wing in 1939. The womens wing was established, in the words
of the associations then vice president, in order to obtain the cooperation of women
in preserving and honoring the national language. The chapter then turns to the
substance of that cooperation, the ideal form of female speech, and the process by
which the contemporary complex of features that were asserted to constitute female
speech was linked to much earlier historical forms of speech associated with women,
specifically nyooboo kotoba court womens language. Washi provides a compel-
ling rationale for female educators to cooperate in disseminating this language con-
struct based on the various associations of female or feminine speech styles with
high social position for women under ideologies of sexual complementarity (or, put
another way, of separate equality); it was only later that Japanese women were to
uncover the trick whereby they were led to espouse speaking patterns thatfar
from affording them separate but equal statusrelegated them to subordinate
positions in the larger social order.
Part I closes with Wim Lunsing and Claire Marees discussion of the implications
of hegemonic gender ideologies for groups of speakers who have been marginalized
by most previous research: lesbians and gays. The relation of gender to sexuality is
a topic that has been largely neglected in Japanese language and gender studies.
Lunsing and Maree challenge the linking of linguistic gender performance in Japa-
nese to dominant heterosexist discourses, arguing for a need to investigate the com-
plex relations between dominant gender norms for language use and individuals who
do not participate in the associated discourse of heterosexual normativity (with its
hentai queer offshoots, the effeminate gay man and the masculine lesbians).
The heart of their chapter is a poignant depiction of the struggles both gay men
and lesbian women make to fit their nonnormative sexual identities with the nor-
matively constrained choices they have for self-reference. Normative (Standard)
Japanese self-referencing practices require that women and men use different sets
of forms. But these gender norms are equally strongly associated with sexual ori-
entation norms; that is, female users of feminine self-reference forms (e.g., watashi)
are assumed to be womanly (hence, heterosexual) and male users of masculine self-
reference forms (e.g., boku) are assumed to be manly (and therefore also hetero-
sexual). The lingering distaste Lunsings and Marees consultants express for the
limitations of the pronoun choices available to them makes abundantly clear the
heteronormative underpinnings of Japanese language and gender ideology. Lunsing
and Maree argue, however, that prescription is never fully reflected in actual lan-
guage use and that self-reference choice is not the result of a speakers gender
identification but a reflexive negotiative strategy for dealing with a sense of multi-
ple selfhood. By focusing on the intersections of gender and sexuality as speakers
strive to make appropriate choices from among a set of stereotypicallythat is, ex-
plicitlygendered and implicitly sexually oriented language forms, Lunsing and
Maree allow us to see, first, the implicit conflation of gender and sexuality in both
linguistic and popular accounts of Japanese gendered language and, second, the
importance of incorporating notions of negotiation and variability into our analyses
of womens and mens language use.
 INTRODUCTION 11

I.3. Linguistic ideologies and cultural models

Part II focuses on the question of how Japanese women and men are conceptualized
and represented in Japanesea particularly crucial exercise for highlighting issues
of gender and language ideology. Each chapter examines how the cultural models
for women and for men are produced and reproduced through the language itself
through the lexical, morphosyntactic, discursive, and performative choices made when
representing women and men in print and other media.
The late twentieth century has been an important period in the reevaluation of
traditional gender relations in Japan; it has also been a period of hypersaturation
of everyday life by (multi)media presence. An understanding of the media messages
Japanese women and men are receivingabout how to be adequate or attractive
women or men, about how to view feminine and masculine social roles, about how
to construe love and marriageis critical to an understanding of the ongoing pro-
duction, reproduction, and transformation of gender relations. Rosenberger, in an
analysis of Japanese womens magazines, argues that print media have some power
over the way women categorize themselves, their wishes for self-actualization and
relationship, and their sexual desires (1995:143144). This is, we argue, equally true
for men. In a variety of ways, print and other media have the power to establish ideas
about what constitutes desired femininity and masculinity with respect to personal-
ity traits, behaviors, and so on. These ideas may not always be internally coherent
but can provide alternatives to the ideas provided by dominant institutions.
The close scrutiny given the representation of gendered selves in novels, maga-
zines, television dramas, and dictionaries provided in the chapters of part II sheds
new light on the role of ideological symbolic forms, forms that have the potential to
produce and reproduce cultural models for women and men in the Japanese social
field. Here, then, we are not concerned with the recognition that individual women
and men create individual female and male personae that may be more or less
stereotypically feminine or masculine but with the linguistic representation of
these stereotypes and its dissemination in society. Compared to the explicit commen-
tary on how Japanese women and men should talk detailed in part I, the linguistic
representations examined in part II are more implicit. Such covert prescriptionsor
implicit pragmatics (Woolard 1998:9)take the legitimacy of their models for
granted. Accordingly, they may be equally or even more effective in serving to regi-
ment women and men into normative femininities or masculinities and to reinforce
the hegemonic control of societal gender arrangements.
Janet S. Shibamoto Smith considers the linguistic representation of cultural
models for women and men by examining Japanese romance fiction. This chapter
is particularly important in bringing issues of emotionality and sexuality, especially
heterosexuality, into Japanese language and gender studies. Shibamoto Smith ana-
lyzes category romance novels to see how (ideal) lovers use language with each other
to convey their femininity and masculinity. Her analyses of first- and second-person
references and sentence-final particles in lovers dialogue show that heroines and
heroes make extensive use of normative gendered speech styles, which may serve as
references for readersthat is, real women and menas they set about falling in
love. Further, her analyses demonstrate that unlike Western romances, in which
12 INTRODUCTION

nonlinguistic attributes, particularly physical features, are emphasized in the depic-

tions of ideal heroines and heroes, language plays an essential role in the cultural
construction of heroes and heroines in Japanese romance novels. Shibamoto Smith
argues that lovers dialogue is a primary site for the expression of gender politics in
these novels. Her study thus illustrates not only the complex relationship of sex and
gender to language but also the role of language in constructing (the images of)
gendered subject positions in a particular activity (i.e., falling in love) through an
appeal to culturally grounded semiotic practice (Kulick 2000).
But, of course, there is more to heteronormative femininity and masculinity than
romance. Momoko Nakamura presents the results of her analysis of discourse con-
structs in two womens and two mens popular lifestyle magazines published between
1999 and 2000. Her analytic framework, a reformulation of Faircloughs (1989) model
of the dialectic between social structures and discourse, gives ideologies a central
mediating role. Within this framework, Nakamura asks how a specific gender iden-
tity not preestablishedor at least, not fully preestablishedin gender ideologies is
invented in discourse practices. Her argument starts with the magazine writers
and producers need to construct a specific identity for a specific set of targeted readers
as they are imagined to be. Using speech styles that are, again, imagined to appeal to
targeted readers, magazine discourse attempts to draw real young women or young
men into the discursively constructed magazine communityand to attract, thereby,
a regular subscriber and purchaser base. Nakamuras extended analysis illuminates
the gender-differentiated ways that such communities are organized, how their readers
are differently constructed as participating in the community, and how topics of
fashion and the like position the members of the respective communities in gendered
ways via language. It effectively shows us that several of the features stereotypically
associated with female patterns of speech (exclamations, inclusive stance marked by
hortatives, interactively oriented sentence-final particles) occur in the young womens
magazines, while stereotypically masculine patterns (absence of affect indicators,
imperatives, use of assertive da forms) occur in the young mens magazines. She
closes with a discussion of the contrastive metaphors for conceptualizing fashion
FASHION AS TASK versus FASHION AS BATTLEmetaphors that themselves align differ-
entially with gender ideologies. Her analysis provides us with a close look at how
the forms implicated in ideologies of gendered language can be and are used to con-
struct finely tuned, very specific images of femininities and masculinities suited to
particular magazines and their communities of readers.
Novels, magazines, comic books, and television inundate Japanese-speaking
subjects with images of suitable feminine or masculine behavior. The tenets for
proper female behavior and the behavioral paradox that these impose on women
are the focus of Laura Millers chapter on burikko phony girl(s). Burikko is a de-
rogatory label used to describe women whose speaking and other behavioral patterns
are seen as exaggerated and false exhibitions of navet, innocence, and cuteness.
Therein, of course, lies the behavioral paradox: many of the prescriptions for proper
female behaviorthat is, the linguistic reflexes of modesty and innocenceover-
lap with traits commonly associated with burikko. Millers chapter first presents a
description of burikko stylistic features, then draws on various popular print and visual
media sources for data on how burikko images are constructed using many of the
 INTRODUCTION 13

same features that are part of the stereotypical womens language. Examining media
representations as well as several ethnographic cases of women performing burikko,
Miller provides a rich analysis of the context in which the term burikko is used to
evaluate this exaggeratedly and childishly feminine form of speaking and acting,
that is, when and why a particular behavior is considered too feminine. These ob-
servations demonstrate that at the same time a Japanese woman is being taught that
being feminine is a desirable social trait, she also is learning that there are relatively
strict limits on the performance of femininity. Burikko is but one of many expressions
that are used to categorize and evaluate women. Millers careful analysis of this term
well illustrates how women are subject to close scrutiny of their behavior. She returns
us, that is, to Inoues point that woman has historically beenand, we see here, con-
tinues to bea regimented category, regimented in this case by the critical gaze and
commentary of other members of womens speaking communities.
In the final chapter in part II, Orie Endo provides us with updated and detailed
information about another regimenting force on the cultural construal of the category
woman: dictionaries. Unlike the popular texts and media productions described in
the previous three chapters, dictionaries are authoritative texts par excellence. Endo
and her associates in Kotoba to onna o kangaeru kai Group for Thinking about Lan-
guage and Women have been studying sexist language in dictionaries and other
authoritative texts (e.g., media and governmental guidelines for nonsexist language)
for over two decades. This chapter is a fine-grained analysis of the treatment of three
representative derogatory expressions related to women, onnadatera (ni) despite
being a woman, rooba old woman/old crone, and oorudo misu spinster (lit., old
miss) in dictionaries, on the one hand, and in newspapers and novels, on the other,
clearly demonstrating the substantial lag of dictionaries compared with popular or
even other official usage (e.g., government publications). Endo contextualizes the
changes found in the use of these words in both popular and (semi)official publica-
tions (such as publishing houses in-house guidelines for nondiscriminatory language
usage)and the unchanging treatment found in many or most of the dictionaries in
her 66-volume corpuswith respect to the various strands of popular protest against
discriminatory language in general, the effects of womens movements in particu-
lar, and national and local-level governmental action. She closes with an expression
of regret that dictionariesthe ultimate authoritative source for words and their
useshave not played a more active role in re-presenting women and their roles in
Japanese life.

I.4. Real language, real people

Part III presents empirical studies of real women and men in real Japanese interactional
contexts, with a primary, although not exclusive, focus on groups of speakers who
are typically not studied in research on language and gender. While previous studies
on (stereotypical) gender differences in speech have tended to focus on speakers of
Standard Japanese, in particular, female speakers who belong to a privileged group,
that is, middle- and upper-middle-class housewives, the speakers examined in these
studies vary widely in their social attributes, such as gender and sexual orientation,
14 INTRODUCTION

region, occupation, and age; they include speakers of regional dialects (Sturtz
Sreetharan, Sunaoshi), socially diverse male speakers (Miyazaki, Ohara, Sturtz
Sreetharan), junior high school students (Miyazaki), women in a variety of occupa-
tions, including farm women (Abe, Ohara, Sunaoshi), and lesbians (Abe), as well as
typical women, or middle-aged housewives in middle-class families (Matsumoto).
The chapters in part III investigate real talk produced by these diverse women
and men vis--vis the cultural models and ideological frames discussed in parts I and
II. They examine actual speech data closely with regard to the use of linguistic fea-
tures that are normatively associated with gender, such as self-reference and address
terminology, sentence-final particles, honorifics, (in)direct speech acts, and phono-
logical features. The results reveal impressive variation and extensive deviations
from normative or stereotypical usage, suggesting that deviant uses are meaning-
ful choices rather than mere exceptions or anomalies. They convince us that speak-
ers do not blindly follow dominant social norms and expectations but rather relate to
them variouslyby adopting them, negotiating them, contesting or resisting them,
or simply disregarding them because of their perceived lack of relevance to some
concrete situationand then choose expressions that they think most appropriate for
a given context. Further, speakers observed in these studies all use stereotypically
gendered expressions to convey not only femininity or masculinity but also a va-
riety of other pragmatic meanings (e.g., solidarity, relative power), suggesting that
indexicality is multiple and indirect and that the interpretation of gendered expres-
sions as feminine or masculine is not given but normative and ideological. These
chapters demonstrate how individual speakers as social actors use language strategi-
cally, by making pragmatic choices that draw simultaneously on their linguistic re-
sources and their beliefs about language use, and how such choices contribute to
constructing their identities and relationships.
Yukako Sunaoshi studies women who have historically been overlooked, a group
of linguistic others, namely, female speakers of a regional dialect living in a farm-
ing community in Ibaraki Prefecture. Sunaoshi analyzes interactions of three farm
women and an Agricultural Extension Adviser, matched with detailed ethnographic
descriptions of the social context, which are essential in understanding these inter-
actions. She identifies two discourse strategiesthe use of Ibaraki dialect and the
treatment of the farm women as the representatives of their householdsas means
for the advisers solidarity building with and empowerment of farm women. The
Ibaraki dialect these women use lacks the morphological, lexical, and phonological
features of the normative construct Japanese womens language, but it serves as
an important linguistic resource for them in constructing their identities and relation-
ships. Sunaoshi then asks if even the ideology of Japanese womens language, as
it is defined in terms of features of Standard Japanese, is shared by these rural women.
Her study suggests that research on Japanese womens speech must consider the extent
to which the dominant gender and linguistic norms are relevant to real women, espe-
cially women who are in some sense linguistic others, and must attempt to look at
their linguistic practices in their own righta suggestion that echoes Okamotos
(chapter 2, this volume) view that it is important to consider individual speakers
apprehensions of linguistic norms for specific local contexts.
 INTRODUCTION 15

Hideko Abe deals with another group of women who have been considered lin-
guistic others, namely, lesbians. Based on ethnographic research, Abe analyzes two
aspects of the discourse of women observed at lesbian bars in Tokyo: the use of ex-
pressions for categorizing themselves and others, such as rezu lesbian and futsuu
ordinary, and their speech styles as they relate to the use of linguistic features stereo-
typically associated with gender (e.g., self-reference and address terminology, sentence-
final particles). Her analyses demonstrate how the speakers discursively construct their
shifting identities and relations through strategic uses of these linguistic means. Par-
ticularly interesting are her observations that these women use masculine expres-
sions extensively, but not exclusively, and that the same speakers shift speech styles
(e.g., between the neutral anta you and masculine omee you) depending on the
context (e.g., interlocutors, speech act types). Her observations show that the speakers
negotiate the dominant gender and linguistic norms and relate to them differently de-
pending on the context and that their use of stereotypically gendered masculine ex-
pressions does not straightforwardly index masculinity (or femininity) but rather
conveys a variety of context-dependent meanings, including solidarity/distance, emo-
tional engagement, and resistance to the dominant gender and sexual norms.
In contrast to Sunaoshi and Abe, Yoshiko Matsumoto looks at the speech of
women who have been commonly assumed to use stereotypical feminine speech
stylesnamely, middle-aged, middle-class Standard Japanesespeaking housewives.
Her analysis centers on the use of expressions that convey different stancesforce-
fulness and delicacynormatively associated with masculine and feminine
speech (Ochs 1993) and reveals that, contrary to the common social stereotype, these
women do not use a traditional feminine speech style exclusively but rather exhibit
wide inter- and intraspeaker variation. Some of them use forceful expressions more
or less than others; further, variation is observed within a single individuals utter-
ances even in the same conversation. Matsumoto explains that forceful and delicate
expressions convey various pragmatic and social meanings, such as friendship, emo-
tion, and deference, and that women use these linguistic resources variably to con-
struct complex and flexible personae and relationships. Thus femininity, argues
Matsumoto, cannot be considered a single concept; rather, depending on speakers social
backgrounds and interactional goals, different women may negotiate traditional gen-
der norms and constraints differently in order to construct diverse female personae.
Yumiko Oharas chapter also addresses the question of how real speakers relate
to Japanese gender and cultural norms when using language in real social situations.
Ohara observes not only women but also menfour company employeesinter-
acting in two kinds of contexts: talking with customers and talking with acquaintan-
ces or friends. Unlike the other chapters in part III, Oharas essay examines the use
of prosody, specifically voice pitch levela feature often associated with gender.
Her analysis shows that this feature, too, cannot be directly linked to the dominant
gender norm. Although a high-pitched voice is normatively associated with polite-
ness and femininity, Ohara found that the women in her study did not always use it
and that men also used it at times. This, however, does not mean that the use of this
feature is haphazard. Taking a closer look at her data, Ohara accounts for her findings
as a reflection of the complex way that speakers treat this featurewhich functions
16 INTRODUCTION

both as a linguistic resource and as a cultural constraint. That is, both women and
men may use a high pitch level pragmatically as a linguistic resource to emphasize
certain parts of their utterances. However, Ohara observes an intriguing phenomenon
in which the female speakers, but not the male speakers, varied their pitch levels
substantially across interlocutors; that is, only the women used a considerably higher
pitch when speaking to customers than when talking to close friends. From this, Ohara
concludes that depending on the context, women but not men may use a high pitch
level according to the cultural constraint that links this feature to politeness and thence
to femininity.
Ayumi Miyazakis chapter investigates the language use of junior high school
students, both female and male, based on extensive ethnographic study. Her close
look at the students use of first-person pronouns reveals a complex and dynamic
process of meaning making that is far beyond the imagination of the fixed, dichoto-
mous picture of the normative usage. These students adopt, contest, and continually
negotiate traditional gender norms, which often brings about nonnormative uses, such
as girls use of the masculine pronouns boku and ore. Examining the relations
among students and the subcultures within the classroom as well as the students
metapragmatic discourses about these pronouns, Miyazaki demonstrates how the
students use pronouns to index a multitude of pragmatic meanings that concern not
only femininity and masculinity but also levels of formality, power, and solidarity.
For example, studious, pro-school girls use the feminine pronoun atashi, whereas
girls who are more nonconformist use masculine pronouns ore and boku; many
girls also prefer the nontraditional pronoun uchi because it is more informal and less
feminine than atashi. Powerful boys always use ore, while powerless boys use boku
when talking with powerful boys but ore when talking with other powerless boys or
girls. These observationspossible only through sustained ethnographic study
eloquently speak of how complex the relation of gender to a linguistic form can be
and how important it is to examine closely the context in which a linguistic form is
used in order to understand its social meaning(s) adequately.
In the final chapter, Cindi Sturtz Sreetharan focuses on mens speech. Rather
than restricting her study to speakers of Standard Japanese, Sturtz Sreetharan includes
speakers of a regional dialect (Kansai, or Hanshinkan, dialect) as well. Her analyses
of the use of sentence-final particles in their conversations exhibit not only individual
and regional differences but also considerable deviation from the normative usage,
which undermines the idea that there is a single Japanese mens language. Based
on her finding that all speakers used masculine forms relatively infrequently, Sturtz
Sreetharan concludes that these men do not utilize sentence-final particles to index
(traditional) masculinity (although they may resort to other means to do so). Rather,
they use sentence-final particles to express a variety of pragmatic meanings; for ex-
ample, masculine, or forceful, expressions are used to signal camaraderie, author-
ity, anger, and so on. Further, speakers of the regional dialect may avoid using forms
in Standard Japanese in order to create a sense of friendliness or solidarity. Sturtz
Sreetharan notes, however, that the metapragmatic discourses of these men suggest
that they are capable of using stereotypical strongly masculine speech, depending
on the context (e.g., in quarreling). Sturtz Sreetharans chapter, in line with those by
Lunsing and Maree, Miyazaki, Ohara, and Okamoto, demonstrates that men also
 INTRODUCTION 17

actively engage in discursive constructions of identities and relationships and use

language strategically according to their interactional goals.
The findings of the studies presented in part III together offer compelling evi-
dence for the importance of examining the social meanings of linguistic forms used
in real contexts. For example, the pronouns boku and ore are both typically charac-
terized as masculine first-person. However, the meanings conveyed by these pro-
nouns in specific contexts are far more complex than this characterization offers. For
example, high school boys may use ore to index relative powerfulness rather than
simple male speakerhood and boku to index relative powerlessness. High school girls
may also use ore (or boku), to reject traditional femininity or to be nonconformist,
but they are not indexing masculinity. Women at lesbian bars may use ore to estab-
lish solidarity with the interlocutor, while at the same time rejecting the dominant
gender and sexual norms, but again their use of ore does not index masculinity. Farm
women in rural areas in Tohoku may use ore regularly (chapter 10, this volume) as
part of their regional dialect, which can serve to establish solidarity among local
people. But this does not mean that they are either rejecting conventional gender and
sexual norms or trying to express masculinity. The meanings of ore and boku are,
we see, highly context-dependent. In like manner, the chapters in part III all demon-
strate that only a close examination of language use in real contexts enables us to
fully understand the complex and dynamic relationship between specific linguistic
choices and their social meanings.

I.5. Concluding remarks

The studies in this volume not only shed new light on issues of Japanese language and
gender but also contribute to deepening our understanding of the complex relationship
of sex, sexual orientation, and gender to language ideologies and language practices in
general. This volume also serves as an illustration of new approaches to the investiga-
tion of linguistic diversity and ideology in sociolinguistic research. Attention to real
Japanese women and men performing linguistic selves is long overdue; this volume
provides a first exploratory look at sexed/sexing, gendered/gendering, and sexually
oriented/orienting persons performing their linguistic lives against the backdrop of
normative ideologies in Japan. The contributions to this volume illuminate the terms
of these ideologies and trace real persons as social agents as they variously relate to
their dictates, by conforming to, negotiating, resisting, openly challenging, or disre-
garding them. The result is a rare glimpse into the heretofore largely hidden situated
linguistic practices of socially diverse Japanese women and men.

References
Abe, Hideko (2001). Speaking of power: Japanese women and their language. Munich:
Lincom Europa.
Befu, Harumi (1993). Nationalism and nihonjinron. In H. Befu (ed.), Cultural nationalism in
East Asia: Representation and identity, 107135. Berkeley: Institute of East Asian
Studies, University of California, Berkeley.
18 INTRODUCTION

(2001). Hegemony and homogeneity: An anthropological analysis of Nihonjinron.

Melbourne: Trans Pacific Press.
Bergvall, Victoria L., Janet M. Bing, and Alice F. Freed (eds.) (1996). Rethinking language
and gender research: Theory and practice. London: Longman.
Bucholtz, Mary (1999). Bad examples: Transgression and progress in language and gender
studies. In M. Bucholtz, A. C. Liang, and L. A. Sutton (eds.), Reinventing indentities:
The gendered self in discourse, 324. New York: Oxford University Press.
Bucholtz, Mary, A. C. Liang, and Laurel A. Sutton (eds.) (1999). Reinventing identities: The
gendered self in discourse. New York: Oxford University Press.
Cook, Haruko M. (1998). Situational meanings of the Japanese social deixis: The mixed use
of the masu and plain forms. Journal of Linguistic Anthropology 8 (1): 87110.
Eades, J. S., Tom Gill, and Harumi Befu (eds.) (2000). Globalization and social change in
contemporary Japan. Melbourne: Trans Pacific Press.
Fairclough, Norman (1989). Language and power. London: Longman.
Hall, Kira, and Mary Bucholtz (eds.) (1995). Gender articulated: Language and the socially
constructed self. New York: Routledge.
Ide, Sachiko (1979). Onna no kotoba otoko no kotoba (Womens language, mens language).
Tokyo: Nihon Keizai Tsushinsha.
(ed.) (1997). Joseego no sekai (The world of womens language). Tokyo: Meiji Shoin.
Ide, Sachiko, Motoko Hori, Akiko Kawasaki, Shoko Ikuta, and Hitomi Haga (1986). Sex
difference and politeness in Japanese. International Journal of the Sociology of Lan-
guage 58: 2536.
Inoue, Miyako (1996). The political economy of gender and language in Japan. Unpublished
Ph.D. dissertation, Washington University, St. Louis, MO.
Irvine, Judith T. (1998). Ideologies of honorific language. In B. B. Schieffelin, K. A. Woolard,
and P. V. Kroskrity (eds.), Language ideologies: Practice and theory, 5167. New York:
Oxford University Press.
Jugaku, Akiko (1979). Nihongo to onna (The Japanese language and women). Tokyo: Iwanami
Shoten.
Komori, Yoichi (1992). Buntai to aidentitii (Style and identity). Gekkan Gengo 21(10): 4855.
(2000). Nihongo no kindai (The Japanese language in modern times). Tokyo: Iwanami
Shoten.
Kroskrity, Paul V. (ed.) (2000). Regimes of language: Ideologies, polities, and identities. Santa
Fe, NM: School of American Research Press and Oxford: James Currey.
Kulick, Don (2000). Gay and lesbian language. Annual Review of Anthropology 29: 243285.
Maher, John C., and Kyoko Yashiro (eds.) (1995). Special issue: Multilingual Japan. Jour-
nal of Multilingual and Multicultural Development 16(1&2).
Maree, Claire (1997). Jendaa shihyoo to jendaa no imisee no henka: Eega Shinjuku Boys ni
okeru onabe no baai (Gender indexicality and semantic shifts in gendered meanings: The
case of an onabe in the documentary film Shinjuku Boys). Gendai Shisoo 25(13): 263
278.
Matsumoto, Yoshiko (1999). Japanese stylistic choices and ideologies across generations. In
J. Verschueren (ed.), Language and ideology: Selected papers from the 6th International
Pragmatics Conference, vol. 1, 352364. Antwerp: International Pragmatics Association.
Miller, Laura (1995). Japanese linguistic diversity. Panel presented at the 9th Annual Meet-
ing of the American Anthropological Association, Washington, DC.
Ochs, Elinor (1993). Indexing gender. In B. D. Miller (ed.), Sex and gender hierarchies, 146
169. Cambridge, MA: Cambridge University Press.
Ogawa, Naoko, and Janet S. Shibamoto Smith (1996). The linguistic gendering of an alterna-
tive Japanese lifestyle: Speech variation in the gay communities of urban Japan. In R. Ide,
 INTRODUCTION 19

R. Parker, and Y. Sunaoshi (eds), SALSA: Proceedings of the Third Annual Symposium
about Language and SocietyAustin, vol. 36, Texas Linguistics Forum, 2840. Austin:
University of Texas, Department of Linguistics.
Ogawa, Naoko, and Janet S. Shibamoto Smith (1997). The gendering of the gay male sex
class in Japan: A preliminary case study based on Rasen no Sobyoo. In A. Livia and
K. Hall (eds.), Queerly phrased: Language, gender, and sexuality, 402415. New York:
Oxford University Press.
Okamoto, Shigeko (1995). Tasteless Japanese: less feminine speech among young Japa-
nese women. In K. Hall and M. Bucholtz (eds.), Gender articulated: Language and the
socially constructed self, 297325. New York: Routledge.
Okamoto, Shigeko, and Shie Sato (1992). Less feminine speech among young Japanese
females. In K. Hall, M. Bucholtz, and B. Moonwomon (eds.), Locating power: Proceed-
ings of the Second Berkeley Women and Language Conference, vol. 2, 478488. Berke-
ley: Berkeley Women and Language Group.
Okamoto, Shigeko, and Janet S Shibamoto Smith (1998). Japanese speaking choices: Real
diversity meets the ideology of homogeneity. Panel presented at the 6th International
Pragmatics Conference, Reims, France.
Reinoruzu [Reynolds]-Akiba, Katsue (ed.) (1993). Onna to nihongo (Women and the Japa-
nese language). Tokyo: Yushindo.
Rosenberger, Nancy (1995). Antiphonal performances? Japanese womens magazines and
womens voices. In L. Skov and B. Moeran (eds.), Women, media and consumption
in Japan, 143169. Honolulu: University of Hawaii Press.
Saft, Scott (2001). Displays of concession in university faculty meetings: Culture and inter-
action in Japanese. Pragmatics 11(3): 223262.
Sato, Shie, and Shigeko Okamoto (1999). Reexamination of Japanese cooperative com-
munication style. In J. Verschueren (ed.), Pragmatics in 1998: Selected Papers from the
6th International Pragmatics Conference, vol. 2, 518527. Antwerp: International Prag-
matics Association.
Shibamoto Janet S. (1985). Japanese womens language. New York FL: Academic Press.
(1987). Japanese sociolinguistics. Annual Review of Anthropology 16: 261278.
[Shibamoto] Smith, Janet S. (1992). Women in charge: Politeness and directives in the speech
of Japanese women. Language in Society 21(1): 5982.
(2003). Gendered structures in Japanese. In M. Hellinger and H. Bussmann (eds.),
Gender across languages, vol. 3, 201225. Amsterdam: John Benjamins.
Silverstein, Michael (1979). Language structure and linguistic ideology. In P. R. Clyne,
W. Hanks, and C. L. Hofbauer (eds.), The elements: A parasession on linguistic units
and levels, 193247. Chicago: Chicago Linguistic Society.
(1985). Language and the culture of gender: At the intersection of structure, usage,
and ideology. In E. Mertz and R. J. Parmentier (eds.), Semiotic mediation: Sociocul-
tural and psychological perspectives, 219259. New York: Academic Press.
Sturtz, Cindi L. (2001). Danseego da zo! Japanese mens language: Stereotypes, realities,
and ideologies. Unpublished Ph.D. dissertation, University of California, Davis.
Sugimoto, Yoshio (2003). An introduction to Japanese society, 2nd ed. Cambridge, MA:
Cambridge University Press.
Sunaoshi, Yukako (1995). Japanese womens construction of an authoritative position in
their communities of practice. Unpublished masters thesis, University of Texas at
Austin.
Takasaki, Midori (1993). Josee no kotoba to kaisoo (Womens language and social class).
Nihongogaku 12(6): 169180.
Weiner, Michael (1997). Japans minorities: The illusion of homogeneity. New York: Routledge.
20 INTRODUCTION

Wodak, Ruth (ed.) (1997). Gender and discourse. London: Sage.

Woolard, Kathryn A. (1998). Introduction: Language ideology as a field of inquiry. In P. V.
Kroskrity, B. B. Shcieffelin, and K. A. Woolard (eds.), Language ideologies: Practice
and theory, 347. New York: Oxford University Press.
Yoshino, Kosaku (1992). Cultural nationalism in contemporary Japan: A sociological en-
quiry. London: Routledge.
 PART I

HISTORICAL AND
THEORETICAL FOUNDATIONS
This page intentionally left blank
 1

SUMIYUKI YUKAWA
MASAMI SAITO

Cultural Ideologies in Japanese

Language and Gender Studies
A Theoretical Review

F or the first two decades of its history, language and gender research could not free
itself from a persistent essentialism that assumed women and men, and womens
language and mens language, as internally undifferentiated categories. It at-
tempted to connect specific linguistic forms or strategies directly to the speakers
sex, theorizing womens speech style as either dominated by men (e.g., Spender 1980)
or culturally different from that of men (e.g., Maltz & Borker 1982, Tannen 1990).
However, in the 1990s a new theoretical framework emerged, based on recent de-
velopments in social and feminist theories and discourse-based methodologies. In-
fluential essays by Eckert and McConnell-Ginet (1992) and Gal (1991) proposed a
new approach to analyzing gender in language use, one that overcomes the essen-
tialism of the earlier studies by focusing on cultural ideologies mediating diverse
language forms and gender meanings. As Susan Gal observed, the study of language
and gender is significantly enhanced by simultaneous attention to everyday practices
on the one hand, and on the other to the ideological understandings about women,
men, and language that frame these practices and render them interpretable in par-
ticular social contexts, historical periods, and social institutions (Gal 1995:180). The
mandate for new research is to analyze the hegemonic power of linguistic ideolo-
gies and the ways in which speakers attempt to parody, subvert, resist, contest, or in
some way accommodate these positioned and powerful ideological framings (Gal
1995).
The new double emphasis of the field on domination and resistance is articu-
lated in one way or another in all the chapters in this volume, the first collection of
essays to reexamine Japanese and gender from the new third-wave language and

23
24 HISTORICAL AND THEORETICAL FOUNDATIONS

gender research perspective (Bucholtz 1999a:viii), which focuses on mechanisms

of gender hegemony on the one hand and the construction of subversive identities
on the other (see Bucholtz 1999b).
This chapter reviews the development of Japanese language and gender studies
(hereafter JLGS) from this perspective. Specifically, it highlights the work of Jugaku
Akiko (1979), who focused on gender ideologies and subversive linguistic practices
long before Western language and gender researchers realized their central impor-
tance. Also, by attending to the metalinguistic discourse generated by linguists of
different schools we hope to reveal when, how, and by whom such influential cate-
gories as Japanese, Japanese womens language, and onnarashisa [Japanese]
womanliness were formed and sustained. In other words, our review attempts to
uncover the cultural ideologies of JLGS.

1.1. National-language studies and the construction

of Japanese womens language

Studies of Japanese and women preceded by many decades the rise of feminism in
Japan in the 1970s. Scholars in kokugogaku national-language studies have de-
scribed and cataloged features of normative womens speech style (personal pronouns,
sentence-final particles, honorifics, and so on) since the beginning of the twentieth
century. As nationalism and militarism increasingly took hold of the country, more
and more scholars alluded to the traditional speech style of Japanese women and
claimed it as part of a valuable cultural heritage that had been maintained for centu-
ries. For example, one leading linguist wrote during World War II, It is now being
noted that the way of Japanese women is beautiful and superb, standing out from the
ways of women throughout the world. Related to the way of Japanese women, Japa-
nese womens language also seems to be a rare phenomenon in the world (Kindaichi
1942, cited in Endo 1997:163).1 Some argued that nyooboo kotoba, the distinctive
register used by court ladies in and after the fourteenth century, formed the founda-
tion of the contemporary womens speech style (cf. Washi 2000, chapter 4, this vol-
ume). These studies assumed that Japanese has a distinct womens speech style,
namely, Japanese womens language, and that all Japanese women (should) speak
it. This assumption and the essentialism behind it were inherited by many strands of
work in JLGS. Thus, the study of womens speech by national-language scholars has
strongly influenced JLGSs course of development.
Many later national-language scholars have also played an important part in
reinforcing the cultural dictates of feminine speech style (cf. Endo 1997, Nakamura
2001). They commented on emerging changes in womens speech style, denounc-
ing them as deviant from beautiful Japanese womens language. Their comments
were very often based on stereotypes that connected womens polite and feminine
speech style to onnarashisa womanliness and respectability, and unfeminine
speech style to improper upbringing. Such comments in turn fed back to commonsense
beliefs, sustaining a language ideology that links a particular speech style with de-
sirable femininity (cf. chapters 6 and 7, this volume).
 CULTURAL IDEOLOGIES IN JAPANESE LANGUAGE AND GENDER STUDIES 25

1.2. The womens liberation movement in the

1970s: Resistance to gender norms in language

It is against such a background that in the 1970s the Japanese womens movement
started to recognize and act against the oppressive force of language. Many of the
newly formed womens groups challenged traditional gender ideologies and called
their own movement uuman ribu (from the English phrase womens lib) to differ-
entiate it from the existing womens movements for peace and democracy.
There is some evidence that the people in the Japanese womens movement,
especially the womens liberation movement in the 1970s, were aware that women
were generally represented in a derogatory manner and that feminine speech style
forced them to speak as respectful subordinates and prevented them from forcefully
asserting themselves as the equals of men. Of particular significance is the fact that
many uuman ribu activists abandoned normative womens language and spoke in an
assertive, nonpolite, and utterly unfeminine style. They dropped honorifics from
their speech and used vulgar vocabulary like gaki hungry devils for kodomo chil-
dren, meshi for gohan meal, and temee for anata you. Their speech style was
part of an attempt to reject prescribed femininity and to revolutionize their own
gendered consciousness.
The womens movement in the 1970s was also characterized by its active ef-
forts to change sexist language practices. The written texts of some activists are scat-
tered with remarks that show their recognition of the patriarchal value system inherent
in Japanese. These remarks include the points that ningen, literally human being,
practically speaking only refers to men and that conjoined phrases such as otoko to
onna men and women reflect a male-centered viewpoint because the term for men
comes first. They also insisted on using the basic term onna to refer to women, which
is often avoided because of sexual connotations, rather than the more tasteful al-
ternatives fujin and josee. In 1975, the womens group, Kokusai fujinnen o kikkake
ni koodoo o okosu onnatachi no kai The International Womens Year Action Group
was founded, consisting of Diet members, lawyers, media critics, teachers, office
workers, and housewives. Their activities included a protest against a hit television
commercial for instant noodle soup that promoted stereotyped gender roles. It fea-
tured a girl saying, Watashi tsukuru hito I am the one who cooks it, joined by a
boy who yelled, Boku taberu hito I am the one who eats it. The commercial was
discontinued because of the protest, an epoch-making event in the history of the Japa-
nese womens movement. The group also demanded that the influential media giant
NHK (Japan Broadcasting Corporation) stop using women in unimportant assis-
tant roles in programs, increase the number of female staff, and refrain from using
sexist words and phrases, including the notorious shujin master for husband.
Past reviews of JLGS (cf. [Reinoruzu [Reynolds]-Akiba 1993:224, Abe 1995:651
652, Nakamura 2001:9) argue that while feminism in the West concentrated its ef-
forts on abolishing sexism in language, feminism in Japan has paid little attention to
matters of language. As we have seen, however, Japanese feminists in the early 1970s
were seriously concerned with oppression through language and were engaged in
26 HISTORICAL AND THEORETICAL FOUNDATIONS

subversive linguistic practices. Their recognition of the important relationship between

language and hegemonic gender ideologies was to be fully articulated by Jugaku
Akikos (1979) monumental treatise on language, gender ideologies, and womens
resistance.

1.3. Jugaku Akikos focus on ideologies

and subversive practices

In Nihongo to onna The Japanese language and women, Jugaku single-handedly

provided a comprehensive theoretical framework for the study of language and gen-
der. Nationally known for her lifelong commitment to postwar democratizing move-
ments, she had also supported the forementioned campaign against the television
commercial. Though her book was written for a general readership, it provided an outline
of an extensive research program with illustrations from her 20 years of research as a
distinguished scholar in national-language studies. Though the significance of the book
in JLGS can be compared to that of Robin Lakoffs Language and Womans Place
(1975) in Western language and gender studies, it oriented later researchers more di-
rectly to the important connection between language and cultural ideologies.
The book explicates the workings of language in sustaining gender ideologies
prevalent in Japanese society up to that time, with analyses of stereotyped images of
women in popular songs and the cultural stock of categories, phrases, and proverbs
used for criticizing women who challenge normative femininity. The core of Jugakus
theoretical framework is the idea that linguistic and cultural ideologies related to
Japanese women severely constrain their behavior and ways of life. Her key concept
onnarashii kotoba womanly language/speech/words (which would ordinarily mean
womanly speech style or womanly expressions), enabled her to go beyond essential-
ism and come to grips with gender ideologies over a decade before Western feminist
linguists did so. She theorized that the concept consists of three components: (1) lan-
guage designed for a female audience, (2) topics chosen for a female audience, and
(3) linguistic strategies used to display that the speaker is a woman. A comparison of
these components with Lakoffs (1975) definition of womens language as (1)
womens speech style and (2) words that describe women shows why Jugaku was
able to tackle the issue of hegemonic cultural ideologies while Lakoff did not go
beyond gender dualism.
The first component of onnarashii kotoba is language used with a female audi-
ence in mind. Jugakus quantitative study of womens magazines had revealed some
distinctive stylistic features, including (1) a greater percentage of exclamatory and
uncompleted sentences, which reflects a belief in womens emotionality and lack of
logical thinking; (2) a smaller number of proper nouns, which implies that women
are less interested than men in learning about the world; and (3) an abundant use of
the second-person pronoun anata you instead of ellipsis2 to directly address the
reader, whichbeing a typical feature of advertisementstreats women as if they
were more susceptible to insincere seduction. These features, Jugaku argues, are part
and parcel of the tendency to think of men as the norm and women as needing spe-
cial consideration (This is too difficult for women to do/understand/etc.).
 CULTURAL IDEOLOGIES IN JAPANESE LANGUAGE AND GENDER STUDIES 27

The second component of onnarashii kotoba involves topics chosen specifically

for a female audience. Womens magazines are filled with articles on fashion, home-
making, and the private affairs of celebrities. Jugaku argues that this limits the range
of topics women read about and discuss, which in turn provides a reason for men to
look down on them.
The third and final component of onnarashii kotoba is linguistic strategies that
serve to display that the speaker is a woman. Jugaku believes that onnarashii kotoba
is part of the general concept of onnarashisa womanliness or idealized femininity
that regulates the thinking and behavior of Japanese women in detail. A womans
ways of doing thingsfrom her way of walking and talking to the course of life she
choosesare subject to criticism based on alleged deviance from these norms of
idealized femininity. Jugaku notes that Japanese is replete with grammatical and
prosodic devices through which women can display their onnarashisa. However, in
contrast to many of her colleagues in national-language studies, she refuses to reify
Japanese womens language. Instead of reviewing the same old linguistic devices,
she emphasizes their arbitrary and intentional nature: Immersed as we are in a lan-
guage like Japanese that manifests an excessive ability to leave traces of onnarashisa
in form, she claims, we may even suffer the illusion that onnarashisa in speech is
indispensable. However, onnarashisa in speech is the result of very carefully ma-
nipulating various features of Japanese to create the appearance of natural femi-
ninity (Jugaku 1979:22).
In light of the constructed nature of womens language, Jugaku focuses on
the subversive acts of women who challenge onnarashisa and onnarashii kotoba in
spite of personal risks. She describes the linguistic practices of a group of farmers
wives (cf. chapter 10, this volume). These women were often told by their husbands
and their parents to shut up and stay out of important matters. Jugaku reports how
they began to speak up to solve personal and community problems. They tried to
say what is difficult to say to the person to whom it is most difficult to say it to
(1979:200) and initiated changes at home and in the community.
Jugaku further notes an interesting subversive linguistic practice emerging among
schoolgirls, that is, their use of boku, a presumed sex-exclusive first-person pro-
noun for boys and men (cf. chapter 14, this volume). She characterizes it as a strat-
egy by which girls make it impossible for boys to make their masculinity salient
through language, and argues that it is evidence for womens advance into mens
world, deserting the speech style that places them in a kind of safety zone (1979:82).
The book ends with a chapter on arguments between married couples, which encour-
ages Japanese women to break their silent obedience and start changing the hierar-
chical gender arrangement at home.
When we look back on the history of JLGS and Japanese womens studies in
the two decades after the book, Jugakus stress on womens agency in the construc-
tion of personal and societal gender relationships is remarkable. She is as much con-
cerned with showing Japanese women what they can do to change their life and the
society as she is with revealing the workings of linguistic and cultural ideologies in
the structural oppression of women. This contrasts with later JLGS and Japanese
womens studies, which largely remained concerned with uncovering and protest-
ing against prevalent sexist practices. Although Nihongo to onna was read quite
28 HISTORICAL AND THEORETICAL FOUNDATIONS

extensively throughout the 1980s, unfortunately it had only a partial influence on

the later development of JLGS. Its influence can be seen in studies of gender repre-
sentation (see section 1.5) but not in sociolinguistic studies of womens speech be-
havior (see section 1.4). The then-emerging field of Japanese womens studies chose
to distance itself from the radical feminism of the uuman ribu movement and turned
for its source of knowledge and authority to the U.S. feminist scholarship of the day
(cf. Ehara 1990). In the field of language and women, this scholarship was intro-
duced into Japan by Ides (1979) discussion of early Western feminist linguistics and
later by Akiba Reynoldss translation of Lakoff (1975/tr. 1985) and Spender (1980/
tr. 1987).

1.4. Essentialism in womens language studies

From the vantage point of the twenty-first century, we see that studies of sex differ-
ences in speech behavior throughout the 1980s and early 1990s failed to advance
JLGS in significant ways. The reasons are both theoretical and methodological, but
underlying them all is an implicit essentialism that caused the researchers to concep-
tualize women and men as homogeneous groups of language users.
Led by Ide, sociolinguistic studies of womens language and linguistic sex
differences emerged in the early 1980s and until recently formed the mainstream of
JLGS. Ide (1993) claimed that she aimed for a scientific study of the linguistic fea-
tures of womens language, not motivated by feminism. She had studied socio-
linguistics in the United States, but her approach did not move JLGS toward studying
the actual speech of women in specific social situations. Instead, she discussed in
abstract terms normative usages of Japanese womens speech assumed to be used by
middle-class women in the Tokyo metropolitan area, neglecting the actual speech of
women, many of whom spoke nonstandard dialects or failed to conform to norma-
tive usages for various reasons. By assuming that all Japanese women speak womens
language, she reinforced the essentialist assumption of many national-language
scholars. In effect, the work of Ide and her colleagues represents part of a long period
in the history of JLGS in which womens language was objectified or taken for
granted.
Ide persistently claimed that Japanese womens language, rather than a lan-
guage of the oppressed sex as feminists tended to characterize it, displays dignity
and that Japanese women have more actual power in domestic matters than men, with
their husbands devoting most of their time and energy to their companies (Ide 1979,
1993). This theory neglected the hierarchical gender structure and its mechanisms
of control, thereby reinforcing hegemonic gender ideologies. However, it remained
influential in JLGS in the 1980s and 1990s, standing in opposition to Reinoruzu-
Akibas argument that represented the feminist camps emphasis on male dominance
and their call for degendering Japanese (e.g., Reinoruzu [Reynolds]-Akiba 1993).
As a sociolinguistic research program, womens language studies carried out
by Ide and her colleagues had some fundamental limitations shared by many re-
search projects in Japanese sociolinguistics. Their research typically assumed sex
differences in politeness level and tried to establish such differences quantitatively
 CULTURAL IDEOLOGIES IN JAPANESE LANGUAGE AND GENDER STUDIES 29

with questionnaires about the participants self-reported linguistic behavior (Ide 1979,
1990). Thus, instead of examining features of actual speech, these researchers really
surveyed speakers knowledge of and attitudes toward normative usage of language,
that is, their language ideologies.
Two methodological explanations can be given for the lack of development in
these sociolinguistic studies of womens speech. The first is that JLGS failed to keep
up with the important shift of focus in Western feminist linguistics from pursuing
abstract sex differences to examining language in the context of actual social inter-
actions through which gender and identity are constructed. Consequently, JLGS was
not able to share important new insights that eventually revolutionized language and
gender studies overseas, for example, the discursive construction of identities, the
inseparability of gender identity from other social identities, and differences among
women. The unfortunate result was that gender dualism was retained in JLGS. Sec-
ond, neither discourse analysis nor ethnography was established as a research method
in sociolinguistics in Japan. Tape-recording social interactions and analyzing their
details from a theoretical perspective is only now becoming part of the methodology
used in JLGS in Japan. As Shibamoto [Smith] (1985:56) observed years ago, JLGS
scholars seldom listened to real people talk in their everyday life.

1.5. Toward eliminating institutionalized

manifestations of gender ideologies

In the mid-1980s, studies of the gender norms reflected in language flourished. They
were carried out by feminist scholars mainly in sociology and national-language stud-
ies who had been encouraged by Jugakus book. Their inquiries into the cultural
ideologies that underlay sexist gender representations developed in the 1980s and
1990s into criticisms of various forms of institutionalized language practices.
Endo Orie and her colleagues (Kotoba to onna o kangaeru kai Group for Think-
ing about Language and Women 1985) examined dictionaries and analyzed how
women and men are represented in definitions and examples. They found that the
definitions of words related to women and men present an image of men as strong,
reliable, taking initiative, and influencing the course of womens lives. The images
of women in these definitions were passive. They were expected to be nice and pretty.
Examples in the dictionaries were found to describe men as agents of actions, whereas
women tended to be depicted only in terms of appearance and sexual promiscuity.
Arguing that such a stereotypical treatment of women and men reproduces asymmetrical
gender arrangements, the group called for dictionaries free from sexism. Some major
publishers have honored their proposal and made revisions. The group has been up-
dating their analyses as new editions come out (chapter 9, this volume). They have
also been examining Japanese textbooks from the same perspective.
Tanaka Kazuko (1993) analyzed newspaper articles and found that they rein-
force a patriarchal viewpoint based on a double standard for women and men. She
identified several conventional sexist practices, including (1) unnecessarily mark-
ing women (female company president, etc.); (2) making the female subject
subordiante by referring to her in relation to a man as, for example, Yoko, the wife
30 HISTORICAL AND THEORETICAL FOUNDATIONS

of a construction worker, Mori Taro, instead of just saying Mori Yoko; and (3)
extensively using stereotypical images that depict women as emotional, considerate,
devoted, and passive. Tanakas analyses of sexist discourse strategies represent an
advance over earlier feminist media criticism, which focused only on obviously dis-
criminatory words and phrases.
Building upon her work and moving further from sexist words to sexist dis-
courses, Media no naka no seesabetsu o kangaeru kai The Workshop on Gender and
Media (1991) focused on the ways media discourse constructs gender ideologies. It
is noteworthy that the group not only analyzed newspaper articles and problematized
sexist representations of women but also made it a rule to provide alternative ex-
pressions and repeatedly met with local newspaper reporters to discuss what could
be changed in the production process to alleviate the problems. More recently, they
have proposed the first guidelines in Japan for gender-equal media discourse, based
on a study of media guidelines in the United States (Ueno & Media no naka no
seesabetsu o kangaeru kai 1996). Their feminist guidelines are reflected in the re-
vised versions of the reporters guidelines issued by such news media as Kyodo News
Agency and Asahi Shimbun, one of the leading national papers (cf. Gallagher 2001:174
175). Many other studies have been carried out that examine the asymmetrical gen-
der ideologies manifested in Japanese and how they can be changed. Notable examples
include (1) studies of the reference term shujin master for husband (Ogino 1992,
Fukuda 1993, Yamaguchi 1998), (2) gendered reference terms in general (Endo 1992),
(3) Sino-Japanese characters related to women (Kawata 2000), and (4) stereotypes
about female and male language use (Takeda 1990, 1991). Saito (1994) and Sasaki
(1994) examined how English textbooks published in Japan reflected feminist lan-
guage reform in English-speaking countries.

1.6. Reconstructing JLGS: Hegemonic ideologies

and resisting identities

As was mentioned in the introduction to this chapter, a new link between recent social
and feminist theories and discourse analytic approaches to identities radically trans-
formed language and gender research in the 1990s. However, JLGS inside Japan took
several years to respond to these new developments in the field in Western coun-
tries. One reason is that there was a persistent tendency to reify Japanese womens
language and dichotomize gendered language use. For example, as late as 1998 Ide
remarked in a published conversation, In Japanese, men and women use different
first-person pronouns. That proves that it has mens language and womens lan-
guage (Ide, Keiko, & Machi 1998:35). Another reason is that Spenders (1980/tr.
1987) central message, that women as a homogeneous group are oppressed by men
as a group, still formed the core of Japanese feminist thinking about the relationship
between language and gender.3
Nevertheless, Nakamura (1993) suggested a new direction for JLGS. She argued
that previous research in JLGS emphasized knowledge of language rather than lan-
guage use, prescriptions rather than descriptions, and a synchronic approach rather
 CULTURAL IDEOLOGIES IN JAPANESE LANGUAGE AND GENDER STUDIES 31

than a diachronic approach. Such tendencies were based on the premise that linguistic
studies were objective. Pointing out that Western feminist linguistics had started to
question objectivism and scientism, she proposed that JLGS do the same. Nakamuras
proposal notwithstanding, research along these lines did not begin until the mid-1990s.
Studies that use this new theoretical framework and new methodologies are
steadily increasing in JLGS both in and outside Japan. These studies are rethinking
both the concept of the Japanese language and the traditional view of gender.
Though it has been generally taken for granted, the category of Japanese or the
national language has been employed to erase linguistic heterogeneity or naturalize
the boundaries between groups of people (cf. Gal & Irvine 1995, cited in Nakamura
2001:216217). Gender has been conceptualized in binary terms, asking only how
women and men speak differently. The new focus of research is the relationship
between identity and ideology, that is, how, and through what linguistic practices
the speakers identity is constructed under the influence of dominant ideologies.
This renewed focus of research in JLGS has been pursued via two different ap-
proaches. The first approach traces how Japanese and gender have been con-
ceptualized, analyzing the history of the metapragmatic discourse that has naturalized
and sustained dominant ideologies. The second approach focuses on agents who have
been marginalized in the earlier studies and examines the strategies with which they
construct new identities in the face of hegemonic gender ideologies. The two ap-
proaches analyze, in other words, how hegemonic ideologies have been maintained
and how they have been negotiated or resisted.
The first approach is represented by the work of M. Inoue (1994), Endo (1997),
Nakamura (1995, 2001), and Washi (2000). Based on a previous study (Komatsu 1988)
that compared colloquial sentence-final forms used in two novels published in 1813
and 1909, Inoue argues that the cultural category of Japanese womens language
emerged in the context of Japans early modernization in the late nineteenth century,
when state formation and industrialization rapidly advanced and both language and
womens role came to be articulated as national issues (Inoue 1994:322). She ob-
serves: The construction and dissemination of womens language is closely linked to
the construction and dissemination of the doctrine of good wife and wise mother and
to the larger political project of the consolidation of the nation-state (1994:325).
Endo (1997) documents the role played in this process by linguists, educators,
and government officials, citing many of their comments on womens language use.
Integrating Inoues and Endos perspectives, Nakamura (2001) examines metaprag-
matic discourse on womens language use throughout Japanese history. She con-
cludes: What is most important about all these comments on womens language use
over several hundred years is not what they specifically encouraged or discouraged
about womens language use, but the maintenance of the concept of womens lan-
guage itself (2001:210). As she has argued in Kotoba to feminizumu Language
and feminism (1995), the category of womens language leads us to believe that
there are characteristic features common to all Japanese womens language use. Thus,
the category on the one hand helps to keep women in submissive silence and on the
other hand seduces them into using the stereotypical feminine speech style, which
most likely protects them from criticism.
32 HISTORICAL AND THEORETICAL FOUNDATIONS

Washi (2000) specifically focused on the function performed by the discourse

in national language studies that connected womens speech to nyooboo-kotoba,
the register of women who served in Japans imperial court in the fourteenth century
and later. As was mentioned earlier, national-language scholars repeatedly claimed
that Japanese womens speech styles had its origin in this register. Nyooboo kotoba
consists of lexical items substituting for ordinary words for foods, clothes, kitchen
utensils, and so fourth. Scholars argued that its use represented refined demeanor
and displayed the privileged status of the speaker (see Shibamoto [Smith] (1985:
2931) and Abe (1995:653654) for descriptions of nyooboo kotoba in English).
Washi examined the discourse of these scholars and concluded that it naturalized
such concepts as Japanese, the national language, and womens speech.
Considering that Japanese or the national language was created as a means of
culturally integrating the country into a modern state (cf. Lee 1996), Washi argues
that the discourse of these national-language scholars performed the functions of
justifying a state centralized around the emperor and of mobilizing women to ad-
here to its prescriptions of their own accord as good subjects of the imperial state
(see also chapter 4, this volume). Washi rightly emphasizes the necessity of look-
ing closely for regional and class-related differences among the Japanese language
and Japanese women, instead of conceptualizing them as internally undifferen-
tiated categories.
Let us now turn to some of the studies that constitute the second approach, which
focuses on how groups of people who had been neglected in past JLGS construct
their identities while coping with hegemonic gender ideologies. Okamoto Shigekos
work on speech style overcomes many of the methodological problems of previous
sociolinguistic studies in JLGS. Okamoto and Sato (1992) tape-recorded conversa-
tions between women and analyzed the sentence-final forms used by women of dif-
ferent age-groups. They found that young women used far fewer stereotypically
feminine forms (e.g., wa, nano, kashira) and more forms that had been regarded as
masculine (e.g., zo, da, dayo). Okamoto (1995) analyzed the pragmatic meanings
of young womens speech styles in specific social contexts, revealing that their un-
feminine speech styles, which had been neglected as exceptions in previous studies,
are actually chosen strategically according to multiple social variables (e.g., age,
marital and occupational status, degree of intimacy, formality level of the situation)
in order to communicate desired pragmatic meanings and to construct identities and
relationships. Her work highlights the variability of the actual language behavior of
Japanese women and shows that Japanese womens language is not reality but
ideology.
Another researcher within this approach, Ohara (1992), measured the pitch level
of bilingual Japanese female and male students when they spoke Japanese and En-
glish. She found that female students but not male students spoke at a significantly
higher pitch when they spoke Japanese than when they spoke English. Her finding
suggests that this aspect of womens speech behavior, which had tended to be con-
sidered wholly biologically determined, is actually influenced by historical and cul-
tural conditions. That is to say, Japanese women choose to speak at a higher pitch
under the influence of a gender ideology that says it is desirable for women to speak
 CULTURAL IDEOLOGIES IN JAPANESE LANGUAGE AND GENDER STUDIES 33

in a high voice (see also chapter 12, this volume). Ohara suggests that the same group
of women speak English at a lower pitch because they are freed from the influence
of Japanese gender ideologies when they speak English.
Other studies focus on how gender ideologies influence the language use of
understudied groups, especially those with nonnormative sexualities or genders.
Ogawa and Shibamoto Smith (1997) examined personal pronouns and sentence-final
particles used by a gay couple in a documentary film and analyzed the relationship
between their speech patterns and gender identities. Similarly, Maree (1997) focused
on different first-person pronouns used in a film about transvestites who lived in
Shinjuku, Tokyo. Her analysis showed that those terms of self-reference (as well as
their nonuse) were strategically chosen according to contextual features like the topic
of the conversation and the nature of relationships and that they represented the per-
formance of new identities different from those of heterosexual men.
Integrating the third-wave language and gender research perspective with
critical discourse analysis (CDA), Saito (2001) reexamined media discourse on the
womens liberation movement in Japan. Although previous studies had stressed that
the movement was marginalized by the media, she pointed out that this view had
been influenced by a research framework that was exclusively committed to expos-
ing the mechanisms by which the media maintained hegemony. Her feminist CDA
showed that the media was not a monolith and that some media discourse actually
constructed subversive gender identities.

1.7. Conclusion

The past decade has seen a burgeoning general academic interest in the workings of
cultural and language ideologies in the course of Japans modernization. Work such
as Lee (1996) and Komori, Kono, and Takahashi (1997) has shed light on how con-
structs such as Japanese, the national language, and Standard Japanese were
deployed as Japan developed into a modern imperial state and how linguists took an
active part in their deployment. In the future, JLGS is expected to strengthen its ef-
forts to reveal the historical processes through which gender and language have been
constructed and sustained.
At the same time, a greater number of studies need to be done to further uncover
linguistic practices in which speakers negotiate and subvert dominant gender ide-
ologies. The studies should focus on groups of people who have been marginalized
in Japanese society, attending carefully to how gender interacts with ethnicity, class,
generation, and sexual orientation in those practices. Blue-collar workers, speakers
of regional dialects, the social outcasts or burakumin,4 transgendered speakers, eth-
nic minorities such as the Ainu and Koreans, and other minority populations need to
be studied. In view of the fact that JLGS itself has long played a part in maintaining
the damaging concept of Japanese womens language and in colluding in the ex-
clusion and marginalization of women and minorities, it is particularly important for
the sound future development of JLGS that it now reexamine its own theoretical
frameworks and methodological assumptions.
34 HISTORICAL AND THEORETICAL FOUNDATIONS

Notes
1. All translations from the Japanese originals are ours.
2. In Japanese sentences, personal pronouns are often deleted when they are obvious
from the context, for example, from the use of honorifics.
3. Introductions to Western language and gender studies published in the 1990s by in-
fluential leaders of Japanese womens studies and JLGS give a false impression that it is
exhausted by the work of Lakoff (1975/tr.1985) and Spender (1980/tr.1987); (Ide 1993,
Reinoruzu [Reynolds]-Akiba 1993, Sakamoto 1993, Ehara 1994, Inoue 1999). It is signifi-
cant that Cameron (1985/tr. 1990) and Nakamura (1995), which reflected later developments
in Western feminist linguistics, were not paid due attention for many years. This reflects the
fact that Japanese feminists had long been comfortable with Spenders simplistic view that
all men dominate all women. The more current history of language and gender research abroad
was not made available in Japanese until recently. See Yukawa (1998) and Nakamura (2001).
4. The exclusion and oppression of these people can be traced back to the thirteenth
century, when some occupations were regarded as impure according to Buddhist and Shintoist
thinking. Discrimination against these groups continues today.

References
Abe, Hideko (1995). From stereotype to context: The study of Japanese womens speech.
Feminist Studies 21(3): 647671.
Bucholtz, Mary (1999a). Series foreword. In M. Bucholtz, A. C. Liang, and L. A. Sutton (eds.),
Reinventing identities: The gendered self in discourse, viiviii. New York: Oxford Uni-
versity Press.
(1999b). Bad examples: Transgression and progress in language and gender studies.
In M. Bucholtz, A. C. Liang, and L. A. Sutton (eds.), Reinventing identities: The gendered
self in discourse, 324. New York: Oxford University Press.
Cameron, Deborah (1985). Feminism and linguistic theory. London: MacMillan. [= 1990,
trans. Momoko Nakamura, Feminizumu to gengo riron (Feminism and linguistic theory).
Tokyo: Keiso Shobo.]
Eckert, Penelope, and Sally McConnell-Ginet (1992). Think practically and look locally:
Language and gender as community-based practice. Annual Review of Anthropology 21:
461490.
Ehara, Yumiko (1990). Feminizumu no nanajuunendai to hachijuunendai (Feminism in the 1970s
and 1980s). In Y. Ehara (ed.), Feminizumu ronsoo: Nanajuunendai kara kyuujuunendai e
(Feminism from the 1970s to the 1990s), 246. Tokyo: Keiso Shobo. [= 1993, Japanese
feminism in the 1970s and 1980s, U.S.-Japan Womens Journal English Supplement 4:
4969.]
(1994). Josee to hyoogen (Women and representation). In Y. Meguro (ed.), Jendaa
no shakaigaku (Sociology of gender), 3844. Tokyo: Hoso Daigaku Kyoiku Shinkokai.
Endo, Orie (ed.) (1992). Josee no yobikata daikenkyuu (Research on how women are ad-
dressed). Tokyo: Sanseido.
(1997). Onna no kotoba no bunkashi (A cultural history of Japanese womens lan-
guage). Tokyo: Gakuyo Shobo.
Fukuda, Mayumi (ed.) (1993). Shujin to yuu kotoba (The expression shujin). Tokyo: Akashi
Shoten.
Gal, Susan (1991). Between speech and silence: The problematics of research on language
and gender. In M. di Leonardo (ed.), Gender at the crossroads of knowledge: Feminist
anthropology in the postmodern era, 175203. Berkeley: University of California Press.
 CULTURAL IDEOLOGIES IN JAPANESE LANGUAGE AND GENDER STUDIES 35

(1995). Language, gender and power: An anthropological review. In K. Hall and

M. Buchettz (eds.), Gender articulated: Language and the Socially Constructed Self,
169182. New York: Routledge.
Gallagher, Margaret (2001). Gender setting: New agendas for media monitoring and advo-
cacy. London: Zed Books.
Ide, Sachiko (1979). Onna no kotoba, otoko no kotoba (Womens language, mens language).
Tokyo: Nihon Keizai Tsushinsha.
(1990). How and why do women speak more politely in Japanese? In S. Ide and
N. H. McGloin (eds.), Aspects of Japanese womens language, 6379. Tokyo: Kuroshio.
(1993). Sekai no joseego, Nihon no joseego: Joseego kenkyuu no shintenkai o motomete
(Womens language of the world, womens language of Japan: Searching for new devel-
opments in womens language research). Nihongogaku 12(6): 412.
Ide, Sachiko, Ochiai Keiko, and Tawara Machi (1998). Zadankai: Kotoba ni miru josee
(Roundtable talk: Portrayal of women in words). In T. J. Zaidan (ed.), Kotoba ni miru
josee (Portrayal of women in words), 1340. Tokyo: Kureyon House.
Inoue, Miyako (1994). Gender and linguistic modernization: Historicizing Japanese womens
language. In M. Bucholtz, A.C. Liang, L. A. Sutton, and C. Hines (eds.), Cultural per-
formances: Proceedings of the Third Berkeley Women and Language Conference, 322
343. Berkeley: Berkeley Women and Language Group.
Inoue, Teruko (1999). Nichijoosee no naka no jendaa to sekushuaritii (Gender and sexuality
in everyday practices). In Kokuritsu Fujin Kyooiku Kaikan Joseegaku Jendaa Kenkyuukai
(ed.), Joseigaku kyoiku/gakushuu handobukku (Handbook of womens studies), 160165.
Tokyo: Yuhikaku.
Jugaku, Akiko (1979). Nihongo to onna (The Japanese language and women). Tokyo: Iwanami
Shoten.
Kawata, Fumiko (2000). Onna to yuu moji, onna to yuu kotoba (Characters and words for
women). Tokyo: Akashi Shoten.
Kindaichi, Kyosuke (1942). Joseego to keego (Womens language and honorifics). In K. Kin-
daichi (ed.), Kokugo kenkyuu (A study of the national language), 293315. Tokyo:
Yakumoshorin.
Komatsu, Hisao (1988). Tokyogo ni okeru danjosa no keesee: Shuujoshi o chuushin to shite
(The form of gender differences in Tokyo dialect: Centering on the sentence-final par-
ticles). Kokugo to Kokubungaku 65(11): 94106.
Komori, Yoichi, Kensuke Kono, and Osamu Takahashi (eds.) (1997). Media, hyooshoo,
ideorogii: Meiji sanjuunendai no bunka kenkyuu (Media, representation, and ideology:
Cultural studies on the Meiji thirties). Tokyo: Ozawa Shoten.
Kotoba to onna o kangaeru kai (The Group for Thinking about Language and Women) (1985).
Kokugo jiten ni miru josee sabetsu (Sexism in dictionaries). Tokyo: Sanichi Shobo.
Lakoff, Robin (1975). Language and womans place. New York: Harper and Row. [1985,
trans. Reynolds Akiba Katsue, Tokyo: Yushindo.]
Lee, Yeounsuk (1996). Kokugo to yuu shisoo (Ideology called kokugo [the national
language]). Tokyo: Iwanami Shoten.
Maltz, Daniel, and Ruth Borker (1982). A cultural approach to male-female miscommunica-
tion. In J. Gumperz (ed.), Language and social identity, 196216. Cambridge: Cambridge
University Press.
Maree, Claire (1997). Jendaa shihyoo to jendaa no imisee no henka: Eega Shinjuku Boys in
okeru onabe no baai (Gender indexicality and semantic shifts in gender meanings: The
case of an onabe in the documentary film Shinjuku Boys). Gendai Shisoo 25(13): 262
278.
Media no naka no seesabetsu o kangaeru kai (The Workshop on Gender and Media) (1991).
36 HISTORICAL AND THEORETICAL FOUNDATIONS

Media ni egakareru josee zoo (Portrayals of women in the newspapers). Toyama: Katsura
Shobo.
Nakamura, Momoko (1993). Gengogaku no kagakuteki gensoku no jissen ni taisuru
feminisuto no hihan (A feminist critique of the scientific principle of linguistics). In
K. Reinoruzu [Reynolds] Akiba (ed.), Onna to nihongo (Women and the Japanese lan-
guage), 6795. Tokyo: Yushindo.
(1995). Kotoba to feminizumu (Language and feminism). Tokyo: Keiso Shobo.
(2001). Kotoba to jendaa (Language and gender). Tokyo: Keiso Shobo.
Ogawa, Naoko, and Janet S. Shibamoto Smith (1997). The gendering of the gay male sex
class in Japan: A preliminary case study based on Rasen no Sobyoo. In A. Livia and
K. Hall (eds.), Queerly phrased: Language, gender, and sexuality, 402415. New York:
Oxford University Press.
Ogino, Miho (1992). Shujin no koogengaku: Nihongo ni okeru otto no koshoo ni tsuite
(Why do Japanese women call their husbands shujin?). Joseegaku Nenpoo 13: 1124.
Ohara, Yumiko (1992). Gender-dependent pitch levels: A comparative study in Japanese and
English. In K. Hall, M. Bucholtz, and B. Moonwomon (eds.), Locating power: Proceed-
ings of the Second Berkeley Women and Language Conference, 469477. Berkeley:
Berkeley Women and Language Group.
Okamoto, Shigeko (1995). Tasteless Japanese: Less feminine speech among young Japa-
nese women. In K. Hall and M. Bucholtz (eds.), Gender articulated: Language and the
socially constructed self, 297325. New York: Routledge.
Okamoto, Shigeko, and Shie Sato (1992). Less feminine speech among young Japanese fe-
males. In K. Hall, M. Bucholtz, and B. Moonwomon (eds.), Locating power: Proceed-
ings of the Second Berkeley Women and Language Conference, 478488. Berkeley:
Berkeley Women and Language Group.
Reinoruzu [Reynolds]-Akiba, Katsue (1993). Gengo to seesa no kenkyuu (The study of sex
differences in language). Nihongogaku 12(5): 224234.
Saito, Masami (1994). Sabetsugo gari de wa nai gengo kaikaku: eego kyookasho, eewa jisho
ni miru (Language reform in English textbooks and dictionaries). Hyuuman raitsu 78:
4049.
(2001). Feminizumu riron ni yoru hihanteki disukoosu bunseki no tenkai: Uuman ribu
undoo no media gensetsu o jiree to shite (Feminist critical discourse analysis: A study
of the media discourse of the womens lib movement in Japan). Unpublished Ph.D.
dissertation, Ochanomizu University.
Sakamoto, Kazue (1993). Kotoba kaidai (Introduction: Language). In S. Kato, K. Sakamoto,
and K. Sechiyama (eds.), Feminizumu korekushon III (Feminism collection III ), 322
326. Tokyo: Keiso Shobo.
Sasaki, Eri (1994). Habikoru josee sabetsu to kokusaijin no yukue: Chuugaku eego
kyookasho no jittai to kongo no kadai (Sexism in English textbooks: How could Japa-
nese possibly survive in the internationalized world?). Joseegaku 2: 121139.
Shibamoto, Janet S. (1985). Japanese womens language. New York: Academic Press.
Spender, Dale (1980). Man made language. London: Routledge and Kegan Paul. [1987, trans.
Reynolds Akiba Katsue, Kotoba wa otoko ga shihai suru: Gengo to seesa (Man con-
trols language: Language and sex difference). Tokyo: Keiso Shobo.]
Takeda, Haruko (1990). Gengo seesa no sutereotaipu: Imada Yuuko e no shikisha no
komento o yomu (Stereotyped images of womens writing: Reading the comments on
the letters to Imada Yuko). Joseegaku Nenpoo 11: 2839.
(1991). Gengo seesa no sutereotaipu (Gender differences in language use). Joseegaku
Nenpoo 12: 5362.
 CULTURAL IDEOLOGIES IN JAPANESE LANGUAGE AND GENDER STUDIES 37

Tanaka, Kazuko (1993). Shimbun ni miru koozooka sareta seesabetsu hyoogen (Sexist ex-
pressions in the newspapers). In K. Reinoruzu [Reynolds], Akiba (ed.), Onna to nihongo
(Women and the Japanese language), 97122. Tokyo: Yushindo.
Tannen, Deborah (1990). You just dont understand: Women and men in conversation. New
York: Morrow.
Ueno, Chizuko, and Media no naka no seesabetsu o kangaeru kai (The Workshop on Gender
and Media) (eds.) (1996). Kitto kaerareru seesabetsugo: Watashitachi no gaidorain
(Nonsexist guidelines for Japanese mass media). Tokyo: Sanseido.
Washi, Rumi (2000). Nyooboo kotoba no imisayoo: Tennoosee, kaisoosee, and sekushuaritii
(The political function of nyooboo kotoba: Symbolizing the imperial system, class con-
sciousness, and sexuality). Joseegaku Nenpoo 21: 1835.
Yamaguchi, Michiyo (1998). Honyaku no naka no shujin (Shujin in translation). Joseegaku
Nenpoo 19: 3642.
Yukawa, Sumiyuki (1998). Kotoba to jendaa to kenryoku no kakawari ni semaru (Exploring
language, gender, and power). Joseegaku 6: 135160.
38 HISTORICAL AND THEORETICAL FOUNDATIONS

SHIGEKO OKAMOTO

Ideology in Linguistic Practice

and Analysis
Gender and Politeness in Japanese Revisited

L anguage ideology plays an important role in affecting both linguistic practice and
analysis. Defining linguistic ideologies as any sets of beliefs about language articu-
lated by the users as a rationalization or justification of perceived language structure
and use, Silverstein (1979:193) stresses that in scientific studies of language one
must distinguish ideology from actual language use. Woolard and Schieffelin (1994)
and Bergvall (1999) emphasize the importance of examining the dominant, or hege-
monic, ideology and its relationship to specific linguistic practices. In this chapter, I
consider the role of ideologies in linguistic practice and analysis with a focus on the
Japanese case, especially issues concerning gender and politeness in Japanese.
In language and gender research, it has commonly been claimed that, women gen-
erally speak more politely, indirectly, and cooperatively and use more standard lin-
guistic forms and prescriptively correct grammar than men (e.g., Trudgill 1972, Lakoff
1975, Brown 1980, Holmes 1995). However, as noted by Eckert and McConnell-Ginet
(1999) and others, such generalizations about gender differences always encounter
exceptions in any community. At the same time, they cannot be easily dismissed as
simple overgeneralizations or stereotypes. For they seem often to relate to dominant
gender norms and expectations in society and may affect actual language practices
and their interpretations in some way, although exactly how is not well understood.
It is thus important to reexamine these generalizations as to what they represent, in
particular, their normative and ideological aspect, and how they relate to actual lan-
guage practices. A close examination of these questions requires us to recognize the
plurality of social meanings, for it is here that we begin to understand the origins of
diversity and change in language practice.

38
 IDEOLOGY IN LINGUISTIC PRACTICE AND ANALYSIS 39

As a case in point, this chapter considers the generalization that women speak
more politely than men. I examine this generalization in relation to the use of Japa-
nese honorificsan important means of expressing politeness. In the ensuing sec-
tions I first discuss how the relationship between honorifics, politeness, and gender
has been treated by researchers (section 2.1) and the general public (section 2.2), both
of whom participate in the construction of linguistic norms and expectations in Japa-
nese society. I then look at some examples of actual honorific uses by both women
and men (section 2.3), and finally, I reconsider the relationship between language,
politeness, and gender in an attempt to account for the diversity in linguistic practice
of Japanese women and men (section 2.4). Before moving to section 2.1, however, a
brief description of the Japanese honorific system is in order.
Japanese honorifics are usually divided into two major categories: the so-called
taisha keego addressee honorifics (AH) and sozai keego referent honorifics (RH).1
Referent honorifics are further subdivided into three types: the so-called sonkee-go
respectful words (RH/R), kenjoo-go humble words (RH/H), and bika-go beauti-
fication words (RH/B; see, for example, Minami 1987). Addressee honorifics are
said to be used to show deference toward the addressee. Among referent honorifics,
sonkee-go and kenjoo-go are used to show deference toward the person being talked
about by elevating that person, her or his belongings or action (in the case of sonkee-
go) or by lowering another person (usually, the speaker), her or his belongings or
action (in the case of kenjoo-go).2 Bika-go are used for making the utterance sound
refined or elegant.
For example, in example (1), the title sensee professor and the auxiliary verb
o-V-ni nar are both referent honorifics or, more specifically, sonkee-go; and the form
-mash is an addressee honorific. In example (2), the prefix o- in o-nimotsu is a refer-
ent honorific, or sonkee-go, the form o-V-sh is a referent honorific, or kenjoo-go in
this case, and the form -mash is an addressee honorific. In example (3) neither a ref-
erent nor an addressee honorific is used.

(1) Tanaka-sensee ga kore o o- kaki-ni nar- imash-ita.

Tanaka-Prof.RH/R SM this OM write-RH/R AH-Pst
Professor Tanaka wrote this.

(2) Watashi ga sensee no o- nimotsu o o-moch-ish-imash-ita.

I SM prof. GN RH/R luggage OM carry-RH/H-AH-Pst
I carried the professors/your luggage (for him/you).

(3) Tanaka ga kore o ka-ita.

Tanaka SM this OM write-NRH/R-Pst.NAH
Tanaka wrote this.

Note that the referent and the addressee may or may not be the same person. For
example, in example (2) the speaker may be talking to a professor about the pro-
fessors own luggage or that of another professor. In the former interpretation, the
referent and addressee honorifics are both used toward this person, but not in the
latter.
40 HISTORICAL AND THEORETICAL FOUNDATIONS

(4) O-yasai mo tabe-te.

RH/B-vegetables also eat-NRH/R-GD
(Please) eat the vegetables, too.

The prefix o- in o-yasai in example (4) is a referent honorific, but it is a bika-go and
not a sonkee-go. The use of bika-go has been linked to stereotypical femininity.

2.1. Previous studies on gender and politeness

in Japanese

One of the most commonly noted gender differences in Japanese is that women gener-
ally speak more politely than men (e.g., Ogino, Misono, & Fukushima 1985, Niyekawa
1991, [Shibamoto] Smith 1992, Kawanari 1993, Shigemitsu 1993). Suzuki (1993:148),
for example, states that the essence of womens language is womens concern about
politeness. Likewise, Ide notes that among various features which make womens
speech feminine, politeness in speech stands out in Japanese (1990: 63). Based on
these observations, it has also been claimed that womens language, or polite speech,
indexes femininity or the female gender (Ide 1993, Suzuki 1993). There are a num-
ber of linguistic devices for making utterances polite (e.g., honorifics, indirect speech
acts, hedges, reactive tokens, and sentence-final particles), but the complex system
of honorifics is considered particularly important. It is commonly claimed that Japa-
nese women tend to use more honorifics and formal expressions than Japanese men,
thereby making their speech more polite than mens (e.g., Ogino, Misono, & Fukushima
1985, Ide 1990, Niyekawa 1991, Kawanari 1993). Shibatani (1990:374), for instance,
notes that more than anything, the politeness in womens speech derives from the
higher frequency of the use of the honorific forms. Similarly, Sugimoto (1997:235)
asserts that womens language has its own unique structure and that this uniqueness
derives from the use of honorifics.
There are a number of different explanations for the claim that women speak
more politely than men by using honorifics and other linguistic devices: (1) biologi-
cally determined traits, such as gentleness (Sugimoto 1997); (2) their relatively lower
social status (Ide 1982; Reynolds 1985, 1990); and (3) their social roles, or their
association with domains that require sociable/civil interactions (Ide 1990). It is
further argued that women, lacking their own status in society, tend to be more con-
cerned about appearance and thus use honorifics, or polite language, to indicate that
they are refined, or members of a higher social class (e.g., Ide 1982, 1990; Reynolds
1985).
While many of these previous studies suffer from methodological weaknesses
by relying on self-report surveys and researchers introspections (see also Shibamoto
1987, chapter 1, this volume), their findingsin particular, the results of self-report
surveysare nonetheless interesting in that they seem to reveal certain normative
expectations for womens and mens speech, although to what extent these expecta-
tions affect actual language practice is unclear. Ide (1990), for example, reports that
when participants were asked to assess the politeness levels of different linguistic
forms expressing the same meaning When do you go? women gave lower levels
 IDEOLOGY IN LINGUISTIC PRACTICE AND ANALYSIS 41

of politeness than men for almost all forms. Ide (1990:67) notes that thus, to ex-
press the same level of politeness, a woman has to use a politer linguistic form than
a man would. Kawanari (1993) reports that when participants were asked what re-
quest forms they would use in hypothetical situations, women generally gave more
indirect and polite expressions, which contained more honorific expressions (see also
Ogino, Misono, & Fukushima 1985). These studies suggest that at the level of lan-
guage norms, it is undeniable that women are expected to speak more politely than
men, using more honorifics and other formal or indirect expressions. Two major
questions at hand are how and why this prescriptive norm is promulgated in Japa-
nese society and how it relates to actual language practices. I discuss the first ques-
tion in section 2.2, and the second in sections 2.3 and 2.4.

2.2. Gender and honorifics: Norms and expectations

As noted by a number of studies (e.g., Miller 1996, Wetzel 1994, Okamoto 1999), in
Japanese society the importance of using correct honorifics to express politeness
is emphasized through various means, in particular, through education and media.
For example, Kokugo-shingikai, a Japanese government council on the national lan-
guage, recommends in its 1996 report that schools continue to improve teaching
children appropriate honorifics according to their developmental levels. There are
also numerous books, magazine articles, and other materials on how to use honorif-
ics correctly, which suggests that knowledge of honorifics is not evenly distrib-
uted in the society and that many native speakers aspire to acquire it, because they
consider it linguistic capital for improving their social status (Miller 1996). Wetzel
(1994) reports that in a subjective reaction test users of honorifics were judged to be
more intelligent, more educated, and more capable than nonusers. These observa-
tions suggest that many Japanese associate the use of honorifics with higher class
status, education, intelligence, and other prestige factors.
Politeness and honorifics are thus considered important for both women and men.
However, as noted earlier, there is a difference in expectations. That is, there is a belief
among many Japanese that women should speak more politely than men, using honor-
ifics and other formal expressions. Media, as well as education at home and school,
seem to play an important role in promoting this idea, as illustrated in the following
excerpt from a self-help book for women, titled Kashikoi hito ni narinasai: Utsukushiku
ikitai anata ni Be a wise woman: To you who want to live beautifully:3

It is often said that young women nowadayswhether they are students or work-
ing womencannot use honorifics well. . . . I sometimes hear female teachers use
the same language as male teachers. . . . Even in a democratic society, its natural
that there are differences in ways of talking based on sex differences, because men
and women have different vocal cords. . . . But women dare to use mens language.
Are they ignorant or lazy, or are they making foolish efforts not to be dominated by
men? . . . Not knowing honorifics is embarrassing. Parents and teachers should teach
that [to children] by showing good models. . . . Even today, . . . when I see such
people [people who use proper and polite language], Im impressed by their good
upbringings. (Tanaka 1986:2933, translated from the Japanese original)
42 HISTORICAL AND THEORETICAL FOUNDATIONS

The book, by a well-known woman writer, has been widely read and was reprinted
73 times between 1986 and 1995. The author assumes in this passage that gender
differences in speech are both biologically and socially based and that it is natural
that women speak womens language and men speak mens language. She then
emphasizes that women should use honorifics properly, assuming that honorifics are
an important part of womens language; those women who do not use honorifics
properly should be socially sanctioned, receiving negative evaluations, such as ig-
norant, lazy, and improperly brought up.
There are numerous books that specifically teach women how they should speak
in order to become attractive and beautiful, as illustrated by the following book titles.

Suteki na anata o tsukuru: Josee no utsukushii hanashi-kata To make yourself nice:

Womens beautiful ways of speaking (Suzuki 1989)
Onna no miryoku wa hanashi-kata shidai Womens attractiveness depends on how
they speak (Kanai 1994)

These how-to books usually contain a section on how to use honorifics correctly
the ability is presented as essential for a sophisticated woman, or kotoba-bijin lan-
guage beauty, as illustrated by the following section heading in one of these books:
Keego o kichin to hanasu koto ga bijin no jooken desu Using honorifics properly
is an indispensable condition for a beautiful woman (Kanai 1994). In his book on
womens beautiful speech, Suzuki (1989), a television announcer and best-seller
writer, recommends that married women speak politely to their husbands, using hon-
orifics (see also chapter 6, this volume):

When a man is asked [by his wife, using honorifics] Nanji-goro o-kaeri ni naru no
What time will you come back?, he feels her gentleness, good upbringing, and
reserved elegance. . . . [At parties, etc.] men are very worried about the impressions
their wives ways of speaking give about their educational levels. (Suzuki 1989:130
32, translated from the Japanese original)

These how-to books thus promote the notion that womens attractiveness depends
on their appearance (beauty), which is in part determined by their good upbringing
and education, including the knowledge of honorifics.
There are also many books on how to write letters, which usually refer to gen-
der differences in styles. One good example is a pair of books on letter writingone
for women and the other for men (Ohashi 1994a, b). Written by the same male au-
thor, these two books provide many sample letters in female and male versions. Com-
pared to the male versions, the female versions are normally longer and written with
more curvy (rather than straight-lined) handwriting, using more honorifics, as shown
in examples (5a) and (5b), taken from Ohashi (1994a, b).

(5a) (a sample letter; from a man to his nephew, congratulating him on his engagement)
Masaki-kun, konyaku-shita soo da ne. Omedetoo. Sore to naku
RH/R NRH/R NAH NAH
kanjite wa ita kedo, konna ni hayaku to wa bikkuri-suru yara
NAH
 IDEOLOGY IN LINGUISTIC PRACTICE AND ANALYSIS 43

odoroku yara, nan to itte o-iwai no kotoba o okur-oo ka. . . .

RH/B NRH/H-NAH
Masaki, I heard that you got engaged. Congratulations! I had a feeling [that you might],
but Im very surprised that it came so fast. What kinds of congratulatory words shall I
give you?

(5b) (a sample letter; from a woman to her nephew, congratulating his marriage engagement)
Haruki-san, go-konyaku omedetoo-gozaimasu. Saikin no anata no
RH/R RH/R AH
yoosu kara sore to naku kanjite wa i- mashita ga, konna ni hayaku
AH
konyaku to wa bikkuri-suru yara odoroku yara, nan to itte o-iwai no
RH/B
kotoba o okur- imashoo ka. . . .
NRH/H-AH
Haruki, congratulations on your engagement! I had a feeling [that you might] from
the way you have been behaving recently, but Im very surprised that you got engaged
so fast. What kind of congratulatory words shall I give you?

Examples (5a) and (5b) have similar content and are both addressed to younger per-
sons. But they differ in that (5b) is more formal and contains more honorifics than
(5a). In addition, the suffix -kun is used in (5a), but a more formal suffix, -san, is
used in (5b).
Further, popular culture materials, such as films, television dramas, novels, and
cartoons, often show female characters, especially middle-aged women in the middle
and upper-middle classes, using more honorifics than male characters. For example,
in one of the most popular television drama series, Wataru seken wa oni bakari
Making it through, two main characters, both middle-aged women in middle-class
families, use honorifics toward their husbands, but not vice versa. Similarly, in the
comic strip Nono-chan, which appears daily in a major newspaper (Asahi Shimbun)
in Japan, honorifics are often used nonreciprocally from one of the main female
characters to her husband.
As these examples demonstrate, the belief that women should use more honor-
ifics, or polite language, than men is widely promoted as a behavioral norm in Japa-
nese society. Further, they suggest that the use of honorifics is linked not only to
gender but also to class status in that it is treated as a sign of good upbringing and
education by many how-to books and is particularly associated with the speech of
middle- and upper-middle-class female roles in popular culture materials.
It seems that the link between polite speech, women, and a higher class status
has long been promoted and sustained as a behavioral norm for women in Japanese
society. Endo (1997) points out that it was during the feudal Edo period (16031868),
a period in which the ruling classes were greatly influenced by Confucian philoso-
phy, that the importance of disciplining women began to be emphasized, as illus-
trated by the emergence of many books that taught womens virtues and stipulated
detailed rules of conduct for women, including prescriptions for speech. Endo (1997)
notes that these prescriptions instructed women to speak gently, politely, and in a
44 HISTORICAL AND THEORETICAL FOUNDATIONS

refined manner and offered many specific example expressions related to nyooboo
kotobaa language used among court ladies-in-waiting since the fifteenth century
(see also Sugimoto 1997). That is, the speech of women in higher social echelons
was thought to be polite, gentle, and refined and hence to be emulated by all women.
At the same time, this emphasis on the use of proper language suggests that many
women, in particular women in the lower classes, did not use such language, as indi-
cated by the speech styles of women in commoners classes represented in novels
written at the time (Endo 1997, Sugimoto 1997).
Upon the termination of the feudal period in 1868, the notion of proper
womens language came to be resituated in the context of modern Japan, in which
the education of women was considered an important part of the development of Japan
as a modern nation-state. Prescriptions of ideal Japanese women, or ryoosai kembo
good wife and wise mother, were disseminated through various channels (e.g., text-
books, magazines). These prescriptions often included instructions about speech,
which recommended that women speak gently, politely, and in a reserved and re-
fined manner (Endo 1997). At this time, the importance of establishing Standard
Japanese as the national language was emphasized and Tokyo-go Tokyo language,
or the speech of the educated, or middle-class, Tokyoites, was promoted through
education to become Standard Japanese (Lee 1996). Media also contributed to its
promotion. In this regard, it has been noted that distinct gendered speech styles as
part of the (presumed) speech of the educated class in Tokyo were disseminated
particularly through their representations in novels (Komori 2000, chapter 3, this
volume), although it is unclear to what extent such speech styles were perceived by
ordinary Japanese women and men, particularly those living in regional Japan, as
models for their own speech.
Washi (2000, chapter 4, this volume) observes that the use of proper womens
language was encouraged through media (e.g., radio, books, newspapers), especially
before and during World War II in the context of rising nationalism, and that nyooboo
kotoba was often given as a model language. For example, in 1935 linguist Yoshida
Sumio gave a radio lecture on womens language in which he said that because gender
differences in speech are biologically based, womens language has always been yuubi
gentle and elegant, containing honorifics, indirect expressions, and the like, as illus-
trated by nyooboo kotoba. Similarly, in a 1936 radio lecture on womens language,
Hoshina Koichi, an expert on national language policies, said that while there is little
gender difference in the speech of the lower and working classes, women in higher
classes, or women with proper demeanor, generally use honorifics, or polite and de-
cent language, like nyooboo kotoba (Washi 2000). These examples also illustrate the
active involvement of scholars in the construction of societal norms for womens speech.
The foregoing discussion is by no means intended to suggest that nyooboo kotoba
(or any other specific speech variety) is the origin of normative womens speech.
Rather, it shows that nyooboo kotoba as an index of femininity and class status played
an important symbolic role in the ideological construction of womens speech. Spe-
cific speech forms regarded as models for womens speech in contemporary Japanese
are, of course, quite different from nyooboo kotoba. In fact, it is questionable if there
is a set of agreed-upon model speech forms in contemporary Japanese. (See section
2.4 for further discussion.) The hegemonic ideology that attempts to regulate womens
 IDEOLOGY IN LINGUISTIC PRACTICE AND ANALYSIS 45

speech may also have had different practical implications and importance in different
historical contexts. However, the belief that women should speak politely, gently, and
in a refined manner as women in higher social strata did seems to have continuously
played a key role in the history of the norm construction for womens speech.

2.3. Variability in the use of Japanese honorifics

in conversations

The notion that women should speak more politely than men is thus widely promul-
gated in Japanese society. However, it may not have a universal effect on women (or
men). Examinations of actual language use show wide intragender differences as well
as intergender similarities. In this essay I illustrate these cases, drawing examples
mainly from a number of recent studies.
With regard to variation in womens speech, both Dunn (1996) and Okamoto
(1996) report that in our research older women used more honorifics than younger
women. For example, in my earlier study, which examined dyadic conversations of
10 college students and 10 middle-aged women, the older women generally used
referent honorifics much more than the younger subjects (65% vs. 12% of the rele-
vant tokens; see Okamoto 1996 for examples). The same kind of variation was ob-
served in the use of the referent, beautification honorific prefix o-. That is, older
women used this prefix more than younger womene.g., o-shooyu soy sauce, said
by an older woman, versus kane money, said by a younger woman. These age-group
differences may also involve differences in roles.
As mentioned earlier, how-to-books on letter writing recommend the use of
honorifics, particularly by women. It has also been noted that in letters one is likely
to use honorifics even toward persons one knows well (e.g., Minami 1987). My analy-
sis of personal letters, which were collected by Yasuda Wakako in 1999, shows that
unlike older people (4060), younger people (1824), both women and men, gener-
ally do not use honorifics when they write to people they know well. This contrast is
illustrated by example (6), a letter from a female college student to her friend, which
uses no honorific, and example (7), a letter from a 59-year-old woman to her younger
sister, which uses both addressee and referent honorifics.

(6) Watashi wa itsumo nagara genki da yo. Shimpai kakechat-ta kedo,

NAH NRH/H-NAH
te wa naot- ta yo.
NAH
Im fine as always. I made you worry, but my hand is cured now.

(7) Wakaba no utsukushii kisetsu desu. Sono go x-san o-kagen ikaga

AH RH/R
deshoo ka. Totsuzen no koto de odoroki mashita.
AH AH
Its a season of beautiful new green leaves. How is X since then? It was sudden and I
was caught by surprise.
46 HISTORICAL AND THEORETICAL FOUNDATIONS

There are also wide individual differences in speech within the same age-group.
For example, in my 1996 study two women in the older age-group used referent
honorifics and the prefix o- much more frequently than the others and one woman
hardly used them at all (see also chapter 13, this volume, for variable honorific use
among middle-aged women). This variation may be related to differences in speak-
ers social backgrounds and their attitudes toward honorific use. Moreover, the non-
use of honorifics by subjects when they spoke to their friends does not indicate lack
of knowledge. Rather, it seems strategic, for when I interviewed subjects in person,
they spoke in formal style, using both referent and addressee honorifics.
Another example of variation in womens use of honorifics comes from my
observation of salespersons in two kinds of retail places in the Kansai area, two major
department stores and two large marketplaces, each housing more than 150 small
shops. According to the canonical rule, a salesperson is expected to use honorifics
toward customers. In fact, in the department stores both female and male salesper-
sons used high levels of honorifics most of the time. However, vendors in the two
marketplaces, including women, often did not use honorifics toward the customers
at all. Example (8) illustrates the use of honorifics by a female salesperson at a de-
partment store and example (9) the nonuse of honorifics by a female vendor (see
also Okamoto 1997).

(8) Irassha- imase. Doozo goran kudasa- imase.

RH/R AH RH/R RH/R AH
Welcome. Please take a look at them.

(9) Koobe-niku yasui yo yasui yo, negittara maketok-u yo.

NAH NAH NRH/R NRH/H-NAH
Kobe beef, its cheap; its cheap. If you haggle, we will make it cheaper.

Similarly, Abe (chapter 11, this volume) observes that employees at lesbian bars often
speak to their customers without honorifics.
Turning now to mens speech, despite the generalization that men use fewer
honorifics than women, there seem to be many situations in which men use honorif-
ics extensively. It is commonly said that lower status persons use honorifics to higher
status persons. This seems to apply to men as well as women. For example, as men-
tioned earlier, I observed that male salesclerks at department stores used high levels
of honorifics for customers in the same way as female salesclerks, as illustrated in
example (10):

(10) (Waiter, at a coffee shop in the department store in Kyoto)

O-sara no hoo o-sage-shiteyoroshii desu ka.
RH/B RH/H RH/B AH
Would it be all right if I took away your plate?

The frequent use of honorifics by male salespersons may be related to their roles.
Takahashi (1996) noted that male instructors on television cooking shows generally
used very formal and polite speech styles, while female instructors showed more
 IDEOLOGY IN LINGUISTIC PRACTICE AND ANALYSIS 47

variation. The speech of male instructors included many instances of the beautifica-
tion honorific prefix o-, as in o-yasai vegetable, o-ajimi tasting, and o-suimono
soup, as well as referent and addressee honorifics, as illustrated in example (11),
taken from Takahashi (1996).

(11) Suton to otoshi te shimaw-areru no ne. Koron to ireru to osshatta desho.

RH/R N-AH RH/R AH
You chop it up and you said you dump them in.

Further, it is often said that in hierarchical situations honorifics are often used
nonreciprocally, from the powerless to the powerful (e.g., Niyekawa 1991). How-
ever, reciprocal uses are quite common. For example, I observed that some custom-
ers, both women and men, reciprocated honorifics, especially addressee honorifics,
to salespersons, while others did not. In example (11), the speaker, a male cooking
instructor in his midfifties, used honorifics toward his young female assistant. In
example (12), the speaker, a male supervisor (in his early sixties), reciprocated hon-
orifics to the addressee, a newly hired female subordinate (in her late forties; see also
Okamoto 1999).

(12) Maa, ano kongo iroiro kyooryoku-shite itadaku to omo-imasu node . . .

RH/H AH
Well, I think I will be asking for your cooperation in various [matters] in the future,
so . . .

2.4. Rethinking the relationship between honorifics,

politeness, and gender

As demonstrated earlier, the use of honorifics by Japanese women and men is di-
verse and does not necessarily conform to the normative expectation. How can we
account for this diversity? Does the norm play any role? How does each speaker
understand the norm in specific contexts? I address these questions in this section by
considering the use of honorifics seen earlier vis--vis three ideologically nonneutral
assumptions that often accompany generalizations and prescriptions about gender
and linguistic politeness.

2.4.1. Social and contextual diversity and the use

of honorifics
As seen in sections 2.1 and 2.2, in both scientific descriptions and nonacademic pre-
scriptions about politeness and gender, women and men are commonly treated as
constituting homogeneous categories. That is, the reasons that Japanese women
should speak more politely than men, using honorifics and other linguistic devices,
are explained on the basis of shared attributes, such as biologically determined traits
(e.g., gentleness), womens relatively lower social status, their social roles, or their
desire to look better (e.g., refined, a member of a higher class). These arguments are
48 HISTORICAL AND THEORETICAL FOUNDATIONS

comparable with the accounts given by a number of researchers (e.g., Trudgill 1972,
Lakoff 1975, Brown 1980) for the womens greater use of standard, or prestigious,
polite linguistic forms in English and other languages (see James 1996 for a critical
review on this topic). They may contribute to highlighting gender inequalities in lan-
guage and society, but their focus on gender differences may also serve to support
and perpetuate the stereotypical gender dichotomy, because they are all based on the
following assumption (see also Cameron 1988, Bucholtz 1999, Nakamura 1995,
among others, for relevant discussions).

Assumption 1: Most women, if not all, share the same attributes (e.g., biologically
determined traits, social inferiority, social roles, concerns about appearance) and
therefore (should) use language in the same way.

However, the examples seen in the previous section as well as the findings of
the chapters in part III of this volume and many other recent studies on gender and
language (Hall & Bucholtz 1995, Bergvall, Bing, & Freed 1996, and Bucholtz, Liang,
& Sutton 1999, among others) suggest otherwise.

Counterproposal 1: Not all women (or men) share the same attributes: they are
socially diverse with regard to age, role, status, and other factors; their conversa-
tional contexts also vary widely with regard to interlocutor, setting, and so on. As
a consequence, not all women (or men) may speak in the same way.

Thus, for example, it is difficult to apply an abstract claim of gender difference in lin-
guistic politeness to the fact that the interactional styles of female vendors are much
less formal than those of male salesclerks at department stores or to the variability in
speaking or writing styles among women of different ages. These examples indicate
that the use of honorifics needs to be examined in relation to social and contextual di-
versity that involves individual speakers, because in addition to gender, a variety of
other factors also seem relevant. For example, as indicated in the foregoing discus-
sion, female students less frequent use of honorifics may be (indirectly) related to their
age, role, type of interlocutor, setting, and other aspects of the social context.
However, while it is important to consider multiple factors for the use of honorifics,
this does not mean that if one can identify all the relevant factors, one can predict the
use of honorifics (see also Cameron 1990 and Eelen 1999 for critical discussions of
correlational studies). This is because it is speakers as social agents, and not contex-
tual features, that ultimately determine the use of honorifics and also because differ-
ent individuals may have different attitudes toward honorific use and associated
ideologies and hence may interpret and use honorifics differently. I consider this issue
in the following two subsections.

2.4.2. Honorifics and politeness:

Ambiguities of linguistic expressions
As mentioned earlier, honorifics are considered to be one of the most important lin-
guistic features that make womens speech polite. This characterization is often based
on the following assumption.
 IDEOLOGY IN LINGUISTIC PRACTICE AND ANALYSIS 49

Assumption 2: Certain linguistic expressions, such as honorifics and other formal

or indirect expressions, are inherently polite.

Assumption 2, also shared by many previous studies of politeness in general (e.g.,

Leech 1983, Brown & Levinson 1987), regards certain expressions (e.g., honorific
expressions) as polite and others (e.g., nonhonorific expressions) as not (or less) polite,
independently of any context.
It may correspond to the commonsense ideology of politeness and influence the
interpretation of honorifics in a significant way. Yet I argue that the interpretation of
honorifics is not unitary. I present this argument as Counterproposal 2.

Counterproposal 2: Certain linguistic expressions, such as honorifics and other

formal or indirect expressions, are not inherently polite; their interpretations may
vary among individuals as well as across contexts, depending on the (ideologically
based) criterion used for evaluating them in specific contexts.

The reasons for this variability are (1) that politeness is a matter of evaluations of
social conduct vis--vis what is understood by the evaluator as the norm (see also
Eelen 1999), (2) that the same person may assume different norms for different situ-
ations, and (3) that the understanding of norms may vary among individuals. As dis-
cussed by Eelen (1999), politeness should be viewed as involving moral judgments
rather than as an objective system of rules, strategies, or principles that is assumed to
govern speakers behavior. It is thus important to examine how and why such judg-
ments are made. Addressing this question requires an examination of the language
ideologies that underlie ones apprehension of what the linguistic norms are.
One interprets (non-)honorific expressions as polite or impolite against some
criterion or what one understands as the norm. However, the same person may use
different criteria for evaluating specific linguistic choices in different situations. For
example, regarding the use of honorifics, what is considered appropriate may not be
the same for vendors at marketplaces and salesclerks in department stores. Thus, the
nonuse of honorifics by vendors toward their customers may be perceived not as rude
but as a sign of friendliness and hence suitable for the speakers role as vendor. Simi-
larly, Abe (chapter 11, this volume) observes that when employees at lesbian bars
speak to their customers without honorifics it is often an attempt to establish solidar-
ity. Thus, the lack of expected honorifics may not necessarily be considered rude.
Analogously, if one believes that women should speak more politely than men, one
may interpret the (im)politeness level of the same linguistic form differently depend-
ing on the gender of the speaker. Accordingly, even if womens speech includes a
higher level of honorifics than mens, this may not make the former more polite than
the latter (see also Ogino, Misono, & Fukushima 1985 and Reynolds 1990).
Further, the understanding of norms themselves may vary among individuals due
to differences in their attitudes toward honorifics and politeness. For example, as men-
tioned earlier, although it is often assumed that honorifics are used nonreciprocally
between unequals, reciprocal uses are quite common. In fact, nonreciprocal uses based
on a hierarchical relationship seem to be decreasing in modern Japanese, while re-
ciprocal uses based on the degree of solidarity are increasing (e.g., Inoue 1989,
Bunkacho 1996). These synchronic and diachronic variations suggest the existence
50 HISTORICAL AND THEORETICAL FOUNDATIONS

of different ideologies about honorific uses, which in turn may bring about different
interpretations (e.g., polite, aloof, rude) of the same honorific or nonhonorific form.
We sometimes encounter native speakers who express differing views about
honorifics. For example, a 68-year-old man, in a letter to the Asahi Shimbun news-
paper (March 3, 1996), criticizes schoolteachers who address students by name with-
out the honorific suffix san. He believes that these teachers are not polite and that
they should reciprocate honorifics to students to reduce the hierachical distance. In
response, a 20-year-old female student expressed disagreement in the same newspa-
per (March 10, 1996), saying that she felt closer to the teachers who did not use san
than to those who used it. Thus, she does not consider the teachers nonuse of hon-
orifics impolite. The two writers express virtually opposite views.
Earlier, we saw examples (11) and (12) in which male higher status persons re-
ciprocated honorifics to lower status persons. Some people, like the 68-year old man
mentioned earlier, may interpret these reciprocal uses as the speakers attempts to
show respect for lower status persons and to reduce the vertical distance. However,
others, like the student who responded to the 68-year-old mans opinion, may inter-
pret them as a sign of (horizontal) distance or unfriendliness. Similarly, some cus-
tomers may perceive vendors nonuse of honorifics toward them as rude, but others
may not, because they have different expectations about honorific uses.
Thus, honorific and nonhonorific expressions cannot be assessed as polite or rude
independently of the context and the speakers ideas about politeness. Accordingly,
one cannot characterize womens speech as more polite than mens in the abstract
simply because the former is presumed to contain more honorifics than the latter.

2.4.3. Politeness and onnarashisa womanliness revisited

The normative expectation that women should speak more politely than men also
implies that polite speech indexes femininity or the female gender. For example, how-
to books instruct women to speak politely, because it can make them appear attrac-
tive as women or make them onnarashii womanly. Reynolds asserts that the notion
that women should behave onnarashiku as expected of women is still predominant.
Thus, Japanese women talk submissively and politely, just as always (1990:129).
Suzuki (1993) notes that in order for a woman to maintain her identity as a woman,
she must choose expressions within the socially expected womens language, which
Suzuki characterizes in terms of concerns about politeness. These studies recognize
the notion of onnarashii language as a behavioral constraint, but they also seem to
essentialize its effect on womens speech without problematizing the link between
particular speech styles (e.g., polite speech) and femininity. This in turn may con-
tribute to perpetuating dominant gender norms.
As pointed out by Jugaku (1979), although Japanese womens behavior seems
to be constrained considerably by the notion of onnarashisa womanliness, it does
not seem to have a universal effect on all women at all times (see also chapter 1, this
volume). In a self-report survey, one of the questions Takasaki (1993:170) asked of
her informants was: Do you sometimes control your speech, thinking that since you
are a woman, you should speak onnarashiku, using honorifics, being reserved, and
speaking in a refined way? In response, the majority answered that they try to speak
 IDEOLOGY IN LINGUISTIC PRACTICE AND ANALYSIS 51

onnarashiku depending on the situation. This answer indicates that they do not try to
speak onnarashiku all the time, and quite a few informants said they try not to do so.
The answers also varied widely, depending on the respondents occupation and edu-
cational level (see Takasaki 1993 for details). The notion of onnarashii language thus
cannot be ignored in accounting for womens speech (cf. Nakamura 1995), but its
ideological underpinnings and its relation to real language practices need to be ex-
amined more closely.
Let us first consider why polite speech is regarded as feminine. As implied by
Reynoldss statement and Takasakis question cited earlier, the answer to this ques-
tion concerns what is expected of (ideal) women, namely, such traits as submissive-
ness, deference, and refinement, which are assumed to be expressed through polite
speech. In other words, underlying the common claims about polite speech and femi-
ninity seems to lie the following assumption.

Assumption 3: Speaking politely, or showing deference and refinement through the

use of honorifics and other formal or indirect expressions, indexes femininity or
the female gender.

Ochs (1993) claims that the relationship between linguistic features and certain con-
textual information, such as gender and status difference, is mediated and constituted
through the pragmatic meanings of linguistic features, such as affective stances, social
acts, and social activities. Assumption 3 illustrates this interpretive duality: honorifics
and other formal or indirect expressions are assumed to index politeness, or deference
and refinement, which is in turn interpreted as indexing femininity or the female gen-
der. However, this construal cannot be taken for granted. Rather, it prescribes the ideal
form of femininity in terms of politeness, involving not only gender arrangements that
assume heterosexuality but also class stratification. Assumption 3 thus derives from
the Japanese hegemonic ideology of language, gender, and class. As a mediating link
between forms of talk and social structure (Woolard & Schieffelin 1994), this ideol-
ogy provides a basis for construing honorific expressions as onnarashii.
It is to be underscored, however, that the hegemonic ideology of politeness
and onnarashisa is not likely to be shared by everyone at all times. As discussed
by Briggs, contestation is a crucial fact of how particular ideologies and practices
come to be dominant (1998:249). Different, often competing, beliefs about polite-
ness and femininity may impart different social meanings to the same linguistic forms.
Assumption 3 does not consider, for example, the possibility that femininity may
be conceptualized and expressed differently by women of different ages and social
strata. Further, it may not be applied to all situations. Accordingly, the use of honor-
ifics, or polite speech, may not always be perceived as onnarashii (see also chapter
8, this volume, for a discussion of how the same linguistic performance may or may
not be perceived as genuinely feminine). Sunaoshi (personal communication) was
told by her Korean students on multiple occasions that, in Korean, formal speech,
including the use of honorifics, sounds stiff and is used more by men, while infor-
mal speech indexes femininity. Although it is unclear whether this interpretation is
typical among Koreans, its contrast with the interpretation of polite speech in Japa-
nese as onnarashii illustrates the subjective and arbitrary nature of the link between
52 HISTORICAL AND THEORETICAL FOUNDATIONS

a particular speech style and femininity (see also Keenan 1974). In other words, the
interpretation of femininity crucially depends on how its relation to linguistic forms
is ideologically conceived. I therefore counter Assumption 3 as follows.

Counterproposal 3: Speaking politely, or showing deference and refinement through

the use of honorifics and other formal or indirect expressions, may not necessarily
index femininity or female gender; its interpretation depends on the (ideologically
based) criterion one uses for the evaluation of womens speech in specific contexts.

This proposal is illustrated by the age difference in the use of referent honorifics
seen earlier. Older womens greater use of honorifics could be an indication that they
are more affected by the normative expectation. That is, for them honorifics may serve
as an expression of onnarashisa. At the same time, however, young womens less fre-
quent use of honorifics may not be interpreted as unfeminine. When talking with a friend
about a higher status person in the latters absence, one may not use honorifics for the
absent person because the formality expressed toward the referent may be taken as a
sign of distance toward the addressee. Young womens nonuse of honorifics may be a
reflection of this concern and not an attempt to speak in an unfeminine manner. In
a similar vein, female vendors nonuse of honorifics may not be perceived as unfemi-
nine, if it is interpreted as a sign of friendliness. However, some may perceive the non-
use of honorifics both by young women and by vendors as unfeminine.
Similarly, the frequent use of honorifics by male salesclerks at department stores
and male cooking instructors on television is not likely to be construed as efforts to
speak like women; rather, it seems to be intended to make their speech fit their roles
and the setting. Likewise, the use of honorifics by the male supervisor toward a fe-
male subordinate in example (12) may be an attempt to present himself as an egali-
tarian person. Although previous studies on politeness have emphasized womens
concerns about their image, these examples suggest that men as well as women are
concerned about how they want to appear to others and use language strategically
(see also chapter 15, this volume, for a discussion of mens linguistic prowess).
In sum, to what extent polite speech, or the use of honorifics, is interpreted as
onnarashii depends on the criteria one uses in specific contexts. Important to con-
sider in this relation are the contexts in which regional dialects are spoken. For ex-
ample, female speakers of regional dialects are sometimes evaluated as rough and
unrefined (Miyake 1995), that is, as not onnarashii. The reference point in this evalu-
ation seems to be stereotypical womens speech in Standard Japanese. However,
speakers of regional dialects may not share such an evaluation. For example, speak-
ers of dialects that have few honorifics (Inoue 1989) may resort to the use of Stan-
dard Japanese honorifics, particularly in relatively formal situations. However, one
may choose not to do so, if one wished to establish solidarity with the interlocutor
who speaks the same dialect (see chapter 10, this volume). In such a case, the nonuse
of honorifics by women does not seem to be considered an unfeminine behavior, at
least by the speakers themselves. This is not to suggest that female speakers of re-
gional dialects are not affected at all by the normative expectation that women should
speak more politely than men. Many of them may be affected by it but may express
politeness differently from the dominant norm. In the academic literature, the so-
 IDEOLOGY IN LINGUISTIC PRACTICE AND ANALYSIS 53

called Japanese womens language is usually characterized in terms of Standard

Japanese features or the (presumed) speech of middle-class women in Tokyo.
However, speakers of regional Japanese may not understand prescriptions such as
Speak onnarashiku only in terms of features of Standard Japanese. It needs to
be investigated to what extent societal norms, such as Speak onnarashikuand
Speak politely, are linked to Standard Japanese or, to put it differently, to what
extent scientists focus on Standard Japanese has contributed to the construction of
a narrow societal norm for womens speech informed by the hegemonic ideology of
class, gender, and region.

2.5. Conclusion

Although honorifics (and other formal and indirect expressions) in Japanese are often
linked to politeness and onnarashisa, I have argued that such a link is not inevitable
but rather is based on the Japanese hegemonic ideology of language, gender, and
class. I have further argued that both linguistic politeness and onna-rashisa involve
evaluations of speech vis--vis what one thinks of as the norms for specific situa-
tions. The hegemonic ideology of politeness and onnarashisa is widely promulgated
in Japanese society and may considerably affect ones conception of the norms. As
a mediating link between language and social structure, this ideology provides a basis
for the interpretation of honorifics as polite and onnarashii, which in turn contrib-
utes to the maintenance of social inequality in regard to gender and class. However,
my analysis suggests that what is understood as the normative usage may not always
be the same. It may vary among individuals and across time; it may also differ de-
pending on the context. Accordingly, honorifics may not always be perceived as polite
and onnarashii. Individuals negotiate the dominant ideology and may or may not
adopt it; they choose expressions strategically based on what they think are most ap-
propriate for specific situations.
The complex indexical process analyzed in this chapter demonstrates that it is
important not only to distinguish language ideology and practice in linguistic analy-
sis but also to appropriately assess the role of ideologies in the interpretation of so-
cial meanings of linguistic forms used in specific contexts. In particular, it illustrates
the importance of considering diversity in the interpretation and use of language as
related to variability in individuals apprehension of speech norms for local contexts.

Notes
I would like to thank Akesha Baron, Sarah Benor, Mary Bucholtz, Misao Okada, Devyani
Sharma, and Yukako Sunaoshi for their valuable comments on the earlier versions of this
chapter. Special thanks go to Janet Shibamoto Smith, my coeditor, and Sumiyuki Yukawa,
who have continuously provided me with numerous valuable comments, suggestions, useful
information, and encouragement.

1. The following abbreviations are used for different types of (non-)honorific forms:
AH = addressee honorific; RH/R = referent honorific-respectful word; RH/H = referent
54 HISTORICAL AND THEORETICAL FOUNDATIONS

honorific-humble word; RH/B= referent honorific-beautification word; NRH = nonreferent

honorific; and NAH = nonaddressee honorific. Other abbreviations used in this study are:
AUX = auxiliary; GD = gerundive form; GN = genitive marker; N = noun; OM = ojbect marker;
Pst = past tense; Q = question marker; SM = subject marker; and V = verb.
2. In the case of verbal expressions, sonkee-go are used for the action of the subject
referent and kenjoo-go for that of the object (or nonsubject) referent.
3. The word hito usually means person, but it is translated here as woman because
it was written with the Sino-Japanese characters women (usually read as josee) but accompa-
nied by interlinear kana syllabary characters that gave its pronunciation as hito.

References
Bergvall, Victoria L. (1999). Toward a comprehensive theory of language and gender. Lan-
guage in Society 28(2): 273293.
Bergvall, Victoria L., Janet M. Bing, and Alice F. Freed (eds.) (1996). Rethinking language
and gender research: Theory and practice. London: Longman.
Briggs, Charles L. (1998). Youre a liaryoure just like a woman! Constructing domi-
nant ideologies of language in Warao mens gossip. In B. B. Schieffelin, K. A. Woolard,
and P. V. Kroskrity (eds.), Language ideologies: Practice and theory, 229255. Ox-
ford: Oxford University Press.
Brown, Penelope (1980). How and why are women more polite: Some evidence from a Mayan
community. In S. McConnell-Ginet, R. Borker, and N. Furman (eds.), Women and lan-
guage in literature and society, 111139. New York: Praeger.
Brown, Penelope, and Stephen C. Levinson (1987). Politeness: Some universals in language
usage. Cambridge: Cambridge University Press.
Bucholtz, Mary (1999). Bad examples: Transgression and progress in language and gender
studies. In M. Bucholtz, A. C. Liang, and L. A. Sutton (eds.), Reinventing identities:
The gendered self in discourse, 324. New York: Oxford University Press.
Bucholtz, Mary, A. C. Liang, and Laurel A. Sutton (eds.) (1999). Reinventing identities: The
gendered self in discourse. New York: Oxford University Press.
Bunkacho (1996). Kokugo-shingikai hookokusho 20 (National Language Council report 20).
Tokyo: Okurasho Insatsu-kyoku.
Cameron, Deborah (1988). Introduction. In J. Coates and D. Cameron (eds.), Women in their
speech communities: New perspectives on language and sex, 312. London: Longman.
(1990). Demythologizing sociolinguistics: Why language does not reflect society. In
J. E. Joseph and T. J. Taylor (eds.), Ideologies of language, 7993. London: Routledge.
Dunn, Cynthia (1996). The construction of identity in conversation: An intergenerational
comparison. Paper presented at the 5th International Pragmatics Conference, Mexico City.
Eckert, Penelope and Sally McConnell-Ginet (1999). New generalizations and explanations
in language and gender research. Language in Society 28(2): 185201.
Eelen, Gino (1999). Ideology in politeness: A critical analysis. Unpublished doctoral disser-
tation, University of Antwerp.
Endo, Orie (1997). Onna no kotoba no bunkashi (Cultural history of Japanese womens lan-
guage). Tokyo: Gakuyo Shobo.
Hall, Kira, and Mary Bucholtz (eds.) (1995). Gender articulated: Language and the socially
constructed self. New York: Routledge.
Holmes, Janet (1995). Women, men and politeness. London: Longman.
Ide, Sachiko (1982). Japanese sociolinguistics: Politeness and womens language. Lingua 57:
357385.
 IDEOLOGY IN LINGUISTIC PRACTICE AND ANALYSIS 55

(1990). How and why do women speak more politely in Japanese? In S. Ide and N. H.
McGloin (eds.), Aspects of Japanese womens language, 6379. Tokyo: Kuroshio.
(1993). Sekai no joseego, Nihon no joseego: Joseego kenkyuu no shintenkai o motomete
(Womens language of the world, womens language of Japan: Searching for new de-
velopments in womens language research). Nihongogaku 12(6): 412.
Inoue, Fumio (1989). Kotobazukai shinfuukee: Keego to hoogen (New views on ways of speak-
ing: Honorifics and dialects). Tokyo: Akiyama Shoten.
James, Deborah (1996). Women, men and prestige speech forms: A critical review. In V. L.
Bergvall, J. M. Bing, and A. F. Freed (eds.), Rethinking language and gender research:
Theory and practice, 98125. London: Longman.
Jugaku, Akiko (1979). Nihongo to onna (The Japanese language and women). Tokyo: Iwanami
Shoten.
Kanai, Yoshiko (1994). Onna no miryoku wa hanashi-kata shidai (Womens attractiveness
depends on how they speak). Tokyo: Yamato Shuppan.
Kawanari, Mika (1993). Irai-hyoogen (Expressions of request). Nihongo-gaku 12(6): 121134.
Keenan, Elinor (1974). Norm-makers, norm-breakers: uses of speech by men and women in
a Malagasy community. In R. Bauman and J. Sherzer (eds.), Explorations in the ethnog-
raphy of speaking, 125143. Cambridge: Cambridge University Press.
Komori, Yoichi (2000). Nihongo no kindai (The Japanese language in modern times). Tokyo:
Iwanami Shoten.
Lakoff, Robin (1975). Language and womans place. New York: Harper and Row.
Lee, Yeounsuk (1996). Kokugo to yuu shisoo (The ideology of the national language).
Tokyo: Iwanami Shoten.
Leech, Geoffrey (1983). Principles of pragmatics. London: Longman.
Miller, Laura (1996). Subversive subordinates or situated language use? A consideration of
keigo ideology and sociolinguistic description. Paper presented at the 48th annual meet-
ing of the Association for Asian Studies, Honolulu.
Minami, Fujio (1987). Keego (Honorifics). Tokyo: Iwanami Shoten.
Miyake, Yoshimi (1995). A dialect in the face of the standard: A Japanese case study. In
J. Ahlers, L. Bilmes, J. S. Guenter, B. A. Kaiser, and J. Namkung (eds.), Proceedings of
the 21st Annual Meeting of the Berkeley Linguistics Society, 217225. Berkeley: Berke-
ley Linguistics Society.
Nakamura, Momoko (1995). Kotoba to feminizumu (Language and feminism). Tokyo: Keiso
Shobo.
Niyekawa, Agnes (1991). Minimum essential politeness: A guide to the Japanese honorific
language. Tokyo: Kodansha International.
Ochs, Elinor (1993). Indexing gender. In B. D. Miller (ed.), Sex and gender hierarchies, 146
169. Cambridge, MA: Cambridge University Press.
Ogino, Tsunao, Yasuko Misono, and Chitsuko Fukushima (1985). Diversity of honorific usage
in Tokyo: A sociolinguistic approach based on a field survey. International Journal of
the Sociology of Language 55: 2339.
Ohashi, Haruo (1994a). Josee kara okuru tegami no kaki-kata (How to write womens let-
ters). Tokyo: Fujin-seikatsu-sha.
(1994b). Dansee kara okuru tegami no kaki-kata (How to write mens letters). Tokyo:
Fujin-seikatsu-sha.
Okamoto, Shigeko (1996). Indexical meaning, linguistic ideology, and Japanese womens
speech. In The Proceedings of the 22nd Annual Meeting of the Berkeley Linguistics
Society, 290301. Berkeley: Berkeley Linguistics Society.
(1997). Social context, linguistic ideology, and indexical expressions in Japanese.
Journal of Pragmatics 28(6): 795817.
56 HISTORICAL AND THEORETICAL FOUNDATIONS

(1999). Situated politeness: Manipulating honorific and non-honorific expressions

in Japanese conversations. Pragmatics 8(2): 5174.
Reynolds, Katsue Akiba (1985). Female speakers of Japanese. Feminist Issues 5: 1346.
(1990). Female speakers of Japanese in transition. In S. Ide and N. H. McGloin (eds.),
Aspects of Japanese womens language, 129146. Tokyo: Kuroshio Shuppan.
Shibamoto, Janet S. (1987). Japanese sociolinguistics. Annual Review of Anthropology 16:
261278.
[Shibamoto] Smith, Janet S. (1992). Linguistic privilege: Just stating the facts in Japanese.
In K. Hall, M. Bucholtz, and B. Moonwomon (eds.), Locating power: Proceedings of
the Second Berkeley Women and Language Conference, 4048. Berkeley: Berkeley
Women and Language Group.
Shibatani, Masayoshi (1990). The languages of Japan. Cambridge, MA: Cambridge Univer-
sity Press.
Shigemitsu, Yuka (1993). Kaiwa no pataan (Conversational patterns). Nihongogaku 12(6):
135140.
Silverstein, Michael (1979). Language structure and linguistic ideology. In P. R. Clyne,
W. Hanks, and C. L. Hofbauer (eds.), The elements: A parasession on linguistic units
and levels, 193247. Chicago: Chicago Linguistic Society.
Sugimoto, Tsutomu (1997). Onna to kotoba ima-mukashi (Women and language, now and
then). Tokyo: Yuzankaku.
Suzuki, Kenji (1989). Suteki na anata o tsukuru: Josee no utsukushii hanashi-kata (To make
youself nice: Womens beautiful ways of speaking). Tokyo: Goto Shoin.
Suzuki, Mutsumi (1993). Joseego no honshitsu: Teeneesa, hatsuwa kooi no shiten kara (The
nature of womens language: Viewing from politeness and speech acts). Nihongogaku
12(6): 148155.
Takahashi, Yoshiko (1996). Whats cooking? An examination of Japanese male language
within media domains. Paper presented at the 5th International Pragmatics Conference,
Mexico City.
Takasaki, Midori (1993). Josee no kotoba to kaisoo (Womens language and social class).
Nihongogaku 12(6): 169180.
Tanaka, Sumie (1986). Kashikoi hito ni narinasai:Utsukushiku ikitai anata ni (Be a wise
woman: To you who want to live beautifully). Tokyo: PHP Research Institute.
Trudgill, Peter (1972). Sex, covert prestige and linguistic change in the urban British English
of Norwich. Language in Society 1: 179195.
Washi, Rumi (2000). Nyooboo kotoba no imisayoo: Tennoosee, kaisoosee, sekushuaritii (In
political function of nyooboo kotoba: Symbolizing the imperial system, class conscious-
ness, and sexuality). Joseegaku Nenpoo 21: 1835.
Wetzel, Patrica J. (1994). Contemporary Japanese attitudes toward honorifics (keigo). Lan-
guage Variation and Change 6: 113147.
Woolard, Kathryn A., and Bambi B. Schieffelin (1994). Language ideology. Annual Review
of Anthropology 23: 5582.
 3

MIYAKO INOUE

Gender, Language, and Modernity

Toward an Effective History of
Japanese Womens Language

J apanese womens language is a socially powerful truth. I mean by this not that
the phrase refers to the empirical speech patterns of women. Rather, I mean that Japa-
nese womens language is a critical cultural category and an unavoidable part of
practical social knowledge in contemporary Japan. The phrase names a space of dis-
course in which the Japanese woman is objectified, evaluated, studied, staged, and
normalized through her imputed language use and is thus rendered into a knowable
and unified object. Doxic statements, such as Women speak more politely than men
and Women are not capable of speaking logically, are commonly heard in daily
conversation. Scholars, too, have produced a highly reflexive and abstractthere-
fore privilegedknowledge of how women speak (differently from men); they have
systematically located female-male differentiation at all levels of the language: pho-
nology, semantics, morphology, syntax, speech acts, discourse (in the technical lin-
guistic sense), as well as prosodic features such as pitch, and they have explained
how female-specific values, attributes, and social roles are registered in speech forms
and in the management of conversation.
Womens language is also a national issue, a reflexive parameter of civil order
and of social change. Nationwide opinion polls are regularly conducted on whether
womens language is becoming corrupted and how much so; national sentiments
over its perceived disappearance are thereby crystallized and circulated in the form of
numbers and statistics.1 This linguistic consciousness of how women speak is closely
connected with notions of culture and tradition in the assumption that womens lan-
guage is uniquely Japanese, with unbroken historical roots in an archetypical Japanese
past, and inescapably linked with an equally traditional and archetypical Japanese

57
58 HISTORICAL AND THEORETICAL FOUNDATIONS

womanhood. For example, Kindaichi Kyosuke, one of the founders of modern Japa-
nese linguistics, noted in his discussion of womens language: Japanese woman-
hood is now being recognized as beautiful and excellent beyond compare with the
other womanhoods of the world. Likewise, Japanese womens language is so good
that it seems to me that it is, along with Japanese womanhood, unique in the world
(1942:293). Kikuzawa Sueo, who is noted as one of the first modern linguists to bring
attention to womens language, observed: Womens speech is characterized by ele-
gance, that is, gentleness and beauty. Moreover, such characteristics correspond with
our unique national language (1929:75). Womens language is also viewed as cul-
tural heritage, on the brink of vanishing and in urgent need of preservation.2 As
Mashimo Saburo, another scholar of Japanese womens language, puts it, We can
not hope for contemporary Japanese women to be as witty and tactful as were those
in the past, but, at least, I would like them to have a sincere and humble attitude and
to preserve the cultural heritage passed down from the ancestors without destroying
it (1969:81). Talk of womens language, thus, necessarily always implicates the
perceived continuing contradiction between Japanese tradition and modernity.
But how and why did some speech forms and functions come to be identified as
womens language? How and why have they become promoted from unselfconscious
sound to a universalized, national symbol that is a both socioculturally and linguis-
tically discrete index? Most important, how did such an indexical practicea link-
age of speech with social structure and cultural meaningcome to be possible to
begin with? Scholars of kokugogaku national language studies3 often date the origin
of womens language as early as the fourth century, and they commonly construct a
seamless narrative of Japanese womens language passed down to the present.
Evidence of womens language is traced in premodern literary works and in records
of terminology used by sequestered groups of feudal women such as court ladies,
Buddhist nuns, and women in the pleasure quarters.4
This primordialist discourse, however, provides an adequate historical-linguistic
account neither of the development of contemporary womens language nor of a
continuous descent from ancient origins. Rather, this discourse merely assumes an
essence of Japanese womens language that teleologically originates at some an-
cient time and descends without interruption or transformation down to the present.
The isolated and discontinuous examples are meant to illustrate the continuous
essence assumed to lie behind them. For our purposes, the point is that because it
denies historical contingency and ignores emergent phenomena this discourse para-
doxically erases the material traces of womens diverse linguistic experience and ends
up simply affirming the transcendental national narrative of culture and tradition. It
hides histories by articulating (teleological) Historywith a capital H.
Bringing this historical threshold into theoretical focus requires a critical method
that allows us to recognize discontinuity in history, a goal Foucault (1977:153155)
calls effective history, whose method is genealogy. This would permit us to locate
not the origin of a transhistorical essence but the emergence of a complex ensemble.
The concept of emergence presupposes neither teleological continuity nor recalci-
trant relativism and requires us to seek the history of the present not in the ideal but
in the material and embodied context that entails multiple social forces in conjunc-
tionin this case, Japans unprecedented capitalist takeoff. In this chapter, I will,
 GENDER , LANGUAGE , AND MODERNITY 59

thus, examine the genealogy of Japanese womens language by locating its one
critical moment of ascendancy at the threshold of Japans modernity during the late
nineteenth and the early twentieth century, when state formation, nationalism, capi-
talist accumulation, industrialization, radical class reconfiguration, colonialism, and
foreign military adventurism were in full efflorescence. It was in this context that
both language and women came to be problematized as national issues and thus to
be political and cultural targets of the state authorities and of intellectuals and entre-
preneurs who represented the progressive classes. I will show how particular speech
forms were carved out, selected, and (re)constructed as Japanese womens language
and how that process was critically linked to a network of diverse institutional and
individual practices bent on modernizationand the particular form Japanese womens
language took in its complex mimicry of, and resistance to, the West.
The significance of this history lies not so much in the emergence of specific speech
forms associated with womens language as in the conditions of modernization and
modernity that, to begin with, made possible and thinkable the practice of the indexi-
cal signaling of women as a nationally regimented category. In other words, history
here involved the opening of a new cultural space where women became objectified
through their language use and thus became the productive site of knowledge of Japa-
nese women that was overdetermined by the production of knowledge of nation, race,
and class. Using Hankss (1996:278) insightful terminology, it is the historical con-
struction of a metalinguistic gaze upon women that is my subject.
In order to zero in on the emergence of womens language, I will focus on the
linguistic modernization movements variously pursued for different goals by the
government agencies, the literary community, the print media, and linguists and
educators from the late nineteenth century to the early twentieth century. These in-
dependent initiatives eventually converged and were led by a literary movement called
gembunitchi speech and writing unification to create a colloquial written Japanese
and to develop modern narrative prose (the novel form, or shoosetsu). The novel is
a distinctively modern representational institution, shaping and shaped by the ad-
vent of industrial capitalism, the rise of the middle class, and the development of
mass print capitalism (Anderson 1983). Most important, however, the critical link-
age between the novel and modernity lies in the latters epistemological commitment
to realism and referentiality or a modernist certitude that language is a transparent
medium that can faithfully and truthfully represent reality. The gembunitchi move-
ment thus engendered a new language ideology (Silverstein 1979, Woolard &
Schieffelin 1994, Schieffelin, Woolard, & Kroskrity 1998) as to what language is
and how language works. This chapter will argue that such a newly developed lin-
guistic consciousness was both the instrument and the critical location of the birth of
Japanese womens language as the subjects modern Japanese women variously
emerged in the state, civil society, and the market.

3.1. Nationalizing women, modernizing women

The late Meiji period, the two decades from 1888 to 1910, was critical for Japans
modern nation-state formation. This period saw the development of heavy and textile
60 HISTORICAL AND THEORETICAL FOUNDATIONS

industries, mass communication and transportation systems, and a legal apparatus

organized through the promulgation of the Meiji Constitution and Civil Code, the
establishment of representative democracy with the opening of the Diet, and the cre-
ation of a direct administrative channel between local and central government. The
development of print capitalism (in the form of mass-circulation newspapers and
magazines) and the instituting of compulsory education further molded a sense of
the population not only as an administrative/political body but also as a nation-state,
where people came to identify themselves as Japanese and to imagine Japan as
more than an administrative unit. Individuals also came to be directlybut un-
evenlyconnected to the state through education, censuses, taxation, and new (bour-
geois) legal rights. At the same time, citizens were differentially positioned, in terms
of class, power, and culture, in the newly emerging mosaic of a capitalist society.
Modern power, as Foucault tells us, is both totalizing and individuating (1982:208
226). These channels of power converged in the reorganization of the person as the
modern (and imperial) subject in the late Meiji period. It was in this context of mod-
ern social power in the form of capitalist development and state centralization that
women as a social category became radically renewed and modern Japanese
women emerged as an articulable social category burdened with new cultural mean-
ings pertinent to its relationship with the nation-state.
One concrete terrain on which women became consolidated as a social category
in conjunction with the historical and cultural complexity of Japans modernity was
womens public education. With womens secondary education incorporated for the
first time with the state-regulated public education system in 1899, the government
actively launched a project to nationalize women and shape their roles vis--vis the
state. Central to this endeavor was the idea of a good wife and wise mother (ryoosai
kembo). The project putatively advocated the traditional virtues and values of ideal
womanhood, such as obedience to father, husband, and, later, eldest male child. Far
from primordial, however, this ideology derived both from the Confucianism espoused
by the ex-samurai class and from the imported Western cult of domesticity. Eclectic
yet decisively Japanese modern, the teaching of good wife and wise mother pre-
pared women to take on a critical gendered role in an anticipated modern capitalist
society, which included an emphasis on motherhood,5 rational and scientific house
management (including hygiene and efficient home economy), and saving.6
Woman, here as elsewhere, came to embody the shifting boundary between tra-
dition and modernity, and her body and sexuality became concrete sites where this
irrevocable binary was negotiated and policed (Mani 1987, Chatterjee 1990, Stoler
1991). Womens language was one powerful effect of cultural work in the realms
of politics, the market, civil society, and personal lifecultural work that sought to
give the nascent signifier, the modern Japanese woman, her voice.

3.2. Write as you speak: Gembunitchi and

the discursive space of the nation-state

For the Meiji elite, language7 was the foremost critical institution to be modern-
ized because it was recognized as an instrument to build a nation-stateto import
 GENDER , LANGUAGE , AND MODERNITY 61

and simulate Western science and technology and to achieve national integration. In
this section, I trace the historical process and various practices of language modern-
ization in the late nineteenth and the early twentieth century, which resulted in a lin-
guistic consciousness intimately connected with the exercise of modern forms of
power. The emergence of the metapragmatic category of womens language is
predicated upon such modern linguistic consciousness, both as a technique with which
new knowledge about (modern Japanese) women was produced and as an epistemo-
logical ground upon which such new knowledge was made intelligible.
The major language modernization movement was in the literary community,
which was called gembunitchi. Gembunitchi unifying speech and writing devel-
oped out of the progressive Meiji writers concern with the lack of an adequate liter-
ary style satisfactory for modern narrative prose as found in the Western realist novel
they saw as a model. They developed the new style of colloquial written Japanese
called gembunitchi-tai gembunitchi style. It is not a coincidence that gembunitchi
took the lead in language modernization. The novelwith its distinct origin in the
late eighteenth- and early nineteenth-century Europewas a technology that was as
constitutive of the modern nation-state as were legislatures, laws, citizenship, po-
liced borders, and standing armies. Its generic framing demands truth telling about
the realities of ordinary people and their daily lives. The novel, thus, attempted to
put on public exhibition the Japanese citizenrycomposed of individual, ordinary
people. This realist metanarrative condition was inseparably mediated by and ines-
capably linked with forces emerging with capitalism in Europe, which involved the
rise of a middle class, the articulation of possessive individualism, the precipitation
of nationalism, and the birth of print capitalism. Through culturally contingent re-
working of the Western realist novel, the Meiji progressive writers encountered a
new idea of language that enabled and was enabled by such elements of western
modernity.
The core of gembunitchi as an effort to create a new colloquial style boiled down
to the stylistic question of how to entextualize8 linguistic excess, the sheer physical-
ity and materiality of the human voice. In an actual face-to-face interaction, we not
only exchange the semantic or referential meaning of what we utter, but also, at the
same time, we also communicate pragmatic meaning, by how we speak, with respect
to elements of the immediate context of the interaction, including the social setting
of the interaction and the social attributes of the participants, such as their gender,
social rank and roles, age, and so on. The gembunichi writers sought to include this
linguistic excess in the text with verb-ending forms. For the writers, verb-ending forms
were the site to deal with such linguistic excess because these function indexically to
mark the authors social and psychological position vis--vis the characters in the
text and the reader in the context. In other words, verb-ending forms are regimented
into an indexical order of different ways of saying the same thing (Silverstein
1996:280). Depending on the verb-ending form one chooses, different pragmatic
effects, with the same referential value, are produced as to how the narrator (speaker)
narrates (talks) to the reader (listener) about the characters and events.
During the time between the Sino-Japanese War (18941895) and the Russo-
Japanese War (19041905), the literary gembunitchi movement was appropriated
by the states nationalist effort to create a national language (kokugo)and, more
62 HISTORICAL AND THEORETICAL FOUNDATIONS

precisely, a standard national language. For the state intellectuals in the early Meiji,
gembunitchi was not a primary concern in language reform. They were concerned
mainly with writing or script reforms for example, whether or not Chinese characters
should be used, for modernization meant Westernization and a break with China.
When the idea of a national language was introduced, however, the terms of the debate
on national-language policy shifted from questions of writing to speech matters.
The gembunitchi movement itself underwent significant transformation as it
became mobilized for the technology of the modern nation-state: Against the origi-
nal advocacy of a spirit of vernacularism, by the 1910s plain verb-ending forms
such as the da-copula had won out as the established literary style. Polite and honor-
ific verb-ending forms, which formally indexed context-bound relationships between
the author and the reader and between the author and the characters, eventually lost
their status in the serious literary style. Concurrent with the predominance of the plain
style, the first-person narrative was superseded by that of the third person. In it, the
narrators presence vanishes from the narrated event and the text ceases to acknowl-
edge the context. Instead, standing outside the narrated event and commanding a
Gods-eye view, the narrator now rationally, objectively, and truthfully represents
the scene. This is the linguistic version of Benthams panopticon (Foucault 1977:195
228). The dictum Write as you speak, then, presented a contradictory task: it de-
manded the speechlike effect of immediacy, transparency, and physicality obtained
with the plain style, but this very plain style evacuated the pragmatic meaning that
would have given the speech polysemy and undecidability. For this new narrative
style, context and audience were no longer time-space bound but became the ab-
stracted and imagined Japan and the Japanese. Thus did a linguistic technology
help to make an imagined community necessary, even if it did not really exist. The
speaking subject of the gembunitchi style literally necessitates and embodies the
modern Japanese citizenit makes him imaginable.
I say him because this narrator, this citizen, is presumed to be (the middle-
class) male9 and he alone had full and legitimate access to the newly emerging bour-
geois public sphere (Harbermas 1989, Fraser 1990, Warner 1990, Calhoun 1992). In
fact, the state language policy designated the speech of the educated Tokyo middle-
class males as the basis of standard language (Okano 1902). The phrase educated
Tokyo middle-class males alluded to the newly emerging petite bourgeoisie of sala-
ried workers in Yamanote, the plateau section of the city of Tokyo.
In short, the gembunitchi style came to narrate the nation (Bhabha 1990). This
new narrating voice as developed in the gembunitchi movement introduced a new
linguistic consciousness: language is a transparent medium, purely and exclusively
referential in its function, according to which nothing comes between language and
the world and there is an exclusive and context-free, one-to-one correspondence
between sound and word, word and meaning, and language and the world. Language
is then simply to reflect what is already out there, always one step behind the world,
docilely ratifying and confirming it. Such a realist conception is inherently ideologi-
cal because it effaces the semiotic work of language in actively mediating and pro-
ducing what is seemingly merely given, reversing the order of things as if the world
existed as it is without the mediation of language. Linked up with the regime of modern
power, it serves to turn things, categories, events, and ideas into a fait accompli.
 GENDER , LANGUAGE , AND MODERNITY 63

In more concrete terms, the new narrating voice functioned at the metalinguistic
level to signal that whatever it narrates, reports, describes, represents, and states is
true, real, serious, and credible and that it speaks not from a particular individuals
point of view but from that of the modern rational and national (male) citizenan
omniscient point of view that masquerades as not being a point of view at all. This
metalinguistic function was facilitated by formal and diacritic devices that separate
the narrating voice and the narrated (Komori 1988). Translating and appropriating
the Western realist novel required gembunitchi writers, for the first time, to develop
subordinated linguistic space in the form of dialogue and reported speech. It is a formal
space where alterity is constructed, highlighted, and neatly kept apart from the self.
The novel thus formally created a hierarchical relationship in which the narrated
whether people, events, or thingsis always already objectified by, represented
through, and subjected to the male gaze of the narrating subject, or of the modern
Japanese citizen. And it is precisely this metalinguistic effect that various Japanese
institutions and projects intent on their own modernization ultimately adopted from
the literary gembunitchi movement. In textbooks, newspapers, magazines, fiction,
scholarly essays, public speeches, legal statements, military orders, advertisements,
and colonial education, the new narrating voice not only provides semantico-refer-
ential information but also functions simultaneously as performative (Silverstein 1979,
Lee 1997), to authenticate and factualize that which is enunciated.10 It is in this lin-
guistic space, a quoted space, objectified, reified, and re-presented by the imbricated
gaze of the male, the national, and the modern, that womens language was pieced
together from heterogeneous origins.

3.3. Quoting womens language/producing

the Japanese woman

Dialogues and reported speech were new linguistic space where the maximum degree
of verisimilitude was logically implied. And this is the space where people heard,
for the first time, modern Japanese women speak. Womens voice was, as mentioned
earlier, to center on the use of final particles. Table 3.1 compares a particular form of
final particle employed in the popular fiction work titled Ukiyoburo The bathhouse of
the floating world, which was written in 1813 by Shikitei Sanba, and the narrative
prose work titled Sanshiro, which was written in 1909 by Natsume Soseki.
Sanba was one of the traditional popular fiction (gesaku) writers of the late Edo
period, before the Meiji Restoration. Ukiyoburo is about the frivolous interactions
of people coming to the bathhouse. It consists of dialogues in which they tease, argue,
compliment, gossip about, and comment on one another. The characters in Ukiyoburo
are diverse in age, gender, social stratification, region, and occupation, according to
which Sanba carefully differentiates and characterizes individual speech styles. What
is glaringly absent in table 3.1 is anything that looks remotely like what contempo-
rary womens language is believed to be.
Soseki was one of the best-known Meiji writers, and Sanshiro was published at
the culmination of the literary gembunitchi movement. By then, whether Soseki
intended to or not, he and other early twentieth-century Japanese writers had
64 HISTORICAL AND THEORETICAL FOUNDATIONS

3.1 The comparison of

final particles attached to verb-
ending form da in two
Japanese novels
Ukiyoburo Sanshiro
(1813) (1909)

da-naa M
da-te M
da-te-na M
da-te-ne M
da-wa-i M
da-wa-su M
da-yoo M
da-ze-e M
da-e F
da-ne-nee F
da-no-ya F
da-yo-nee F
da-yo-noo F
da-mono-o B
da-na B da-na M
da-ne B da-ne M
da-nee B
da-no B
da-noo B
da-su B
da-wa B da-wa F
da-wa-e B
da-wa-na B
da-wa-sa B
da-yo B da-yo M
da-ze B da-ze M
da-zo B da-zo M
da M
da-koto F

M: used by male characters only. F: used

by female characters only. B: used by both
male and female characters.
The data is drawn from Komatsu (1988).
The reader should note that I have rear-
ranged Komatsus data to draw my own
conclusions and that I am using his data
for purposes other than those he intended.

inescapably become deeply involved with narrating the nation and its modern sub-
ject. Sanshiro depicts the lives of modern women and men, new characters that did
not exist prior to Meiji. The significance of Sanshiro is that the voices assigned to
these women and men contain both the female-exclusive and male-exclusive final
particles, identical with those of contemporary Japanese. Table 3.1 compares the final
particles attached to the verb-ending form da in Ukiyoburo and in Sanshiro. It is
 GENDER , LANGUAGE , AND MODERNITY 65

noteworthy, first of all, that the gender-neutral final particles in Ukiyoburo have
become gendered into either female-exclusive or male-exclusive in Sanshiro. Sec-
ond, the final particles in Ukiyoburo, on the one hand, are not gendered but rather
are idiosyncratic: there is a plurality of individual voices in the use of final particles,
and final particles are, thus, not in a position to index anything, perhaps because they
index everybody. In Sanshiro, on the other hand, final particles have become sys-
tematized and standardized so as to index gender in the modern nation. By this time,
female and male Japanese subjects are imaginable components within the nation, and
the modern Japanese novel significantly flattens out the individual grain of the voice,
so that one Japanese woman (or man) is interchangeable with another.
Table 3.2 compares another set of final particles in Ukiyoburo and in Sanshiro.
The left-hand column lists the female-exclusive and male-exclusive final particles
that appear in Sanshiro, and the right-hand column is drawn from Ukiyoburo. The

3.2 The comparison of final particles between

Ukiyoburo and Sanshiro
Sanshiro Ukiyoburo
(1909) (1813)

da-koto F B
no F B
NOM. + yo F B
*wa F B
da(COP.) M B
da-ne M B
da-yo M B
na M B
sa M B
zo M B
*na-no F / No example
*no-ne F / No example
*NOM. + nee F / No example
*wa-ne F / No example
*wa-yo F / No example
*da-wa F B Not used by samurai-class females
*no-yo F B Not used by samurai-class females
da-na M B Not used by samurai-class females
da-ze M B Not used by samurai-class females
da-zo M B Not used by samurai-class females
ze M B Not used by samurai-class females
zee M B Not used by samurai-class females
NOM. + ne F B Only few examples seen
ya M B Various usages
M: used by male characters only. F: used by female characters only. B: used
by both male and female characters.
COP.Bcopula
NOM.Bnominal
The data is drawn from Komatsu (1988). The reader should note that I have
rearranged Komatsus data to draw my own conclusions and that I am using
his data for purposes other than those he intended.
66 HISTORICAL AND THEORETICAL FOUNDATIONS

final particles marked with an asterisk are identified today as quintessentially female-
exclusive final particles. While the final particle wa appeared in Ukiyoburo, the other
female-exclusive final particles of contemporary Japanese womens language
na-no, no-ne, nominal plus ne, wa-ne, and wa-yodid not appear at all in the pre-
Meiji work. The next two asterisked final particles, da-wa and no-yo, which are par-
ticularly salient feminine particles in the present, are not assigned to the female
characters in the elite (samurai) class. Da-wa and no-yo were in fact considered to be
vulgar and low-class as late as the end of the nineteenth century, and educators
and others strongly advised parents and teachers not to let their daughters use them.
We hardly have here a seamless history of a traditional Japanese womans voice.
My point is not that Sanshiro represented womens language because that is how,
in actual fact, Japanese women had come to speak in 1909.11 The modern narrating
voice may tell us metalinguistically that what is being reported is merely that which is
actually spoken out there, but we should not be deceived by this metalinguistic
whispering. Reported speech (Voloshinov 1973, Bakhtin 1981) entails the authorial
(and social) act of (re)creating womens voice within a socially producedand self-
reproducingknowledge of how women speak. As Voloshinov (1973:82) argues,
reported speech is a social phenomenon, a historical product that wells up from a
complex social field in ideological and political flux.
The degree to which the speech of female characters in the modern novel was
not, in fact, naturalistically represented during the early stage of the gembunitchi
movement is evidenced by the recollections of the gembunitchi writers themselves.
For example, writer Sato Haruo notes: In those days, womens speech in daily conver-
sation was invented by certain writer(s) (though I do not remember who he or she
was) and came to be of general use (1941:18). Tsubouchi Shoyo also recalled late
in his life the difficulties of writing the speech of female characters, observing: In
those days, the language used by women in the middle class and beyond was filled
with so many honorifics that one could not possibly manage to use it for translation
(1930:7). Many gembunitchi writers developed their modern narrative prose by using
Japanese that had been translated from (and thus filtered through) Western languages.
Futabatei Shimei, for example, wrote his first novel, Ukigumo, while translating
Turgenevs A Sportsmans Notebook into Japanese. Tsubouchi also translated into
Japanese a large body of Shakespeares work. The irony is that these writers needed
the authentic speaking voice of modern Japanese women in order to represent that of
white women.
What these writers hit upon to solve the problem of womens reported speech was
the speech of jogakusee schoolgirls that they claimed that they overheard on the street.
Tsubouchi thus exclaimed, There were unimaginable obstacles and difficulties that
writers who were used to hearing jogakusee-kotoba the speech of schoolgirls from
the end of Meijitodays onna-kotoba female speech stylecould not have even
dreamed of. Oh, how blessed contemporary writers are! (1930:7). The modern edu-
cation policy made it possible for daughters of the elite family to go to girls high schools.
Jogakusee12 thus represented a new social category of female: they were neither pro-
ducers (workers) nor reproducers (mothers). As jogakusee as a cultural construct be-
came increasingly recognized, objectified, and imagined as a metonymy for Japans
modernization, so did their speech. Schoolgirls were reported to use a set of distinctive
 GENDER , LANGUAGE , AND MODERNITY 67

final particles, including te-yo, da-wa, no-yo, and others, many of which are the essen-
tial linguistic features identified today as womens language.
The final particle te-yo first appeared in a novel in 1888 (used by a young woman
to a man and to her female friend), no-yo in 1885 (used by a young woman to a man),
and da-wa in 1886 (used by a young woman to her maid and in her monologue;
Ishikawa 1972). In female characters speech, final particles such as te-yo, da-wa,
and no-yo were thus initially restricted to the speech of young women or schoolgirl
characters. Later, however, they came to be used in a wider variety of dialogues, for
example, those between wife and husband, or between daughter and father (Ishikawa
1972). By the early twentieth century, these particles, claimed by writers to have
derived from schoolgirls speech, had become elevated in writing to womens lan-
guage, through the work of writers actively to indexicalize the speech of female
characters as genericyet increasingly feminizedfemale speech. The elevation
of schoolgirls speech to generic womens language, the voice of the modern gendered
subject, was thus far from being a natural history in any way. Schoolgirls speech
originally raised intellectuals and educators eyebrows as vulgar and low-class.
For example, writer Ozaki Koyo (1994 [1888]: 45) warned in 1888 that, eight or
nine years previously, a certain speech style with strange sorts of verb endings such
as te-yo, no-yo, and da-wa had first occurred among elementary schoolgirls between
close friends and seemed to be spreading among high school girls and even grown
women. He insisted that sensible ladies would never use these verb endings because
they were originally part of the vulgar speech used by daughters of low-rank samu-
rai families. Intellectuals and educators further developed this origin narrative by
locating the original speakers of these verb-ending forms in the pleasure quarters and
teahouses. These intellectuals and educators claimed that the speech of these women
was adopted by daughters from the low-rank samurai families and, later, by the middle
class and elites.
Indexicalization here involved not just the active linking of a sign to a referent but
also the active construction of the referent itself. In other words, the speech of school-
girlsas a constructed voicecould not be (re)signified to become womens language
unless the discursive and disciplinary space of the modern Japanese woman existed.
This critical move took place in the discursive space where the states intended sur-
veillance of women, print capitalisms interest in women as a market, and women
themselves as new consumers intersected at a particular conjunction in Japanese his-
tory. This discursive space was also largely without a head. While it could be ar-
gued that the state had power because of its power to censor or that consumers
had power because they could vote with their feet, this discursive space was, as
Foucault said of disciplinary power, multiple, automatic and anonymous power
that functioned from top to bottom, but also to a certain extent from bottom to top
and laterally (1977:176). In fact, it is the very lack of an actor clearly in control
within this discursive space that helps to erase the extent to which the Japanese
woman was socially produced.
This process is most visible in kateeshoosetsu the domestic novel, a particular
form of the Japanese novel that appeared in the early twentieth century. The domestic
novel was originally serialized in newspapers, as a technique to expand readership
among nonelite readers in order to create a mass market for newspapers. Particularly
68 HISTORICAL AND THEORETICAL FOUNDATIONS

new among the readers was an increasing number of literate women. With this read-
ership in mind, what would become the female final particles were extensively em-
ployed in the domestic novel (Morino 1991:247248), to the extent that the literary
genre of kateeshoosetsu became associated with the excessive use of female final
particles such as teyo and dawa. Writer Uchida Roan (1984[1894]), for example,
scornfully refers to kateeshoosetsu as teyo-dawa novels. Many of these novels had
schoolgirls or young women as main characters and exhibited both explicit and im-
plicit allusion to the virtue of a good wife and wise mother. The goal of the do-
mestic novel is perhaps best explained by Kikuchi Yuho, one of the most successful
male domestic novel writers.

I wanted to write a story which would be a little more secular than the regular novel,
not pretentious, but sophisticated, with good taste. I wanted to write it in such a
way that it could be read in the family circle, that anyone could understand it, and
that no one would blush. I intended to write a novel which would contribute to the
joy of home and help to develop good taste. (Kikuchi 1971[1903]:89)

Kikuchi further emphasized that a good female character in his novel represented
the ideal Japanese woman expected by society (Kikuchi 1971[1903]:90). As the
contemporary literary critic Komori Yoichi (1992) has rightly argued, the domestic
novel was in a way complicit with the state apparatus to nationalize women and to
enable readers to imagine the modern Japanese womanhood.
But at the same time, the stories of the domestic novel were interesting enough for
women to consume. Many featured upper-class womenan unknown yet seductive
imaginary for ordinary people. The stories covered a wide variety of dramatic plots
that involved, for example, romantic love, bearing illegitimate children, extra- and
premarital affairs, elopement, suicide, murder, money and power, deception and
betrayal, mother-daughter relationships, and so on. They provided the detail of the
imagined urban middle-class sociality, dispositions, and material goods. By the early
twentieth century, final particles such as te-yo, da-wa, no-yo, and others, which were
once viewed as vulgar, had come to be increasingly ascribed in the novels to middle-
to-upper-class women. In combination with elaborate honorifics, these final particles
were thus instituted as the voice of those who were depicted as haikara high-collar,
that is, modern, well educated and sophisticated, urban, and of good upbringing.
The fact that the novels were consumed and not simply imposed as some kind
of elite or state-articulated ideology is critical because of the normalizing power of
consumption. This was critical in dissemination. It was not long before the female
final particles were not just reproduced in the novels but also circulated in some young
womens magazines in the form of letters from real Japanese women. These maga-
zines embodied a gender-specific counter public sphere; they constituted a virtual
speech community where virtual friends communicated with one another through
letter writing. Girls wrote to the editors and their magazine friends that they had
never met, using female final particles as if they were speaking to each other, thus
practicing their own write as you speak. Womens language thus came to be no
longer a mere quoted voicea process in which the active voice lies with the one
who quotesbut a quoting voice, as young women claimed their new modern Japa-
 GENDER , LANGUAGE , AND MODERNITY 69

nese identity and constructed their virtual speech community. Letters came from all
over Japan (and its colonies) where the actual dialectal difference might have fatally
fractured any sense of common (gendered) Japanese modernity. But in this virtual
speech community, everyone spoke the speech style of modern Japanese women
and thus claimedand was allowed to performthe subjectivity of modern Japa-
nese woman. In the emerging young womens counter public sphere made possible
by print capitalism (and thus disciplined by both the market and the state), young
women staked claim to a new identity.
We do not know for certain who read these magazines, much less who wrote the
readers letters. Various surveys on reading and readership (Nagamine 1997) show,
however, that the magazines were read not only by daughters of the urban middle-
class family but also later by factory girls (Tsurumi 1990) and young women in peasant
families (Smith & Wiswell 1982, Tamanoi 1998). These readers were not, in fact
speakers of the magazines haikara speech style. The critical point is that the only
place that those readers in Japans periphery heard or perhaps spoke womens
language was in print mediaserialized novels and letters in girls and young
womens magazines. In other words, for them the copy was the original. And the
way they experienced womens language was by consuming it as a metonymy of the
modern, the urban, the nationaleverything that they were not. It was the consump-
tion of womens language that enabled participation in the imagined national
(speech) community. The dissemination of womens language, for these women,
had a lot to do with class and region as these became punctuated within the nation-
alist and capitalist project.

3.4. Conclusion: Toward an effective history

of language, gender, and modernity

I have sought to account for the genealogy of Japanese womens language and to
examine the historical process by which the practice of representing gender mean-
ings through speech was brought into being in the early twentieth century in Japan
during its collective yet uneven experience of modernity. I have also argued that the
social genesis of the metapragmatic category of womens language is not a natural
or evolutionary outcome of indexing gender through womens repeated and sponta-
neous use of certain speech forms. Instead, it is a hazardous effect of modernity in
which, far from being relics of feudal Japan, both gender and language became
problematized as targets of national and capitalist interest and social reform, and both
were significantly reconfigured as various domains of society responded to the project
of modernity and as its attendant social formationscapitalism, nationalism, and
colonialismprofoundly transformed the contour of social relations. Historical be-
ginnings are derisive and ironic, to use Foucaults (1984:79) words. Precisely the
same new modern conception of language that enabled and was enabled by the
development of rational bureaucracy, the universal education system, nationalism,
the military, the print media, colonial education, science and technology, and mimetic
apparatuses such as stenography, the photograph, and the phonograph made possible
the emergence of womens language, the sign of Japanese culture and tradition. The
70 HISTORICAL AND THEORETICAL FOUNDATIONS

representational technique of write as you speak, which was instrumental in creat-

ing the speech forms of womens language, was made necessary by nothing less than
the engagement with the Western realist novel.
The experience of womens language was and has been intensely modern and
national. Indexing gender in early twentieth-century Japan involved imagining the
voice of (the yet to be imagined) modern Japanese women. Indeed, it was in the speech
of schoolgirls that the Meiji writers discovered the linguistic forms with which to
represent the voice of modern women. These forms became elevated to the rank of
Japanese womens language only after they were displaced, grafted, quoted, recycled,
and circulated in the network of newly available representational genres and media. In
this sense there is no original or authentic speaking body that utters womens language.
It is no ones language; indeed, it is disembodied language. What we find is a series of
entextualization processes (Bauman & Briggs 1990) in which the empty signifier of
the modern Japanese woman became (imaginatively) fleshed out and she was given
her voice. The rise of womens language is thus inescapably connected with the
development of consumer culture. In short, the majority of women experienced
womens language not so much as its producer but as its (gendered) consumer.
They heard modern women speaking in the novels and the magazines and saw
them in the cover pictures of beautiful girls, the advertisements for cosmetics and
perfumes, the letters from the readers, and the recipes for exotic Western dishes.
More important, however, the process of constructing womens language, in
its critical articulation with the emergent discourses of nation, race, and class, as well
as with those of culture and tradition, opened up a new discursive space where the
cultural meanings of women are produced and processed and turned into a concrete
object that is knowable and transparent and is readily available for social, not to
mention self-, control. This is precisely the condition under which indexing gender
itself is possible and thinkable and under which statements such as Women speak
more politely than men or ontological questions such as Why do men and women
speak differently? make sense. Paradoxically, however, this new space simulta-
neously allows for strategic appropriation and subversion. The utopian speech com-
munity constructed in the readers column in young womens magazines attests to
this point.
The genealogy of Japanese womens language that I have outlined aims at a
historical narrative that problematizes what is claimed to be natural and obvious in
the present. The stability of Japanese womens language as a discourse derives pre-
cisely from the claim of its ancient origin and of its continuity and linearity up to the
present. A genealogy historicizes the past, and by doing so the present is problematized
(Scott 1996) in a way that makes visible the potential existence of heterogeneous
temporalities, spaces, and experiences. By locating the emergence of what has been
essentialized as womens language in its historical context and by treating it as
one of the historically contingent (yet powerful) events of objectifying the relation-
ship between femaleness and language, we can recognize diverse linguistic experi-
ences as they are situated in and refracted by class and other parameters of social
force. The point is not, however, simply to celebrate subaltern voices and resistance
but rather to recognize the way in which womens linguistic experiences are shaped
by the larger social system of the real world, in which access to womens language
 GENDER , LANGUAGE , AND MODERNITY 71

is just as unevenly distributed as other forms of capitalof both the cultural and the
more familiar kind (Bourdieu 1977, 1991). A focus on indexing forces us to think
historically. It is in this sense that historicizing the practice of indexing as I have
outlined here echoes a larger and growing concern in linguistic anthropology to bring
political economy (social power and its orchestrating and organizing potentialities)
into linguistic analysis (Gal 1989, Irvine 1989). A metapragmatic category such as
womens language is never pregiven but is contingent upon historically specific
social arrangements, in which linguistic forms are motivated and regimented to be-
come an index by being mediated through broader political and economic processes.
My aim in this chapter has also been to demonstrate the social power of the
indexicality of language. Very often, language does not wait until the category it refers
to or indexes is out there. The case of the development of womens language
shows that indexical practice was involved with the construction of modern Japa-
nese women right from its inception. Indexicality constitutes reality not by naming
and pointing to a preexisting object but by inverting the order of the indexed and
indexing to make it appear as if the indexed preceded the indexing. Finally, a focus
on indexing forces us to think critically and counterfactually, because the process by
which a particular speech form is selected or negotiated, out of multiple competing
voices and interpretations, to become an institutionally discrete index of female-
ness is an inescapably political process. The index is inherently unstable and more
a process than a thing, and its reproduction is a perennially political matter of self-
naturalization. It is, however, precisely this processual and productive nature of
indexicality that allows us to see how much any established structure of linguistic
rules in a very rule-governed context is in fact saturated with individual strategizing,
cultural remaking, politics, and historicity.

Notes
This chapter is a shortened and slightly revised version of Inoue (2002). Reproduced by per-
mission of the American Anthropological Association from American Ethnologist 29(2). Not
for sale or further reproduction. I would like to thank Shigeko Okamoto and Janet S. Shibamoto
Smith for encouraging me to participate in this edited volume.

1. NHK (Japan Broadcasting Corporation) and The Agency for Cultural Affairs (Bunkachoo)
in the Ministry of Education, for example, regularly conduct surveys on language consciousness.
2. See Ivys (1995) important analysis of the reflexive projection of an unsullied es-
sence into the past. In the case of womens language, contradictorily enough, any state-
ment about perceived corruption then functions to affirm the ontology of the essence by
implying that there was once a pure womens language.
3. National-language studies (kokugogaku) refers to a domestic scholarly circle for the
study of the Japanese language. It is aligned institutionally and conceptually more or less with
the governments national language policies.
4. See Ide (1994) and Ide and Terada (1998) for a concise introduction to the study of
womens language in National Language Studies.
5. See Koyamas (1991, 1999) excellent discussion on the historical transformation of
the idea of a good wife and wise mother particularly in its linkage with modernity and
modernization. Koyama (1991) shows, for example, that the emphasis on motherhood is
72 HISTORICAL AND THEORETICAL FOUNDATIONS

relatively absent in premodern primers. From a similar critical standpoint, Muta (1996) com-
pellingly argues that the notion of the family (ie) and womens role within it became qualita-
tively discontinuous from that of premodern Japan. Both authors represent the increasing body
of recent critical historical studies that challenge the idea that such institutions as the family
and gender continuously and linearly evolve.
6. On the importance of saving as part of womens domestic duties, see Nolte and
Hastings (1991). For a rich ethnographic account of the ways in which women in isolated
villages in the mid-1930s became nationalized, see Smith and Wiswell (1982).
7. I put quotation marks around the term language to indicate that what I am dealing
with here under the name of language is neither a system nor an object. Language here
can be best understood as an assemblage of various statements, practices, and activities to
produce a variety of knowledge about what counts as language and what does not. In dis-
cussing language modernization, therefore, it does not mean that some kind of structured
object, premodern language, underwent a structural or systematic transformation into a
modern one. Here I follow Sakais approach: Rather I look for various differentiations
and oppositions and their interactions, which, when put together, circumscribe an area in
human activities called language (1992:8). I will dispense with the quotation marks around
language for the remainder of the chapter, but the reader should assume that unless other-
wise indicated, the term language as used in this chapter carries the implication specified
here.
8. See Hanks (1989), Bauman and Briggs (1990), Briggs and Bauman (1992), and
Silverstein and Urban (1996) for further theoretical discussions on the concept of entextualization.
9. This, of course, is the gender of the genre, not that of the author, insofar as the modern
Japanese language interpolates its subject as male. Warners (1990) studies of the develop-
ment of the public sphere in eighteenth-century America presents a similar situation, in which
writing was integrated into a mode of being as white male bourgeois. Exclusion of women,
then, is not simply a matter of access but one of cognitive split; as Warner notes, women
could only write with a certain cognitive dissonance (1990:15).
10. For an insightful discussion on the metalinguistic construction of publicity, as well
as the semiotic explication of Habermass and Andersons notions of community (public sphere
and imagined community, respectively), see Lee (1997).
11. My point here is that it was possible for the author and the reader at that time to
socially imagine that women spoke, ordinarily spoke, or should have spoken in a
particular gendered way. My view of the emergence of womens language is similar to
Foucaults conception of the emergence of disciplinary power: The [actual] automatic func-
tioning of power, mechanical operation, is absolutely not the thesis of Discipline and Punish.
Rather it is the idea, in the eighteenth century, that such a form of power is possible and de-
sirable (Foucault 1980:20)that it could be socially imagined.
12. See Honda (1990) and Kawamura (1993) for recent social and cultural studies of
jogakusee.

References
Anderson, Benedict (1983). Imagined communities: Reflections on the origin and spread of
nationalism. London: Verso.
Bakhtin, Mikhail (1981). The dialogic imagination. Austin: University of Texas Press.
Bauman, Richard, and Charles L. Briggs (1990). Poetics and performance as critical perspec-
tives on language and social life. Annual Review of Anthropology 19: 5988.
Bhabha, Homi K. (1990). Nation and narration. New York: Routledge.
 GENDER , LANGUAGE , AND MODERNITY 73

Bourdieu, Pierre (1977). The economy of linguistic exchanges. Social Science Information
16(6): 645668.
(1991). Language and symbolic power. Cambridge, MA: Harvard University Press.
Briggs, Charles L., and Richard Bauman (1992). Genre, intertextuality, and social power.
Journal of Linguistic Anthropology 2(2): 131172.
Calhoun, Craig (1992). Jurgen Habermas and the public sphere. Cambridge, MA: MIT Press.
Chatterjee, Partha (1990). The nationalist resolution of the womens question. In K. Sangari
and S. Vaid (eds.), Recasting women in India: Essays in colonial history, 233253. New
Brunswick, NJ: Rutgers University Press.
Foucault, Michel (1977). Discipline and punish: The birth of the prison. New York: Vintage
Books.
(1980). Power/knowledge. New York: Pantheon Books.
(1982). The subject and power. In H. L. Dreyfus and P. Rabinow (eds.), Michael
Foucault, beyond structuralism and hermeneutics, 208226. Chicago: University of
Chicago Press.
(1984). Nietzsche, genealogy, history. In P. Rabinow (ed.), The Foucault reader, 76
100. New York: Pantheon Books.
Fraser, Nancy (1990). Rethinking the public sphere: A contribution to the critique of actually
existing democracy. Social Test 8-9(3-1): 5680.
Gal, Susan (1989). Language and political economy. Annual Review of Anthropology 18: 345
367.
Habermas, Jurgen (1989). The structural transformation of the public sphere: An inquiry into
a category of bourgeois society. Cambridge, MA: MIT Press.
Hanks, William F. (1989). Text and textuality. Annual Review of Anthropology 18: 95127.
(1996). Language & communicative practices. Boulder, CO: Westview Press.
Honda, Masuko (1990). Jogakusee no keefu (The geneology of jogakusee). Tokyo: Seidosha.
Ide, Risako, and Tomomi, Terada (1998). The historical origins of Japanese womens speech:
From the secluded worlds of court ladies and play ladies. International Journal of
Sociology of Language 129: 139156.
Ide, Sachiko (1994). Womens language in womens world. Paper presented at the Third
Berkeley Women and Language Conference, University of California, Berkeley.
Inoue, Miyako (2002). Gender, language, and modernity: Toward an effective history of Japa-
nese womens language. American Ethnologist 29(2): 392422.
Irvine, Judith T. (1989). When talk isnt cheap: Language and political economy. American
Ethnologist 16(2): 248267.
Ishikawa, Sadayuki (1972). Kindaigo no teyo, dawa, noyo (Teyo, dawa, noyo in modern
Japanese). Kaishaku 18(10): 2227.
Ivy, Marilyn (1995). Discourses of the vanishing: Modernity, phantasm, Japan. Chicago:
University of Chicago Press.
Kawamura, Kunimitsu (1993). Otome no inori: Kindai josee imeeji no tanjoo (The maidens
prayer: The birth of the image of modern women). Tokyo: Kinokuniya Shoten.
Kikuchi, Yuho (1971[1903]). Chikyodai. In S. Senuma (ed.), Meiji katei shoosetsu shuu (Meiji
family novel collection), vol. 93, 89240. Tokyo: Chikuma Shobo.
Kikuzawa, Sueo (1929). Fujin no kotobano tokuchoo ni tsuite (On the characteristics of
womens language). Kokugo Kyooiku 14(3): 6675.
Kindaichi, Kyosuke (1942). Zooho kokugo kenkyuu (A study of the national language, addi-
tional supplement). Tokyo: Yakumoshorin.
Komatsu, Hisao (1988). Tokyogo ni okeru danjosa no keesee: Shuujoshi o chuushin to shite
(The form of gender differences in Tokyo dialect: Centering on the sentence-final par-
ticles). Kokugo to Kokubungaku 65: 94106.
74 HISTORICAL AND THEORETICAL FOUNDATIONS

Komori, Yoichi (1988). Kozoo to shiteno katari (Narrative as a structure). Tokyo: Shinyosha.
(1992). Buntai to aidentitii (Style and identity). Gekkan Gengo 21(10): 4855.
Koyama, Shizuko (1991). Ryoosai kenbo to yuu kihan (The norm of a good wife and wise
mother). Tokyo: Keiso Shobo.
(1999). Katee no seesee to josee no kokuminka (The creation of home and the nation-
alization of women). Tokyo: Keiso Shobo.
Lee, Benjamin (1997). Talking heads. Durham: Duke University Press.
Mani, Lata (1987). Contentious traditions: The debate on sati in colonial India. Cultural
Critique 119158.
Mashimo, Saburo (1969). Fujingo no kenkyuu (The study of womens language). Tokyo:
Tokyodo Shuppan.
Morino, Muneaki (1991). Joseigo no rekishi (The history of womens language). In H. Miyaji
(ed.), Nihongo no rekishi: Kooza nihongo to nihongo kyooiku (The history of Japanese:
Series on the Japanese language and Japanese language education), vol. 10, 225249.
Tokyo: Meiji Shoin.
Muta, Kazue (1996). Senryaku to shite no kazoku: Kindai nihon no kokumin kokka keesee to
josee (Family as a strategy: The modern Japanese nation-state and women). Tokyo:
Shinyosha.
Nagamine, Shigetoshi (1997). Zasshi to dokusha no kindai (The modernity of magazines and
their readers). Tokyo: Nihon editaa sukuuru shuppanbu.
Nolte, Sharon H., and Sally Ann Hastings (1991). The Meiji States policy toward women,
18901910. In G. L. Gernstein (ed.), Recreating Japanese women, 16001945, 151
174. Berkeley: University of California Press.
Okano, Kyuin (1902). Hyoojungo ni tsuite (On standard language). Gengogaku zasshi 3(2):
3240.
Ozaki, Koyo (1994[1888]). Hayarikotoba (Trendy language). In M. Oka (ed.), Koyo zenshu
(The complete collected works of Koyo), vol. 10, 45. Tokyo: Iwanami Shoten.
Sakai, Naoki (1992). Voices of the past: The status of language in eighteenth-century Japa-
nese discourse. Ithaca, NY: Cornell University Press.
Sato, Haruo (1998[1941]). Shoosetsu (The novel). In T. Sakuragi (ed.), Kokugo bunka kooza,
vol. 4, 7198. Tokyo: Asahi Shimbunsha.
Schieffelin, Bambi B., Kathryn Ann Woolard, and Paul V. Kroskrity (1998). Language ide-
ologies: Practice and theory. New York: Oxford University Press.
Scott, Joan Wallach (1996). After history? Common Knowledge 5(3): 926.
Silverstein, Michael (1979). Language structure and linguistic ideology. In P. R. Clyne,
W. Hanks, and C. L. Hofbauer (eds.), The elements: A parasession on linguistic units
and levels, 193247. Chicago: Chicago Linguistic Society.
(1996). Indexical order and the dialectics of sociolinguistic life. In R. Ide, R. Parker,
and Y. Sunaoshit (eds.), SALSA: Proceedings of the Third annual Symposium about
Language and SocietyAustin, vol. 36, Texas Linguistic Forum, 266295. Austin:
University of Texas, Department of Linguistics.
Silverstein, Michael, and Greb Urban (eds.) (1996). Natural histories of discourse. Chicago:
University of Chicago Press.
Smith, Robert John, and Ella Lury Wiswell (1982). The women of Suye Mura. Chicago:
University of Chicago Press.
Stoler, Ann Laura (1991). Carnal knowledge and imperial power: Gender, race, and morality in
colonial Asia. In M. di Leonardo (ed.), Gender at the crossroads of knowledge: Feminist
anthropology in the postmodern era, 51101. Berkeley: University of California Press.
Tamanoi, Mariko (1998). Under the shadow of nationalism: Politics and poetics of rural
Japanese women. Honolulu: University of Hawaii Press.
 GENDER , LANGUAGE , AND MODERNITY 75

Tsubouchi, Shoyo (1930). Kaki no heta (The navel of persimmons). Tokyo: Chuokoron.
Tsurumi, E. Patricia (1990). Factory girls: Women in the thread mills of Meiji Japan. Princeton,
NJ: Princeton University Press.
Uchida, Roan (1984[1894]). Bungakusha to naru hoohoo (How to become a literary scholar).
In T. Nomura (ed.), Uchida Roan zenshuu, shohan (The complete collected works of
Uchida Roan), vol. 17, 175299. Tokyo: Yumani Shobo.
Voloshinov, V. N. (1973). Marxism and the philosophy of language. New York: Seminar
Press.
Warner, Michael (1990). The letter of the republic: Publication and the public sphere in
eighteenth-century America. Cambridge, MA: Harvard University Press.
Woolard, Kathryn A., and Bambi B. Schieffelin (1994). Language ideology. Annual Review
of Anthropology 23: 5582.
76 HISTORICAL AND THEORETICAL FOUNDATIONS

RUMI WASHI

Japanese Female Speech

and Language Policy in the
World War II Era

L anguage evolves naturally over the course of human communication. However,

certain aspects of language are molded artificially. France and Germany provide clas-
sic examples of the artificial modification of language: each, in the course of build-
ing a nation-state, established its own national tongue purified through the elimination
of dialectal elements and foreign loanwords. A similar process began in Japan at the
end of the nineteenth century, when developments in Germany and the rest of Europe
were followed with interest. The state, specifically the Ministry of Education, launched
a series of surveys to lay the groundwork for the establishment of a standardized
national language.
The form of modern Japanese employed principally by women is in some
respects an artificial construct. One argument is that female education in the pre
World War II years produced and ingrained gender differences. Sugimoto asserts that
from the 1890s to 1912 in the Meiji period every effort went into making women
more feminine, to a greater extent even than in the Edo Period (16031867), al-
though women gained some social rights in the Meiji Period. The acquisition of a
single predetermined feminine form of speech was relentlessly pursued as part of
schooling(1997:35).1 He seems to be alluding to the instructional content of such
school subjects as Shuushin Morals and Sahoo Manners.2 This argument treats
instruction in female speech as part and parcel of female education, the goal of
which was restricted to providing training to women in their role as wives and mothers
on the basis of the ryoosai kembo shugi ideology of a good wife and wise mother.3
Another view holds that the spread of female speech was accelerated by the teach-
ing of Standard Japanese. In an interview with Usami Mayumi, Endo Orie remarks

76
 JAPANESE FEMALE SPEECH AND LANGUAGE POLICY IN THE WORLD WAR II ERA 77

that the development of female speech may be connected with the trend toward stan-
dardization of the language in the Showa era (19261989) (1997:7071). The con-
tention here is that, during the war-racked Showa period, beginning in the 1920s,
so-called national-language classes started to provide more direct instruction in Stan-
dard Japanese, including female-oriented forms of speech. Both views assert that
language policy played a role in the development of female speech through edu-
cation. Or, to change perspective, girls were trained by the state to use female
speech, that is, the norms of the Japanese language for women.
This chapter explores the relationship between the establishment of norms of
language for women and language policy by considering how those norms were fos-
tered through education. Specifically, it traces the formulation and propagation of
Reehoo yookoo Essentials of etiquette, a set of guidelines on instruction in good
manners compiled by the Ministry of Education in 1941, and examines the collabo-
rative relationship that emerged between female educators and the Kokugo Kyookai
National Language Association (NLA), which campaigned for language policy
during the years in question. My investigation covers the period from the 1920s to
1945, years during which the attempt to standardize the language for women was
especially intense. Furthermore, during this period the Emperor Showa was inaugu-
rated and the Japanese government embarked on a total war for the first time and
lost. As we shall see, World War IIs character as an all-out conflict was one of the
primary factors that motivated the establishment of a new curriculum in etiquette and
prompted the emergence of a cooperative relationship between the NLA and female
educators.
In this chapter, the terms Japanese and national language both denote what
is now referred to as kyootsuugo common language, which is based on a Tokyo
dialect4 spoken by middle-class people.5 I have excluded channels used to educate
women whose native tongue was not Japanese, such as women in the Korean penin-
sula, then a Japanese colony.

4.1. Language policy

4.1.1. Definition
Language planning is defined as an attempt by a group or individual to solve a prob-
lem relating to the use of language with a specific objective in mind. The entity doing
the planning need not necessarily be the state (Kaplan & Baldauf 1997:34). Lan-
guage planning typically assumes concrete shape in several stages: survey work, a
survey report, establishment of policy, formulation of a plan of action, and imple-
mentation (Kaplan & Baldauf 1997:106). However, in order for a problem to be re-
garded as such, some type of ideal model must already exist in the mind of the
planners. Therefore, it would be wise to assume that a provisional policy is postu-
lated before the survey work even begins.
In Japan, the state first embarked on language planning at the beginning of the
twentieth century. In 1903, the Kokugo Choosa Iinkai National Language Research
Council was established under the Meiji government. The following definition of
language policy appears in the Kokugogaku jiten Dictionary of Japanese language
78 HISTORICAL AND THEORETICAL FOUNDATIONS

studies (Kokugo Gakkai 1955), the first compilation of the cumulative results of
research by the Kokugo Gakkai Society for the Study of Japanese Language to be
offered to the public after World War II. The writer is Ishiguro Yoshimi, who was
also a member of the NLA.

[Language planning is a] method whereby a government or other public agency

attempts to reform, rationalize and improve, or propagate language with a particu-
lar intention . . . In formulating language policy, an organization may be established
in order to carry out research, engage in deliberations, and oversee implementa-
tion. . . . The policies established are implemented in the form of legislation or
through education, newspapers, radio, and so forth. (Ishiguro 1955:331332)

Language planning strategies at the national level were not essentially different from
those of today.

4.1.2. Language policy for women

As Ishiguro (1955) observes, the language policy that the state disseminates through
education and the media is generally the end result of state-sponsored surveys and
policy deliberations. However, in the case of female speech, no single public in-
stitution dedicated itself to the task of establishing and propagating norms. The Min-
istry of Education was directly involved in the Kokugo Shingi-kai Council for
National Language, which played a role in language policy and in education. Dur-
ing the period in question, linguistic gender differences and female speech fell
beyond its purview. However, the 1930s witnessed mounting criticism of the form
of language employed by women in such media sources as radio, newspapers, and
magazines. The issue of what form of speech women should adopt came up repeat-
edly in debates among national-language scholars (e.g., Yoshida 1935), linguists (e.g.,
Shinmura 1938), and commentators (e.g., Hasegawa 1941). A consensus about the
ideals and standards of national language for women emerged as the result of their
criticisms of the language employed by women.6 There were, therefore, multiple
channels that prescribed ideals and standards for what form female speech should
take, that selected specific grammar and vocabulary for it to employ, and that pro-
vided education about female speech usage, and these channels functioned together
as a composite. The consensus was reflected in the classroom, and a propaganda
campaign was launched to foster the spread of the new norms. The state, for its part,
in taking over certain elements of this debate, incorporated standards of female
speech into documents distributed to members of the public. One of these was Reehoo
yookoo.

4.2. Reehoo yookoo Essentials of etiquette

In April 1941, the Ministry of Education released Reehoo yookoo, which was origi-
nally a new textbook for secondary schools (boys middle schools, girls high schools,
and technical schools).7 This manual defined standards of reegi-sahoo etiquette and
 JAPANESE FEMALE SPEECH AND LANGUAGE POLICY IN THE WORLD WAR II ERA 79

manners for people of all classes as well as for those who were attending secondary
schools. In section 5 on Kotobazukai language use, it explicitly prescribed gendered
forms of standardized speech.

2. For the first person, watakushi I should normally be used. In addressing a su-
perior one may on occasion use ones surname or given name. Men may use boku
I when addressing a social equal, but not when addressing a superior.
3. For the second person, when addressing a superior one should use an honorific
appropriate to rank. When addressing an equal one should normally use anata you.
Men may also use kimi you. (Ministry of Education 1941:6)

Since this text was written not only for boys middle schools or girls high schools
but also for the education of the nation, the exhortation that watakushi I should
normally be used applies to both women and men. These stipulations on first- and
second-person usage came in the wake of criticisms in the media and by the Kokumin
Seeshin Soodooin Chuuoo Renmee Central Federation for the General Mobiliza-
tion of the National Spirit that female students in Tokyo were using the terms kimi
you and boku I, which were considered proper only for men. (See the article Onna
wa onna-rashikukeego o wasureru na Ladies should be ladylikedont forget your
honorifics in the Tokyo Asahi Shimbun August 8, 1938.)8 The next month saw
publication of Reehoo yookoo kaisetsu Commentary on the essentials of etiquette
(Reehoo Kenkyuu-kai 1941). This commentary prescribed in detail the speech style
appropriate for women. As a matter of principle, it stated, men should generally use
masculine language, while women should use feminine language (Reehoo Kenkyuu-
kai 1941:65). In other words, both this commentary and Kokumin Seeshin Soodooin
Chuuoo Renmee claimed that women should use more polite and humbler expres-
sions than men. The state was not attempting merely to control language usage. The
governments aim, it is clear, was to reinforce hierarchical social relations and gen-
der roles by regulating language. That is plain from a proposalto be discussed
latermade at the Kyooiku Shingikai Education Council by Tokugawa Yoshichika,
who oversaw the compilation of Reehoo yookoo, and from the states overall policy
of strengthening hierarchical social relations under the emperor in preparation for
total war.
The concept of Reehoo yookoo became embodied when a proposal for the spread
of etiquette was made in the Education Council by Tokugawa, a marquess and a
member of the House of Peers, who had also chaired the Etiquette Research Asso-
ciation. The Education Council was formed in 1937 by the cabinet of Prime Minister
Konoe Fumimaro for the purpose of reorganization of educational policy, as he was
trying to establish a general national mobilization in order to win the war. Tokugawas
speech to the third general meeting of the Education Council held in January 1938
asserted that reform in education had to be reform in the inculcation of the national
spirit. However, only the general mobilization of the nations spirit was appealed to,
and no concrete plan was made. One concrete way of putting this appeal into prac-
tice was to propagate etiquette throughout Japan. In order to correct the spirit, it was
held to be important to propagate certain aspects of etiquette, such as being dutiful
to parents or being faithful to masters, that can be learned in the family (Ishikawa
80 HISTORICAL AND THEORETICAL FOUNDATIONS

1957:545549). Tokugawa considered etiquette the expression of spirit, made it a

goal to have the nation behave as instructed, and proposed control over the Japanese
nation by establishing a comprehensive social hierarchy and clear sex role distinc-
tions. The Ministry of Education carried out this plan.
Propaganda for the spread of Reehoo was widely disseminated. According to
Shoseki nenkan Shoowa 17 nen ban Books almanac, 1942 edition, during a single
year1941six publishers produced Reehoo yookoo as well as 13 types of explana-
tory commentaries (Kyodo 1942). Womens magazines gave extra space to special
feature articles on this manual, and interpretive articles appeared serially in the news-
papers. Tokugawa and other members of the committee spoke on nationwide radio
broadcasts.
Nonetheless, it should be noted that in reality a small percentage of the female
population of Japanese were directly taught female speech on the basis of Reehoo
yookoo, because in the 1940s people who could speak Standard Japanese were in the
minority (Kokuritsu Kokugo Kenkyuujo 1951). Also, the instruction in female
speech that was offered at educational institutions at the secondary level and above
on the basis of this textbook was directed exclusively at girls of elite background,
because in 1941 the number of students enrolled in secondary educational institu-
tions, both young women and young men, was equivalent only to about 20% of the
number of pupils who were attending primary school.9
Reehoo yookoo, however, appears to have exercised considerable influence on
the teenagers of the daythe generation that formed the backbone of postwar Japa-
nese society. This is a fact that previous research (e.g., Sugimoto 1997) has almost
completely overlooked.
I now turn to a detailed investigation of one official (or semiofficial) body and
its role in mediating between state policy and women and in gaining the support of
prominent women, including educators who taught Reehoo yookoo, for the imple-
mentation of a state language policy.

4.3. The NLA and female educators

4.3.1. The Kokugo Kyookai National
Language Association
The NLA was founded in 1930 with the goal of supporting the language policies of
the Ministry of Education. According to article 2 of the Regulations of the National
Language Association, the NLAs objective was to strive for the rationalization
and improvement of the national language and lovingly preserve it (Kokugo Kyookai
1937:back cover). The NLAs activities are described thus in article 3: In pursuit of
this objective, [the NLA] will conduct research, organize various types of meetings,
publish magazines and books, and undertake other suitable endeavors (Kokugo
Kyookai 1937:back cover). The NLA was thus devoted expressly to implementing
language policy. Articles 4 and 17 stipulated that the associations expenses were to
be defrayed from membership fees, donations, business revenue, and so forth. Super-
ficially, then, it had the trappings of an economically independent private body. How-
ever, in a Ministry of Education directive dated November 7, 1945, immediately after
 JAPANESE FEMALE SPEECH AND LANGUAGE POLICY IN THE WORLD WAR II ERA 81

the war, the NLA is described as an external organ of the Ministry of Education
(Ministry of Education 1980). Its president was Konoe Fumimaro, who became prime
minister in the year of its establishment. The vice president was Minami Hiroshi, then
head of the Council for National Language. One of the members of the board, Hoshina
Koichi, likewise belonged to the Council for National Language. In reality, there-
fore, the NLA would more accurately be described as an extension of government
that lay halfway between the Ministry of Education and the public and put language
policy into practice while assimilating the sentiments of both the Ministry of Educa-
tion and the public.
What vantage point did the NLA membership adopt in carrying out its programs?
One leading figure in the association, Shimose Kentaro, an army physician with the
rank of major general, wrote on the necessity of improving the national language
in the NLAs monthly bulletin. Noting the abstruseness of the diction used in an-
nouncements by the military authorities and in speeches for the general mobiliza-
tion of the national spirit, Shimose declared the need for language that anybody could
understand. As grounds for this contention, he cited the fact that the vast majority
of the [Japanese] people85%barely have a primary-school education, and so their
reading abilities are quite rudimentary (1937:1). The NLAs members believed that
they, as intellectuals, had a mission to guide the semi-illiterate, poorly educated masses
in what they themselves regarded as the right direction.

4.3.2. Establishment of the womens wing

In March 1939, the NLA commenced preparations for establishing a womens wing.
Table 4.1 summarizes developments between the NLAs foundation and the inaugu-
ral ceremony of the womens wing, including changes in the course of the womens
movement and government approaches to female educators and the movements
leaders.
The womens wing was established not in response to the wishes of female
members but rather at the encouragement of the NLAs leaders. During the March
meeting at which the preliminary arrangements were made, Minami, the associations
vice president, remarked that he had extended the invitation in order to request the
understanding and obtain the cooperation of women in preserving and honoring the
national language and with regard to the NLAs work (Kokugo Kyookai 1939:49).
Hoshina, who was on the board, spoke on the subject of women and language and
emphasized the important role that women had to play in this area.
The launch of a womens wing by the NLA was of considerable significance. It
suggests that the association had fallen into step with other ministries such as Finance
and Welfare, which were increasingly encouraging women to participate in the
nations policies as the whole country mobilized for war. The fact that the organiza-
tion of the wing and the recruitment of leading female members occurred in parallel
leads to the same conclusion. The women who joined the association or were appointed
coordinators were prominent figures who held official positions at the national level as
committee members with the Central Federation for the General Mobilization of the
National Spirit, for example, or as Ministry of Finance instructors. Ichikawa was the
leader of the movement for womens suffrage. Ishimoto admired Margaret Sanger
82 HISTORICAL AND THEORETICAL FOUNDATIONS

4.1 Evolution of the NLA

1937
June The NLA is founded. General meeting.
July 7 War breaks out between Japan and China.
Sept. 28 The League of Japanese Womens Organizations is established through an amalgam-
ation of leading womens groups, including Gauntlett Tsunes Japan Womans
Christian Temperance Union and Ichikawa Fusaes Fusen Kakutoku Doomee Alliance
for Womens Suffrage.
Oct. The Kokumin Seeshin Soodooin Chuuoo Renmee Central Federation for the General
Mobilization of the National Spirit is launched. (Yoshioka Yayoi is appointed to the
Central Federation council.)

1938
Aug. 5 Minister of Education Araki attends a roundtable discussion with the Central
Federation for the General Mobilization of the National Spirit, during which he
criticizes the speech of female students in Tokyo for not being ladylike. He arouses a
considerable response.

1939
March 2 Ichikawa Fusae, Gauntlett Tsune, Muraoka Hanako, and others are appointed Ministry
of Finance Women Instructors for the Encouragement of Savings.
March 14 A meeting is held to make arrangements for establishing the womens wing of the
NLA. (In attendance: eleven women, including Yoshioka Yayoi, Inoue Hide, and
Kiuchi Kyo, and seven men, including Minami, the vice president; Hoshina Koichi;
and Ishiguro Yoshimi.)
March Dan Michiko, Ishimoto Shizue, Ichikawa Fusae, and several other women join the
association.
July 14 Inaugural ceremony of the womens wing of the NLA. Five coordinators are appointed
for the womens wing, among them Inoue Hide and Otsuma Kotaka.
July Inoue Hide, Kiuchi Kyo, and other elite female leaders join the association.

and promoted the movement for birth control. The NLA recruited an especially large
number of female educators. Yoshioka was the founder of Tokyo Womens Medical
University. Inoue was headmistress of Japans Womens University, Kiuchi of an
elementary school. Otsuma established Otsuma Womens University. Authors such
as Muraoka also joined up, as did the singer Dan. As Kiuchi endeavored to improve
the status of female teachers, many of the women who joined in the womens divi-
sion of NLA were also involved in the movement for womens liberation and the
equality of women and men. However, the NLA did not simply follow the lead of
the government ministries. Aware of the role that mothers played in language edu-
cation, it chose as members educators and authors capable of both teaching other
women and serving as models for them. Female campaigners for the rights of work-
ing women like Yamakawa Kikue and Oku Mumeo did not join the organization even
though they, too, advocated improving the status of women. The association, it can
 JAPANESE FEMALE SPEECH AND LANGUAGE POLICY IN THE WORLD WAR II ERA 83

be conjectured, was not primarily concerned about members of the working class
content to work outside the home; rather, through exemplars of the same sex it sought
to drive home language policy among middle- and upper-class women who were
dedicated to becoming mothers and aspired to the ideals both of onnarashisa femi-
ninity and ryoosai kembo good wife and wise mother.
The womens wing did not prove very active, perhaps because of the escalation
of the war. It did, however, carry out studies of household language. For example,
at a meeting that was attended by the national-language expert Yoshida Sumio, au-
thor of Fujin no kotoba Ladies language (Yoshida 1935), Miyake Takuro, and
Yoshioka Yayoi (Kokugo Kyookai 1941: 27), the discussants debated what term was
appropriate for a husband to use when addressing his wife.

4.3.3. The ideal of female speech

The NLA, which was not overly enthusiastic about the question of female speech,
nevertheless increased its exposure in the mass media after establishing the womens
wing, targeting educated middle-class women and propagating the ideal of female
speech. Before moving on to the question of the NLAs propagatoin of female
speech, however, let us examine the ideal of female speech itself.
The studies of female speech that started to appear in the late 1920s directed
their attention primarily to the yuujogo courtesans speech of the Edo period (1603
1867) and the nyooboo kotoba court ladies language of the Muromachi period
(fourteenth to sixteenth centuries). Yuujogo courtesans speech is the unique lan-
guage spoken by yuujo in red-light districts where the houses kept young women
who entertained men with singing, dancing, serving drinks, and providing sex.
Yuujogo was adopted in Shimabara, the red-light district in Kyoto, and then spread
to the Yoshiwara, its Edo (Tokyo) counterpart. Many yuujo were from poor families
in towns or villages other than Kyoto or Edo. The distinctive characteristics of yuujogo
were its unique first- and third-person pronouns and verb endings. In example (1), a
represents Standard Japanese and b yuujogo.

(1) a. Watashi no kimono wa soko ni arimasu.

b. Watchi no kimono wa soko ni arinsu.
I GEN kimono TOP there LOC is
My kimono is there.

Both of the preceding examples mean my kimono is there, but in yuujogo the first-
person pronoun is watchi instead of the more usual watashi. The end of the verb arimasu
is is changed to n from ma. The form arinsu is is considered typical of yuujogo.
Nyooboo kotoba first appeared in fourteenth-century records.10 It began to be
used by court ladies, the daughters of noblemen who attended the emperor. The du-
ties of court ladies included looking after the emperor and sometimes doing clerical
work for court events. Some court ladies even gave birth to babies of emperors to
provide successors. The main characteristics of nyooboo kotoba were the frequent
use of honorific prefix o- with nouns and various neologistic formations that replaced
nouns for many domestic items. In examples (24), b represents nyooboo kotoba.
84 HISTORICAL AND THEORETICAL FOUNDATIONS

(2) Abbreviation
a. matsutake matsutake mushroom
b. matsu

(3) Adding -moji to the end of an abbreviation

a. tako octopus or kite
b. ta-moji

(4) Adding -mono thing to descriptive words for things.

a. tofu bean curd
b. shiro-mono white thing (i.e., bean curd)

Some nyooboo kotoba such as oishii delicious and oden Japanese stew have be-
come common words in modern Japanese.
Kikuzawa Sueo (1929) was the first scholar to consider female speech an object
of national language research. He regarded nyooboo kotoba as the ideal of female
speech, describing it as courteous and feminine. This was an ideology (i.e., a body
of knowledge turned to a particular political purpose) of female speech rooted in
language policy. In subsequent research, he went on to compare nyooboo kotoba with
yuujogo, not only analyzing their grammar and vocabulary but also passing judg-
ment on them. The former, he declared, exuded genuine feminine grace, while the
latter possessed a vulgar beauty (Kikuzawa 1933:39). For example, tamoji, the
nyooboo kotoba equivalent of tako octopus, and arinsu, the yuujogo version of
arimasu to be, elicited quite contrary appraisals, the grounds of which cannot be
traced to the words themselves. His assessment merely reflects the stereotypes at-
tached to the court ladies and courtesans who used the terms. Yet Kikuzawas analy-
sis was to be reiterated later by other national-language scholars, such as Yoshida
(1935), who advanced the theory that the original characteristic of female speech
was its elegance and that elegant female speech had originated in the language of
court ladies. He emphasized that the language peculiar to women was created in the
court and spread to the lower classes. His argument was that female speech was a
symbol of both femininity and upper-class status, so that women in the lower classes
should learn female speech, too. In the context of Japans emperor system, a dis-
course that held the language of court ladies to be the origin of female speech had
the potential to serve as a tool in integrating women into the national culture and
thus making them more likely to submit to the ultranationalistic program in wartime
(Washi 2000).

4.3.4. Propaganda for the ideal of female speech

by the NLA members in mass media
The subject of female speech was taken up by various commentators in newspa-
pers and magazines in the wake of the 1938 roundtable discussion between Educa-
tion Minister Araki and the Central Federation for the General Mobilization of the
National Spirit (see table 4.1). The majority of national-language scholars who dealt
with the issue were the NLA members. As shown in table 4.2, many of their views
 4.2 Publications and commentaries on female speech by NLA members
1929
Jan. Kikuzawa Sueo: Radio broadcast Fujin kooza: Fujin no kotoba no tokuchoo ni tsuite
Course for ladies: On the characteristics of ladies speech.
March Kikuzawa Sueo. Fujin no kotoba no tokuchoo ni tsuite On the characteristics of ladies
speech. Kokugo kyooiku National-language education.

1935
March *Yoshida Sumio: Radio broadcast Kotoba no kooza: Fujin no kotoba Language
course: Ladies speech.
May *Yoshida Sumio. Fujin no kotoba Ladies speech. Kotoba no kooza 2 Course
in language 2. Japan Broadcast Publishing.
Dec. *Hoshina Koichi: Radio broadcast Gendaigo kooza: Fujin no kotoba to kodomo no
kotoba Modern language course: Ladies speech and childrens speech.

1936 *Hoshina Koichi. Fujin no kotoba to kodomo no kotoba Ladies speech and childrens
speech. In Kokugo to nihon seeshin The national language and the Japanese spirit.
Jitsugyoo no Nihon Sha.

1938
Nov. Kokugo kokuji mondai no kenkyuu Studies on the question of the national language
and national script, a series of occasional articles in Fujo shimbun Ladies news.

1939
Jan. *Takakura Teru. Josee to kokugo no mondai Women and the national-language
question. Fujo shimbun.
July The NLA launches its womens wing.

1940
Oct. *Ishiguro Yoshimi begins a fifty-installment series titled Kokugo no wadai Topics in
national language in Fujo shimbun.
Dec. *Ishiguro Yoshimi. Kokugo no wadai 8: Josee no kotoba/10: Nyooboo kotoba Topics in
national language 8: Womens speech/10: Court ladies language. Fujo shimbun.

1941
Jan. *Ishiguro Yoshimi. Kokugo no wadai 13: Fujin to kotoba Topics in national language
13: Women and language. Fujo shimbun.
April Ministry of Education. Reehoo yookoo.
Dec. *Yanagi Yae. Fujin no kotoba Ladies speech. In Kokugo bunka kooza 5: Kokugo
seekatsu hen Course in national language culture 5: National language in life. Asahi
Shimbun-sha.

1943
July *Ishiguro Yoshimi. Utukushii nihongo Beautiful Japanese. Kofukan. Jogakusee no
kotoba-zukai Language use of female students. Shoojo no tomo Girls companion.
Aug. *Ishiguro Yoshimi. Jogakusee no kotoba-zukai. Shoojo no tomo.
Sept. *Ishiguro Yoshimi. Jogakusee no kotoba-zukai. Shoojo no tomo. *Dan Michiko. Josee
to kotoba Women and language. Sakuragi Shobo.
Nov. *Inoue Kiyoshi. Josee no kotoba: Keego no tsukaikata Womens language: How to use
honorifics. Bokutaku-sha.
Asterisks indicate NLA members.
86 HISTORICAL AND THEORETICAL FOUNDATIONS

appeared in the weekly Fujo Shimbun Ladies News, a publication for educated
middle-class female readers. Fujo Shimbun for its part spotlighted the so-called ques-
tion of national language and national script in the belief that this issue was of deep
concern to women.
While Ishiguro was especially prolific, Hoshina, the announcer Yanagi, and
Yoshida likewise produced discourses that enshrined nyooboo kotoba as the ideal of
female speech. Ishiguros serialized column in Fujo Shimbun makes no mention
of yuujogo, nor does Yoshida. Repeatedly the point was driven home that the origins
of Japanese female speech lay in nyooboo kotoba. Although from about 1943 on,
as paper shortages grew more serious, the publication of books required governmental
approval (Kyodo Shuppan 1942), Dan and K. Inoue were still able to bring to press
their work that promoted female speech. This favorable treatment can be ascribed
to the fact that, being NLA insiders, they wrote in accord with the states policy of
controlling women, as members of the Japanese nation, through language.

4.3.5. Female educators and female speech

Finally, I consider the implications of the establishment of a collaborative relation-
ship between womens leaders and the NLAs womens wing, a vehicle of official
language policy. Why were female educators so prompt to respond to the rallying
cry of this organization dedicated to honoring and improving the national language?
Why did they adopt a cooperative stance by conducting studies of household lan-
guage and recruiting new members? One reason is that their membership in the
association was another instance of yokusanka collaboration with and absorption
into the regime (see Suzuki 1997), just like their decision to help the national war
effort by merging the various existing womens movements into one or the assump-
tion of official posts by female educators.
Another reason is that the higher educational institutions for girls run by NLA
members pursued the same policy of instructing women to use feminine language. It
was natural, then, that womens leaders cooperated with the vehicle of language
policy. If anything, they appear to have regarded language as their own proper sphere
of competence. Female speech was sometimes discussed in terms of the Edo pe-
riod (16031867) concept of fugen ladies speech, the ideal language to which
women were supposed to aspire. The following are the words of Shimoda Jiro (1872
1938), who was active in female education in the Meiji period. Women should dis-
play femininity in their speech and demeanor. They should be gentle and modest,
not blunt like men. These constitute the qualities of fugen feminine speech and fuyoo,
[two] of the shigyoo four ways (Shimoda 1977[1904]:359). Shigyoo refers to the
four doctrines of Confucianism. Fuyoo means the courteous behavior proper to
women. Both fuyoo and fugen were used by Kaibara Ekken, a Confucian scholar of
the seventeenth century. Kaibara employed fugen in the sense of words of obedi-
ence to men. However, as the meaning of words changes with their context, the fugen
that Shimoda used in the passage quoted earlier has a different significance from the
fugen that Kaibara employed. Shimoda explains as follows: The duties of a woman
are, first, to observe the code of morals incumbent on her sex; second, to be a good
wife; and third, to be a wise mother. Fourth, she should undertake suitably feminine
 JAPANESE FEMALE SPEECH AND LANGUAGE POLICY IN THE WORLD WAR II ERA 87

employment: People today, whether male or female, should not practice blind obe-
dience (Shimoda 1977[1904]: 358359). Ryoosai kembo shugi the ideology of a
good wife and wise mother was more than a mere vestige of feudalism (Koyama
1991; see note 3). It was a concept that aimed to achieve what might be termed sepa-
rate but equal status for womena concept that assimilated Western ideas of fam-
ily (see Muta 1992). Female educators such as Yoshioka, Otsuma, and H. Inoue, the
founders of todays womens universities, emerged from educated middle-class circles
brought up in this tradition of ryoosai kembo shugi. Most of them collaborated with
the war effort, and most were readers of and contributors to Fujo Shimbun, which
molded educated middle-class female opinion.
But the ideal of female speech that these female educators upheld differed
slightly from that propounded by national language scholars, the viewpoints of Reehoo
yookoo and of the Central Federation for the General Mobilization of the National
Spirit. Otsuma, for whom use of language was a prime concern, wrote as follows.
(The term beranmee kotoba that appears in the quotation refers to the lower-class
speech of the Tokyo dialect, which Otsuma viewed as lacking in refinement.)

Say there is a lady. Say she possesses the beauty of a flower, looking as one imag-
ines Yang Guifei or Cleopatra to have been in days of yore. . . . What would it be
like if that lady then opened her mouth and spoke in beranmee kotoba? Alas! her
lovely countenance and arching brow would lose their luster; her very garb would
appear a sham. Thus decorous language is more beautiful still than fair looks.
(Otsuma 1929:12)

These words express pride in femininity as a human quality different from those pos-
sessed by men. In other words, Otsuma insisted here that female speech is an impor-
tant tool for sexual charm in the same way as beautiful looks. However, as we have
seen in section 4.2, both the commentary of Reehoo yookoo and the Central Federation
for the General Mobilization of the National Spirit claimed that women should employ
more polite and humbler expressions than men. They did not instruct women to use
female speech to express their sexual charm. Nonetheless, both the female educa-
tors and upper- and middle-class women agreed that female speech expressed achieve-
ment of a higher status in society. National-language scholar Hoshina analyzed the
attitude of the educated middle-class women of the day as follows.

In middle-class and higher society, a lady of breeding will generally use honorifics,
that is, reverential language . . . While there may be no obstacle to her employing
rough speech when addressing an inferior, that would be demeaning to her, and so
she will naturally use refined, courteous language. (1936: 227228)

The following instance supports Hoshinas analysis. Otsuma, principal of a girls high
school, liked the farewell phrase gokigenyoo good-bye, which was employed by
members of the imperial families and the nobility, and made her students use it. For
them and their upper-class counterparts, female speech was a representation of
femininity, that is, sexual charm and social status.
Already highly regarded by women themselves, female speech was then af-
firmed as an element of the states national language policy with scholarly backing
88 HISTORICAL AND THEORETICAL FOUNDATIONS

from national language research. Therein lies the significance: the establishment of
a cooperative relationship between the states language policy and womens leaders.
As far as female students and educated middle-class women in Tokyo were concerned,
we may surmise, the idea was implanted that their own speech was the standard to be
followed (see chapter 5 of Washi 2003).

4.4. Conclusion

The objective of this chapter has been to demonstrate the existence of channels for
educating women as an element of Japanese language policy. I have recounted the
history of Reehoo yookoo in order to demonstrate that the Ministry of Education
propagated gendered speech for women. This manual, released by the Ministry of
Education in 1941, officially prescribed gendered forms of standardized speech for
the first time. Reehoo yookoo was originally a new textbook of etiquette for second-
ary schools, but the state propagated it as appropriate for all citizens regardless of
class and especially directed it toward women through various forms of mass media.
Girls were thereby instructed by the state to use female speech. Second, I traced
the activities of the NLA, which served as the organization of language policy, to
demonstrate the fact that some educated women even cooperated actively with the
NLA to promote female speech. As we have seen, the NLA took advantage of
womens media and female leaders to foster the idea of preserving the national
language and to encourage the use of female speech among women. The estab-
lishment of a cooperative relationship with language policy-making bodies was, for
female leaders, part of a broader pattern of collaboration with the war effort. It also
meant that education in language use pursued at higher institutions of learning for
women was now officially recognized by the state as an element of language policy
backed up by national language research.
As I pointed out earlier, in reality only a small percentage of the female popula-
tion of Japanese was directly taught female speech, which was simply Standard
Japanese for women, because in the 1940s people who could speak Standard Japa-
nese were in the minority. It was not until after the war that female speech gained
widespread currency. Yet it is crucial to recognize that female speech is an artifi-
cial construct serving state interests in an ideology of femininity, a construct that was
developed largely as a state project during the war years but which was also strongly
supported by elite women educated during this period.

Notes
I thank Professor Oso Mieko for her invaluable suggestions and help with my English. I also
thank Suzui Junko and Naito Kikue for their kind help in its preparation.

1. All the translations from Japanese into English are mine.

2. Shuushin as a school subject is equivalent to religion in schools of some European
countries. Shuushin was based on National Shintoism, one of the Japanese religions. Sahoo
Manners is equivalent to comportment, which was taught, for example, in convents in some
 JAPANESE FEMALE SPEECH AND LANGUAGE POLICY IN THE WORLD WAR II ERA 89

European countries. The Meiji government studied European manners and introduced some
European elements into its nationalist project of Sahoo modernization.
3. Koyama (1991) examines the ideal of ryoosai kembo in detail. Here I restrict myself
to a summary based on Koyama (1991) to assist in understanding the ideas of the female
educators discussed later. The ryoosai kembo ideal of womanhood emerged at the end of the
nineteenth century. It is neither peculiar to Japan nor a relic of Edo-period Confucianism.
Rather, it is a modern concept designed to justify gender roles, one that evolved continuously
right up to the period in question. Ryoosai means a good wife capable of doing housework
and managing household affairs. Unlike in the Edo period, more was expected than mere
obedience to ones husband. Kembo wise mother refers to the qualities of mind needed to
bring up children properly, child rearing being thought to be the womans instinctive role. In
the Edo period women were seen as stupid and inferior to men, and mothers in samurai house-
holds were not entrusted with the education of their children.
As Koyama (1991) points out, this view of womanhood was underpinned by the belief
that women and men were polar opposites both physically and mentally. Given their differ-
ent makeup, it was thought only natural that men should work outside the home while women
took care of the housework and children. Furthermore, although in reality women were by no
means equal to men, they were regarded as being so due to the complementary roles of the
two sexes, and over the course of time educators came to extol the ideal of onna rashisa femi-
ninity as the badge of womanhood. Shimoda Jiro, who came under the influence of Havelock
Ellis during his studies abroad (see section 4.3.5, Female educators and female speech),
was among the first to advocate this view of gender differences. The ideal is thus a modern
one that, if anything, resembles Victorian thinking on the issue.
During the years under examination this ideal of ryoosai kembo lay at the core of
middle-school education for girls, which the Ministry of Education tailored very much to
the middle class.
4. The Japanese government chose the dialect spoken in the capital to make a national
language that would serve as a tool to integrate the nation. But the Japanese case was not a
special one. For example, it is well known that Standard French is based on the dialect that
upper-class people spoke in Ile-de-France, with Paris at its center (e.g., Jespersen 1925).
5. Both the denotation and connotation of the term middle-class before World War II
are different from those at present in Japan. Before the war, people termed middle-class were
not a majority of the population, as they are today, but were rather considered by the Japa-
nese government to be leaders of society; therefore, before World War II, the term middle-
class referred to wealthy upper-class people below the nobility and imperial families. (For a
discussion of the nobility in Japan, see Lebra 1993.)
6. While a host of conflicting views arose on points of detail, it was agreed that stan-
dard spoken Japanese for women should be based on the language of cultured women of the
middle and higher classes in Tokyo. Therefore, debate on female speech from the outset
excluded distinctive female locutions in the regional dialects, except in special cases. It was
restricted in scopegeographically, socially, and with respect to age and level of schooling
to the language of cultured adult women with at least a secondary education who dwelled in
Tokyo and belonged to the middle class. See chapter 1 of Washi 2003.
7. Boys went to middle school for five years, while girls went to girls high school for
four years. The content of education in girls high schools was different for some subjects
from that in boys middle schools and was easier.
8. The cabinet of Prime Minister Konoe launched Kokumin Seeshin Soodooin Undoo
The General Mobilization of the National Spirit Movement to prepare to prosecute total
war in 1937. Kokumin Seeshin referred to the spirit of the nation, whose citizens would sac-
rifice themselves for the state regardless of class, sex, or age.
90 HISTORICAL AND THEORETICAL FOUNDATIONS

9. Secondary educational institutions for girls were girls high schools, technical schools,
and the like. In 1941, 5,119,638 girls and 5,248,241 boys attended elementary school, repre-
senting an enrollment rate of 99.73% and 99.70%, respectively (Nihon Tookee Kyookai
1988:213). In the same year, 454,423 girls were enrolled at girls high schools and 364,486
boys were enrolled in middle school (Nihon Tookee Kyookai 1988:243).
10. Nyooboo kotoba was used not only by court ladies but also by the nobility, includ-
ing men. Nyooboo kotoba is also known as kuge kotoba the nobilitys language. It seems
reasonable to suppose that the name kuge kotoba represents the actual status of the jargon
more properly than the name nyooboo kotoba.

References
Dan, Michiko (1943). Josee to kotoba (Women and language). Tokyo: Sakuragi Shobo.
Hasegawa, Nyozekan (1941). Kokugo to shakai seekatsu (National language and social life).
In T. Sakuragi (ed.), Kokugo bunka kooza 5: Kokugo seekatsu hen (Course in national
language culture 5: National language in life), 1734. Tokyo: Asahi Shimbun-sha.
Hoshina, Koichi (1936). Fujin no kotoba to kodomo no kotoba (Ladies speech and childrens
speech). In K. Hoshina (ed.), Kokugo to nihon seeshin (National language and the Japa-
nese spirit), 227228. Tokyo: Jitsugyoo no Nihon-sha.
Inoue, Kiyoshi (1943). Josee no kotoba: Keego no tsukaikata (Womens language: How to
use honorifics). Tokyo: Bokutaku-sha.
Ishiguro, Yoshimi (19401941). Kokugo no wadai (Topics in national language). Fujo
shimbun 2112: 13, 2114: 15, 2117: 25.
(1943a). Utsukushii nihongo (Beautiful Japanese). Tokyo: Kofukan.
(1943b). Jogakusee no kotoba-zukai (Language use of female students). Shoojo no
tomo 36(7): 104107, 36(8): 9698, 36(9): 9295.
(1955). Gengo seesaku (Language policy). In Kokugo Gakkai (The Society for the
Study of Japanese Language) (ed.), Kokugogaku jiten (Dictionary of the Japanese lan-
guage studies), 331332. Tokyo: Tokyodo Shuppan.
Ishikawa, Ken (1957). Kindai Nihon kyooiku seedoshi dai 14 kan (History of the educational
system in modern Japan), vol. 14. Tokyo: Dainippon Yuben-kai Kodansha.
Jespersen, Otto (1925). Mankind, nation and individual from linguistic point of view. Lon-
don: Williams and Norgate.
Kaplan, Robert B., and Richard B. Baldauf Jr. (1997). Language planning from practice to
theory. Clevedon, UK: Multilingual Matters.
Kikuzawa, Sueo (1929). Fujin no kotoba no tokuchoo ni tsuite (On the characteristics of
womens language). Kokugo Kyooiku 14(3): 6675.
(1933). Kokugo kagaku kooza 3. Kokugogaku Kokugo isoo ron (Course in national
language science 3: National-language studies: The phasic theory of national language).
Tokyo: Meiji Shoin.
Kokugo Kyookai (1937). Kokugo kyookai no kisoku (Regulations of the National Language
Association). Kokugo undoo 1(1): back cover.
(1939). Kokugo kyookai ni fujinbu ga dekita (Womens wing established in the
National Language Association). Kokugo undoo 3(9): 49.
(1941). Dai 5 kai kokugo kyookai sookai (Fifth general meeting of the National Lan-
guage Association). Kokugo undoo 5(7): 27.
Kokuritsu Kokugo Kenkyuujo (National Language Research Institute) (1951). Gengoseekatsu
no jittai (Language survey in Shirakawa City and its neighboring rural district). Tokyo:
Shuei Shuppan.
 JAPANESE FEMALE SPEECH AND LANGUAGE POLICY IN THE WORLD WAR II ERA 91

Koyama, Shizuko (1991). Ryoosai kenbo to yuu kihan (The norm of a good wife and wise
mother). Tokyo: Keiso Shobo.
Kyodo, Shuppan (1942). Shoseki nenkan Showa 17 nen ban (Books almanac, edition of the
17th year of Showa). Tokyo: Kyodo Shuppan.
Lebra, Takie Sugiyama (1993). Above the clouds: Status culture of the modern Japanese
nobility. Berkeley: University of California Press.
Ministry of Education (1941). Reehoo yookoo (Essentials of etiquette). Tokyo: Noogyoo Tosho
Kankoo-kai.
(ed.) (1980). Shuusen kyooiku jimu shori teeyoo dai 1 shuu (Summary of educational
administrative procedures at the end of the war), vol. 1. Tokyo: Bunsendo Shuppan.
Muta, Kazue (1992). Senryaku to shite no onna: Meiji, Taisho no onna no gensetsu o megutte
(Woman as strategy: On womens discourses in the Meiji and Taisho periods). Shisoo
812: 211230.
Nihon Tookee Kyookai (Japan Statistical Association) (1988). Nihon chooki tookee sooran
dai 5 kan (Compendium of long-term statistics on Japan), vol. 5. Tokyo: Japan Statisti-
cal Association.
Otsuma, Kotaka (1929). Nichijoo jooshiki giree sahoo (Etiquette and manners as daily knowl-
edge). Tokyo: Okamura Shoten.
Reehoo Kenkyuu-kai (Etiquette Research Association) (1941). Reehoo yookoo kaisetsu (Com-
mentary on the essentials of etiquette). Tokyo: Kokoku seenen kyooiku kyokai.
Shimoda, Jiro (1977[1904]). Joshi no kyooiku (Female education). In M. Tametomo (ed.),
Nihon fujin mondai shiryoo shuusee 4 kan. Kyooiku (Anthology of materials on womens
issues in Japan), vol. 4: Education, 359. Tokyo:Domesu Shuppan. [=1904, Joshi no
kyooiku (Female education). Tokyo: Kinkodo.]
Shimose, Kentaro (1937). Atarashii kokugo ishiki e (Toward a new sense of national lan-
guage). Kokugo undoo 1(5): 1.
Shinmura, Izuru (1938). Josee no kotoba (Womens speech). Fujin no tomo 32(9): 9698.
Sugimoto, Tsutomu (1997). Onna kotoba ima-mukashi (Women and language, now and then).
Tokyo: Yuzankaku.
Suzuki, Yuko (1997). Shinban Feminizumu to sensoo: Fujin undooka no sensoo kyooryoku
(New version feminism and war: The cooperation of women activists in the war effort).
Tokyo: Maruju-sha.
Takakura, Teru (1939). Josee to kokugo no mondai (Women and the national-language ques-
tion). Fujo shimbun 20: 12.
Usami, Mayumi (1997). Kotoba wa shakai o kaerareru (Language can change society). Tokyo:
Akashi Shoten.
Washi, Rumi (2000). Nyooboo kotoba no imisayoo: Tennoosee kaisoosee sekushuaritii (The
political function of nyooboo kotoba: Symbolizing the imperial system, class conscious-
ness, and sexuality). Joseegaku Nenpoo 21:1835.
(2003). Jendaaka shita Nihongo: Keeseekatei oyobi shoochooteki imi to seejiteki kinoo
(Gendered language in Japan: Its development, symbolic meaning, and political func-
tion). Unpublished Ph.D. dissertation, Graduate School of Nagoya University.
Yanagi, Yae (1941). Fujin no kotoba (Ladies speech). In Kokugo bunka kooza 5: Kokugo
seekatsu hen (Course in national-language culture 5: National language in life), 248
268. Tokyo: Asahi Shimbun-sha.
Yoshida, Sumio (1935). Fujin no kotoba (Ladies speech). In Kotoba no kooza 2 (Course in
language 2), 145156. Tokyo: Japan Broadcast Publishing.
92 HISTORICAL AND THEORETICAL FOUNDATIONS

WIM LUNSING
CLAIRE MAREE

Shifting Speakers
Negotiating Reference in Relation
to Sexuality and Gender

I n what ways, if any, does a speakers personal sense of gendered and sexualized
self influence language use? What gender and sexuality norms police the borders of
Japanese language use? How do speakers negotiate these borders? How can we at-
tend to the particularities of Japanese language, gender, and sexuality paradigms while
at the same time problematizing prejudices within the sociocultural realm? These
issues form the entry points of our inquiry into the intertwining of language, gender,
and sexuality in spoken Japanese.
Due to the widespread conflation of gender with sexuality, notions of gender-
appropriate language simultaneously invoke notions of gender-appropriate sexuality.
Consequently, socially conceptualized and shared norms of gender-appropriate lan-
guage remain overwhelmingly heterosexist. These norms are, however, both shared
and contested, overarching and contradictory. Research on gender and language thus
far has produced an eclectic discourse, which both problematizes and solidifies no-
tions of gendered language performance. Since the 1990s, research that contests the
supposed intrinsic social link between gender and sexuality has increased. Collected
volumes include Leap (1995), Harvey and Shalom (1997), and Livia and Hall (1997).
Work on gender, sexuality, and language in Japanese includes Ogawa and [Shibamoto]
Smith (1997), Maree (1997, 1998), and Vanbaelen (1998). Issues of linguistic gender
performances in Japanese, however, remain to be unraveled from dominant discourses
of compulsory heterosexuality intertwined with overt and covert cultural misogyny.
In order to build on the body of current literature, it is imperative to show how
issues of gender and sexuality are entwined in the language use of Japanese speakers.
At the same time, it is necessary to untangle gender norms from dominant heterosexist

92
 SHIFTING SPEAKERS 93

discourses of gender. In this chapter, which constitutes an exchange between two

researchers of differing academic backgrounds and methodological approaches, we
address these issues in our discussions of real Japanese spoken by lesbian and gay
people. In section 5.1, we discuss the conflation of gender and sexuality and what
strategies people may employ to confront this. Section 5.2 discusses uses and abuses
of terminology for male homosexuality and issues of self referencing in relation to
sexual activity, based on Lunsings work with gay men, lesbians, and others in Japan.
Discrepancies between various experiences of terminology and reference provide
grounds for contesting heterosexist discourse as well as possibilities for self-assertion.
Section 5.3 discusses the linguistic negotiation of self-reference by a lesbian, based
on Marees work with lesbians and gay men in Tokyo. Personal metadiscursive ac-
counts of language use provide insights into the multiplicity of meanings indexed by
everyday self-reference terms and into the negotiating strategies individual speakers
employ in everyday interactions. Section 5.4 offers conclusions illustrating the extent
to which prescription is never fully reflected in actual language use, where speakers
negotiation of gender and sexuality norms intersects with their individual sense of
self and contextual/situational pressures.

5.1. Disentangling the entwined

Sedgwick (1990) and Rubin (1993) have clearly demonstrated the need to disentangle
theories of sexuality from those of gender. While this demarcation is necessary, it is
simultaneously essential to emphasize the ways in which gender and sexuality inter-
relate as social phenomena in the popular mind. The complex connection between
sexuality and gender has been articulated both in Japanese gay and lesbian studies
and in popular lesbian and gay writings. For example, Sunagawa (1999:145) argues
against the idea of heterosexuality as an identity, claiming that few Japanese overtly
or explicitly identify themselves as heterosexual. Instead, Japanese hetero-society
pivots on strict gender norms and roles whereby women have romantic relationships
and sex with men and vice versa. Thus, from a heterosexist viewpoint, men who love
men and women who love women are in effect transgendered (Lunsing 2001b).
Although, as Butler (1993:238) suggests, a non-causal and non-reductive con-
nection between sexuality and gender is . . . crucial to maintain, gender is not the
direct cause of sexuality and it cannot merely be reduced to sexuality. In fact, there
is no natural relation between gender and sexuality other than the conflation in many
peoples minds, similar to the manner in which sex and gender are commonly
conflated (Lunsing 2001a). Under compulsory heterosexism (to borrow Richs 1980
term), both sex and gender are constructed as natural. This belief effectively obstructs
understanding sex/gender as a naturalized social custom (Butler 1990). In the con-
temporary Japanese social context, with its widespread conflation of heterosexual
desire with gender and its idealization of heterosexual romantic love, homosexuality is
implicated as trivial and abnormal and is often greeted with the phrase kimochiwarui
sickening. Furthermore, jooshiki common sensea term Lunsings lesbian, gay,
straight, and other consultants used very often in dismay at what they were up against
deems heterosexual marriage the natural (shizen na) way to live and by implication
94 HISTORICAL AND THEORETICAL FOUNDATIONS

denaturalizes relational possibilities other than heterosexuality (Lunsing 2001b).

Homosexuality, thus, is regarded as hentai, alternately translated as perverse
(Frhstck 1997), obscene (Pflugfelder 1999), and queer (Lunsing 1997).
These issues have implications for the question of gendered language. Current
research has documented wide-ranging synchronic and diachronic variation within
womens speech styles in Japan, thereby contesting the notion of a monolithic womens
language (joseego/onna-kotoba), and consequently the notion of a monolithic mens
language (danseego/otoko-kotoba). For example, Abe (1993, chapter 11, this volume),
Matsumoto (chapter 13, this volume), Miyazaki (chapter 14, this volume), Okamoto
(1994), and Sunaoshi (chapter 10, this volume) document womens language use that
diverges from normative usage. Similarly, Inoue (1994, chapter 3, this volume), Endo
(1997), and Washi (chapter 4, this volume) discuss the historical construction of
(normative) womens language. As Okamoto (1994:578) writes, Japanese womens
choice of speech styles is a complex process involving the simultaneous consider-
ation of multiple social attributes associated with identity and relationships. Based
on their understanding of themselves in specific relationships and contexts (Kondo
1990), Japanese women strategically choose particular speech styles to communi-
cate desired pragmatic meanings and images of self.
As will be demonstrated in this chapter, the language use of lesbians and gay men
is highly strategic. Since there are prevalent gendered and sexualized stereotypes of
language use and other behavioral aspects that speakers must negotiate, lesbians and
gay men may alternately choose to comply with the existing stereotypes of masculine
lesbians and effeminate gay men or remain invisible as lesbian or gay to those who
do not see past these stereotypes by adhering to more normatively feminine or mascu-
line speaking patterns (Valentine 1997). It is necessary to have some understanding of
this complicated process and its interrelationship with heterosexist gender norms when
discussing issues of gender, sexuality, and language use in Japanese.

5.2. Gay male speech, sexual activity,

and homophobia

This section investigates various aspects of gay male speech that are often the source
of controversy within gay circles: meanings of words identifying gay men, effemi-
nate speech, and self-reference terms. These aspects are discussed in relation to sexual
activity and homophobia, based on data acquired by extended participant observa-
tion; in-depth interviews, most of which were audiotaped; group discussions; sur-
veys; and vernacular written sources. Methodologies are discussed in detail in Lunsing
(1999b, 2001a, 2001b).

5.2.1. Between okama and gei

In the Edo period (16031868), male homosexuality in Japan was characterized ei-
ther by transgenderism, in which case the penetrated person possessed characteris-
tics associated with the female gender, or by transgenerationality, in which case the
penetrated person possessed youthful characteristics (Furukawa 1994, Pflugfelder
 SHIFTING SPEAKERS 95

1999). The penetrated person yielded to the penetrating person, analogous to the struc-
ture of heterosexual relations, in which the woman was supposed to yield to the mans
wishes. During this period, the term okama, a type of pot, came into use to refer to
homosexual men and, more specifically, the penetrated person. The adult man was
understood to have the option to penetrate either women or feminine or youthful men
and hence stayed outside of the category of okama. The penetrated man was not
expected to enjoy anal penetration per se, as it was supposed to be painful (Pflugfelder
1999:4142). Therefore, ways of enticing the okama were necessary. The okama was,
for instance, supposed to love the inserter enough to endure the pain. Power imbal-
ance was another means to make the okama agree to insertion. Most important, how-
ever, was the okamas financial reward for engaging in sex work.
The term okama is still one of the most popular terms to refer to a homosexual
man, usually in a derogatory manner. The term is used to refer to an effeminate man
who makes use of a language close to stereotyped womens speech. In 1971, one of
the most vocal Japanese gay men of the twentieth century, Togo Ken, ran for the first
time as a candidate for Parliament. Dressed in traditional kimono and wearing make
up, as he usually does on festive occasions, he introduced himself to the media with
the words: Watashi wa okama no Togo Ken desu I am Togo Ken, the okama. By using
the derogatory term okama1 in self reference, Togo confronted his prejudiced audi-
ence, who view gay men as effeminate and interpret his clothing choice as transves-
tismand him as thus somehow less worthy than other men. Most Japanese gay
organizations, which originated in the 1980s, distanced themselves from Togos strat-
egy and instead started using borrowed English terms like homo (from homosexual)
and the now most popular gei gay. Instead of Togos watashi wa okama, they say
boku wa gei, whereby the use of the normatively masculine pronoun boku instead of
the normatively feminine watashi also stresses masculinity. The explanation the gay
activist group Occur gave for this trend is that okama and, to a lesser extent, homo are
derogatory terms (Lunsing 1999a). It seems that the use of gei is propagated due to its
relation and homophony with the English gay and the relatively positive implications
it has in Anglo-American contexts. However, activists choice of the term gei has its
own flaws. Gei was, in the past, used by straight people interchangeably with the term
gei booi gay boy, which was typically understood to refer to transvestite gay men
whose transvestism was part of their occupational activities, be it as sex workers, as
theatrical performers, or as singers, such as Miwa Akihiro.2 Nyuuhaafu new half is a
newer term to refer to a similar transgendered category, and it is equally commonly
confused with gei by straight people and others who have not sufficiently acquainted
themselves with gay and new-half scenes in Japan. Male homosexuality is generally
confused with transgenderism, as reflected in popular interpretations of reference terms.
Similarly, gei baa gay bar(s) is a term that, to the general public, refers to bars
where transvestites perform a type of transgender act to entertain a predominantly
straight patronage. In gay circles, such bars are referred to as kankoo baa tourist
bars, bars for straight people being tourists in the gay world; gay men refer to their
own bars as gei baa. Gay patrons of gei baa overwhelmingly choose styles and
clothes commonly regarded as masculine. For straight people gei is a male perfor-
mance of femininity, which confirms their misperception of male homosexuality
as transgendered.
96 HISTORICAL AND THEORETICAL FOUNDATIONS

At a performance in 1993 with the classically trained female singer Haruka Mari,
the drag queen singer Shimoonu Fukayuki (Simone Deepsnow) explained to the mixed
gay and straight audience that he was not a woman but a gay man. In this context,
Fukayuki used the term gei. Only later in life did he come to use the term okama,
which he had come to avoid after negative experiences in his youth. He wrote in a
short essay on queer culture that by using the term okama he felt that he finally had
overcome the self-oppressive feelings so common among gay men in Japan (Fukayuki
1996)and elsewhere.

5.2.2. Onee kotoba older sister speech

Togos self-introduction to the media watashi wa okama is an example of onee
kotoba older sister speech, effeminate camp language used by gay men (and some
lesbians; see section 5.3.1). In Japan (as elsewhere), young gay men making first
contacts with other gay men tend to feel awkward about those who use onee kotoba,
who are perceived as more queer than supposedly normal gay men. Projects aimed
at a broader gay audience hardly use onee kotoba at all. An example is the Kamigata
DJ Club tape project, in which a group of mostly gay men made, without resort to
onee kotoba, cassette tapes that resembled radio programs on gay issues and sent
them to gay men throughout Japan. In order to appeal to isolated gay men, com-
mon masculine speech is used in such contexts, as it is in most gay contexts. In
many gay circles in Japan, onee-type men are scorned. Gay men often profess to
prefer real men over effeminate ones, and speech style is a focus of contention.
This concerns more than just words. Users of onee kotoba typically accompany
their words with intonational contours and gestures that are also seen as effemi-
nate. In addition, their clothing may be relatively flamboyant or they may wear
womens clothes.
While some gay activists in Japan demand that gay men be depicted as just as
masculine as any other men in order to counteract the effeminate stereotype, many
gay men do not behave accordingly. This repeatedly leads to controversy, such as
the one that followed the 1991 documentary film Rasen no sobyoo Rough sketch of
a spiral, used by Ogawa and [Shibamoto] Smith (1997) as the basis for their analy-
sis of gay language. In the film, one of the gay partners assumes the womans role,
including the use of effeminate language, or onee kotoba. The other partner assumes
the mans role and uses masculine language. Schoolteacher Hirano Hiroaki, who also
appeared in the film, later wrote that he was very unhappy with yet another depic-
tion of a gay relationship that mimics a stereotypical heterosexual one, wondering
who had murdered the film (Hirano 1994). The depiction in the film was, how-
ever, an accurate rendering of the actual relationship of the two men.
Onee kotoba is not, moreover, simply effeminate speech. It often distinguishes
itself from conventional female speech by its crudity, as in the following example: Atashi
ima karee kuttara geri da wa If I ate curry now, Id get diarrhea. Here the topic itself
is uncouth as is the verb kuu for eating, while the pronoun atashi and the ending da wa
suggest femininity. The cruder-sounding term okama, instead of gei, is more likely to
be used by those who employ onee kotoba than by those who do not.
 SHIFTING SPEAKERS 97

Such crudity, often combined with sarcasm, can easily be misunderstood. Miwa
Akihiro, as well as the gay twin television personalities Osugi and Piiko, who are
not transvestites, use onee kotoba. Miwa once remarked that he had never experi-
enced sex with women, followed by the remark: Watashi wa kiree I am clean
(quoted in Ueno 1991[1986]). This is a typical example of onee kotoba. Consultants
say things like Atashi wa kitanai I am dirty when they have had sexual intercourse
with men and Saikin no atashi wa kiree Lately I have been clean when they have
had a period without sex. Within the Japanese feminist framework, Miwas comment
can be interpreted as misogynous, as Ueno Chizuko (1991[1986]) once did. How-
ever, it may also be seen as a typical example of onee kotoba. Making jokes about
women, like Osugi and Piiko often do, is an integral part of onee kotoba, as is mak-
ing jokes about almost anyone or anything available. Two lesbian consultants, who
were also ardent feminists, said that they appreciated Osugi and Piikos performance,
even if it could be interpreted as misogynous. To them, it was of more importance
that Osugi and Piiko ridiculed heterosexism.
Usage of onee kotoba is largely determined by environmental circumstances.
One has to learn to employ it. Those who have learned it usually feel the need to
employ it only in the right situations, which may not be easy, as noted by a consul-
tant, who was horrified to find himself employing it at his office.

5.2.3. Self-referencingbetween masculine and feminine

Self-referencing is an area of contest not only when people employ onee kotoba. As
Miyazaki (chapter 14, this volume) observes, in popular understanding watashi I
is regarded as a feminine or perhaps neutral term for self-reference, but definitely
not as masculine. In the earliest mainstream publication by a gay man in Japans so-
called gay boom of the early 1990s (Lunsing 1997), Fushimi Noriaki wrote that in
his youth he had been ridiculed for using watashi (Fushimi 1991). It was not that he
wanted to appear feminine but rather that he could not bring himself to use the mas-
culine informal equivalent boku, a more common term for boys. Likewise, Hirano
Hiroaki had the same problem with his inability to use the even more masculine ore
(Hirano 1994). As first-person pronouns are normatively gendered in Japanese, un-
easiness about self-reference is a common phenomenon, in particular among young
gay men. This may be why, instead of watashi, boku became the most commonly
used self-reference term in most gay activist circles in the 1980s (Lunsing 1999a),
though many other gay men feel uncomfortable using boku and persist in calling
themselves watashi or atashi, which is even more explicitly feminine. Fushimi re-
ports that he eventually came to use boku but more recently has reverted to watashi
or atashi in many cases and even uses the female name Noriko to refer to himself as
part of a performance of femininity. While in his youth he felt compelled to over-
come his emotional resistance to boku, later in life he adopted watashi again as part
of a performance he deemed useful for his purposes. It appears that he came full circle
and ended up reappraising his youthful emotions.
In 1999, one of us interviewed Shimoonu Fukayuki and we discussed, among
other issues, his self-reference choices as a gay man who performs in drag. On stage,
98 HISTORICAL AND THEORETICAL FOUNDATIONS

Fukayuki refers to herself as watashi. During the interview, however, the interviewer
spoke with her nonperforming male alter ego. In this situation, Fukayuki also used
watashi and the more formal watakushi for self-reference. When asked whether he
ever used boku, he replied that he often did in his daily life but that in the interview
situation he felt that watashi was the correct term, being more deferential and polite
than boku. Indeed, in many contexts both straight and gay men feel obliged to make
use of watashi rather than boku for simple reasons of politeness or other contextual
factors.
In the same period, the coauthor interviewed D. K. Uraji, who also performs as
a drag queen. Uraji consistently referred to himself by the term watashi. He told me
that he always does and that he has never seen himself as a boy or a man. In fact, he
said that it was not until he reached the age of 18 that he was first confronted with
the idea that he was a man. His mind wrapped up in fairy tales like Snow White,
Uraji had come to identify with the heroine and declared his love to a male classmate
he saw as fitting the role of the hero. Urajis classmate wrote a letter that said such a
relationship could not be, because Uraji was a man. Totally shocked by this confron-
tation, Uraji entered a period of major depression, which lasted for about 10 years.
He felt that his very existence was denied. It was not until he made contact with a
group of drag queens that he found a way to live with himself.
In performing as a woman onstage in front of an audience, Uraji found a way to
express his being. However, he said that when doing drag performances onstage he
feels more masculine than at any other time. He maintains that in drag performance
masculinity has to show through in order to make the performance interesting. It is
the imperfection in his impersonation of a woman that makes it enjoyable to watch.
In his daily life, however, he is less masculine. As he sees it, he performs various
roles, that of a son to his parents and that of a hardly gendered illustrator in his occu-
pational capacity. In public contexts, he usually passes for a woman. He has found
love and lovers, most of whom are heterosexual men, thus fitting the type of hero he
needs to complete his love fantasy. He prefers to avoid calling himself gei, because
that does not fit Uraji. When talking of homosexuality in general, however, he uses
the terms okama and gei interchangeably. When people are young, self-reference
terms are usually chosen because of their feelings. Once confronted with reactions
from their environments, they may develop strategies of employing varying terms
for varying occasions.3

5.2.4. Speech and sexuality

The language use of gay men is often considered indicative of their preferred sexual
activities, but this is not necessarily the case. Japanese gay men distinguish one
another based on a number of qualities, of which age is a major one. When a gay
man is discussing what type he feels attracted to, toshishita younger and toshiue
older are often the first determinant. Another important dichotomy is that of tachi
top and neko bottom. Though commonly used among gay menand lesbians
these terms are not reliable as references to what sexual activities actually take place.
A neko is supposedly sexually passive in the sense that he is the one who is pene-
 SHIFTING SPEAKERS 99

trated by the tachi. In the Japanese heterosexist power structure this would mean
that the neko is also the more feminine. The case of a younger man would be simi-
lar. In actual practice, however, individual sexual attraction precedes the question
of whether someone is younger rather than older or bottom rather than top. In the
words of a consultant: If there is love (in this case, koi), then [these categories]
matter little.
Indeed, while the general image of homosexual activity is that of anal penetra-
tion, in Japan this activity is only engaged in by a minority of gay men (e.g., Za Gei
Henshuubu 1992, Lunsing 2001b: 284285). Uraji does not like sex characterized
by what he calls pisuton undoo piston motion, a common Japanese term for the up-
and-down movement that characterizes anal (and vaginal) intercourse. He does not
like penetrative sex at all. Many okama or gei who behave in a relatively more femi-
nine manner do not. At the same time, Lunsing found quite a number of instances in
which, contrary to the expectation based on heterosexist patterns, the one who em-
ployed onee kotoba actually became the penetrator, be it oral or anal, during sex.
There is no simple direct link between a gay mans use of language and his sexual
activities. A watashi can be just as sexually aggressive as a boku or an ore, and the
term okama, when used for self-reference, is no indication of the type of sex the
speaker desires other than that it is with a man. Mainstream depictions of homosexu-
ality may influence the language use of gay men, but such depictions have much less
influence on their actual sexual activities.
Although the heterosexist language system leads toward a binary system within
which one would expect gay men to divide into active/masculine and passive/femi-
nine categories, in fact such a division is not related in a direct manner to the use of
language. How people perform in bars or other social contexts does not predict what
they do in bed. Furthermore, the historical discourses introduced earlier (section 5.2.1)
must be approached cautiously. They appear to be largely informed by heterosexist
discourses in which power relations are supposed to be evident. Much of these dis-
courses can simply be regarded as prejudice, prejudice that remains largely intact
today.
It seems that those who have changed their language use and come to refer to
themselves by the term okama feel a self-confidence that helps them find their place
in Japanese society and that homophobia is found more in those who avoid the use
of okama, which does not mean that those who use gei or homo are necessarily al-
ways more homophobic or ill at ease with their sexuality. Similarly, gay men who
use watashi in peer or gay contexts without feeling awkward may in some cases be
more likely to have come to grips with their sexuality than those who use boku or
ore. By using okama, (w)atashi, or onee kotoba, gay men may identify with a cultur-
ally deeply entrenched construction of homosexuality and thereby confront its in-
herent prejudices, rather than avoid them. Among the many people Lunsing has
spoken with, those who used ore in gay or peer contexts tended to be young and often
did not regard themselves as gay. Typically, they also had a strong dislike of people
who employed onee kotoba. Nevertheless, the words people use cannot be taken as
simple indications of their feelings toward their sexuality. Much depends on the in-
dividual case.
100 HISTORICAL AND THEORETICAL FOUNDATIONS

5.3. Negotiating compulsory heterosexist gender

and language: A lesbians personal history
of discourse

The preceding discussion points to the diversity of gay mens relationships to

language and to the fact that language use in itself is not indicative of individual sexual
activity or vice versa. This section focuses on the speech of women who identify as
lesbian and who, in a continuous process of shifting identification with same-sex
sexuality, negotiate the terrain of heterosexist gendered language norms. Derived from
a study of linguistic negotiative strategies employed by Japanese speakers (Maree
2002a), the data (in which Maree occupies the position of both researcher and par-
ticipant) comprise two hours of conversation among seven speakers in a roundtable
discussion about language and sexuality convened by the researcher, an informal
dinner conversation that followed the roundtable discussion, and follow-up inter-
views.4 In focusing on lesbian-identified language users, the aim is not to demonstrate
their supposed uniqueness, or deviation from the norm, but to indicate how all speak-
ers of Japanese negotiate spoken communication and to trace the ways in which nego-
tiation processes are remembered and repeated in the speech acts of all speakers.
The key to the following discussion is this very notion of negotiation (Maree
2002a,b). Japanese speakers, like all speakers, repeatedly and reflexively negotiate
complex contexts in which they position and perform multiple identities. Speakers
position their multiple identifications within interactive acts via an ongoing process
of negotiation (see Bucholtz, Liang, and Sutton 1999 for further discussions of
gendered self in discourse). The term multiple identities invokes the notion of a to-
tality of multiplicity, which the speaker negotiates at every point of the speech act
and in all speech situations. A speakers negotiation of complex contexts is aided by
her or his creative use of language. This creativity is founded in individual experi-
ences of language and personal histories in which discourse can uncover speakers
conscious negotiation of coercive language norms. In particular, metadiscursive
narrative, a speakers talk of personal language experiences, offers insights into ne-
gotiation processes. Like other narratives, metadiscursive narratives both are born
of and shape experiences, for, as speakers, we come to know ourselves as we use
narrative to apprehend experiences and navigate relationships with others (Ochs &
Capps 1996:21). In these narratives, speakers elaborate on and explain incidents and
personal feelings that are pertinent to their understanding of language norms and
attitudes toward them.
Narratives featured frequently in the aforementioned roundtable discussion, and
the topic of self- reference was extensively discussed. The Japanese self-reference
system is generally explained according to factors such as gender, age, region, the
vertical social relationships and in-group/out-group relationships between interlocu-
tors, the level of formality of the exchange, and so on.5 These factors are said to in-
fluence choice of first-person pronouns and other self-reference terms. (In casual
conversation it is common to refer to oneself not only using pronouns, e.g., atashi,
boku, but also using kinship terms, e.g., okaasan, lit. mother; occupational titles,
e.g., sensee, lit. teacher; or ones own name or nickname, e.g., Junko or Junchan.)
 SHIFTING SPEAKERS 101

In short, self-reference terms in Japanese not only directly index the person who
currently speaks but also simultaneously constitute elements of the social. (For a
general overview of social deixis see, for example, Foley 1997.) It is no surprise that,
when we focus on the entwining of gender and sexuality in Japanese, self-reference
is highlighted as a linguistic strategy subject to continued (re)negotiation.

5.3.1. The coercive force of watashi

Participants in the roundtable discussion each spoke of their individual language
experiences. Of the participants, Sayuri, Oka, and Marina spoke in detail of their
experiences with self-reference terms. Sayuri, a young woman who uses onee kotoba
in her personal interactions with friends and with her partner, recounted shifts in her
use of self-reference terms. In her teens, Sayuri accommodated her straight-identified
girlfriend by making use of the normative masculine pronoun boku and avoiding
normative feminine forms. However, when Sayuri formed a relationship with a
lesbian-identified woman, she avoided using boku and switched to using the more
stereotypically feminine form watashi. Oka, Sayuris partner at the time of this re-
search, also spoke extensively of her use of onee kotoba. Oka, who describes herself
as not being overtly feminine,6 prefers to use the more informal atashi because in
doing so she can let people know that she is a woman/female (atashi o tsukau to
yatto mina-san ni atashi wa onna da to yuu koto o wakatte moraeru no ka na to ka
omou by using atashi I think I can probably kind of finally let everyone know that
I am a woman/female). For both Oka and Sayuri onee kotoba is not a reproduction
of conventional feminine speech but a parody of stereotypical femininity that counter-
acts the stereotype of the butch lesbian and at the same time constitutes affinity with
gay male friends (Maree 2002b). In this context, Okas and Sayuris metadiscursive
explanations of their self-reference choices illustrate the complex relationship be-
tween the speakers image of self, the gendered norms of the self-reference system,
and cultural expectations. In what follows, we focus on the self-referencing strate-
gies of Marina, who spoke in considerable detail of her memories of childhood lan-
guage use, and investigate her negotiation of heterosexist language norms. Marinas
metadiscursive narratives clearly show the layers of meanings speakers traverse in
real-life interactions. Self-reference choice is not the result of a speakers gender
identification but a reflexive negotiative strategy.
As she explains in the roundtable discussion, Marina refers to herself as Marina
and does not use first-person pronouns. Although she says she does use the so-called
masculine pronouns ore, boku, and washi when she is joking with intimates, she is
particularly careful to avoid using the so-called feminine pronouns watashi and atashi.
She explains that she constantly struggled to overcome the gender expectations held
by her parents and recounts unpleasant experiences with first-person pronouns from
early childhood. In the narrative in example (1), she recalls her experiences of being
reprimanded for using the boyish pronoun boku.

(1) Marinas experiences of boku7

1 : . . . ue ni futari no, ani ga iru desho? soosu to mazu, see nanka
2 wakaranai toshi kara, mazu kotoba o oboeru ja nai? sosu to anikitachi
102 HISTORICAL AND THEORETICAL FOUNDATIONS

3 ga sa, boku da no boku da no ttsu tte itte, de issho ni natte asonderu wake
4 dakara, sore ga kotoba ga chan to hanaseru yoo ni natta toki ni (a daijoobu
5 yo), boku mo- to ka tte mazu yuu ja nai? boku mo boku mo to ka,boku mo-
6 to ka itta toki ni (bashitto kita) (3) sono-, boku ja arimasen yo, watashi tte
7 iinasai tte iwareta toki ni, pokkaan to suru to ka ne,
. . . above me, two, I have two brothers right, so first of all, I mean you learn
language before you have any understanding of sex/gender right, so my brothers
would, OK, say boku this and boku that, and because Im playing together with
them [and] the thing was, when I could speak properly (oh, its OK), I would say
first boku mo boku mo [me too, me too] and like that, when I said boku, too, (it
floored me) (3) when I was told like it isnt boku, say watashi, I was left stunned,
with my mouth hanging wide open,

As Marina indicates here, she was required to modify her speech before she became
aware (or was made aware) of sex/gender itself (lines 12). She expresses her lack
of understanding and her surprise as a child by describing her reaction with the ono-
matopoeic pokkaan to suru to be left mouth wide open, to be shocked speechless
(line 7). Later, when Marina is questioned on whether she has specific memories of
being reprimanded for using boku, she recounts an incident when she echoed her
brothers use of it. Her reaction to this reprimand was one of total surprise: moo
mawari no otonatachi to ano anikitachi mo issho ni natte, dame tto ka iwarete,
pokkaan to shite well, the surrounding adults and my brothers joined together, tell-
ing me no, I was shocked speechless. In both instances in which she uses the phrase
pokkaan to suru, the extended vowel [a] and the geminated [k], which shape the mouth
into an expression of astonishment, emphasize Marinas expressions of inability to
understand demands to modify her speech. She estimates that she was still a young
child at the time: yoochien e iku mae ka yoochien kurai probably just before or after
I started kindergarten.
A little further into the roundtable discussion, Marina recounts that she contin-
ued to refer to herself by her pet name Maachan upon entering preschool. While use
of a pet name (generally comprising the childs first name or part thereof and the
suffix -chan) is not unusual among young children, it is generally maintained that as
children enter school and experience group socialization their use of self-reference
terms alters from pet names to first-person pronouns (Ide 1990). Marina, however,
reports that she resisted the pressure adults placed on her to use particular pronouns.
Marina relates that her continued use of the pet name Maachan was met with
frequent warnings from other speakers. In fact, through laughter and repetition she
stresses that as she grew from a preschool to a primary-school-aged child, when she
continued to call for inclusion using the phrase Maachan mo, literally, Maachan,
too, she received warnings such as moo iikagen ni shiro and iikagen ni shinasai
enough is enough, already. The difference in style between the verb forms shiro
(stereotypically masculine) and shinasai (stereotypically feminine) in these phrases
indicates that multiple speakers warned Marina about her language use on multiple
occasions. The phrase iikagen ni shiro/shinasai and the adverb moo already both
indicate that the act in question, in this instance childish talk and, by extension, the
refusal to speak in gender-appropriate ways, has persisted beyond the dictates of
 SHIFTING SPEAKERS 103

common sense. Marina forges ahead (gooin ni stubbornly) with her use of Maachan
even though she is aware of, or indeed in spite of, the response it provokes. The
strength of confrontation Marina reports that she encountered is implied in the reit-
eration of warnings cited in her narrative and alerts us to the difficult terrains nego-
tiated in language.
In situations of great contextual pressure to use watashi, Marina concedes, she used
the form, however, gritting her teeth all the while. In example (2) she recalls one such
ocasion when she was summoned to the teachers room and punished. In a compara-
tively formal situation, in which an imbalance of power exists between reprimanding
teacher and punished student, she reports that she temporarily ceased using Maachan.

(2) Relenting to watashi

8 : gakkoo wa sasuga ni ne, moo shokuinshitsu ni itte shikarareru yoo na toki wa
9 ne? watashi tte dooshite mo tsukawanakya ikenai toki ya, ha kuishibatte
10 itteta n dakedo,
11 : (sonna ni iya datta no?)
12 : hontoo ni iya datta no,
13 : a soo?
14 : de ima wa ne, hontoo ni ne, heeki de, kore wa moo, ima wa ne, anoo, tsukaeru
15 yoo ni natta no, [sore
16 : [moo itsu mo wa Marina da yo ne? ima wa,
17 : soo
8 : well at school it was like you know, when I was called into the teachers room
9 and told off and stuff, at times when I just had to use watashi, I gritted my teeth
10 and said it but,
11 : (you hated it that much?)
12 : I really hated it,
13 : really?
14 : well now you know, really, without hesitation, this is like, now you know,
15 well, Ive become able to use it, [it,
16 : [now you always use Marina, dont you? now
17 : right

Although she no longer uses the nickname Maachan, Marina repeatedly used the expres-
sion Marina wa as for Marina in the roundtable discussion, the subsequent dinner,
and the follow-up interview. In fact, the only time she used watashi was in her highly
formulaic introduction at the beginning of the discussion session. Sayuris interrupting
comment (line 16) indicates that Marinas use of Marina is well known and accepted.
In this interaction, Marina at first says that she is now able to use the first-person
pronoun watashi (lines 1415). It is over 40 years since she first experienced dis-
comfort with watashi. However, even though she can now articulate the word, she
concedes she cannot collocate it with her real or honest opinions. In another utter-
ance spotted with hesitancy, truncated phrases, rephrasing, and intrautterance verifi-
cation seeking, Marina stresses that she definitely cannot use watashi when expressing
her true self: de mo sore de mo ne? yappari, anoo, jibun no chan to shita iken to yuu
ka kan ano nani? zenbu jibun dashitai toki ni wa watashi nante kotoba wa (zettai
104 HISTORICAL AND THEORETICAL FOUNDATIONS

ni) dasenai but you know even so? all the same, for my own real opinions and stuff,
feel what? when I want to express the real me I (definitely cant) use a word like
watashi. It is, therefore, impossible to dismiss Marinas refusal to use watashi as
merely an aspect of a childs process of language socialization.
As example (3) indicates, to Marina, watashi is not merely a first-person pro-
noun; it is an expression of heterosexist femininity (line 20) and patriarchal space
(line 2224). Marina understands from experience that watashi produces and repro-
duces social relations anchored in heteropatriarchal norms. Rejecting these coercive
norms entails rejecting the use of watashi. For Marina, to use watashi is to impose
on herself, and to create through her personal discourse, a social self she can neither
embrace nor control. Consequently, she selects alternative expressions of self.

(3) Multiple meanings of watashi

18 : tada, [Marina no ano jidai ni wa watashi to yuu kotoba ni wa,
19 : [<@hahahaha@>
20 : [<F hetero noF>, on onna no,aa, soshite (prefecture name [where Marina grew
up])=
21 : [un, un
23 : =to yuu <F hijoo ni kafuchoosee no tsuyoi F> <@<Hx>@>,
24 : un,
25 : [basho no, [imi ga subete komerarete te,
26 : [un, [un, un,
27 : sono <F dono ichi F> ni mo zoku zoku suru ki wa arimasen deshita to yuu
28 koto de sa-,
18 : but [still, in Marinas time the word watashi,
19 : [<@hahahaha@>
20 : [ <F hetero F>, wo women, um, and, <F the extremely patriarchal (prefecture
name [where Marina grew up])
21 : [yeah, yeah
20 :
21 grew up] F>=
22 : yeah,
23 : =<@<Hx>@>, [place, [contained all of those meanings
24 : [yeah, [ah-ha, ah-ha
25 : and I didnt want to be a pa part of any of those categories,

In the preceding passage, Marina articulates her strong desire to pull away from
outside pressure to conform to heteropatriarchal femininity. Her continued reitera-
tion of nonpronominal self-reference is a relatively simple linguistic strategy. She
does not want to be inadvertently implicated in heteropatriarchal femininity via use
of watashi, so she negotiates her way around it. This is less a function of desire than
of ability. In the follow-up interview, when asked how she responded to societal
pressures to speak more like a woman, Marina replies, Yarenai to yuu koto <@de
norikoeta hahaha@> I surmounted it by not being able to do it. By ridiculing it,
the laughter that accompanied this utterance underscores her powerful denial of het-
erosexual femininity. It is not simply that Marina does not conform to societys ex-
 SHIFTING SPEAKERS 105

pectations of feminine speech but that she cannot, and subsequently will not, impli-
cate herself in traditional femininity.
Marinas negotiative stance is molded by situational context and interpersonal
relationships. In the past, her negotiation of situational pressures such as teacher-
student power relations resulted in her use of watashi. In such a situation, contextual
pressure invokes dominant gender norms. This invocation does not signal Marinas
subjugation to those norms and the temporary suspension of her agency but, as her
account of gritting her teeth illustrates, manifests intentional linguistic negotiation.
As contexts develop, speakers respond to complex interactions of self, interlocutor,
formality, situation, and so on. Similarly, Marinas current usage of watashi in con-
versations with intimates indicates that, perhaps where there is no threat of being im-
plicated in heterosexist femininity, she is able to negotiate parts of her speech to
incorporate watashi. However, as her contradictory utterances imply, even though
Marina now allows herself to use watashi, it seems clear that when discussing her
true opinions she will avoid the term as much as possible.
Marinas metadiscursive narratives make it clear that the distaste she feels toward
watashi is anchored in the coercive heteropatriarchal femininity posited by Japanese
socioculture. At the same time, her narratives offer examples of situations where the
nonthreat of heteropatriarchal femininity enables her to use watashi. Marinas cur-
rent speech demonstrates the struggle between wider societys expectations of the
feminine (= heterosexist femininity) and her desire to express herself as Marina. She
negotiates the language terrain in front of her, her creative language use enabling
her not to become bogged down in the restrictive norms that lie there.
As briefly discussed in the beginning of this section, Marinas negotiative self-
referencing strategies are not shared by all the speakers who participated in the
roundtable discussion, dinner conversation, and follow-up interviews. Although there
is no space to examine them here, all the speakers employ self-referencing strategies
that are reflexive choices anchored in their current image of self and continuing lin-
guistic performance of gender and sexuality.

5.4. Conclusion

Heterosexism in the Japanese language poses a problem for people who feel that they
do not fit into this binary structure, lesbian and gay people foremost among them.
They are confronted with surroundings that ask them to be either heterosexual women
or heterosexual men. Given that homosexuality is understood effectively as an act of
transgenderism within this heteronormative system, this means that they are continually
forced, or at least pressured, to hide their sexual preference.8 A way of subverting this
pressure linguistically is to employ types of what is generally seen as transgendered
language, for instance using onee kotoba in the case of gay men or using transgressive
language, as in the case of Marinas persistent avoidance of watashi in self-reference.
People can master highly individuated language practices. Both Marina and Uraji
as children used the self-reference term attributed to the other sex and both were
shocked to be confronted with the fact that others did not see them as they saw them-
selves (in Urajis case) or that adults prescribed self-reference use according to gender
106 HISTORICAL AND THEORETICAL FOUNDATIONS

(in Marinas case). Marina employs various negotiative strategies to avoid using
watashi unless in situations of great contextual normative pressure, while Uraji uses
watashi and establishes himself as feminine in those contexts in which he wants to
be seen as such. It appears that Uraji met with less resistance than Marina, which
may be partly attributed to the fact that watashi is not exclusively feminine, whereas
boku is perceived as clearly masculine (but see chapter 14, this volume). Further-
more, Marinas avoidance of pronouns clashes with sociocultural expectations for
adult speakers to use pronouns rather than other reference terms (e.g., kinship terms
or nicknames) in nonintimate situations. It may also be that transgressive males are
perceived as less threatening than transgressive females.
Gender restrictions seek to normalize sexual relations between female and male
and, as part of that project, to prescribe differences in language use for women and
men. However, when we view the speech of real Japanese speakers, we witness the
extent to which prescription is never fully reflected in actual language use. Speakers
negotiations of language prescription and gender/sexuality norms intersect with their
sense of multiple selfhood and with contextual/situational pressures. This process
results in creative uses of language that may exploit available sociolinguistic and
cultural rules. In our discussion, we have focused on the intersections of gender and
sexuality within a social environment that conflates the two. In conclusion, we em-
phasize how important it is to incorporate notions of negotiation and speakers rela-
tionships with gender and sexuality norms, in future research.

Notes
We are sincerely thankful to all who collaborated in data collection and who have allowed
their words to be used here. Their comments, observations, and support were invaluable.

1. Togo maintains, based solely on his own imagination, that the term okama stems from
the Sanskrit karma and therefore is a beautiful term, which, again according to him, means love.
2. Miwa Akihiro (born in 1935) made his debut as a singer in 1952 and had his first
megahit, Mekemeke, in 1957. Since then, he has remained active as a television personal-
ity (Miwa 1992).
3. The development of varying strategies for self- and other-reference is a dynamic process.
For example, Uraji is referred to in the masculine here, because in the network of friends in
which he is known to one of the authors, he does not pass for female nor make any effort to do
so and is regarded as male, though obviously not a particularly masculine one. Here he is
foremostly known as Uraji, the person, and neither his gender nor sex is the object of special
scrutiny.
4. Although it is standard in sociolinguistic studies to provide detailed information about
the interlocutors, this study problematizes the simplistic mapping of language to social iden-
tity, and therefore no such supporting data is offered. See Maree 2002a for a full discussion
of this issue.
5. For scholarship that discusses self-reference and gender in Japanese see, for example,
Ide (1990) and Kanamaru (1993).
6. In this section, Oka refers to herself as being frank (atashi kekkoo sabasaba shitete
Im quite frank), as not having feminine characteristics ( joseeteki to iwareru yoo na yooso
wa nai dont have so-called feminine characteristics), and as not being feminine internally/
 SHIFTING SPEAKERS 107

mentally (naimen naimen ga sugoi joseeteki ja nai kara because Im internal-internally

very not feminine).
7. Transcription conventions are: : truncated word; [: simultaneous speech; =: latch-
ing; ?: rising intonation; .: falling intonation; ,: brief pause; -: extended sound or syllable;
<@@>: laughter; <Hx>: exhalation; <F F>: loudness; (): unintelligible or uncertain speech;
(1): pause in seconds.
8. The term preference here keeps open the question of whether people choose their
sexuality or not, as the word itself does not stipulate how the preference came about.

References
Abe, Hideko (1993). The speech of urban professional women. Unpublished Ph.D. disserta-
tion, Arizona State University, Tempe.
Bucholtz, Mary, A. C. Liang, and Laurel A. Sutton (1999). Reinventing identities: The gendered
self in discourse. Oxford: Oxford University Press.
Butler, Judith (1990). Gender trouble: Feminism and the subversion of identity. New York:
Routledge.
(1993). Bodies that matter: On the discursive limits of sex. New York: Routledge.
Endo, Orie (1997). Onna no kotoba no bunkashi (A cultural history of Japanese womens
language). Tokyo: Gakuyo Shobo.
Foley, William A. (1997). Anthropological linguistics. Oxford: Blackwell.
Frhstck, Sabine (1997). Die Politik der Sexualwissenschaft: Zur Produktion des sexologischen
Wissens in Japan 19081941 (The politics of sexology: On the production of sexological
knowledge in Japan 19081941). Vienna: Universitt Wien, Institut fr Japanologie.
Fukayuki, Shimoonu (1996). [Untitled]. In Kuia Sutadiizu Henshuu Iinkai (ed.), Kuia sutadiizu
96 (Queer studies 1996), 108109. Tokyo: Nanatsumori Shokan.
Furukawa, Makoto (1994). Sekushuariti no henyoo: Kindai nihon ni okeru dooseeai o meguru
mittsu no koodo (The changing nature of sexuality: The three codes framing homosexu-
ality in modern Japan). Nichibei Josei Jaanaru 17: 2955.
Fushimi, Noriaki (1991). Puraibeeto gei raifu: Posuto renairon (Private gay life: Post love
theory). Tokyo: Gakuyo Shobo.
Harvey, Keith, and Celia Shalom (eds.). (1997). Language and desire: Encoding sex, romance
and intimacy. London: Routledge.
Hirano, Hiroaki (1994). Anchi-heterosekushizumu (Anti-heterosexism). Tokyo: Pandora
Books.
Ide, Sachiko (1990). Person references of Japanese and American children. In S. Ide and
N. H. McGloin (eds.), Aspects of Japanese womens language, 4362. Tokyo: Kuroshio.
Kanamaru, Fumi (1993). Ninshoo-daimeeshi, koshoo (Personal pronouns and address terms).
In S. Ide (ed.), Joseego no sekai (The world of womens language), 1532. Tokyo: Meiji
Shoin.
Kondo, Dorinne (1990). Crafting selves: Power, gender, and discourse of identity in a Japa-
nese workplace. Chicago: University of Chicago Press.
Leap, William (ed.) (1995). Beyond the lavender lexicon: Authenticity, imagination, and
appropriation in gay and lesbian languages. Buffalo: Gordon and Breach.
Livia, Anna, and Kira Hall (eds.) (1997). Queerly phrased: Language, gender, and sexual-
ity. New York: Oxford University Press.
Lunsing, Wim (1997). Gay boom in Japan: Changing views of homosexuality? Thamyris:
Mythmaking from Past to Present 4(2): 267293.
(1999a). Japan: Finding its way? In B. Adam, J. W. Duyvendak, and A. Krouwel
108 HISTORICAL AND THEORETICAL FOUNDATIONS

(eds.), The global emergence of gay and lesbian politics: National imprints of a world-
wide movement, 293325. Philadelphia: Temple University Press.
(1999b). Life on Mars: Love and sex in fieldwork on sexuality and gender in urban
Japan. In F. Markowitz and M. Ashkenazi (eds.), Sex, sexuality, and the anthropologist,
175195. Urbana: University of Illinois Press.
(2001a). Between margin and centre: Researching non-standard Japanese. Copenhagen
Journal of Asian Studies 15: 81113.
(2001b). Beyond common sense: Sexuality and gender in contemporary Japan. Lon-
don: Kegan Paul.
Maree, Claire (1997). Jendaa shihyoo to jendaa no imisee no henka: Eega Shinjuku Boys ni
okeru onabe no baai (Gender indexicality and semantic shifts in gendered meanings: The
case of an onabe in the documentary film Shinjuku Boys). Gendai Shisoo 25(13): 263
278.
(1998). See no aru kotoba (Sexual words). Gendai Shisoo 26(10): 122135.
(2002a). See(sa) no gengotai: Kotoba no kojinshi kara yomitoru jendaa, sekushuariti,
fukugoo-identiti (The language ecology of gender: Personal narratives of gender, sexu-
ality, multiple identities). In H. Ishida and Y. Komori (eds.), Gengogaku to gengotai,
vol. 5: Shakai no gengotai Linguistics and language style, vol. 5, 117135. Tokyo: Tokyo
University Press.
(2002b). Nihongo to jendaa oyobi sekushuaritinegoshieshonjibun ga jibun [atashi,
boku, ore] de iru tame ni (Negotiation of Japanese language, sexuality, gendermain-
taining (my)self). Unpublished Ph.D. dissertation, University of Tokyo.
Miwa, Akihiro (1992). Murasaki no rirekisho (A purple curriculum vitae). Tokyo: Sui
Shoboo.
Ochs, Elinor, and Lisa Capps (1996). Narrating the self. Annual Review of Anthropology 25:
1943.
Ogawa, Naoko, and Janet S. Shibamoto Smith (1997). The gendering of the gay male sex
class in Japan: A preliminary case study based on Rasen no Sobyoo. In A. Livia and
K. Hall (eds.), Queerly phrased: Language, gender, and sexuality, 402415. New York:
Oxford University Press.
Okamoto, Shigeko (1994). Gendered speech styles and social identity among young Japa-
nese women. In M. Bucholtz, A. C. Liang, L. A. Sutton, and C. Hines (eds.), Cultural
performances: Proceedings of the Third Berkeley Women and Language Conference,
569581. Berkeley: Berkeley Women and Language Group.
Pflugfelder, Gregory M. (1999). Cartographies of desire: Male-male sexuality in Japanese
discourse 16001950. Berkeley: University of California Press.
Rich, Adrienne (1980). Compulsory heterosexuality and lesbian existence. Signs: Journal of
Women in Culture and Society 5(4): 631660.
Rubin, Gayle S. (1993). Thinking sex: Notes for a radical theory of the politics of sexuality.
In H. Abelove, M. A. Barale, and D. M. Halperin (eds.), The lesbian and gay studies
reader, 344. New York: Routledge.
Sedgwick, Eve Kosovsky (1990). Epistemology of the closet. Berkeley: University of Cali-
fornia Press.
Sunagawa, Hideki (1999). Nihon no gei/rezubian sutadiizu (Japanese gay/lesbian studies).
Queer Japan 1: 135153.
Ueno, Chizuko (1991 [1986]). Onna to yuu kairaku (The pleasure of womanhood). Tokyo:
Gakuyo Shobo.
Valentine, James (1997). Skirting and suiting stereotypes: Representations of marginalised
sexualities in Japan. Theory, Culture and Society 14(3): 5785.
 SHIFTING SPEAKERS 109

Vanbaelen, Ruth (1998). Dansee dooseeaisha no kotobazukai: Intabyuu no bunseki o chuushin

ni (Language use of male homosexuals: From interviews). Shakaigengokagaku Gakkai
Dai 2kai Kenkyuukai Yookooshuu (Synopsis of Papers from the 2nd Meeting of The Japa-
nese Association of Sociolinguistic Sciences, Kyoto University), 1722. Tokyo: The
Japanese Association of Sociolinguistic Sciences.
Za Gei Henshubu (1992). Daigokai Za Gei ankeeto (Fifth survey of The Gay). Za Gei, July:
4252.
This page intentionally left blank
 PART II

LINGUISTIC IDEOLOGIES
AND CULTURAL MODELS
This page intentionally left blank
 6

JANET S . SHIBAMOTO SMITH

Language and Gender in

the (Hetero)Romance
Reading the Ideal Hero/ine through Lovers
Dialogue in Japanese Romance Fiction

R omantic category fiction (love stories) provides a rich source of data about
womens and mens romantic expressivity. Such texts offer cultural models of how
heterosexual couples falling in love are represented as thinking and speaking about
their feelings. This chapter centers around a linguistic analysis of the dialogue be-
tween female and male protagonists in a set of nine romances drawn from the three
major subgenres of Japanese romance fiction.1 The goal of this chapter is not to de-
scribe how real Japanese women and men act as they embark on the projects of court-
ship and seduction but to begin to understand what cultural models are available to
at least some of themthat is, the largely female readership of romance novelsas
they set about falling in love. This chapter, then, brings issues of emotionality and
sexuality, specifically heterosexuality, into Japanese language and gender studies.
Or, more properly, since it is not my intention to conflate either a particular femi-
nine or masculine emotional condition or sexuality with gender, it brings these
issues into juxtaposition with Japanese language and gender studies in ways that
facilitate shifting the focus of our research from fixed gender identities to the role of
language in producing gendered subject positions in particular activities through
appeal to culturally grounded semiotic practice (as called for in a somewhat differ-
ent context in Kulick 2000:272273). The particular activity I address here is fall-
ing in (true) love.
Examining how heterosexual desire and romantic love are communicatedor,
to be more precise, how such feelings are represented as being communicatedis
the goal of this chapter. In particular, I argue that Japanese readers of romance fic-
tion respond to the sounds of true love as represented in the lovers dialogue

113
114 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

([Shibamoto]Smith 1999, Shibamoto Smith & Occhi n.d.). I argue that the dialogue
between the women and men encountered in the texts of my corpus constitutes a
primary site for the expression of one aspect of gendered language politics. I further
argue that the very gendered patterns in the spoken words of love in these novels can
tell us a great deal about prospective lovers as they are depicted as attractive and
attracted and about how they use language in dialogue with each other to convey
their femininity or masculinity. One of the messages Japanese romances convey is
where (in language) and how (by being maximally feminine or masculine) to
locate and enact attractiveness. This study, then, not only sheds a new light on how
language ideologies play out in the cultural construction of ideal women and men
engaged in an activity central to doing womanhood or manhoodheterosexual
pair-bondingbut also contributes significantly to issues of Japanese language and
gender through the investigation of the complex relationship of sex/gender/sexual-
ity to language ideologies and representations of language practices in a heretofore
unstudied context.
Along with other aspects of identity, the formalisms of love and desire, that
is, the frameworks for romance, are instances of socially-mediated, institutionally
regimented and regimenting, metapragmatic discourses . . . Whether implicitly or
explicitly these discursive frames indicate how persons should calculate and cali-
brate the stakes, pleasures and risks of being in the particular type of formed space
called the heteroromance (Povinelli 1999:3). In examining Japanese category romance
fiction for evidence of how women and men are held properly to engage in projects
of courtship and seduction, I begin to uncover some of the normative standards being
represented to which real women and men may make reference when they are in a
true love relationship. To borrow further from Kulick, who writes rather more di-
rectly about desire, being in love is signaled in culturally codified ways; and the
range of semiotic codes used to convey being in love are recognizable as convey-
ing desireor in the case of the present study, being in lovebecause they are
iterable signs that continually get recirculated in social life (Kulick 2000:273).

6.1. The data

I report here the results of a linguistic analysis of the dialogue between female and male
protagonists in nine representative novels, three from each subcategory, from a corpus
of novels that spans the three major subgenres of Japanese romance fiction published
in the period 19801999. All novels in the corpus center around stories of heterosexual
romance. The domestic romance field of Japanese mass literature is not as clearly di-
vided from other popular genres as it is in North America (Mulhern 1989), and although
a number of Harlequin-like romance series have been initiated since the 1979 intro-
duction of Harlequin romance translations to Japan, none has succeeded in becoming
a permanent part of the mass literature scene. Thus, the criteria for inclusion of novels
in this category are adjusted slightly from those used by Radway (1984) in her study of
American romance texts and readers. The status of the genre notwithstanding, writers
of romance fiction are well known and very prolific; and renai shoosetsu romance
novels have a history that stretches back to the early years of the Meiji period (1868
 LANGUAGE AND GENDER IN THE ( HETERO ) ROMANCE 115

1912; Noguchi 1987, Atsumi, Arimitsu, & Kobayashi 1991). Contemporary renai
shoosetsu fall into one of three subgenres: tales of blighted love (hiren), domestically
produced romances with happy endings (happii-endo), and Western, Harlequin-style
romances in Japanese translation. In previous research, I have looked primarily at the
metaphoric models of true love and found some differences between domestically
produced romances and translated Harlequin-type romances from the United States and
England (Shibamoto Smith 1997, 1999, forthcoming).
In this chapter I instead begin to elucidate the verbal process of identifying an
attractive prospective romantic partner and of presenting oneself as attractive. To
this end, dialogue between the romancing pair is extracted from each text and ana-
lyzed for the presence of gendered language forms. Verbal interactions between the
heroines and heroes of the novels in my study are traced from first meeting through
the happy (or unhappy) end to their romance in order to examine gendered patterns
in spoken words of love as they are inscribed in fictional form. I suggest, in fact, that
one salient way Japanese heroines and heroes signal heterosexual attractiveness is
through their verbal behaviors and that this stands in marked contrast to the Harlequin-
style translations, where attractiveness is depicted in other ways. In particular, I sub-
mit that it is through the intensified use of privileged forms of verbal femininity,
termed joseego womens language by many linguists and disseminated as part of
twentieth-century nationalist discourse (see chapter 3, this volume, chapter 4, this
volume), and its much less fixedbut fairly well knowncounterpart, mens lan-
guage, or danseego (Sturtz 2001, chapter 15, this volume), that this attractiveness
is most clearly portrayed. Joseego establishes a complex of features stereotypically
associated with women that stands in opposition to a complex of forms associated
with men. That this normative understanding of Japanese female and male personhood
excludes much of the citizenry of Japan is a point that many of the chapters in this
volume make abundantly clear; that the terms joseego and danseego themselves may
be part of that exclusionary project may be less clear, but to avoid the potential taint
of these labels, I here refer to these complexes of features as normative womens
language and normative mens language, respectively. The elements of these
gendered language norms/stereotypes examined in the romance text dialogues are
outlined in subsequent sections. The following analysis links dialogue found in ro-
mance texts to gendered language use as indexical of normative femininity and
masculinity and, hence, of heterosexual attractiveness.

6.1.1. Looking like a lover, Harlequin style

How, then, do the fictional women and men in my corpus of romances know that
they are the ones for each other? In particular, how are their sexualities, their at-
tractiveness, and their emotional appeal represented to the reader?
Radway (1984) describes ideal heroes and heroines for Western romances. The
Harlequin-style hero is spectacularly masculine, sexually (hyper)experienced, re-
served (or bad-tempered), and seemingly indifferent, but with some small feature
that prefigures tenderness or softness, such as a love of animals or small children. A
great deal is made of his masculine good looks, both on the book covers and in textual
descriptions. In example (1), one hero of a Japanese Harlequin translation, Mitcheru
116 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

Mitchell, is first encountered by Guen Gwen as she sits dazed in her automobile
after the two have collided on a narrow, winding mountain road.

(1) Dorodarake no kawa no buutsu ga me ni hairu. Hidoku ooki na ashi da. Sore wa kaaki-
iro no zubon ni tsuzuite, hosoi kawa no beruto no ue ni wa buruu no furanneru no hiroi
mune ga atta. Soshite sono ue ni wa, hiyake shita nodo no hifu to, chikarazuyoi ago.
Teire no ii kuroi hige wa, hageshiku musubareta kuchibiru no sen o kakushite inai. . . .
Tankisoo na hanasuji no ue de, tsuyoi hikari o tataeta me ga, Guen no shisen o toraeta.
Sono mabuta no iro wa . . . soo, yuugure no hikari no naka de kirameku sekitan no
katamari no yoo da. Sore mo, kyoo yama de mitsuketa yawarakai rekiseetan de wa naku,
motto kooshitsu no muentan
Muddy leather boots enter her line of vision. [On] very big feet. Following upon khaki
pants, [and] above a narrow leather belt, appeared a broad blue-flannel [shirted] chest.
Then, above that, a sunburned throat and a strong chin. A well-kept black beard [fails]
to hide the line of tightly compressed lips. . . . Above a sharp nose, bespeaking a quick
temper, flashing eyes captured Gwens gaze. The color of the eyes . . . yes, [they] were
like [the color of] lumps of coal glittering in the twilight. Not the soft bituminous coal
that [she] had found today in the mountains, but the harder anthracite. (Browning/
Nakagawa 1985/1987:8)

All heroes may be tall, but not all are dark; 31-year-old Chaaruzu Charles, the
new CEO of a major urban department store, is blond and something of a tease. He
is rich and powerful (ooganemochi de kenryoku mo aru), he dresses well (e.g., he
wears expensive-looking [kookasoo na ] suits and tasteful [shumi (no) ii] silk neck-
ties), and his resolve to push forward (tsukisusumu) toward the achievement of his
goals is a commanding aspect of his personality (rippa na mono da). Harlequin-style
heroes have munage chest hair, which peeks out from unbuttoned shirts and the
like, hard muscles (hikishimatta kinniku), narrow waists, nice hips, and, when the
occasion arises, impressive [evidence(s) of] arousal (takaburi [no shooko]).
Radway also provides a characterization of the ideal Western heroine: she is
unusually intelligent or unusually honest/moral/true or both; she may or may not have
a fiery disposition but is generally given to impulsive but well-intentioned actions;
she invariably has a childlike innocence and inexperience when it comes to love no
matter how professional or sophisticated she is otherwise. Although she need not be
virginalshe can, in fact, be a mother, like Kasandora Cassandra in (2b)she just
cannot truly have enjoyed sex before encounters with her true love. Above all, she
is possessed of an extraordinary but unselfconscious beauty (see ex. 2). Although
she may not start out this way, she will, by the end of each novel, dress pretty well,
too, maybe even somewhat provocatively, although this is usually contrived to be
accomplished without her agency.

(2a) Ooki na masshuruumu no yoo na booshi o kaburu to, Joojiina no haato-gata no kao
wa odoroku hodo hikitatte mieta. Nagai kuri-iro no kami o booshi no naka ni
takushikonde ita toki, genkan no beru ga natta. Nagai matsuge ni fuchidorareta Joojiina
no tankasshoku no mabuta ni fuan no iro ga yogitta.
 LANGUAGE AND GENDER IN THE ( HETERO ) ROMANCE 117

When she put on the large, mushroom-shaped hat, it brought out [the beauty of]
Georginas heart-shaped face to a surprising degree. As she tucked her long chestnut
hair up into the hat, the doorbell rang. An uneasy look filled Georginas light brown
eyes framed by long lashes. (Lawrence/Yamanouchi 1997/1998:5)

(2b) Kinkasshoku no mabuta to ii, shikkoku no nagai kami to ii, masa ni uri futatsu da.
Tada kaodachi wa to yuu to, Besanii no hoo wa kodomorashiku fukkura to shite i[ru]
. . . Ippoo, Kasandora no hossori to shita kao ni tadayou no wa yuubi na otona no
utsukushisa datta.
Whether it was their chestnut gold eyes or their long jet black hair, they were truly
two peas in a pod. Only their faces [differed, with] Bethany (Cassandras daughter)
having the full face of a child. . . . For her part, Cassandras delicate face had an ele-
gant, adult beauty. (Mortimer/Hirae 1993/1994:6)

True heroines, then, are marked by the unselfconscious, unaware possession of such
attributes as heart-shaped faces (haato-gata no kao), long chestnut hair (nagai kuri-iro
no kami) that flows down their backs like rivers of flame (honoo no kawa no yoo ni),
and elegant (yuubi na) rather than showy (hanayaka na) beauty (utsukushisa), the
latter being characteristic of false loves. They also, it is generally revealed some-
what later in the texts, have long, slender legs, tiny waists, and perfectly shaped breasts,
fleetingly visible through inadvertently opened blouse buttons or V-necked evening
gowns.

6.1.2. Doesnt s/he sound loverly?

Japanese ideal heroines and heroes are a bit different from their Western counter-
parts; Japanese romances rarely have covers that depict the principal characters, and
the text descriptions are by no means as detailed as in the Harlequin-style texts. Cloth-
ing is not, in my corpus, ever used as a tool of seduction, and no lovers furtive glances
at escaping chest hair or heaving breasts or quick but intoxicating looks at glimpses
of beautiful long legs through the slits of skirts help our Japanese lovers know that
they have found Ms. or Mr. Right.
Japanese heroes, when they are described, are generally depicted as more aver-
age than the Harlequin heroes; and, in fact, are often depicted as vulnerable in a va-
riety of ways, making mistakes, errors of judgmenteven, according to Mulhern
(1989), more mistakes than the heroines. Certainly Japanese heroes sometimes are
special; they are often rich, they may be tall or robust (e.g., possessing supootsu
de kitaeta ii taikaku a good physique built up by sports), and they may even seem
arrogant or sardonic. But virtually all heroes, even the tall, rich ones, have a sense of
humor; and most of them are just ordinary, nice men (ex. 3a), acting in relatively
ordinary or even awkward ways (ex. 3b).

(3a) Kushoo shita kao ga asaguroku, doko ka hitonatsukkoi kanji no suru seenen de aru.
His wryly smiling face was swarthy; he was a youth with an amiable look about him.
(Hiraiwa 1982:9)
118 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

(3b) Kare wa shashinka no donarigoe ni, tekipaki to ugoite iru. Totemo adaruto bideo o
otoshite atafuta shite ita otoko to dooitsu jinbutsu to wa omoenai.
He was moving briskly [in obedience to] the photographers barked orders. It was
totally hard to believe that this was the same man who got flustered over dropping an
adult video. (Yuikawa 1997:5455)

Heroes are often characterized as seenen youth(s), as in example (3a); they are
employees, as in (3b), as well as employers or independently wealthy men. Flattering
physical descriptions, while not absent from the domestic texts,2 are generally much
abbreviated. And descriptions are often less than flattering. In example (3a), for ex-
ample, the hero is described as swarthy as distinguished from tanned; this is not a
compliment (Ashikari 2000). Wealth, power, and social position are also not guaran-
teed our heroes, although there is a systematic glossing over of socioeconomic status.
Tooru, the young man in example (3b), who is described simply as otoko [the] man
for the first several chapters of his story, is a photographers apprentice, photogra-
phy being an interest he turned to when his parents cut off support after he was dis-
missed from college after eight years. Other heroes range from company employees
(sarariiman) to company owners, with a fair sprinkling of artists and men from the
katakana shokugyoo kai world of Western/contemporary/foreign occupations.3
None, however, match the Harlequin-style imports in their glorious physical specimens
of overtly sexual masculinity with money, position, and power to spare.
The Japanese heroine, too, as noted by Mulhern (1989) is neither socially iso-
lated, as Western heroines often are, nor sexually repressed. She is generally self-
confident, as in example (4), often enjoys a career, and frequently has had (or takes
during the course of the romance) lovers other than her true love. And, most im-
portant to my argument here, she is not necessarily a great, sexually compelling
beauty, as in example (5). Japanese heroines looks are no more often described than
are those of the heroes, and often in sketchy fashion. When heroines are described in
some detail (as in ex. 4b), they are as often described in terms of relatively realistic
self-evaluation, which suggests a more down-to-earth self-awareness on the part of
the Japanese heroine than is the norm for the Harlequin-style woman, with her shin-
ing beauty cloaked in a veritable fog of unselfconsciousness.

(4a) Yoofuku erabi wa, Chikako nari ni jishin o motte iru.

When it came to choosing clothes, Chikako had her own self-confidence. (Yuikawa
1997:8)

(4b) Kagami no mae ni tatta. Ookiku wa nai ga, warenagara ii kakkoo da to omoeru chibusa,
uesuto wa shimatte iru shi, koshi mo maamaa marukute onnappoi. Jiipan ga niau kara
to itte, soo suteta mon de mo nai to Sayaka wa moderu no yoo na poozu o totte mita.
She stood in front of the mirror. Not large, but nicely shaped breasts, even if she
says so herself; a small waist, and her hips are passably round and womanly. Think-
ing Just because jeans suit me, Im not so bad, Sayaka struck a models pose.
(Hiraiwa 1982:43)
 LANGUAGE AND GENDER IN THE ( HETERO ) ROMANCE 119

(5a) Sore ni shite mo irokebusoku de aru koto wa inamenai.

Even so, [she/Sayaka] could not deny that she lacked sex appeal. (Hiraiwa 1982:43)

(5b) . . . bijin taipu de mo nai. . . . Dare mo ga kookan o motsu, atama no yosa o hikerakasanai,
shitashimiyasui hodohodo no bijo. . . . Yuri wa, korera no jooken ni pittari atehamatte
ita . . .
[She] wasnt a beauty. . . . [She was someone whom] everyone liked, a so-so beauty
who didnt show off her intelligence and whom everyone found easy to be friends
with. . . . Yuri fit these criteria perfectly. (Hanai 1989:9)

But no matter that the ideal Japanese heroine is competent and socially fulfilled,
she wont be a romance heroine unless she falls in love and is fallen in love with.
She need not, readers may be relieved to know, be a great beauty; but to attract her
man, she will have to participate in the heterosexual marketplace in ways that will
signal her qualifications and her willingness to be partnered if she is, in the end, to
be happily partneredor, since one subgenre is the tragic love story, unhappily but
truly partnered. In Harlequin-style translations, this signaling is accomplished for
both women and men via a particular sexually compelling kind of physical beauty.
Women are completely unaware of it, as described earlier. For men, such signaling
involves rugged handsomeness, great wealth, and a superficial coldness that covers
raging sexual desires. The domestic romance texts in my corpus do not contain sig-
nals of this sort. Heroines are much more likely to be characterized by sunaosa, that
cheerful enthusiasm and unresisting participation in work or social activities described
by Peak (1989) as sought after in preschool children in Japan, than they are by great
beauty, andtruth be knownso are the heroes.
If the Japanese heroines and heroes of category romances are not glorious physi-
cal specimens of overtly sexual femininity and masculinity, if they are not lush, al-
beit unselfconscious and innocent, beauties and gorgeous men whose demeanor
situates them just this side of the sexual predator, where is their femininity or mascu-
linity being encoded? My preliminary survey of a small number of texts suggested
that the signaling of a prospectively successful heterosexual attractiveness/attraction
came less in the descriptions of the lovers than in the dialogue between them. One
way Japanese heroines and heroes signal heterosexual attractiveness may be through
their verbal behaviorsin particular, in their use of normative womens language
(see Shibamoto Smith 2003 for discussion of normative womens language) and
normative mens language (chapter 15, this volume).

6.1.3. Normative womens language/normative

mens language
I examined several elements of normative womens language and normative mens
language in the romance text dialogues. I report on two here: (1) first- and second-
person pronominal and other self- or other-referring forms and (2) sentence-final
particles.
120 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

Reference and address. First- and second-person referring forms have been one of
the centerpieces of the literature on language and gender in Japanese. Pronouns are
the most often cited gendered forms. As is widely reported in the literature, norma-
tive usage in standard Japanese would have it that women and men share the formal
first-person pronouns watakushi and watashi, although the contexts in which they
are used by speakers of either sex differ (Ide 1979, Shibamoto 1985, Shibatani 1990,
Shibamoto Smith 2003). In less formal contexts, male-speaker-associated first-
person pronouns are boku, ore, jibun, and washi, the latter two of which are some-
what specialized (but see chapter 11, this volume); female-speaker-associated
first-person pronouns are atakushi, atashi,4 and atai, which last form we also will
not encounter in my data. Table 6.1 summarizes these forms. It is crucial to the
goal of this volumeintended as a demonstration that ideology plays an impor-
tant shaping but also significantly constraining and exclusionary role in understand-
ing the relationship of sex/gender to language practiceto stress that the following
tables outline normative usage; data in chapters 5, 10, 11, and 14 all shows quite
different usages.
Second-person pronouns are shown in table 6.2; women and men share the for-
mal anata and the less formal, less classy anta. Men have two forms, kimi and omae,
that are relatively to very informal or intimate, as well as two very aggressive or vulgar
forms, kisama and temee. We should not expect to see any of these latter, since our
guys are busy loving women, not fighting with other men.
There are also conventions for the ways wives and husbands address each other.
Today most young couples call each other by their first names, without -san, or by
nicknames until they have children; however, older conventions still seem to obtain
to a reasonable degree in these novels, and these are shown in example (6). Since
our lovers are not parents, we should not expect to see forms for mother and fa-
ther in our corpus.

(6) Husband Wife Wife Husband

first name first name + san
kimi otoosan father
omae anata
okaasan mother
(from Lee 1976)

In sum, then, Japanese first- and second-person reference is significantly marked for
gender and for formality.

TABLE 6.1. First-person pronominal forms

Context
FormalInformal

Women watakushi watashi atashi

(atakushi) (atai)
Men watakushi watashi boku ore
( jibun) (washi)
 LANGUAGE AND GENDER IN THE ( HETERO ) ROMANCE 121

6.2 Second-person pronominal forms

Context
FormalInformal

Women anata anta

Men anata kimi omae kisama
temee
(anta)

Pronoun use is also, unlike English, characterized by the use of a zero form
wherever possible. Speakers also refer to others by titles, names, or other devices in
preference to second-person pronominal forms. We may conclude, then, that the
appearance of these forms is quite marked. And so it was of some interest to me to
note, in a small sample of romantic text material, relatively high frequencies of first-
and second-person reference compared to natural conversations. It seems reasonable
to suggest, then, that, among other gendered forms, terms of self-reference and address
may be serving to signal or represent heteronormative attractiveness or mutual attrac-
tion in these texts. The results of my quantitative analysis may be seen in tables 6.3
and 6.4, where overt first- and second-person reference frequency is measured.
Relatively high frequencies of overt first- and second-person reference are ob-
served in the texts. This is hardly surprising for the Harlequin novels, which we might
expect to follow a pattern that accommodates the grammatical requirements of En-
glish. But, especially in the case of first-person reference, we see pronouns and a
handful of other forms used at similarly high levels in the Japanese texts as well.
And they are used in unusual ways, as in example (7), where instead of the expected
zero first-person reference there is a repetition of watashi.

(7) Watashi, jihyoo dashita no yo. Watashi, Makoto-chan ni warui koto o shite shimatta n
da mono. Watashi wa . . . kono mama Makoto-chan to kao o awasete iku wake ni wa
ikanai no yo.
I submitted my resignation. I did a bad thing to Makoto, you know. I . . . [just] cant go
on facing him as before. (Kamata 1991:159160)

Comparisons of first-person reference frequency across genres and sex of

speaker5 (table 6.3) show that hiren blighted romance lovers produce slightly fewer
first-person references than other lovers. This result is entirely due to the fact that
men in blighted-love stories produce fewer first-person pronouns than men either in
happii-endo romances or Harlequin translations.
More interesting than the absence of bokus and ores in the blighted-love stories,
however, is the fact that the Japanese men in happii-endo novels produce as many
such forms as Harlequin men do, as in example (8).

(8) Tonari no onna to wa, heya ni yotta dake dakara, ore wa nani mo shinai. Ore wa ano
onna to wa, nani mo shinai. Ore, kokoro o irekaeru kara. Hoka no onna to wa, moo
zettai ni tsukiawanai. Ore, kore kara NHK dake ni suru.
122 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

TABLE 6.3 First-person reference by women and

men, by text type
Text Type Women Men

Hiren 42.9% 67/156 35.3% 79/224

blighted atashi 6 ore 38
romance FN 3 boku 36
watashi 53 jibun 4
other 5 kotchi 1
Happii- 43.8% 110/251 45.9% 144/314
endo atashi 1 ore 89
happy uchi 1 boku 46
ending FN 3 jibun 6
watashi 97 kotchi 1
other 8 LN 2
Harlequin 43.1% 88/204 42.9% 102/238
translations watashi 88 boku 100
jibun 2
All genres 43.4% 265/611 41.9% 325/776

FN = first name; LN = last name.

[I] just stopped by the apartment of the girl next door; I didnt do anything. I didnt do
anything with her [lit., with that woman]. Ill become a new man. [I] absolutely wont
go out with other women. Ill stick with just NHK [the Japan Broadcasting Corpora-
tion] from now on. (Kamata 1991:15)

There could be a number of reasons for this, and more qualitative analysis is required.
Heroines, unlike their partners, exhibit a similar pattern of first-person reference
across the three subgenres; although women in hiren novels produce very slightly
fewer first-person referents than those in happii-endo or Harlequin novels, the dif-
ference by no means reaches significance. There is, it should be noted, more vari-
ability across the texts in the hiren category, which suggests that more attention is
required to variability in what constitutes tragedy. And it is important to keep in
mind the close similarities of the happii-endo novels to the Harlequin translation,
our benchmark for high first-person pronoun use.6
Further, I suggest that the particular first-person referent forms chosen signal
something along the lines of heterosexual attractiveness or, at least, heterosexual
attentiveness. In the Harlequin translations, women refer to themselves as watashi,
men as boku; use is so close to categorical that one might speculate that corporate
guidelines as to appropriate pronoun choice were provided to translators, a specula-
tion that I was unable to verify through Harlequin Tokyo. In the other two subgenres,
women use watashi most of the time, with a sprinkling of the somewhat more feminine
atashi, uchia new pronoun used by girls (see chapter 14, this volume)and first
name, another self-reference practice associated with young girls (see chapter 5, this
volume). This sprinkling is a bit thicker in the hiren than in the happii-endo novels
(13.4% vs. 4.5%),7 which suggests that atashi, uchi, and first name, if used too often,
are bad news. Women in love should stick to watashi rather than turning to more
 LANGUAGE AND GENDER IN THE ( HETERO ) ROMANCE 123

feminine forms of self-reference. Certain kinds of hyperfemininity are, it seems,

not desirable (see chapter 6, this volume).
Men, however, never use watashi; even on a first and occasionally rather for-
mal meeting, men use a mixture of boku and ore, except for the Harlequin heroes,
who never use anything but boku. Choice between boku and ore seems to be some-
what context sensitive; except in the case of one happii-endo text from the 1990s,
the more masculine ore appears to be used more often in scenes of heightened emo-
tion or after a sexual relationship has been established. More ore use occurs in happii-
endo texts than in hiren texts (61.8% vs. 48.1%).8 Successful male lovers, apparently,
say it with ore.
A similar, though not identical, situation obtains for the second-person forms
(table 6.4). First, it seems most plausible to assume that, due to translation effects,
Harlequin heroines and heroes produce significantly more second-person forms
all pronounsthan do any other lovers.9 Second, women in the domestic genres make
significantly fewer overt second-person references than men. This tendency is highly
significant (c2 = 8.94, p 0.01). One might expect this, since second-person pro-
nouns are avoided by subordinate members of status-asymmetric dyads. It is hardly
surprising that the cultural model of male superiority over women is evidenced in
these texts, leading women to be represented in both blighted love and happy ending
texts as avoiding second-person pronouns, in just the way subordinate members of
dyads properly should. And throughout, the actual forms used, when they are used,
are consistent with the status-asymmetric forms held to be used by husband-wife
couples, as seen in (6). Since these are pre-husband-wife couples (except in the hiren

6.4 Second-person reference by women and

men, by text type
Text Type Women Men

Hiren 35.7% 51/143 46.3% 93/201

blighted anata 22 omae 23
romance pet name 1 kimi 24
FN 4 FN 38
LN-san 24 FN-san 2
Anata 4
LN-san 1
other 1
Happii- 28.0% 53/189 42.7% 112/262
endo anata 26 omae 15
happy FN-san 1 kimi 41
ending FN 1 FN 17
LN-san 16 FN-chan 1
LN-kun 1 an(a)ta 2
pet name 8 LN-san 35
other 1
Harlequin 58.1% 86/148 57.0% 135/237
translations anata 72 kimi 118
FN 12 FN 14
pet name 2 pet name 3
All genres 39.6% 190/480 48.6% 340/700
124 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

subgenre), the data suggest that first- and second-person reference prescriptions of
normative womens and mens language, at least, form a moderately productive site
for the encoding of attraction.
One form that deserves special mention is the male-speaker-associated pro-
noun omae. This form is not used by Harlequin heroes and is used somewhat spar-
ingly in the hiren and happii-endo texts (24.7% vs. 13.4%, respectively). This
difference turns out to be highly significant (c2 = 18.25, p 0.001). Omae is an
old-fashioned second-person pronoun and one with more condescending connota-
tions than the competitor kimi; heroes who hope for a happy ending are advised to
minimize its use.

Sentence-final particles. Like pronouns, sentence-final particles (SFPs) are central

to ideologies of gendered Japanese, despite the existence of diverse use in real speech
(Okamoto 1995; see also chapters 10, 13, and 15, this volume). They belong to a
larger category of sentence extenders. Japanese is a verb-final language, and sen-
tence extenders are elements that occur postverbally. As a group, sentence extenders
serve to index a speakers stance or attitude toward the propositional information
conveyed in the preverbal to verbal portion of an utterance. One type of sentence
extender is the sentence-final particle. Figure 6.1 provides a continuum of norma-
tively masculine to normatively feminine SFPs.10
The following examples illustrate some of the normatively gendered SFPs
drawn from my data; normatively feminine SFPS used by heroines are given in ex-
ample (9) and normatively masculine SFPs used by heroes in example (10).

(9a) Anata no kokoro no naka de utsu no.

you GEN heart GEN in at hit
Youll hit [me] in your heart.
(Mori 1986:170)

(9b) Masaka, hitori nan desu mono.

really one person NOM COP
Really [of course not, you silly], I live alone.
(Setouchi 1996:40)

(9c) Shun-chan, mata kita wa.

Shun again came
Shun, I came [back] again.
(Tachihara 1970:49)

zo ze da yo neg IMP na sa V + yo ne kashira no NP~AN mon(o) wa

na + yo

masculine relatively neutral relatively feminine

masculine feminine

6.1. Sentence-particle continuum

 LANGUAGE AND GENDER IN THE ( HETERO ) ROMANCE 125

(10a) Ato ikkagetsu da zo.

later one month COP
Its just one more month.
(Kamata 1991:35)

(10b) Totemo omoe- nai na.

utterly can think [that] NEG
I just dont think so.
(Setouchi 1996:194)

(10c) Dotchi ga katte nan da yo.

which SU selfish NOM is
Whos the one whos being selfish?!
(Hanai 1989:104)

Given the centrality of these SFPs in gendered language ideologies, coupled with
their saliencederived from their utterance-final positionand their availability for
inclusion in a much broader range of utterances than personal pronouns, I argue that
we should expect to find in the representations of the speech produced by proper
romantic heroines and heroes a high frequency of sentence particles, which, so long
as they are stereotypically gender-differentiated, will tell the reader as well as the
romantic heroine or heros prospective romantic partner that the speaker is an ap-
propriately feminine or masculine potential true lover. Thus, we may expect to find
that romances fictional pairs would display their attractiveness/attraction to each other
through both a very high use of interactional particles and a very feminine or mascu-
line selection from among particle options. Results of the first aspect of this analysis
are presented in table 6.5.
The immediate conclusion to be drawn is that our novels true lovers do, indeed,
produce a very high frequency of SFPs when in dialogue with each other. Women
use interactional particles more than men (c2 for women vs. men across all texts =
44.19, p 0.001), especially if they are foreign (that is, a Harlequin heroine, c2 =
47.16, p 0.001) or if their love is doomed (c2 = 18.80, p 0.001). Neither the at-

6.5 Sentence-final particles, by text type

Women Men

Text Type Year % SFP/All Utterances % SFP/All Utterances

Hiren 1980s 46.1% 106/203 33.6% 111/330

blighted 1990s 56.9% 83/146 38.9% 81/208
romance All 50.3% 189/376 35.7% 192/538
Happii-endo 1980s 32.1% 63/196 47.6% 109/229
happy 1990s 53.8% 177/329 42.1% 188/447
ending All 45.7% 240/525 43.9% 297/676
Harlequin 1980s 75.4% 144/191 53.8% 113/210
translations 1990s 66.8% 139/208 43.5% 123/283
All 70.9% 283/399 47.9% 236/493
126 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

tractiveness of the heroine nor her attraction to the hero is signaled by elevated fre-
quencies of SFP use relative to her hero in love affairs with happy endings (c2 = 0.32,
not significant). The profile of lesser overall SFP use for men seems to derive in part
from the relatively infrequent use of SFPs by hiren heroes (c2 for hiren vs. happii-
endo men and Harlequin men are 8.13, p < 0.01, and 15.19, p 0.001, respectively).
We recall that these heroes do not make much use of the most masculine first-person
pronoun, ore, either.
What has been reported up to this point is simply the frequency of SFP use.
However, claims of SFP use in the construction of womanly women and manly men
will not hold if the actual SFPs used do not conform to the ideological alignments of
certain particles with womanly women or manly men. It is, therefore, necessary to
look at the SFPs selected by our heroines and heroes. I examined a selected subset of
heroines use of strongly feminine SFPs (wa, noninterrogative no, mono, and yo,
along with the single occurrence in a 1996 hiren text of -te) measured against he-
roes use of a similar subset of masculine SFPs (zo, ze, da yo, dai, negative impera-
tive na, and certain other na-type SFPs). Results are presented in table 6.6.
The production rates of strongly feminine or masculine SFPs by our heroines and
heroes is most striking when compared to real speaking data. Okamoto (1995, 1996)
reports only 4.5% strongly feminine SFP use by the 10 students in her studythe women
who most closely resemble our heroinesand 18.1% for 10 older women. And young
men may use strongly masculine SFPs even less than this; Sturtz Sreetharan (chap-
ter 15, this volume) reports 0% strongly masculine SFP use in her Kansai data and
only 5% in the Tokyo data. Of course, the contexts of speaking in these studies are
quite different from the contexts in which fictional lovers encode their mutual attrac-
tion. Nonetheless, the differences are striking; these SFPs are clearly used much more
to construct gendered images in fiction than they appear to be in real life.
Table 6.6 also shows that our heroines are, across the board, constructing femi-
ninity with feminine SFPs much more than their heroes are constructing masculinity
with masculine SFPs (c2 = 126.23, p 0.0001). This may simply have to do with the
analytic choice to examine only SFPs rather than including a larger set of sentence-
final forms. But these results are also in line with Eckert and McConnell-Ginets
argument that women have to work harder at inhabiting social categories than do
men (Eckert & McConnell-Ginet 1995).
Another point with respect to feminine SFP use is that happii-endo women use
significantly fewer than do hiren and Harlequin women (c2 = 13.19 and 20.58, respec-
tively, p 0.001). This, in a more detailed comparison across the two decades of this

6.6 Feminine and masculine sentence-final

particles, by text type
Womens Use of Mens Use of
Strongly Feminine Strongly Masculine
Text type % SFPs/All Utterances % SFPs/All Utterances

Hiren 69.8% 132/189 29.2% 56/192

Happii-endo 52.1% 125/240 30.0% 89/297
Harlequin 71.7% 203/283 38.6% 91/236
All 64.6% 460/712 32.6% 236/725
 LANGUAGE AND GENDER IN THE ( HETERO ) ROMANCE 127

study, turns out to be due to very low use of feminine SFPs in the novels from the 1980s
(22.2%); happii heroines in the texts from the 1990s produced feminine SFPs at rates
very similar to heroines in the other subgenres (62.7%). And, although it did not create
a significant difference in the production of masculine SFPs by genre overall, this in-
crease in the femininity encoding via SFPs seen in the happii women is echoed by a
large increase in the masculinity encoding work via SFP done by their men (from 10.1%
in the 1980s to 41.5% in the 1990s). What accounts for this major shift is as yet unclear.
It seems, then, that SFPs are less a part of the verbal package of masculinity in
romance novels than of femininity.11 Whether the larger group of sentence extenders
such as evidentials and the like play such a role for heroes as well as heroines, how-
ever, must still be investigated further. And it is also possible that bare sentences,
particularly those that end in plain verbal forms, which were not included in table 6.6,
may be precisely where masculinity is located. When utterances that end in da is,
the plain nonpast form of the copula to be that is strongly associated with masculin-
ity, are added to the preceding numbers, men may be seen to be constructed as manly
at sentences end in ways more similar to womens construction as feminine than
they appear when SFP forms are singled out as the only possible resource for so doing.
Adding da, bald imperatives, and other forms strongly associated with masculine
speech to the mixalthough beyond the scope of this workshould demonstrate
that heroes, too, use the ends of sentences to hint at their masculinity and their at-
traction to the heroine.
The use of interactional particles does, nonetheless, add significantlyand sa-
lientlyto the hypergendered nature of the dialogue between each of the romancing
pairs and, coupled with their pronoun use, indexes the interactions between the pairs
in each novel as love talk. Love talk, I suggest, plays a role in both Japanese and
translated category romances similar to that of looking love(r)ly (that is, having an
appropriately sexually appealing physical appearance) in the Western trash romance.

6.2. Conclusions

Japanese women who read romance fiction read romantic narratives built around
women and men who are suitable candidates for being true lovers. In a penetrat-
ing analysis of Japanese womens magazines, Rosenberger (1995:143144) argues
that print consumed by women has some power over the way women categorize
themselves and their sexual desires for particular Others. Readers establish a dis-
course of ideas and practices that, while not always coherent within itself, presents
an alternative to the discourse of ideas and practices available to them through domi-
nant institutions and, in doing so, construct imaginary communities to which they
may belong (see chapter 7, this volume). This is equally true of romance novels. It
is important, therefore, to understand how the heroines and heroes of these novels
are presented for reader consumption. One of the messages romances convey is where
to locate and enact attractiveness. In the Japanese category romance, beauty is in the
ears, not the eyes, of the about-to-be-significant romantic Other.
Being in love is signalled in culturally codified ways, and the range of semiotic
codes used to convey being in love are recognizable as being in love because they
128 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

are iterable signs that continually get recirculated in social life (Kulick 1999:10).
Japanese category romance fiction provides evidence of how women and men are
held properly to engageverbally, at leastin being in love. My analysis uncovers
some of the normative standards to which real women and men may make reference
when they judge whether they are or are not in a true love relationship and provides
us with a glimpse of how ideologies of gender and ideologies of heterosexuality re-
late to each other as they are simultaneously manifested in language.

Notes
This research was supported through a 19992000 University of California Faculty Devel-
opment Award and a 20002001 University of California Faculty Research Grant. I would
like to thank Debra Occhi, Cindi Sturtz Sreetharan, and, most of all, my coeditor, Shigeko
Okamoto, for their many helpful comments on this work.

1. See [Shibamoto] Smith and Schmidt (1996) and Shibamoto Smith (1999) for a de-
scription of the major subgenres and the metaphoric models of true love found therein.
2. See, for example, Norikos first glimpse of Goo: Kochira e yatte kuru futari no otoko
o mitsuketa. Hitori wa se ga takaku wanryoku ga tsuyosoo de, ugoki no hayai ashi to, arakezuri
na kokkaku o motta seenen [de aru]. [I] caught sight of two men coming this way. One
[was] tall and muscular-looking, a fast-walking young man with a rough-hewn physique
(Tanabe 1978:114).
3. The term comes from the fact that words for such occupations are written in katakana,
the Japanese syllabary used to encode foreign words.
4. Uchida (1997:89) notes a change since the 1970s; atashi is apparently now gaining
ground on or even superseding watashi for use in all but the most formal contexts in standard
Japanese (cf. chapter 14, this volume).
5. Comparisons were made using 2 x 2 c_ tests for homogeneity, each with a single degree
of freedom, with Yatess correction.
6. It is worthwhile to inquire whether this is the only result of translation effect in the
case of the Harlequin texts or these forms function even further to heighten the (hyper-)
sexual attractiveness of the Harlequin lovers, but this question must remain for future
investigation.
7. c2 = 3.38; thus, this finding is significant at the .07 level or, perhaps more prop-
erly, suggestive of a real but not fully characterizable textual meaning.
8. Oddly, here, too, c2 = 3.38; the similarity between this resulta tendency to-
ward hypermasculine self-reference in happy romances for men, one might sayand the
tendency for hyperfemininity to associate with bad outcomes is interesting but, for the
moment, inexplicable.
9. c2 = 7.98, 14.6 for the women and men of Harlequin vs. hiren texts and 22.45, 19.18
for the women and men of Harlequin vs. happii-endo texts across all utterances. In the first
case, p 0.01; in all the rest, p 0.001.
10. Abbreviations are as follows: IMP = imperative, neg = negative, NP = noun phrase,
AN = adjectival noun, NOM = nominalizaer, COP = copula, and SU = subject.
11. This is, perhaps, to be expected, as the problems of the hypermasculine SFPs si-
multaneous association with masculinity and with an undesirable working-class identity make
them less available for use in a linguistic construction of male attractiveness than their coun-
terpart hyperfeminine SFPs.
 LANGUAGE AND GENDER IN THE ( HETERO ) ROMANCE 129

Source Texts
Browning, Dixie/Nakagawa, Reiko, tr. (1985/1987) Something for herself/Tomadoi no kisetsu
(The lost season). Tokyo: Harlequin.
Hanai, Aiko (1989). Futari jikan (Time for the two of us). Tokyo: Shueisha.
Herter, Lori/Minami, Rutsu, tr. (1996/1998) How much is that couple in the window?/Shoo
uindoo no hanayome (The show window bride). Tokyo: Harlequin.
Hiraiwa, Yumie (1982). Kekkon no toki (Time for marriage). Tokyo: Kodansha.
Kamata, Toshio (1991). Koishite mo (May I love you?). Tokyo: Kadokawa.
Lawrence, Kim/Yamanouchi, Fumie, tr. (1997/1998). Wedding-night baby/Kekkonshiki no
yoru ni (On our wedding night). Tokyo: Harlequin.
Mori, Yoko (1986). Onnazakari (In the prime of her life). Tokyo: Kadokawa.
Mortimer, Carole/Hirae, Mayumi, tr. (1993/1994). Hunters moon/Kiseki no uedingu (Miracle
wedding). Tokyo: Harlequin.
Setouchi, Jakucho (1996). Kawaku (Thirst). Tokyo: Kodansha.
Tachihara, Masaaki (1970). Takiginoo (Torchlight Noh). Tokyo: Kadokawa.
Tanabe, Seiko (1978). Iiyoru (Approach). Tokyo: Bungei Shunju.
Yuikawa,, Kei (1997). Kisu yori mo setsunaku (Crueler than a kiss). Tokyo: Shueisha.

References
Ashikari, Mikiko (2000). Urban middle-class Japanese women and the white faces:
Gender, ideology and representation. Unpublished Ph.D. dissertation, University of
Cambridge.
Atsumi, Takako, Takashi Arimitsu, and Sachio Kobayashi, (eds.) (1991). Renai no kindai
bungaku (The modern literature of love). Tokyo: Sobunsha Shuppan.
Eckert, Penelope, and Sally McConnell-Ginet (1995). Constructing meaning, constructing
selves: Snapshots of language, gender and class from Belten High. In K. Hall and
M. Bucholtz (eds.), Gender articulated: Language and the socially constructed self, 469
507. New York: Routledge.
Ide, Sachiko (1979). Onna no kotoba, otoko no kotoba (Womens language, mens language).
Tokyo: Nihon Keizai Tsushinsha.
Kulick, Don (1999). Language & gender/sexuality. Language and Culture: Review Language
and Culture Symposium #6. Online: http://www.language-culture.org/colloquia/sympo-
sia/kulick-don/.
(2000). Gay and lesbian language. Annual Review of Anthropology 29: 243285.
Lee, Motoko Y. (1976). The married womans status and role as reflected in Japanese: An
exploratory sociolinguistic study. Signs: Journal of Women in Culture and Society 1(4):
991999.
Mulhern, Chieko Irie (1989). Japanese Harlequin romances and transcultural womans fic-
tion. Journal of Asian Studies 48(1): 5070.
Noguchi, Takehiko (1987). Kindai nihon no renai shoosetsu (Romance fiction in modern
Japan). Osaka: Osaka Shoseki.
Okamoto, Shigeko (1995). Tasteless Japanese: Less feminine speech among young Japa-
nese women. In K. Hall and M. Bucholtz (eds.), Gender articulated: Language and the
socially constructed self, 297325. New York: Routledge.
(1996). Indexical meaning, linguistic ideology, and Japanese womens speech. The
Proceedings of the 22nd Annual Meeting of the Berkeley Linguistics Society, 290301.
Berkeley: Berkeley Linguistics Society.
130 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

Peak, Lois (1989). Learning to become part of the group: The Japanese childs transition to
preschool life. Journal of Japanese Studies 15(1): 93123.
Povinelli, Elizabeth (1999). Subject of desire. Language and Culture Review: Language and
Culture Symposium #6. Online: http://www.language-culture.org/archives/mailing-lists/
l-c/199911/msg00005.html.
Radway, Janice A. (1984). Reading the romance: Women, patriarchy, and popular litera-
ture. Chapel Hill: University of North Carolina Press.
Rosenberger, Nancy (1995). Antiphonal performances? Japanese womens magazines and
womens voices. In L. Skov and B. Moeran (eds.), Women, media and consumption in
Japan, 143169. Honolulu: University of Hawaii Press.
Shibamoto, Janet S. (1985). Japanese womens language. New York: Academic Press.
Shibamoto Smith, Janet S. (1997). Koishi, Aisuru: Competing images of Japanese men in
love. Paper presented at the American Anthropological Association 96th Annual Meet-
ing, Chicago.
(1999). From hiren to happii-endo: Romantic expression in the Japanese love story.
In G. B. Palmer and D. J. Occhi (eds.), Languages of sentiment: Pragmatic and concep-
tual approaches to cultural constructions of emotional substrates, 147166. Amsterdam:
John Benjamins.
[Shibamoto] Smith, Janet S. (1999). The speaking patterns of Japanese true love: Loves
dialogue as seen in the romance novel. Paper presented at the Talkin Gender and Sexu-
ality Symposium, Aalborg, Denmark.
Shibamoto Smith, Janet S. (2003). Gendered structures in Japanese. In M. Hellinger and
H. Bussmann (eds.), Gender across languages, vol. 3, 201225. Amsterdam: John
Benjamins.
(forthcoming). Translating true love: Japanese romance fiction, Harlequin-style. In
J. Santaemilia (ed), Gender, sex and translation: The manipulation of identities. Manches-
ter: St. Jerome Publishing.
[Shibamoto] Smith, Janet S., and David L. Schmidt (1996). Variability in written Japanese:
Towards a sociolinguistics of script choice. Visible Language 30(1): 4671.
Shibamoto Smith, Janet S., and Debra J. Occhi (n.d.). It sounds like love to me: The speaking
patterns and mimetics of Japanese true love in romance fiction and manga. Unpub-
lished manuscript.
Shibatani, Masayoshi (1990). The languages of Japan. Cambridge, MA: Cambridge Univer-
sity Press.
Sturtz, Cindi L. (2001). Danseego da zo! Japanese mens language: Stereotypes, realities,
and ideologies. Unpublished Ph.D. dissertation, University of California, Davis.
Uchida, Nobuko (1997). Kaiwa koodoo ni mirareru seesa (Sex differences seen in conversa-
tional interaction). In S. Ide (ed.), Joseego no sekai (The world of womens language),
7493. Tokyo: Meiji Shoin.
 7

MOMOKO NAKAMURA

Lets Dress a Little Girlishly!

or Conquer Short Pants!
Constructing Gendered Communities
in Fashion Magazines for Young People

L ike all forms of mass communication, fashion magazines for young people use
language to construct the identity of their readership. This identity is both imaginary
and specific. Texts for mass audiences such as magazines often assume target readers
and use language so as to communicate to them. In the process, the imaginary reader is
constructed as having particular interests, values, and experiencesthat is, as having
a particular identity. Actual readers have choices to make about whether they will use
the identity of an imaginary reader as a resource to construct their own identities. In
order to compete with other fashion magazines, furthermore, each magazine needs to
make very specific the identity of its imaginary reader. This chapter focuses on ques-
tions of how magazine producers construct specific feminine and masculine identities
of imaginary readers and how magazine discourse makes these imaginary identities
meaningful to actual readers. Focusing on the notion of a magazine community, I
demonstrate that magazine discourse constructs gendered communities by incorporat-
ing stereotypical gender organizations and conceptual frameworks.
The data is drawn from two fashion magazines for young women, Junon and
Non-no, and two fashion magazines for young men, Popeye and Mens Non-no. Junon
started in 1973 and had a circulation of 380,000 in 1999. Popeye started in 1976 and
had a circulation of 220,000 in 1999 (Media Research Center 2000). Non-no started
in 1971 and had a circulation of 1,040,000 in 1999. Its mens counterpart, Mens
Non-no, started in 1986 and had a circulation of 370,000 in 1999 (Media Research
Center 2000). They were selected because fashion is one of the few topics shared by
both womens and mens magazines. The data comes from issues of these magazines
published in 1999 and 2000. I first analyze one article on hairstyles in Junon (March

131
132 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

1999; fig. 7.1) and one article on hair removal in Popeye (February 1999; fig. 7.2),
both on the topic of hair. By focusing on similarities between womens and mens
magazines, I will present three strategies magazine writers employ in constructing
magazine communities. I then compare the captions that appeared on the fashion pages
in Non-no (May 2000) and Mens Non-no (June 2000). The comparative analysis
will reveal how gender ideologies are used, and reproduced, to engender magazine
communities.

7.1. A dynamic model of language

and gender studies

Language and gender studies have developed dramatically by integrating the recent
feminist conception of gender not as predetermined by essential features but as a
dynamic process we actively perform in our interactions with others. The objectives
of the field have turned from how women and men use language differently to how
subjects construct gender identities in discourses. Women are redefined as active
language users, who reinvent identities as well as reproduce given ones (Bucholtz,
Liang, & Sutton 1999). The focus of analysis, too, has shifted from normative inter-
actions to subversive and resisting performances (Butler 1990), which transform
existing gender relations.
Along with the benefits of a conception of gender as constructed in discourses,
however, some problems emerge (Nakamura 2001:123124). The first problem is
related to the diversity of gender identities. The constructive view of gender denies
the dichotomous female/male distinction and emphasizes the diversity of female
identities. But if the differences among women are so wide, we are forced to ask
whether we should dispense with the single category of woman, which has united
the various forms of the feminist movement (Weedon 1997[1887]:177). The second
problem is how we can characterize the difficulties speakers face in their attempts to
subvert existing gender relations. To redefine women as active language users em-
phasizes the agency of each individual woman and assumes that the choice of
whether to reproduce or subvert gender relations is available to her. In order to
explore the difficulties of performing subversive practices, we need to consider
what Gal (1991:177) has called symbolic domination, ideological control on the
macrosocietal level. Finally, the third problem is how to integrate studies that focus
on different aspects of language and gender, such as sexist lexical structures (e.g.,
Miller & Swift 1979[1976]), the stereotypes and norms of womens language use
(e.g., Lakoff 1975), and sociohistorical dimension of gender construction (e.g., Fou-
cault 1976). Thus, a dynamic and integrative model of language and gender studies
is required, which will incorporate (1) a poststructural view of gender as constructed
in discourses and a process of subversive identity construction, (2) both the single
category of woman and a variety of female identities, (3) a model of ideological
control through discourse practices, and (4) both sociohistorical constructions of
gender representations and discursive constructions of gender identities.
In my previous work (Nakamura 1995:236), I reformulated Faircloughs model
of the dialectic relationship between social structures and discourse (1989:38) and
 LET S DRESS A LITTLE GIRLISHLY ! OR CONQUER SHORT PANTS ! 133

7.1. Junon, March 1999 p. 88. Courtesty of Shufu to Seikatsu Sha

claimed that those dialectic relationships are mediated by ideologies. If we are to

account for the language use of Japanese women, which, like the language use of all
speakers, constantly changes and varies, it is crucial to distinguish the stereotypes
and norms of womens language use from womens actual language use (Nakamura
1990:77). Womens language is then conceptualized as a set of abstract norms to
which actual usage may conform to a greater or lesser extent (Milroy & Milroy
1985:23), that is, the ideology of womens language (Silverstein 1979:1193). The
134 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

7.2. Popeye, February 1999 p. 114. Courtesy of Magazine House

 LET S DRESS A LITTLE GIRLISHLY ! OR CONQUER SHORT PANTS ! 135

ideology of womens language mediates between male dominance in social struc-

tures and Japanese womens discourse practices. It redefines the diversity in womens
language use as the result of individual womens active choices, that is, their varied
responses to the norms of womens language, as well as to many other norms re-
lated to the speakers age, occupation, sexuality, and ethnicity.
Thus, a model for language and gender studies that focuses on the gender aspect
of these dialectic relationships can be outlined (see fig. 7.3). Figure 7.3 indicates that
the dialectic relationships between gender relations and gender identities constructed
in discourse practices are mediated by gender ideologies. Gender identities refers to
a variety of gender-related identities that subjects actively (re)construct in discourse
practices. Gender ideologies include gender-related categories, social subjects, so-
cial relationships, and conceptual frameworks, which have been historically consti-
tuted and ordered by previous discourse practices. Gender relations refers to the
gender-related power structures in society.
The dialectic relationships between gender ideologies and gender identities con-
structed in discourse practices are represented by the two arrows and . In the di-
rection indicated by the downward arrow , gender ideologies both restrict and
provide resources to discourse practices. Gender ideologies restrict the ways indi-
vidual discourse practices are performed. At the same time, gender ideologies offer
categories, subject positions, social relations, and conceptual frameworks, which can
be used as resources in discourses (e.g., a Japanese speaker can perform a variety of
identities by using or not using feminine language precisely because the notion of
womens language exists in Japanese gender ideologies). In the direction indicated
by the upward arrow , discourse practices produce, reproduce, and subvert existing
gender ideologies, which, in turn, construct, maintain, and transform existing gen-
der relations. Gender ideologies in a particular society have been historically con-

Gender Relations (Social Structures)

Gender Ideologies (Discourse Orders)

[restrict and provide resources to] [produce, reproduce and subvert]

Gender Identities (Discourse Practices)

7.3. Dynamic model of language and gender studies (Nakamura 2002:29)
136 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

structed by discourse practices. We reproduce gender ideologies, for instance, by

taking subject positions provided in dominant ideologies (e.g., a woman may per-
form a feminine identity by following the norms of womens language). At the
same time, discourses subvert gender ideologies in the sense that no discourse prac-
tice is a perfect reproduction of ideologies (e.g., even if a woman intends to follow
the norms of womens language, her practice more or less deviates from the norms).
The transformation of gender ideologies ultimately subverts gender relations in social
structures. Reflecting the dynamic relation between ideologies and practices, I call
my model the dynamic model of language and gender studies.
The model resolves the three problems brought out by the new conception of
gender discussed earlier. First, the diversity of female identities is captured on the
level of discourse practices, while the norms, stereotypes, and categorizations ap-
plied to the category of woman are captured on the level of gender ideologies. For
example, the norms of womens language in gender ideologies are applied to all
womens language usethat is, to all members of the category of womansince,
on this level, women are believed to form a homogeneous group whose behavior is
determined by their female sex (Nakamura 1990:109). But on the level of discourse
practices, an individual woman responds to the norms differently depending on the
kind of identity she intends to construct.
Second, the difficulties concomitant with performing subversive discourse prac-
tices can be captured on the level of gender ideologies. We can examine the difficul-
ties by studying the contradictions among gender ideologies. For instance, the category
of womens language creates the twin standards of standard (male) language /
deviant (female) language and, for women only, good (feminine) language / bad
(unfeminine) language, which put women in the situation of a double bind (Lakoff
1975:61). The difficulties can also become evident through analyzing the relations
between gender ideologies and other ideologies (e.g., womens language is rein-
forced by other ideologies such as the beautiful Japanese tradition and the unique
character of Japanese language; Nakamura 1995:263).
Third, and finally, in the dynamic model the contributions of previous studies
on sexist lexical structures are incorporated into the analysis of gender ideologies. I
consider sexist lexicons the extreme case of dominant gender ideologies, ideologies
that have been naturalized enough to be represented in lexical structures. Sexist lexi-
cons provide resources for sexist discourse practices. Furthermore, by defining gen-
der ideologies as constituted by discourse practices the model productively connects
sociohistorical studies of the construction of gender ideologies to linguistic analyses
of discourse practices.
This chapter investigates how the model accounts for the process of inventing a
specific gender identity by analyzing magazine discourse. Identity construction often
has been conceptualized as the process by which a person constructs a variety of
gender identities by taking subject positions within gender ideologies (Gough &
Talbot 1996).1 A person constructs herself (or himself) as a feminine woman, for
instance, by speaking womens language, that is, by taking the subject position of a
feminine woman based on the distinction feminine woman / unfeminine woman and
the connection between feminine woman and womens language in gender ide-
 LET S DRESS A LITTLE GIRLISHLY ! OR CONQUER SHORT PANTS ! 137

ologies. Similarly, the process of creating a new identity is conceptualized as the

process either of resistance to taking a subject position or of combining subject po-
sitions (Fairclough 1989:39). A girl who speaks boys language, for example, is
constructing a new identity by resisting the subject position of a traditional girl asso-
ciated with womens language. In any case, subject positions are presupposed before
discourse practices are performed. In producing a youth fashion magazine discourse,
however, writers need to construct the specific identity of a Junon girl or a Popeye
boy. Instead of simply taking subject positions in gender ideologies, they have to
invent new identities not preestablished in gender ideologies. This chapter analyzes
how magazine discourse constructs a specific gender identity for the reader and how
the dynamic model accounts for the process.

7.2. Constructing magazine communities

Since magazine discourse communicates to mass audiences, its meaning varies depend-
ing on the variety of ways audiences receive magazine texts. Thus, magazine produc-
ers first make informed guesses about the targeted audience, that is, the implied reader,
and then set out to formulate their messages in the form most appealing to this imagi-
nary reader. The distinctive ways language is used in magazine discourse create the
identity of the imaginary reader. Furthermore, analyses of a British teen magazine
(Talbot 1992) and the U.S. Shopping Channel (Bucholtz 2000) have demonstrated that
media discourse often constructs an imaginary community, which offers potential con-
sumers membership in the community. In this section, I demonstrate that both womens
and mens magazines construct imaginary communities for their implied readers.

7.2.1. Constructing interactions between the writer

and the readers
The texts in youth fashion magazines for both women and men are full of linguistic
features that construct interactions between the writer and the readers. For example,
the use of informal style of speech, along with the frequent use of exclamation marks,
increases the sense of interaction.
Another way this effect is achieved is through the use of linguistic features that
presuppose imagined responses from the receivers. Typical examples are sentence-
final particles such as ne and yo ne, which have functions similar to tag questions.
Example (1) demonstrates how the addition of yo ne to the simple statement togechatta
implies responses such as yes or thats right from readers.2

(1) Kamisama, Moo Sukoshi Dake de wa odoroku hodo

God more a little only in TOP surprise extent
no henshin o togechatta yo ne (Junon, March 1999)
GEN transformation OBJ accomplished
In [the television show] God, Just a Little More, [actress Fukada Kyoko] has accom-
plished a surprising transformation of herself, hasnt she?
138 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

Interaction is also implied by the use of the imperative (ex. 2) and hortative (ex.
3) forms, which directly address the readers.

(2) Imperatives
hitsudoku! (Popeye, February 1999)
must-read-IMP
Required Reading!

(3) Hortatives
chokkotto suiito de ikoo! (Non-no, May 2000)
a little sweet by lets-go
Lets dress a little girlishly! [lit., in a sweet little way].

In addition to constructing interaction between the reader and writer, magazine

discourse involves the reader in the process of production by deleting whole clauses
for the reader to fill in.

(4) daisuki wanpiisu de odekake shitara . . . (Non-no, May 2000)

like-very-much dress in go-out DO-if
If [I] go out in my favorite dress, . . .

In example (4), the reader is expected to be able to supply the whole main clause, as
in: If I go out in my favorite dress, something wonderful will happen. Through the
act of completing the text, the reader transforms herself from a mere recipient of the
message into a co-participant in the production of the message.

7.2.2. Constructing the readers voice

Along with constructing interactions, magazine discourse actively constructs the
readers utterances. The readers voice is directly presented within double quotations,
as in example (5).

(5) dokusha minna kara no mane shitai to yuu

readers everybody from GEN imitate want-to QT say
atsuui rabu kooru (Junon, March 1999)
hot love call
[There have been] many phone calls [lit., hot love calls] from all the readers saying,
[I] want to imitate [her].

The readers voice is also indirectly implied as in example (6), where the term watashi
I indicates that the statement is represented as voiced by the reader.

(6) motto kyooun na watashi ni naru (Non-no, May 2000)

more good fortune I GOAL become
[I] become a more lucky person [lit., a more lucky myself].
 LET S DRESS A LITTLE GIRLISHLY ! OR CONQUER SHORT PANTS ! 139

Another indirect way to present the readers voice is to use the conversational
structure of adjacency pairs (Talbot 1992:182). Example (7), for instance, can be
interpreted as the writers answer to an unstated question from the reader, as shown
in example (8).

(7) kami o agetari, maku dake de

hair OBJ do-up curl only
soku otona no kao ni! (Junon, March 1999)
instantly adult of face to [become]
Simply by pinning up or curling your hair, you instantly look more mature!

(8) [Reader: How can I become more mature-looking?]

Writer: Simply by pinning up or curling your hair, you instantly look more mature!

By presenting only the response element of the adjacency pair question-response,

the text makes the reader infer the question and, thus, constructs an active role for
the reader as the utterer of the question.
The use of the desiderative morpheme tai want to and the verb hoshii want
in subject-deleted structures also produces readers voices. Hayashi (1997:368)
points out that tai is used only for the first person, since it expresses the speakers
own desire. In order to make coherent interpretations of examples (9) and (10),
therefore, the reader needs to interpret the deleted subject as the first-person I or
we.

(9) Desiderative morpheme tai want to + subject deletion

de motte, onna no ko ni mote -tai! (Popeye, February 1999)
and girls by be-liked want-to
And [I, we] want to be liked by girls.

(10) Verb hoshii + subject deletion

iro T-shatsu ga hoshii (Non-no, May 2000)
color T-shirt SUBJ want
[I, we] want a colored T-shirt.

The writer sometimes even speaks out on behalf of the readers. Example (11)
asks Dr. Izawa, a doctor in an aesthetic orthopedic clinic, how to remove hair.

(11) oshiete, Izawa-sensei (Popeye, February 1999)

teach Izawa doctor
Please teach [me, us], Dr. Izawa.

This has two possible readings depending on who utters it. One interpretation is that
example (11) is uttered in the readers voice (Teach me). But of whom may the
reader ask the question? Possibly the writer, because the writer, not the reader, is
assumed to know the expert. So the other interpretation of example (11) is that it is
140 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

uttered in the writers voice (i.e., Teach us). The two readings position the writer
as a representative of the readers (Hayashi 1997:369). In order to make the writer
a representative of the readers, moreover, example (11) must presuppose a commu-
nity of men who find value in hair removal.
The same effect is achieved by interrogative forms with unstated subjects. Most
interrogative forms in magazine discourse are not used to elicit answers from readers.
Rather, they are used to introduce the topic as a problem and to ask for the solution
of the problem, which will be given by the magazine. Example (12) can be uttered
both in the writers voice as a representative of the readers (we) or in the readers
voice (I).

(12) sate kono taimoo, doo ni ka shori

now this body hair somehow get-rid-of
dekinai mono ka !? (Popeye, February 1999)
DO Q
Now, how can [we, I] get rid of this body hair?

The strongest evidence for the construction of the magazine community, how-
ever, is the voice(s) of other readers.

(13) sore dare? tte kikareru hayamimi kyara o

this who QT ask PASS rapid-ear character OBJ
sagase! (Non-no, May 2000)
look for-IMP
Find a new animation character, [so that you will] be asked, Who is it?

Example (13) encourages the reader to find a T-shirt printed with a new animation
character and to wear it, so that people will be impressed and ask the reader who
the character is. Thus, Who is it? in this case is not a mere request for informa-
tion but conveys approbation for finding a new character so quickly. This question
and the attendant implied approval would only be voiced by someone who shares
with the reader the value in finding a new character, that is, the other members of
the Non-no community.

7.2.3. Constructing shared practices, knowledge, and values

These strategies of constructing interaction and the readers voice function to create
specific practices, knowledge, and values as shared by the writer and the readers. In
addition to these, magazine discourse has other strategies to accomplish a sense of
shared community. One is reference to other popular media. In example (1), the tele-
vision program God, Just a Little More is simply mentioned as if the readers of Junon
shared not only the practice of watching the series but also the knowledge and the
value of the television stars transformation.
In the same article, minna everybody (i.e., all the readers of Junon) is used to
create the same effect. It presupposes that the readers share an interest in Kyokos
hairstyle and a desire to imitate it, as in ima, minna ga daichuumoku na no ga Fukada
 LET S DRESS A LITTLE GIRLISHLY ! OR CONQUER SHORT PANTS ! 141

Kyoko-chan no kamigata Now, what everybody pays attention to is the hairstyle of

Fukada Kyoko; minna ga mane shitai Kyoko sutairu [the] Kyoko style everybody
wants to imitate. The readers even share the act of making a phone call to the writer
shouting, Mane shitai! [I] want to imitate [her]! as in example (5).
Another way knowledge is constructed as shared in magazine discourse is
through the mention of technical terms and the names of commodities without any
explanation. In the Junon article on hairstyle, technical terms such as beesu hea
base hair, shagii shaggy, reiyaa katto layered cut, and suraido katto slide cut
are simply mentioned as if all the readers already know what each term means.
Names of commodities necessary to create hairstyles, such as haabu uootaa herb
water, sutoreeto airon straight iron, and wakkusu wax, also appear abruptly,
as if all the readers already possess them, know how to use them, and actually use
them daily.
New expressions and novel uses of existing words also increase the sense of
shared knowledge. For instance, the ideas expressed by the compound nouns iro
T-shatsu a colored T-shirt in example (10) and hayamimi kyara rapid-ear charac-
ter in example (13) are usually expressed by relative clause + noun constructions,
such as iro no tuita T-shatsu a T-shirt that has color or dare mo mitsukete inai kyara
an animation character nobody has found. By being expressed in compound nouns
without explanation, the ideas are presented as presupposed categories. Iro T-shatsu
establishes colored T-shirts as one category in the semantic field of T-shirts. Hayamimi
kyara establishes new animation characters as a category distinctive from other char-
acters already printed on T-shirts. These categories are not known to those who do
not read the magazine. Accordingly, they serve as secret codes that unite members
of the magazine community. The process of interpreting these new expressions is a
process of learning and aligning oneself with these secret meanings of the magazine
community.
It is worth noting here that magazine discourse positions the members of the
community in different ways. In the fashion magazine communities I analyzed, tele-
vision-stars, fashion models, and celebrities are positioned as core members. They
represent community membership in its idealized form. Readers are positioned as
more or less peripheral members who are oriented toward the community and are in
the process of learning its practices (Wenger 1998:167). The writer is positioned as
the mentor who mediates between the core members and the readers. An actual reader
who reads a magazine for the first time may position herself or himself as a periph-
eral member, while others, who have read the magazine before, may position them-
selves closer to the core members. The simulation of heterogeneity gives a sense of
progress toward increased membership in the community and personal improvement
on the part of the actual readers.
Therefore, the community constructed in the discourse of fashion magazines for
young people is an imaginary interactive community. Constructing magazine com-
munities creates a set of membership identities, a process that transforms the infor-
mation in the magazines into resources for constructing those identities. In this way,
a T-shirt is transformed from just a piece of clothing into a symbol of membership in
a particular magazine community. A magazine community provides identity resources
to the actual readers in the form of membership.
142 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

7.3. Gendering magazine communities

Thus far, I have shown the similar strategies of both young womens and mens fash-
ion magazines in creating a community. This section examines the differences be-
tween womens and mens magazine discourses and analyzes how stereotypical
gender ideologies are incorporated into them. I use as data the captions on the cover
pages, the content pages, and the first 57 pages (except advertisement pages) of Non-
no and Mens Non-no. The total number of caption sentences analyzed is 318 from
Non-no and 264 from Mens Non-no.
The analysis shows that magazine communities are gender-differentiated in the
ways that community members are organized and participate in the community and
in the ways that magazine communities conceptualize the topic of fashion and posi-
tion their members.

7.3.1. Gendered organization of community members

Non-no and Mens Non-no differ from each other in several respects. Some of the
differences are absolute (some styles are restricted to either Non-no or Mens Non-
no), but many of the differences are a matter of frequency (one uses a particular style
more often than the other). These relative differences combined together, however,
make the magazines very different.
The discursive styles that I have observed much more often in womens maga-
zines than in mens magazines are (1) exclamation marks; (2) the use of the sentence-
final particles da ne, da yo ne, and da yo rather than da, and (3) more frequent use of
hortatives and interrogatives and less frequent use of imperatives. These differences
contribute to constructing different kinds of relationships between the writer and the
readers, namely, egalitarian relationships in the womens magazines and hierarchi-
cal relationships in the mens magazines.
The differences found in the use of exclamation marks and sentence-final par-
ticles show that there is less interaction between the writer and the readers in mens
magazine communities. First of all, the number of exclamation marks is 123 (38.6%)
in Non-no but only 60 (22.7%) in Mens Non-no. As noted earlier, exclamation points
simulate the effect of informal interaction.
Sentence-final particles are also used differently, as shown in table 7.1, which
shows that the use of the simple declarative da form in Mens Non-no is more fre-
quent than in Non-no (4.1%, 1.8%), in part because Non-no prefers to add sentence-
final particles ne, yo ne (1.8%), and yo (2.8%). Mens Non-no does not add these
sentence-final particles at all (0%).3 Both ne and yo strongly require responses from
the interpreters (Maynard 1993:111). The addition of ne, yo ne, and yo after da in
womens magazines, therefore, creates a sense of interaction.
Similar differences are observed in the relationship between the readers and the
core members of the respective magazine communities. In Junon, the television star
Fukada Kyoko confesses that because my hair is curly, coarse, and full, I take the
greatest caution not to make my head look too big. Here even an ideal member of
the community suffers from the same mundane problems as the readers. The equal
relationship is established by emphasizing similarities. Mens Non-no, however,
 LET S DRESS A LITTLE GIRLISHLY ! OR CONQUER SHORT PANTS ! 143

7.1. The occurrence of sentence-final particles in Non-no

and Mens Non-no
Non-no (n = 318) Mens Non-no (n = 264)

da 6 (1.8%) 11 (4.1%)
da ne, da yo ne 6 (1.8%) 0 (0%)
da yo 9 (2.8%) 0 (0%)

describes a television star, Kimura Takuya, as the man who outshines everybody in
his love of T-shirts and wears T-shirts stylishly. Here his perfection and confi-
dence are emphasized, which foreground the distance and difference between him
and the readers.
The occurrences of hortatives, interrogatives, and imperatives are also different
between Non-no and Mens Non-no. Table 7.2 shows that Non-no uses more hortatives
and interrogatives, while Mens Non-no uses more imperatives. These differences
construct different interactions between the writer and the readers. The interactions
constructed by hortatives and interrogatives in Non-no, on the one hand, are mutual
and equal. They present the writer as a friendly sister of the readers (Talbot 1995).
The interactions invoked by imperatives in Mens Non-no, on the other hand, are
unilateral and hierarchical.4 Thus, the stereotypes of womens egalitarian relation-
ships and mens hierarchical relationships are incorporated into magazine discourses
in the different ways magazine communities are organized.
The different organizational relationships seen in womens and mens magazine
communities imply different forms of reader participation in these communities. The
reader of the womens magazines enters into the community, expecting friendly help
from an older sister. The reader learns the shared practices of the community in
which intimacy and similarity are fostered. The reader of the mens magazine, in
contrast, enters a community led by a confident superior. The reader will be ordered
to acquire the shared practices of the hierarchically structured community. Thus,
stereotypical gender ideologies are drawn into magazine discourses in the gendered
structures of participation in the womens versus the mens magazine communities.

7.3.2. Gendered frameworks for conceptualizing topics

Comparison of the womens and mens magazine discourses reveals that they adopt
distinctive conceptual frameworks of their primary topic: fashion. One of the major
sets of contrastive frameworks is that of in womens magazines

7.2. The occurrence of hortatives, interrogatives, and

imperatives in Non-no and Mens Non-no
Non-no (n = 318) Mens Non-no (n = 264)

Hortatives 10 (3.1%) 1 (0.3%)

Interrogatives 17 (5.3%) 2 (0.7%)
Imperatives 2 (0.6%) 11 (4.1%)
144 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

versus in mens magazines. These different frameworks posi-

tion the writer and the readers in different relationships.
Womens magazine discourse defines fashion as a difficult task of dressing well
imposed on the reader. Non-no frequently uses kokufuku get over (ex. 14), tsuyoi
mikata strong supporter, and kowai mono nashi nothing to fear. For the reader
who struggles to get over the task of dressing well, the magazine gives informa-
tion about what clothing can be her strong supporter (ex. 15) and what clothing
can make her feel there is nothing to fear (ex. 16).

(14) mitame nenrei kokufuku meiku juku (Non-no, May 2000)

look age get over make-up seminar
The Make-up Seminar to Look Appropriate to Your Age [lit., The Make-up
Seminar to Get Over Your Age (Which People Determine) Based on How You Look]

(15) sappari shatsu ga tsuyoi mikata (Non-no, May 2000)

tidy shirt SUBJ strong supporter
A tidy shirt is [your] strong supporter.

(16) kore sae areba kowai mono nashi (Non-no, May 2000)
this only BE-if frightening thing BE-NEG
[If I/you have] this, there is nothing to fear.

Here fashion is conceptualized as a set of tasks to fulfill and a burden or problem

to solve. The reader is positioned as a frightened, nervous learner who needs the
magazines help. The writer is positioned as an intimate, helpful sister who tells the
reader maruhi aitem top secret items and anshin aitemu relief items.
The framework of and the positioning of the reader as a ner-
vous learner in the womens magazine function to feminize the topic of fashion.
Women are born with female bodies, but they need to work and rework their bodies
in order to remain women, since ideologically womens bodies are always imper-
fect (Smith 1988:47). For the reader of the youth fashion magazine who is begin-
ning to enter the world of femininity, the conception of places fashion
magazines in a helpful, counseling genre. The egalitarian relationships found in the
womens magazine community reinforce this conception of womens magazine as
the genre of advice.
Mens magazine discourse, in contrast, conceptualizes fashion as something to
conquer and dominate as in examples 17, 18, and 19.

(17) shooto pantsu o kooryaku suru! (Mens Non-no, June 2000)

short pants OBJ conquer DO
Conquer short pants!

(18) shuchoo no aru T-shatsu de kokoroshite

assertion of BE T-shirt by carefully
mukaeutsu (Mens Non-no, June 2000)
shoot-to-repulse
 LET S DRESS A LITTLE GIRLISHLY ! OR CONQUER SHORT PANTS ! 145

Fight back by wearing a T-shirt that expresses the real you [lit., (We/You/I) shoot the
(attacking enemy) in the T-shirt with an assertion.]

(19) natsu no 3dai aitemu kanzen seiha (Mens Non-no, June 2000)
summer of 3-big items perfect conquer
The Perfect Conquest of Summers Big 3

These examples draw upon the framework of to conceptualize

fashion. The reader is positioned as a fighter or a soldier who is expected to conquer
fashion. The writer is a commander who gives the orders and leads the troops (readers)
into battle. By applying the framework, mens magazine dis-
course masculinizes fashion. The hierarchical relationships among the members of
the mens magazine community reinforce this conception of fashion, because armies
are the organizations most characterized by strict hierarchy.

7.4. Conclusion

My analysis demonstrates that (1) youth fashion magazine discourse constructs an

imaginary, interactive, and gendered magazine community and (2) the specific identity
of magazine readers such as Junon girls or Popeye boys is invented as member-
ship in the community. The notion of magazine community plays an important role
in two respects. First, it enables the writers to construct the specific identity of the
implied reader as a community member. Second, it encourages actual readers to refer
to community membership when they construct their own identities, by providing a
variety of membership positionings in the community and a sense of progress from
the margin to the center of the community. One of the ways to invent a specific gen-
der identity, therefore, is to construct a gendered community.
My analysis demonstrates that magazine discourse attempts to construct specific
reader identities, such as Junon girls and Popeye boys, which are defined by
consumption of specific styles of fashion as well as gender. These specific identi-
ties, however, cannot be constructed simply by taking the subject positions of young
women and young men predetermined by gender ideologies. In order to construct
identities specific to each magazine, magazine producers create a gendered maga-
zine community by drawing upon other resources contained within gender ideolo-
gies, that is, gendered social relationships and conceptual frameworks. In other words,
instead of directly taking preestablished subject positions for young people in gen-
der ideologies, magazine discourse invents a magazine community membership iden-
tity by using other stereotypical gender ideologies.
The result of my analysis supports the dynamic model discussed in the intro-
duction to this chapter in demonstrating that, even in inventing a new identity, some
aspects of gender ideologies are reproduced. My analysis shows that magazine dis-
course contributes to both reproducing and altering gender ideologies. It reproduces
stereotypical gendered organizations and conceptual frameworks in gender ideolo-
gies. At the same time, it transforms the subject positions of young women and men
preestablished in gender ideologies in the sense that it invents identities defined by
146 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

consumption of specific fashions. The dynamic model accounts for the mechanism
by which the invention of new identities often accompanies the reproduction of ste-
reotypical gender ideologies, rather than simply replacing them.

Notes
I am most grateful to Shigeko Okamoto, Janet Shibamoto Smith, and Mary Bucholtz for helpful
comments and to Janet Shibamoto Smith and James Placzek for assistance with English.

1. I distinguish identity constructed in discourse practices from subject position in gen-

der ideologies. The notion of gender identity, on the one hand, refers to an infinite array of
gender identities, constantly constructed in discourse practices. The notion of subject posi-
tion, on the other hand, refers to the gender identities that have been produced over a consid-
erable range of discourse practices and are ordered in gender ideologies. One can invent, create,
and improvise a variety of gender identities in discourse practices. But in order to be included
in gender ideologies as subject positions the identities need to be performed iteratively over
a variety of discourses.
2. Abbreviations for grammatical items are: BE (various forms of the verb be), DO (vari-
ous forms of the verb do), GEN (genitive marker), GOAL (goal marker), IMP (imperative
marker), NEG (negative marker), OBJ (object marker), PASS (passive marker), Q (question
marker), QT (quotative marker), SUBJ (subject marker), and TOP (topic marker).
3. This does not mean that these sentence-final particles never occur in mens magazines.
In fact, the clinic doctor in the hair removal article in Popeye used ne seven times. The use of ne
functions to shorten the distance between the middle-aged doctor and young readers.
4. Construction of the less interactive and more hierarchical relationships in mens
magazines may be considered to contradict the notion of community. Tannen (1993:170)
correctly points out, however, that hierarchy does not necessarily preclude closeness and that
the Japanese relationship is characterized by a hierarchical interdependency; i.e., the mem-
bers lower in rank are protected by the members higher in rank and they unite to cope with
odds (Hayashi 1997:362).

References
Bucholtz, Mary (2000). Thanks for stopping by: Gender and virtual intimacy in American
shop-by-television discourse. In M. Andrew and M. M.Talbot (eds.), All the world and
her husband, 192209. London: Cassell.
Bucholtz, Mary, A. C. Liang, and Laurel A. Sutton (eds.) (1999). Reinventing identities: The
gendered self in discourse. New York: Oxford University Press.
Butler, Judith (1990). Gender trouble: Feminism and the subversion of identity. New York:
Routledge.
Fairclough, Norman (1989). Language and power. London: Longman.
Foucault, Michel (1976). The history of sexuality. New York: Random House.
Gal, Susan (1991). Between speech and silence: The problematics of research on language
and gender. In M. di Leonardo (ed.), Gender at the crossroads of knowledge: Feminist
anthropology in the postmodern era, 175203. Berkeley: University of California Press.
Gough, Val, and Mary Talbot (1996). Guilt over games boys play: Coherence as a focus
for examining the constitution of heterosexual subjectivity on a problem page. In
 LET S DRESS A LITTLE GIRLISHLY ! OR CONQUER SHORT PANTS ! 147

C. Caldas-Coulthard and M. Coulthard (eds.), Texts and practice: Readings in critical

discourse analysis, 214230. London: Routledge.
Hayashi, Reiko (1997). Hierarchical interdependence expressed through conversational styles
in Japanese womens magazines. Discourse & Society 8(3):359389.
Lakoff, Robin (1975). Language and womans place. New York: Harper and Row.
Maynard, Senko K. (1993). Kaiwa bunseki (Conversation analysis). Tokyo: Kuroshio Shuppan.
Media Research Center (2000). Zasshi shimbun soo katarogu (Japans periodicals in print).
Tokyo: Media Research Center.
Miller, Casey, and Kate Swift (1979[1976]). Words and women: New language in new times.
Harmondsworth: Penguin.
Milroy, James, and Lesley Milroy (1985). Authority in language: Investigating language
prescription and standardization, 2nd ed. New York: Routledge.
Nakamura, Momoko (1990). Onna no gengo shiyoo kaitai: Shakai gengogaku ni okeru seisa
kenkyu kara (The deconstruction of womens language use: Critical review of socio-
linguistic studies on sex differences). Shizen, Ningen, Shakai 11:45123.
(1995). Kotoba to feminizumu (Language and feminism). Tokyo: Keiso Shobo.
(2001). Kotoba to jendaa (Language and gender). Tokyo: Keiso Shobo.
(2002) The dynamic model of language and gender studies. Shizen, Ningen, Shakai
32:126.
Silverstein, Michael (1979). Language structure and linguistic ideology. In P. R. Clyne,
W. Hanks, and C. L. Hofbauer (eds.), The elements: A parasession on linguistic units
and levels, 193247. Chicago: Chicago Linguistic Society.
Smith, Dorothy (1988). Femininity as discourse. In L. G. Roman and L. K. Christian-Smith
(eds.), Becoming feminine: The politics of popular culture, 3758. New York: Falmer
Press.
Talbot, Mary (1992). The construction of gender in a teenage magazine. In N. Fairclough
(ed.), Critical language awareness, 174199. London: Longman.
(1995). A synthetic sisterhood. In K. Hall and M. Bucholtz (eds.), Gender Articu-
lated, 143165. New York: Routledge.
Tannen, Deborah (1993). The relativity of linguistic strategies: Rethinking power and soli-
darity in gender and dominance. In D. Tannen (ed.), Gender and conversational inter-
actions, 165188. New York: Oxford University Press.
Weedon, Chris (1997[1887]). Feminist practice and poststructuralist theory. Oxford: Blackwell.
Wenger, Etienne (1998). Communities of practice: Learning, meaning, and identity. Cam-
bridge: Cambridge University Press.
148 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

LAURA MILLER

You Are Doing Burikko!

Censoring/Scrutinizing Artificers
of Cute Femininity in Japanese

A journalist writing for the New York Times (Kristof 1995) offered the following
description of a female elevator operator at the Mitsukoshi department store in Ginza:
The Voice is as fawning as her demeanor, as sweet as syrup, and as high as a dog
whistle. Any higher and it would shatter the crystal on the seventh floor. The author
is reporting on the high-pitched voice frequently considered a stereotypical feature
of the burikko, a derogatory Japanese label used to describe women who exhibit
feigned navet. The word is derived from the term buru to pose, pretend, or act
and the suffix -ko, used for child or girl, to mean something like fake child or
phony girl. This chapter will argue that the burikko designation does not simply
reflect the uniform affectation of a childlike persona but rather is primarily estab-
lished through situated social judgment elicited by a combination of speaker, recipi-
ent, and setting. A variety of linguistic and nonverbal phenomena are assessed as
constituting the makeup of a burikko performance. I will look at some folk percep-
tions and media representations of the burikko and will also provide examples of
socially contextualized occasions of talk in which burikko features in its evaluation.
One outcome of a burikko performance is that it downplays or masks the adult sexu-
ality of the woman doing it. An exploration of burikko supports a growing scholarly
interest in the interconnections between linguistic ideology and gender performativity
and also contributes to the literature on labeling practices.
According to Cherry (1987), the term burikko was invented by female singer
Yamada Kuniko on a television program in 1980. Whether or not this is true, by at
least 1981 burikko was commonly used in colloquial conversation, and the editors
of an encyclopedia of postwar culture provided it on their list of trendy new words

148
 YOU ARE DOING BURIKKO ! 149

that were popular that year (Sasaki et al. 1991:1049). It certainly reached mainstream
status by 1982, as evidenced by the song title for the debut single by the pop music
group Grease, Burikko Rock n Roll. Although use of the word these days has
decreased, Inoue (1986) found that during the years 19831984 burikko was one of
the most popular new coinages used among young people.
Burikko was one of several neologisms formed during the first half of the 1980s
through the suffixing of -ko. Yonekawa (1996:150) mentions gameko, coined in 1980
and clipped from gametsui ko a chintzy, money-grubbing girl (Kansai dialect). There
was also kidoko, from kidotte iru ko a girl who puts on airs, and yumeko, from yume
miru ko a girl who dreams, for a romantic girl or woman. Once burikko was estab-
lished, there was a gradual semantic broadening of the concept that allowed new
coinages, including majime burikko child who pretends to be serious and burikko
joshidaisee cutesy-coeds, clipped from joshi daigakusee female college students.
Most recently we find burikko kogyaru phony KoGal. Kogyaru originated as a media
term used for young women who have bleached hair, loose socks (knee-length socks
worn hanging around the ankles), and big shoes (Miller 2000). The term kogyaru is
most likely a clipped version of kookoosee gyaru high school girl, and is not de-
rived from the morpheme ko, which means small, as many folk etymologies claim.
Kogyaru has overtones of rebellious insolence and unchecked sexuality. The burikko
kogyaru is therefore not a real KoGal at all but only someone pretending to be naughty
and cool.
One problem connected to the burikko tag is that it has become synonymous, in
some writing, with young woman. When the subject of womens language arises, the
burikko label may worm its way into the discourse, especially among male writers.
In Inoues (1989) report on linguistic changes in the speech of women, he discusses
the shift in pronunciation of the syllable shi to si as a phenomenon named burikko
hatsuon burikko pronunciation. This label is used because only young women
employ the new pronunciation, but it has the effect of classifying all young women
as burikko.
Despite its use in numerous other contexts, the core meaning of burikko remains
that of a woman who displays bogus innocence. Now the frequent object of ridicule,
only a few decades ago the burikko was the natural product of the cult of cuteness so
deliciously described by Kinsella (1995) and McVeigh (2000). The reified cute aes-
thetic that Treat (1996:283) once termed a celebration of vapidness was perfectly
expressed by women who acted the part of the contriving maid. The master and pro-
totype for exemplary burikko style was 1980s singer Matsuda Seiko. Wearing the
hair of Gidget and the petticoats of Marie Antoinette, Matsuda with her pigeon-toed
impersonation of a 14-year-old is still remembered as the epitome of the type more
than 20 years later, even though she has since reinvented herself as a more mature
celebrity. In my interviews, young people only a few years old when Matsuda first
appeared still offer her first when asked to name burikko. Matsuda was part of the
stream of childlike performers, called aidoru idols, who populated the 1980s pop
music scene. Although these days aidoru tend to be much more womanly, a modi-
fied style of ultrafeminine cuteness is still a fashion option. For example, a womens
magazine categorizes burikko as a trendy style obtainable through wearing pastels
and lace (Tokyo go dai GAL sutairu zukan 2002).
150 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

Scholars have historically had an interest in how women are labeled and how
these labels may reflect sexist, stereotypical viewpoints. A critique of such words is
important because they often serve as vehicles for assumptions that are uncritically
accepted as normal and therefore go unchallenged. As Cameron (1990:12) asserted,
Like other representations, linguistic representations both give a clue to the place
of women in culture and constitute one means whereby we are kept in our place. A
feminist perspective assumes that use of denigrating labels for women is not simply
the reflection of an individuals nasty opinion or attitude but is also the manifesta-
tion of patriarchal social structure (Flax 1979). One goal of a feminist analysis is to
explore the ways in which labeling and naming fortify a patriarchal system.
Although negative labels for women have been a focus of research in many lan-
guages, discussions of terms rarely incorporate an analysis of the contexts of their
use and their evaluation. Researchers of language and gender note the scholarly his-
tory of ignoring womens experience in studies of language in society (Eckert &
McConnell-Ginet 1992, Ochs 1992, Bucholtz 1999). While this erasure has been
addressed in current research on Japanese, the speech of women characterized as
burikko is still avoided as silly, inconsequential, and embarrassingly unworthy of
academic attention. Stimulated and aided by the work of [Shibamoto] Smith (1992),
Okamoto (1995), and others, I have been thinking about the nature of the burikko
designation, what it means, and how it relates to a Japanese cultural ideology about
proper female behavior. My interest in this topic was also aroused when I observed
Japanese women I admired for their intelligence, good sense, and capability occa-
sionally performing behaviors judged as burikko-ish. In the same manner in which
Okamoto (1995) questions the category Japanese womens language, I want to view
the burikko not as a fixed personality or character type but as an evaluative interpre-
tation of the behavior exhibited through linguistic, paralinguistic, prosodic, and non-
verbal means in specific social settings. In other words, although some people are
dismissed for being burikko, they are in fact simply doing burikko. Indeed, one
often hears the expression burikko suru to do burikko. This phrase indicates that
there is a certain level of awareness on the part of scrutinizers and observers that
burikko is a gender performance.
This approach to social identity means that individuals are not viewed as lin-
guistic versions of the social dope (Giddens 1976) or as unthinking types who
unthinkingly carry and broadcast their social identities at all times. Rather, we should
see speakers as consciously and unconsciously drawing on linguistic and nonverbal
repertories that reflect idealized norms. In this case, some women who select too
generously or inappropriately according to cultural norms from a menu of femi-
nine and childish indexical forms will be negatively sanctioned through labeling
as burikko. This is not to say that there is some absolute threshold level of linguistic
forms that will trigger the evaluation or that there is a set amount of frilly femininity
that will automatically doom the behavior as burikko. Because they draw from the
same gendered cultural system, many of the features that typify a burikko perfor-
mance are also part of acceptable female gender presentation or innocent girls talk.
When a woman is regarded as doing burikko because she is displaying overly femi-
nine, innocent, or cute childishness in a specific situation, it is because these other-
wise valued traits are thought to be inappropriate for her or for the situation at hand.
 YOU ARE DOING BURIKKO ! 151

What is considered too much or inappropriate femininity or cuteness will also de-
pend on the evaluators stance. Young men might view this through jaded postmodern
acerbity as impishly fake, while an older man thinks it sweetly girlish. In short, the
same display of cuteness, childishness, or femininity can be seen as real or fake,
depending on who is the actor and who is doing the evaluation.
Before offering examples of folk representations, I will begin with a description
of some features that are thought to mark feminine or childish speech and commonly
accompany the manifestation of burikko behavior.

8.1. Vocal drag and other features considered part

of a burikko performance

In addition to the grammatical features of talk considered part of the female register
(Shibamoto 1985, Ide & McGloin 1990), the performance characterized as that of a
burikko may include specific vocal attributes. A combination of linguistic, prosodic,
paralinguistic, and nonverbal features contributes to an interpretation of speech as
burikko-like. Although representations and interpretations of burikko do not always
distinguish these as separate indices, the most readily associated traits are the fal-
setto voice and a glissando movement through a pitch range. Additional features
include a nasalized delivery, use of a baby-talk register, a sprinkling of amusing
coinages, and mannerisms such as covering the mouth when smiling or laughing.
Vocal pitch is only partially the product of anatomical differences between women
and men (Mattingly 1966). Pitch also reflects a pattern of cultural training that funnels
speakers voices into expected grooves. Differences in adult voices are due to uncon-
scious pitch selection that approximates a vocal image that reflects culturally expected
attributes. Although the manipulation of the vocal tract in order to correspond to gen-
der expectations is usually learned early, transgender and transsexual media often in-
cludes books and tapes on voice passing. In a sense, the burikko voice is a similar form
of vocal drag. For decades foreign observers of Japan have commented on the notice-
ably high voices of some female train station announcers, department store clerks, ele-
vator girls, and television commercial narrators. The heightened pitch combined with
formulaic politeness phrases used by these women is sometimes called the service
voice. Ohara (chapter 12, this volume) discusses the change of pitch heights in the
service voice according to interlocutor. Moreover, despite superficial impressions,
Japanese womens public voices show a range of variation and are not uniformly high-
pitched. Morita Miyuki, an evening news anchor for Japans public broadcasting cor-
poration, was first rejected when she entered the broadcasting business because it was
thought that her voice was too low, and an esteemed member of the Japanese Diet, Doi
Takako, is often characterized as having a voice on the extreme low end of the scale.
The seemingly marked nature of womens vocal pitch in Japan led many research-
ers to embark on more empirical analyses, with the aim of examining actual pitch
variation. Yamazawa and Hollien (1992) review three different studies that reported
a higher average pitch for Japanese women than for Euroamerican women. Loveday
(1981) measured pitch in two Japanese womens speech while uttering formulaic
expressions. While the average pitch range for American and English women is
152 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

214 hertz (van Bezooijen 1995:253), Loveday found soaring pitch peaks between
310 and 450 hertz for the Japanese women. We should note that most often these
studies are carried out in foreign laboratory settings, where subjects are usually from
elite social backgrounds and are speakers for whom a particular version of feminin-
ity impression management is important. In another study, Ohara (1992) found that
Japanese women modify their pitch when reading English sentences, producing
speech in a slightly lower pitch range. The average pitch frequency for her subjects
was 19 hertz greater when speaking Japanese. The outcome was explained as un-
conscious, learned behavior: Displaying femininity is an automatic process in most
cases. Controlling the larynx to produce higher-pitched sound may be a part of the
automatic process (Ohara 1992:474). Van Bezooijen (1995), working in Japan, asked
8 Japanese women to read a narrative text and used a recording of their voices as the
stimulus for ratings by 30 other Japanese subjects. She did not find the expected high
pitches in the speakers she used in her study. However, she did find that listeners
associated womens higher pitch with the values of femininity, weakness, meekness,
and dependency. She also found that listeners considered a medium and a high pitch
more attractive in women than a low pitch.
When a woman is performing burikko, her pitch will often rise to a level above
that she normally uses in casual speech. This burikko voice, with its air of ineffectu-
ality, is thought to most often be elicited in the presence of powerful males. It an-
nounces that the speaker is unsure, weak, or less powerful. In my interviews and in
my reading of womens magazines, I frequently encounter the belief that it is otoko
no mae in front of men that the burikko voice is most frequently manifested. In
order to stay within the bounds of gender expectations, Japanese women have been
described as using a variety of techniques when faced with conflicting role respon-
sibilities, such as managerial positions in corporate life. They may, for instance, use
a combination of listening behaviors, repetition, sentence particles, intonation pat-
terns, and prefacing to buffer managerial directives. [Shibamoto] Smith (1992) has
suggested that use of a mother register, or Motherese, is one strategy women in
positions of authority employ when dealing with male subordinates. Burikko talk may
be exploited as another type of linguistic strategy, this one used on the part of subor-
dinate women for dealing with male authority. A wheedling pitch is used for the same
reason we find it in other language communities, because speakers calculate that it
is most likely to bring the ends they desire (McConnell-Ginet 1983:83). A woman
who takes on the social role of a burikko is asking to be given the same lenience and
indulgence accorded to an unschooled child. Of course, the speaker herself must be
aware of her fake performance for it to be considered intentional. She is consciously
placing herself in the role of the innocent, reflecting cultural expectations that women
not be knowledgeable about certain cultural domains.
It is not just a high pitch that leads to an interpretation of the speaker as simulat-
ing innocence but also a melodic swoop over the vocal cords (physiologically, the
speed of the vibration of the vocal folds varies from low to high). The importance of
this element became clear to me when I watched a Japanese television series named
Hotel with a Japanese female friend.1 In one scene a prostitute mistakes one of the
main male characters for a customer. When he explains the misunderstanding, the
woman riffs through a protestation of surprise, exclaiming, with a particularly sweet
 YOU ARE DOING BURIKKO ! 153

and sugary voice, Nan da. Anata mita toki rakkii to omota no ni Oh, gee. When I
saw you I thought I got lucky. As she watched, my friend sneered, Burikko! at the
screen. Here, despite the fact that the characters interlocutor demonstrated knowl-
edge of her activities as a sex worker, the actress presented her character as a pristine
pixie expressing disappointment at a failed date. She did this through the manipula-
tion of the features of pitch range often attributed to the burikko character.
The doing of burikko may also be accomplished through use of a nasalized deliv-
ery that accompanies the high-pitched voice. I located an example in a television pro-
gram titled Tokyo erebeetaa gaaru Tokyo Elevator Girls, a soap opera series about
the young women who greet and assist customers in posh department stores.2 Inter-
spersed throughout the narrative are scenes of the elevator girls being trained in how to
properly address customers with just the right words and cheery intonation, the so-called
service voice. In one scene, an elevator girl (played by actress Miyazawa Rie, consid-
ered something of a burikko herself by some critics) is out visiting a shrine on an illicit
date with a married coworker. Although they have been standing next to it for some
time, she suddenly notices a box that sells fortunes and exclaims in a high-pitched
voice, Ahh . . . ne, omikuji ikoo, omikuji ne Hey! Lets go for a fortunea fortune,
OK? After paying her fee, she reaches into the box to grab a fortune paper while im-
ploring the gods to be kind, chanting, Ii no ga demasu yoo ni Let a good thing come
of it, a formulaic way of requesting a happy wish, yet here uttered with emphasized
nasality. The viewer of this performance knows that the character has seen the fortune
box prior to the exclamation of noticing it, that she is not in a happy situation (she sneaks
around to see her married lover), and that her childlike expressions and innocent vocal
qualities are not at all reflective of her actual status.
When manga comic book artists want to illustrate burikko behavior, they often
use drawn out syllables in speech bubbles to represent the high-pitched, nasalized voice.
For example, figure 8.1 (Moritsugu 1998:458) features a 21-year-old OL character (OL
or Office Lady is used to mean a female clerical worker), who responds to a question
about how her lover is doing with, Iya~~ne, himitsu yo Oh, dear . . . its a secret.
Elongation of the syllable ya is indicated with a wavy line. The hand-to-cheek gesture
also suggests that she is overacting the role of an unsophisticated girl.
Another important feature attending the burikko display is a baby-talk register
that casts the speaker in the role of a credulous innocent who ought to be the recipi-
ent of the listeners indulgence. Features characteristic of baby talk are avoidance of
Chinese loans, use of the honorific prefix o-, phonological modifications, use of
onomatopoeic words, and reduplication. Avoidance of Chinese loans entails rejec-
tion of erudite kango, words of Sinitic origin that suggest complex concepts beyond
the burikkos hazy scope. An example would be using uso lie but never kyogi false-
hood. The use of the honorific prefix o-, often considered polite and refined and
hence a feature of the female register (Shibamoto 1987), is extended to everyday
words such as otete in place of te for hand and omeme for eye rather than me.
There are often phonological modifications, such as katchoi, from kakkoii cool, and
yappashi, from yappari after all. Similar forms are kawayuui for kawaii cute (Inoue
1986) and utchoo in place of uso lie (Horiuchi & Omori 1994). Onomatopoeic words
and mimetic nouns are used liberally, such as wanwan bowwow for dog (the un-
marked form is inu) and chu chu for small bird (rather than kotori). In her research
154 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

8.1. Comic book representation

of burikko behavior

on Japans cultural aesthetic of cuteness, Kinsella (1995) suggests that some favor-
ite burikko words derive from the deliberately contrived speech of pop idol Sakai
Noriko. During 1987, Norikos coinages, uttered with an endearing lisp and called
Norippi-go Noriko language, were widely emulated by others. These cute labialized
forms included ureppi, from ureshii, meaning happy, and kanappi from kanashii
for sad. Yonekawa (1996:83) also includes okabii from okashi funny and batchishi
from batchiri right on or no sweat on his list of Noriko words. Reduplication is
often used for animals, body parts, or indelicate concepts, as in nenne in place of
neru to sleep, which has a euphemistic connection to sex, and babatchi from baba
poo-poo for icky poo or yucky.
 YOU ARE DOING BURIKKO ! 155

Other features that imbue speech with burikko-ness involve special lexical and
stylistic forms, such as extended use of diminutive or familiar title suffixes with names,
novel lexical clipping, and display of hesitation or uncertainty. For example, address
terms are used for inanimate objects, animals, or parts of the body. One example is
taku-chan Little Mr. Taxi for taxi. This is derived from a clipped form of takushii
taxi combined with the diminutive address term -chan (Yoshioka 1993). Indeed,
both the elfin -chan and the familiar -kun have proven to be very productive for the
creation of cute lexical items. Other combinations I have heard used by adult women
include wanchan Mr. Bowwow to refer to a dog and the appellation tsuru tsuru-
kun Mr. Smooth used in reference to young men who use hair removal products on
their chests in order to be more attractive (in accordance with new beauty norms; see
Miller 2003). Talk may be heard as girlishly cute when it contains an abundance of
amusing coinages or charming constructions of the wasee eego Japan-made English
type, which frequently involve novel lexical clipping (Miller 1998b). For example,
there is pii suru for PHS o kakeru or use a cell phone, PHS is an acronym for per-
sonal handyphone system. Another amusing coinage is kenta-kun Mr. Kentucky
Fried Chicken, used to refer to relatively young men who have beards and are sort of
fuddy-duddyish (Yonekawa 1991). This is created through a clipped form of Kentucky
combined with the intimate address term kun and is presumably derived from the
Kentucky Fried Chicken logo, which features the companys goateed founder, Colo-
nel Sanders. Although use of new or trendy lexical items in and of itself does not
transform the speaker into a burikko, their use in otherwise formal situations may
have that result. A type of syntactic clipping, in which sentences are left incomplete
or drift off into inaudibleness, will also mark speech as timorously artificial.
Nonverbal correlates of the burikko performance include a head tilt to the side
and clasping the hands in front of the body when one hand is not demurely covering
the mouth. The use of the hand to shield the mouth when smiling or laughing is a
custom traced to at least the Edo period (Casal 1966), and most likely is related to an
underlying interpretation of teeth display as an aggressive signal. Covering the teeth
is therefore seen as a form of submissiveness. These days many young women are
expressing opposition to the enduring custom of enshrouding the female mouth by
emphasizing their white teeth and by refusing to cover their mouths when laughing
or giggling. The result is that when they are doing burikko this aspect of the per-
formance is highlighted.
An assemblage of vocal features, linguistic usage, and nonverbal behaviors as
described earlier contributes to a gendered performance that may therefore become
the object of criticism or commentary. These reactions to the burikko display fre-
quently surface in books, magazines, and other media.

8.2. Media and folk assessments

of burikko performances

An etiquette book titled Anthology of OL Taboos (Zennikku Eigyohonbu Kyoiku-

kunrenbu 1991:181) admonishes female clerks to avoid loading their office desks
with character goods, such as Mickey Mouse pencils or Hello Kitty staplers
156 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

(fig. 8.2). A woman who surrounds herself with such items is said to be yoochi in-
fantile and incompetent. Advice like this tells us that while cuteness is thought to be
an approved aspect of femininity display, it is nevertheless sanctioned within certain
age/status limits or locales.
In an interview with Harvard Magazine, an anthropologist in the United States
classified burikko as a cultural style of consumption and self-expression. According
to him, the quintessential burikko collects toys, wears girlish clothing, and is heard
giggling and squealing in a high-pitched voice (Hodder 1997). But his description
suggests that burikko is an all-or-nothing identity, rather than a social evaluation that
arises from situated behavior. His characterization, in fact, is more properly that of
the shoojo young miss rather than the burikko. Treat (1996:281) describes the shoojo
as one whose sexual energy is directed toward stuffed animals and cute artifacts. The
concept of the shoojo implies an adolescent postpubescent space that revels in all
that is cute, pink, fluffy, and adorable, with an absence of heterosexual experience
(but not necessarily of homosexual experience; see Robertson 1989:59). In some cases
then, women too obviously beyond the shoojo age-group who perform childlike
behaviors will be termed burikko. The display meant to neutralize adult sexuality,
when performed by those clearly beyond an age of innocence, unmasks the artifice
of the maneuver. A disjuncture between the speakers age and her display of cute-
ness is a critical aspect of many cases of burikko ascription. This fact is illustrated by
reactions to womens use of an orthographic fashion popular among schoolgirls, the
use of writing called maru moji round characters (Yamane 1986; see also Kataoka
1997). Some women retain this innovation from girls culture as they enter the

8.2. Cute goods in the office. Published with permission of Goma Shobo
 YOU ARE DOING BURIKKO ! 157

workforce, where it may become an object of commentary and ridicule or mark them
as doing burikko. In conduct literature such as Anthology of OL Taboos (Zennikku
Eigyohonbu Kyoikukunrenbu 1991:100), female clerks are advised not to use round
script in memos to coworkers, even when they do so because they think it will make
the information easier to understand. In figure 8.3, a male colleague expresses exas-
peration at the use of round characters he finds in a memo from an OL, which says:
Nakamura-san kara Tel arimashita. Keeko Theres a telephone message from
Mr. Nakamura. Keiko.
Complaints about burikko can be found in Japanese womens magazines, which
have become a forum where expressions of anger or dissatisfaction with others may be
safely vented. In one feature article that discusses disliked behaviors of coworkers,
people complained about women they termed burikko (Uchi no kaisha no kanchigai
OL 1996). One writer deplores a woman who uses nicknames on the telephone, saying
things like Waa . . . Kin-chan Ohh! . . . Kinnie! In this case, it is felt that use of a
diminutive nickname and the diminutive address term -chan is evidence of babyishness
and immaturity. Another reader provided the example of a 30-year-old woman in her
office who uses a kittenish, obsequious voice to say things like Gomen nachyaa Ooohh,
Im sorryy! The hearers reaction is Kimochi warui It gives me the creeps.
In popular media and in interviews, burikko are characterized as both shallow
airheads and crafty flatterers. People often say a distinctive feature of burikko talk is
that they always utter hazukashii Im embarrassed even though its clear they arent
the least bit mortified. And as noted earlier, most commentators suggest that it is otoko
no mae in front of men that the burikko is most likely to make her appearance. Former
pornographic film star Ikuma Ai once performed a classic burikko drill on TV when

8.3. Cute round characters. Published with permission of Goma Shobo

158 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

she covered her mouth while giggling and saying, Hazukashii! How embarrass-
ing! in response to a talk show hosts impertinent questions. When young men were
asked to name the top five celebrities they would not want to be married to, they
included actress Miyazawa Rie (Otoko hyakunin ga yuu onna no fushigi rankingu
1996). Miyazawa represents a burikko type because early in her career she publicly
presented the image of a sheltered upper-class daughter of the suburbs but in fact
has done some racy things behind the scenes, such as posing nude and having a love
affair with a Sumo wrestler (Schilling 1992).
Readers of a womens magazine asked to provide examples of really nauseat-
ing burikko OLs (choo mukatsuku burikko OL) offer examples that illustrate how
the switch to counterfeit behavior is often triggered by the presence of men (Uchi no
kaisha no kanchigai OL 1996). One clerk writes to complain that a coworker uses
amaeta koe bootlicking voice, a sweet, smarmy tone used to wheedle things out of
silly male section heads and managers. A 27-year-old man writes in to gripe about a
woman who aggressively chastises juniors and coworkers, yet her voice goes up
one octave in front of male seniors (jooshi no mae de wa ichi okutabu koe ga takaku
naru). Another woman reports on a coworker who graphically talks about anything
and everything imaginable among girlfriends when they are in private areas such as
the ladies room, but if theres naughty talk in front of men she modestly exclaims,
Ya daa! Hazukashii Thats disgusting! Im so embarrassed. Describing a woman
who does burikko in her office, someone else writes: She uses a saccharine voice to
butter up male superiors. I cant tell you how disgusting it is when I see her stroking
his arm while saying things like, Section head, the shirt you have on today is really
fab (Chokuzoku no otoko no jooshi ni amattarui koe de obekka o tsukau Kachoo
kyoo no Y-shatsu suteki desu ne to ude no atari o surisuri shite iru tokoro o mita
toki ni wa, akirete mono ga ienakatta).
That some women will use a burikko performance to manipulate an older male
coworker points up another dimension to interpretation. Not all observers will agree
on whether or not a given performance is burikko-like. It is clear that there will be
different attitudes on the part of those from different age and social backgrounds and
that interpretation also hinges on the social situation in which burikko behavior oc-
curs. In other words, one persons burikko is anothers proper well-bred miss. Con-
temporary young people, both female and male, react negatively to extreme forms
of docility and cuteness, but older Japanese men may still admire and endorse the
subservient pose of those who do burikko. Even if an older man discerns the fakeness
of a burikko stance, he may still like what the pose represents. As feminist writers
remind us, the cute infantilization of women places them in a weaker social position.
According to a male critic, while seeing burikko behavior as somewhat silly, some
men will still consider it ear massage (Hosokawa Shuhei, personal communica-
tion). Since the 1990s, there has been a change in younger mens attitudes and women
who are coy, overly docile, and indecisive have been disparaged (Miller 1998a). When
asked which type of women they least like, young men polled by a mens magazine
ranked burikko number 9 on their list of top 10 peeves (Kore dake wa yurusen! 1999).
The concept of the burikko, then, has much to do with the perceiver and her or
his ideas about gendered talk. Conflicting evaluations of speech marked as femi-
nine illustrates the way in which norms have been changing. During her first offi-
 YOU ARE DOING BURIKKO ! 159

cial public interview after her marriage to the Crown Prince, Princess Masako, an
articulate woman who attended Harvard University, sat demurely and meekly, say-
ing very little. On the one hand, I heard older observers exclaim, Shes marvelous,
approving of her display of modesty and reserve. On the other hand, young women
who now have revised ideas about female propriety suggested that Masako was a
burikko sellout.
When one of the university students I interviewed teasingly urged his female
classmate to burikko yatte mite try doing burikko, she stonily refused, not find-
ing it an amusing request at all. The New York Times Japan correspondent Kristof
(1995) interviewed a 15-year-old girl who said, When girls speak in really high
voices, I just want to kick them in the head. Its totally fake and really annoying. It
gives me a headache. McVeigh interviewed a female university student who ex-
pressed a similar view.

I hate this word [burikko]. Girls in Japan tend to be different when they are with
girls or with boys. Their voice changes. I have seen my sister, so I know. When the
telephone rings, they clear their throats, and I dont know where this voice comes
from, but a cute little voice comes out from somewhere. I hate this moment. They
are lying to the people they are talking to . . . Burikko women are hated by most
women and liked by only a few men. I just want them to stop pretending. (2000:147)

Despite these negative assessments, such burikko behaviors as high pitch are
connected to prescribed cultural norms of femininity. Conduct literature, ethno-
graphic research (Lebra 1984, McVeigh 1996), and other sources have provided a
clear picture of the relationship between Japanese gender ideology and canons for
appropriate female behavior and language for middle-class women. One of the traits
valued as an aspect of femininity is cheerfulness. For example, in a how-to book
titled Fundamental Checklist of OL Manners (Nakamura 1993:137), readers are cau-
tioned about proper vocal attributes: Pay attention to the volume, tone and speed of
your voice. For the person with a low-pitched voice, rather than using your ordinary
voice, its probably better to speak in a slightly higher pitch. That way you will im-
part a cheery, bright image. Although the behavior and mannerisms advocated by
Nakamura may work for young women in some contexts, these rules for good
speech behavior may also fall flat, depending on the speaker and the degree to which
her sincerity is seamlessly accepted. When speech is delivered in such a manner by
a woman of the wrong age, status, or class, it could be negatively evaluated.
We also find that conduct literature advocates behaviors that in other instances
are denigrated as insincere meekness. Conflicting messages directed at women are
not at all uncommon in popular writing (Miller 1998a). These discrepancies are
sometimes the result of media targeting readerships from different age or class back-
grounds, but in some instances the contradictory information is presented in the same
book or article. For example, while handing out advice on how to display proper
deference to male superiors and to raise ones vocal pitch in order to be more cheery,
Nakamura (1993) also admonishes readers not to be too self-effacing or unsure of them-
selves. Likewise, in a womens magazine article on dating behaviors that aggravate
men, women who only utter platitudes along the lines of Oh, that was interesting
or who keep saying, Anywhere is fine, while on a date are severely chastised as
160 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

graceless ticklebrains: Dont you know any other words? ask commentators
(Hajimete no deai de, kirawarechau hito, daisuki ni sasechau hito 1994). While they
may be reading this type of advice in one place, women will be told the opposite in
etiquette manuals such as Manners for Age Twenty and Over (Tanaka 1993), where
they are cautioned not to express their ideas or opinions and to play it safe by falling
back on bland generalities.

8.3. Evaluations of situated burikko performances

Performing ultrafeminine behaviors reveals a desire to showcase a decorous ojoosama

proper young miss upbringing, but it could have unfortunate consequences for those
who misjudge the recipients of their act or the setting in which it is performed. Even
in cases in which a display of modesty is expected, the speaker may nevertheless fall
victim to burikko labeling. A look at a few instances of situated talk in which the
woman was described as doing burikko illustrates how, while aspects of their per-
formance accord with acceptable feminine conduct, other attributes confound or dis-
turb this meaning.
On one occasion I witnessed a burikko performance at the building where I was
housed with other foreigners associated with a Japanese engineering university. Stu-
dents frequently came by to hang out upstairs on a balcony. One day I went there to
talk to a small group about mens esute body aesthetics, or male beauty work. This
school had once been an all-male university, and there are still very few female stu-
dents. The women who are there have entered by virtue of superior math and science
skills that enabled them to beat out male competitors. All the women I met were highly
intelligent, straightforward, serious, and unpretentious. As I talked with them on the
balcony, an American male student came up and, not understanding what we were
discussing, asked for a translation. When I explained that many Japanese women find
body hair on men unattractive, which has led to the development of new products
and services for male body hair removal, he was incredulous. He refused to believe
me, claiming that chest hair in particular indicates that one is a real man, and that
women universally dig it. I suggested that different aesthetic sensibilities are in
operation, but he continued to protest against such an idea. Just then Naoko, a fe-
male student as bright as any I met, came up to join our group. The American chap,
deciding to simply test my theory empirically, lifted his shirt to display his hairy chest,
asking her what she thought. Naoko screamed, Iya da! How hateful! in a shrill
voice and ran to hide behind a door while periodically peeking out to whimper at the
unspeakable sight. One of the Japanese men who witnessed this muttered in puzzle-
ment, Dooshite burikko shiteru no? Why is she doing burikko? It took me by
surprise to see someone I never expected had an ounce of artifice in her instantly
portray the precious imp. Yet it served to illustrate the degree to which burikko be-
havior may be elicited by situations that demand a public display of modesty (or
legitimate revulsion) and is therefore not a persona a woman adopts or continually
performs. Indeed, had Naoko openly stared at the naked male chest, she would have
been the recipient of other denigrating labels. Some days later I asked Naoko her
opinion of burikko. Theyre disgusting, she told me.
 YOU ARE DOING BURIKKO ! 161

This instance reveals that there are settings or topics of conversation that are likely
to create a burikko-like response. According to a discourse of heightened feminin-
ity, there are aspects of the world a proper woman ought not to possess knowledge
of, just like a young and unsophisticated child. Hence, when sex and sexuality are
introduced into talk, squeak-toy disclaimers are to be expected. The former porn star
who claims a sense of disgust at sexual innuendo and the college-aged woman who
acts as if shes never seen a bare male chest are both expected to show modesty yet
both are accused of doing burikko because there is good reason to suspect that these
are not truthful responses.3
I offer a final example of a burikko performance, although not a real one, that
was negatively sanctioned. One evening I went with my coworkers, two other female
professors, to a local restaurant for something to eat after work. Two of us went ahead
in one car to be joined later by the other, the director of the program. We arrived
before her and were talking when I mentioned that I was wondering about the nature
of burikko. As a joke, my colleague Hiromi went into a burikko imitation. Raising
her pitch level a few octaves, she began uttering inanities. In the midst of her act the
director walked in and, hearing her subordinate, yelled at her, Nan de burikko shiteru
no?! Why are you acting like a burikko?! Of course, we explained the situation
and laughed about it, and Hiromi resumed her normal speaking manner.
However, on other occasions I witnessed this same chastising director, a woman
in her thirties, perform self-acknowledged burikko behavior. While discussing the
problems that women still face in the workplace, especially in academic settings, she
told me of how they must be careful to display deferential behavior to their male
colleagues to avoid threatening them. Contemplating her own career trajectory, she
said (in English), I have to learn to do burikko better to get ahead. It is interesting
that she would view burikko behavior done to manipulate men as necessary yet see
burikko behavior performed in a women-only group as totally unacceptable. A few
occasions on which her female colleagues reported that she affected burikko man-
nerisms included while she was speaking to male principals during a trip to a local
school and greeting male dignitaries at a formal reception. At these times she report-
edly assumed a lacy vocal singsong uplift (distinguished by a thin, light voice qual-
ity, giving an impression of immaturity) and spoke with numerous self-corrections
and hesitation markers. She also covered her mouth while smiling or giggling.
In searching for the burikko, I found that tenets for proper female behavior place
some women in a behavioral paradox. While a woman is taught that docility, inno-
cence, femininity, childishness, and cuteness are desired commodities, if such a display
doesnt suit her age or status or the situation, her behavior will be judged as phony
burikko pretense. Like Naokos dilemma when faced with the foreigners bare torso,
this could be a frank instance of damned if she does and damned if she doesnt. As we
have seen from magazines, television, and other representations, as well as from situated
instances, a disjuncture between what we know about someone and the sort of femi-
nine behavior she is pressured to exhibit can sometimes result in fleeting burikko scenes.
As Ochs (1992) has pointed out, indexing of gender is not a simple process of
hooking up linguistic forms to speaker sex. Instead, speakers draw from a menu of
stereotypical forms to enact social identities. Could it also be that, like a bad actor,
the woman evaluated as a burikko is really just caught doing gender red-handed?
162 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

Is it the case that she deploys indexical forms unbefitting her situation and is dis-
liked because the manipulation of expected gender traits becomes just too transpar-
ent, too camp? Perhaps a burikko performance makes us uncomfortable because it
asserts a hard truth about gender roles in general, which is, as American drag queen
RuPaul put it, that were all born the same and the rest is drag.
From the study of burikko we may detect a rising consciousness that the way
one speaks is intimately tied to social position. There is increasing evidence that a
change in gender roles and expectations is changing speech behavior. Many younger
women are eschewing features of language considered markers of the feminine
and are incorporating aspects of the male register into their own speech (Okamoto &
Sato 1992, Okamoto 1995). There have been enough of these latter types that a nega-
tive label was coined to categorize them. This is the oremeshi onna me-food woman,
a term that parodies an autocratic husbands command. Oremeshi onna are said to
totally reject features of talk characteristic of the burikko in favor of hypermasculine
forms. In the worldview of younger, contemporary Japanese, the type of sugary child
exhibitionism displayed in burikko performances is viewed as deceitful and creepy,
and they would rather risk being called oremeshi onna than burikko. Although this
change suggests that ideas about womens speech are opening up a space for varia-
tion, it also tells us that women continue to be objects of intense social surveillance.
Whether they are conforming to gender norms, challenging them, or overdoing them,
Japanese women continue to be the recipients of derogatory labeling.

Notes
A version of this chapter was first presented at the 1997 meeting of the American Anthropo-
logical Association, in Washington, D.C. I would like to thank Jan Bardsley, Mary Bucholtz,
Yuka Fukunaga, Yuko Hoshino, Shuhei Hosokawa, Shigeko Okamoto, and Janet S. Shibamoto
Smith for their comments and suggestions. All translations and mistakes are my own.

1. Hotel was a five-part Tokyo Broadcasting System television drama that first began
airing in 1990, about workers and guests at a luxury hotel.
2. Tokyo erebeetaa gaaru, written by Komatsu Eriko, aired on the Tokyo Broadcasting
System during Winter 1992. The English loan girl has two forms in Japanese, gaaru and gyaru.
This TV program used the former.
3. The line between acceptable feminine behavior and burikko categorization may also
relate to other factors besides age or innocence, particularly class status. A display of pro
forma sweetness might be tolerated when done by debutante girls from elite family back-
grounds, yet the woman or girl who lacks cultural capital and displays her class origins will
have her failed efforts described as burikko-ish.

References
Bucholtz, Mary (1999). Why be normal? Language and identity practices in a community
of nerd girls. Language in Society 28(2): 203223.
Cameron, Deborah (1990). Introduction: Why is language a feminist issue? In D. Cameron
(ed.), The feminist critique of language, 128. London: Routledge.
 YOU ARE DOING BURIKKO ! 163

Casal, U. A. (1966). Japanese cosmetics and teeth blackening. Transactions of the Asiatic
Society of Japan 9: 527.
Cherry, Kittredge (1987). Womansword: What Japanese words say about women. Tokyo:
Kodansha International.
Eckert, Penelope, and Sally McConnell-Ginet (1992). Think practically and look locally:
Language and gender as community-based practice. Annual Review of Anthropology 21:
461490.
Flax, Jane (1979). Women do theory. Quest 5(1): 2026.
Giddens, Anthony (1976). New rules of sociological method. New York: Basic Books.
Hajimete no deai de, kirawarechau hito, daisuki ni sasechau hito (Detestable people and people
who are really likable in first encounters) (1994). Say 137: 196200.
Hodder, Harbour Fraser (1997). Right now: Tokyo Disney. Harvard Magazine, JulyAugust.
Online: http://www.harvard_magazine.com/issues/ja97/right.disney.html.
Horiuchi, Katsuaki, and Yoshiko mori (1994). Wakai josee no kotoba no gokee, gogi no
tokushoku (Characteristics of word form and word meaning in young womens speech).
Nihongogaku 13: 7280.
Ide, Sachiko, and Naomi Hanaoka McGloin (eds.) (1990). Aspects of Japanese womens lan-
guage. Tokyo: Kuroshio.
Inoue, Fumio (1986). Sociolinguistic aspects of new dialect forms. International Journal of
the Sociology of Language 58: 7389.
(1989). Shiin no hatsuon no henka (Changes in the pronunciation of consonants). In
S. Miyoko (ed.), Nihongo no onsee, onin (The voice and vocal sounds of Japanese),
109134. Tokyo: Meiji Shoin.
Kataoka, Kuniyoshi (1997). Affect and letter writing: Unconventional conventions in casual
writing by young Japanese women. Language in Society 26: 103136.
Kinsella, Sharon (1995). Cuties in Japan. In L. Skov and Brian Moeran (eds.), Women, media
and consumption in Japan, 220254. Honolulu: University of Hawaii Press.
Kore dake wa yurusen! 99 Saisho no rankingu (Just this is not permitted! Our first ranking
of 1999) (1999). Rankingu Daisuki, March, 88.
Kristof, Nicholas (1995). Japans feminine falsetto falls right out of favor. New York Times,
December 13.
Lebra, Takie Sugiyama (1984) Japanese women: Constraint and fulfillment. Honolulu: Uni-
versity of Hawaii Press.
Loveday, Leo (1981). Pitch, politeness, and sexual role: An exploratory investigation into
the pitch correlates of English and Japanese politeness formulae. Language and Speech
24: 97107.
Mattingly, Ignatius (1966). Speaker variation and vocal tract size. Journal of the Acoustical
Society of America 39: 1219.
McConnell-Ginet, Sally (1983). Intonation in a mans world. In B. Thorne, C. Kramarae, and
N. Henley (eds.), Language, gender and society, 6988. Boston: Heinle and Heinle.
(= 1978 Signs 3: 541559).
McVeigh, Brian (1996). Life in a Japanese womens college: Learning to be ladylike. London:
Routledge.
(2000). Wearing ideology: State, schooling and self-presentation in Japan. Oxford:
Berg.
Miller, Laura (1998a). Bad girls: Representations of unsuitable, unfit, and unsatisfactory
women in magazines. U.S.-Japan Womens Journal 15: 3151.
(1998b). Wasei eigo: English loanwords coined in Japan. In J. Hill, P. J. Mistry,
and L. Campbell (eds.), The life of language: Papers in linguistics in honor of William
Bright, 123139. The Hague: Mouton/De Gruyter.
164 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

(2000). Media typifications and hip bijin. U.S.-Japan Womens Journal 19: 176205.
(2003). Male beauty work in Japan. In J. Roberson and N. Suzuki (eds.), Men and
masculinities in contemporary Japan: Dislocating the salaryman doxa, 3758. New York:
Routledge.
Moritsugu, Yahiro (1998). Kaihoo no shirushi: Hana to yume (The scars of liberation: Flowers
and dreams), 456501. Tokyo: Hakusensha.
Nakamura, Yoichir (1993). OL manaa kihon chiekku (Fundamental checklist of OL man-
ners). Tokyo: Isetan.
Ochs, Elinor (1992). Indexing gender. In A. Duranti and C. Goodwin (eds.), Rethinking con-
text: Language as an interactive phenomenon, 335358. Cambridge: Cambridge Uni-
versity Press.
Ohara, Yumiko (1992). Gender-dependent pitch levels: A comparative study in Japanese and
English. In K. Hall, M. Bucholtz, and B. Moonwoman (eds.), Locating power: Proceed-
ings of the Second Berkeley Women and Language Conference, 469477. Berkeley:
Berkeley Women and Language Group.
Okamoto, Shigeko, (1995). Tasteless Japanese: Less feminine speech among young Japa-
nese women. In K. Hall and M. Bucholtz (eds.), Gender articulated: Language and the
socially constructed self, 297325. New York: Routledge.
Okamoto, Shigeko, and Shie Sato (1992). Less feminine speech among young Japanese fe-
males. In K. Hall, M. Bucholtz, and B. Moonwomon (eds.), Locating power: Proceed-
ings of the Second Berkeley Women and Language Conference, 478488. Berkeley:
Berkeley Women and Language Group.
Otoko hyakunin ga yuu onna no fushigi rankingu (Strange ranking of women according to
100 men) (1996). Say 153: 7.
Robertson, Jennifer (1989). Gender-bending in paradise: Doing female and male in Japan.
Genders 5: 5069.
Sasaki, Takeshi, Tsurumi Shunsuke, Tominaka Kenichi, Nakamura Masanori, Masamura
Kimihiro, and Murakami Yichir (eds.) (1991). Sengoshi daijiten 19451990 (Ency-
clopedia of postwar Japan 19451990). Tokyo: Sanseido.
Schilling, Mark (1992). Worshiping the naked Goddess: The media, mores, and Miyazawa
Rie. Japan Quarterly 105(23): 218224.
Shibamoto, Janet S. (1985). Japanese womens language. New York: Academic Press.
(1987). The womanly woman. In S. Philips, C. Tanz, and S. Steele (eds.), Language,
gender and sex in comparative perspective, 2649. Cambridge, MA: Cambridge Uni-
versity Press.
[Shibamoto] Smith, Janet S. (1992). Women in charge: Politeness and directives in the speech
of Japanese women. Language in Society 21(1): 5982.
Tanaka, Toshiyuki (1993). Hatachi kara no manaa (Manners for age twenty and over). Tokyo:
Shobunsha Shuppan.
Tokyo go dai GAL sutairu zukan (Picture book of five big Tokyo Gal styles) (2002). Popteen,
July, 7682.
Treat, John (1996).Yoshimoto Banana writes home: The shjo in Japanese popular culture.
In J. Treat (ed.), Contemporary Japan and popular culture, 275308. Honolulu: Uni-
versity of Hawaii Press.
Uchi no kaisha no kanchigai OL (The clueless OLs in my company) (1996). Can Cam 174:
419423.
Yamane Kazuma (1986). Hentai shoojo moji no kenkyuu (Research on abnormal girls char-
acters). Tokyo: Kodansha.
Yamazawa, H., and H. Hollien (1992). Speaking fundamental frequency of Japanese women.
Phonetica 49: 128140.
 YOU ARE DOING BURIKKO ! 165

Yonekawa, Akihiro (1991). Wakamono-go jiten 4 (Dictionary of young peoples speech 4).
Gengo 20(8): 19.
(1996). Gendai wakamono kotobakoo (Analysis of current youth language). Tokyo:
Maruzen.
Yoshioka, Yasuo (1993). Wakamono-go inobshon (Innovations in young peoples speech).
Gengo 22(1): 110.
van Bezooijen, Renee (1995). Sociocultural aspects of pitch differences between Japanese
and Dutch women. Language and Speech 38(3): 253265.
Zennikku Eigyohonbu Kyoikukunrenbu (ed.) (1991). OL tabuushuu (Anthology of OL taboos).
Tokyo: Goma Shob.
166 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

ORIE ENDO
JANET S . SHIBAMOTO SMITH , TRANSLATOR

Women and Words

The Status of Sexist Language in Japan
as Seen through Contemporary Dictionary
Definitions and Media Discourse

O ne often hears the argument that in order to eliminate sexist language, it is first
necessary to eliminate sexism; as long as there is sexism, it is useless to problematize
sexist language alone. However, I argue that the existence of sexist language sup-
ports and reproduces sexist attitudes and it is, thus, essential to problematize sexist
language itself. In this chapter, I focus on three of the many derogatory words used
to describe women in Japanese. I examine their treatment both in a sample of Japa-
nese language dictionaries and, more briefly, in the media. While the use of sexist
expressions in the media has decreased considerably over time, in dictionaries the
treatment of many sexist words remains largely unchanged. Although media have
their own shaping effects on the public understanding of gender (see chapters 6 and
7, this volume), they must, ultimately, attract and retain their audience; not surpris-
ingly, then, popular media are found to be more responsive than dictionaries to so-
cial change. Lexicographers, however, are experts who authenticate the words and
word meanings we use (Gal 1989). As authenticators of symbolic capital (Bourdieu
1977), they have a different relation of symbolic power vis--vis the public and tend
toward a conservative vision of word meaning and usage. One thinks, in this vein, of
dictionaries as offering such authoritative information about lexical items as cor-
rect pronunciation, meaning, orthography, etymology, usage, and the like. How-
ever, if lexical items are defined without regard for the discriminatory nature of sexist
terms, they do not provide appropriate information to dictionary users. Vigilant scru-
tiny of these authoritative texts is, therefore, imperative.
This chapter is organized as follows: Section 9.1 discusses the discriminatory nature
of the three sexist terms chosen for this study; section 9.2 examines the changes in the

166
 WOMEN AND WORDS 167

use of these three terms in postwar media; section 9.3 then assesses how dictionaries
do or do not reflect these changes. Finally, section 9.4 reviews the social context in
which the changes in usage have occurred in popular media but not in most dictionaries.

9.1. Sexist terms in Japanese: Onnadatera (ni),

rooba, and oorudo misu

Sexist language exists at two levels in Japanese. First, at the lexical level, there are
terms such as busu ugly [woman] and urenokori old(er), unmarried woman (lit.,
unsold merchandise), the semantic representations of which contain discrimina-
tory elements. Then, at the level of the utterance, there are terms that are sexist when
used in particular social contexts or in particular fixed phrases. Examples of these
include uchi no onna no ko ni ikasemasu Ill have my girl go, uttered by an em-
ployer about one of his female employees, and onna no kuse ni namaiki da youre
pretty uppity for a woman. The discriminatory content of these forms includes evalu-
ations of the looks or age of a woman, criticism or ridicule for womens delay or
failure to marry, presupposition of womens inferiority to men, and denial of womens
human rights, which derives from the old ie seedo household system.1
In this chapter, I examine three terms as representatives of sexist words, selected
from each of the three tiers of the Japanese lexicon, wago [native] Japanese words,
kango Sino-Japanese words, and gairaigo loanwords, the last most commonly
from Indo-European languages. These terms are: (1) onnadatera (ni) despite [being]
a woman, (2) rooba hag/crone, literally, old woman, and (3) oorudo misu old
maid, literally, old miss. An understanding of how these terms do or do not appear
in contemporary texts and dictionaries will give us a basis for assessing the status of
sexist language in Japanese more generally. I first discuss the three terms in ques-
tion, drawing on examples both of dictionary definitions and uses in newspapers and
novels.

9.1.1. Onnadatera (ni) despite [being] a woman

The term onnadatera ni, a wago native word, appears in two forms in dictionary
entries: onnadatera and onnadatera ni; there is no difference in meaning and I treat
them together here. This term comprises two morphemes, with the suffix -datera
attaching to the word onna woman. The suffix -datera is defined as follows.

Definition 1: Following a word that expresses sex, status, or position (tachiba),

[datera] expresses the sense that [something is] inappropriate or unbecoming to that
sex, status, or position. Onna_____ni despite [being] a woman], hooshi____ despite
[being] a Buddhist priest. (DJS 1995:1656)2

Onnadatera thus is a form used to rebuke or criticize a woman when her behavior
falls outside the framework of what men think is appropriate for women.
In the 1980s, examples such as (1) appeared in newspapers, showing that women
had not, up to that time, been considered fit to be politicians.
168 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

(1) Josee seejika o onnadatera ni to hyoosuru no wa hansee suru

hitsuyoo ga aru.
We need to stop thinking of women politicians as being so despite being women. (Asahi
Shimbun, July 12, 1987)

Even very recently, a student reported to me that when as a high school student she
aspired to college, her grandmothers reaction was Onnadatera ni daigaku nanzo
itte doo suru what will you(despite being) a womando going to college?
(personal communication, June 27, 2000). Such statements are clear expressions of
bias against women.

9.1.2. Rooba old woman (old crone)

Rooba old woman/hag is a kango Sino-Japanese word used to disparage or
reject a woman on the basis of her age. The basic meaning of rooba is found in de-
finition 2: other definitions (definition 3) offer a slightly different, more negative
interpretation.

Definition 2: An old woman, lit., a woman who has aged (SGS 1998: 1351)

Definition 3: A woman who clearly shows her age (SMKJ:5 1997: 1498)

The morpheme ba of rooba is also an element in such polymorphemic words as onibaba

witch, ogress and yaritebabaa procuress/Madam Go-getter, which are negative or
derogatory; ba is also repeated in the derogatory word babaa old woman/bitch. The
morpheme ba carries negative connotations, present in the word rooba as well.
A search of CD-ROM-ban shinchoo bunko no hyakusatsu A hundred volumes
of Shincho-paperback editions on CD-ROM (hereafter, Shincho CD) yielded ex-
amples such as ex (2).

(2) Shige wa mukashi no oku zutome no jochuu no hitori de atta ga, ima wa chiisaku
shinabita, kanari kenkai na rooba to natte ita.
Formerly, Shige was one of the inner house maids, but now she had become a with-
ered, petty, narrow-minded old woman. (Nireke no Hitobito The House of Nire, 1964
[1568]; Shincho CD)

Examples where rooba is used to create negative images of people support the claim
that it is an offensive word.

9.1.3. Oorudo misu old maid

The term oorudo misu old maid is a Japanese-made construction based on English
loanwords that conveys disdain for women who remain single after a certain age. It
has a long history, having been used since the Meiji period (18681912). It was an
entry in Meekai kokugo jiten (MKJ) (1997 [1943]), the first dictionary to focus on
contemporary Japanese (Muto 1997).
 WOMEN AND WORDS 169

Definition 4: (fr. Eng. old + miss) (made in Japan) A woman past marriageable
age. An old maid. High miss. [Translators note: High miss is another term de-
rived from English but made in Japan, also meaning old maid]. (KSK 1981: 136)

Hongo notes that when [a woman] remains single past a certain age, society sud-
denly turns against her, and starts calling her oorudo misu, urenokori, and the like
(1985: 80).

9.2. Onnadatera (ni), rooba, and oorudo misu

in the media

9.2.1. Onnadatera (ni)

A search of the novels in the Shincho CD yields examples such as ex (3) and (4).

(3) Ginko ga ichiban kurushinda no wa yahari onnadatera ni, tanshin, otoko dake no gakkoo
ni norikonda koto ni atta.
What Ginko was most troubled by was that, despite being a woman, she had pushed her
way into an all-male school. (Hanauzumi Buried in Flowers, 1970: 348; Shincho CD)

(4) Onnadatera ni kenjutsu no hoo wa sootoo na mono da ga . . .

Despite being a woman, [she] was pretty good at fencing and kendo . . . (Kenkaku
Shoobai The Swordsman Business 1973: 235; Shincho CD)

All such examples were produced nearly 30 years ago, and the novels were set in the
earlier Edo (16001868) and Meiji (18681912) periods.
In todays newspapers, onnadatera ni . . . is not used to criticize womens ac-
tivities. A search of the 19931998 Mainichi Shimbun Mainichi Newspaper based
on CD-Mainichi Shimbun CD-Mainichi Newspaper (Nichigai Asoshieetsu 1993,
1994, 1995, 1995, 1996, 1997, and 1998) yielded examples such as ex (5).

(5) Tora-san wa onnadatera ni e nan ka kaite ru yatsu ni, roku na no wa nai n da to kankan.
Tora-san was furious, [ranting,] Women who paint [lit., women who paint despite being
women] are just scribblers. (Mainichi Shimbun, September 8, 1996)

This, however, is a quotation from Tora-san (the main male character in a popular
film series) and not something a journalist wrote on her or his own. There are no
direct examples of a journalist using this word to characterize a woman in the entire
Mainichi corpus. Journalists and newspaper publishers, it appears, recognize the
sexism of this word.
Newspaper publishers issue usage handbooks, and in the eighth edition of one
such handbook, Kyodo Tsushinshas Kisha handobukku Journalists handbook
(2000), discriminatory or offensive terms to avoid are specified in various catego-
ries, such as disability and race. Under the heading seesabetsu Sex Discrimina-
tion, two words, joryuu female, woman and joshi Madame, are listed as forms to
170 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

avoid, with the additional comment that joketsu heroine, outstanding woman, jojoofu
tough and dependable woman, onnadatera ni, and the likewhich stress that the
person being referred to is femaleshould be avoided to the extent possible (Kyodo
Tsushinsha 2000:85). However, the category of sexist language did not appear until
the seventh edition, which appeared in 1997. Avoidance of sexist terms in news-
paper writing is a very recent phenomenon.

9.2.2. Rooba
Unlike onnadatera (ni), rooba is used even in contemporary newspapers when char-
acters in serialized novels are described (ex. 6), but such uses do not seem to have
negative connotations.

(6) Kinoshita Keesuke Kantoku Narayama Bushikoo (58nen) wa Tanaka Kinuyo ga

sashiba o 4-hon nuite rooba o enjita.
In Director Kinoshita Keisukes Ballad of Narayama (1958), Kinuyo Tanaka took out
four dental implants and played an old woman. (Mainichi Shimbun, March 5, 1996)

However, the use of rooba to describe older women, focusing attention on their
age, in actual news reports has changed greatly from 1945 to the present. Endo
(1988) examines the terms used to describe elderly women in the headlines from
Mainichi and Asahi newspaper metropolitan pages and regional editions over the
postwar years. In the late 1940s, rooba was used as in examples (7) and (8). But
from the late 1960s, usage shifted to terms such as roojo old woman or roofujin
old lady, as in examples (9) and (10).

(7) Rooba o koroshi, umeru

An old woman killed and buried (Asahi Shimbun, March 15, 1948)

(8) Roobagoroshi torawaru

The killer of an old woman captured (Mainichi Shimbun, September 16, 1948)

(9) Roojo futari hanerare shinu

Two old women struck by a car and died (Asahi Shimbun, evening edition, March 13,
1978)

(10) Roofujin no ie, tochi nottori

Old lady swindled of home, property (Mainichi Shimbun, June 4, 1978)

And today such persons are denoted by age, as in example (11).

(11) Nanajuuni-sai josee no itai suukagetsu hoochi

Body of a seventy-two-year-old woman undiscovered for several months (Mainichi
Shimbun, evening edition, June 7, 2000)

And, as in example (12), news articles appear in which only the term josee woman
is used in the headline and the age of the woman is noted only in the text.
 WOMEN AND WORDS 171

(12) Sakeobi jiko josee shinaseru

. . . oodanchuu no mushoku Uta Tami-san (78) ga . . . no kuruma ni
hanerare sokushi shita.
Alcohol-related accident kills woman
. . . Uta Tami-san (78, unemployed) was hit by a car driven by . . . and killed instantly.
(Asahi Shimbun, June 1, 2000)

In NET Iikaeshuu, an unpublished collection of alternatives to discriminatory

words circulated in 1974, rooba was described as discriminatory under section 1,
Words That Violate Human Rights, article c, Derogatory Words; replacement
of rooba with roojo or roofujin was recommended. And even in the 1975 edition of
Torikimeshuu Usage agreement, issued by Asahi Shimbun, one of Japans major
national newspapers, replacement of rooba with roojo or roofujin is specified. In the
face of these various cautions and guidelines, journalists no longer use rooba to de-
note elderly women.

9.2.3. Oorudo misu

Two examples of this word were found in the Mainichi CD database. In both ex-
amples, oorudo misu appears as a quotation and not as the words of the writer her- or
himself, as in example (13).

(13) Chuugokujin no kankaku de wa, kekkon tekireeki o mukaeta danjo no uchi, mikonsha
wa josee ni ooi to miraregachi de, yoron no doojoo mo oorudo misu ni atsumatte
ita ga, genjitsu wa igai ni mo gyaku datta to yuu wake.
The Chinese perception was that more women remained unmarried after a mar-
riageable age (kekkon tekireeki) and there was considerable public sympathy for these
old maids, but the reality, surprisingly, was the opposite. (Mainichi Shimbun,
October 23, 1993)

In the Shinchoo CD database, there are examples such as (14). This example is
drawn from the background description in a novel written in the 1950s but set in
premodern times.

(14) Korera no imeeji wa sono mama sakuhin no mochiifu to nari, sono mama hitori no
ningen no sugata o shoochoo shi, sorezore, shuui no hakugan ni taeru oita mekake,
rinshoku de kenkai na oorudo misu no sugata to kasanariau.
These images, just as they were, became the motif of the work, and, thus, came to
symbolize the figure [sugata] of a single being; the old mistress bearing up against
societys disdain and the picture of a miserly, narrow-minded old maid, respectively,
merged. (Asunaro Monogatari 1954: 378)

Gendai yoogo no kiso chishiki A basic encyclopedia of current terms (hence-

forth Gendai kiso), an annual collection of neologisms published by Jiyukokuminsha
provides another measure of a words popular currency. Appearance in Gendai
kiso represents a public acknowledgment of new terms currency and stability in
172 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

contemporary discourse. Until the 1960s, oorudo misu did not appear there. In the
1970s, there was a change of editor and a new chapter, Neologisms and Loanwords
in the Mass Media, was added. Oorudo misu was found among these loanwords
and neologisms and was defined as follows.

Definition 5: Old miss (Japanese-produced form). Spinster (a woman who remains

unmarried beyond marriageable age); correctly, old maid (Gendai Yoogo no kiso
1972 1972: 1346)

The entry remains unchanging through the 1990 edition, but in the 1991 edition
oorudo misu is defined differently, as follows.

Definition 6: * Old miss. Old failure; a colloquialism meaning old wound. (Gendai
Yooga no kiso 1991 1991: 1389)

We see, then, the following shift. Gendai kiso began publication in 1948 as a
collection of current terms, practical (common) terms, popular words, slang, loan
words, and abbreviations (from the preface to the first edition; Hasegawa 1948:1).
Oorudo misu was neither a new nor a current term and did not appear. After a change
of editors, it was included in the 1972 and subsequent editions until 1991. In that
edition, the definition was changed to the obscure old wound. After 1994, this entry
again disappeared.

9.3. Dictionary treatments

There is very little research on the relation between Japanese dictionaries and sexist
language (Endo 2000), but the role of dictionaries in combating sexist language should
not be overlooked. Decisions not to include sexist words in dictionaries may help
reduce these words vitality. Failing to find a word in the dictionary can send the
message that it is used so seldom as not even to merit an entry. Alternatively, an entry
that explicitly details the sexist nature of the word and explicates its connotations in
usage can send the message that the word should be used only cautiously or not at
all. Unfortunately, few of todays Japanese language dictionaries grapple seriously
with biased language.
How unfortunate the consequences of this failure can be became clear in April
2000, on the occasion of the sabetsu hatsugen racist remarks of Tokyo governor
Ishihara Shintaro in an address to the Nerima Corps of the Ground Self-Defense
Forces. Ishihara intentionally used the obsolete discriminatory word from Japans
colonialist past sangokujin third country people to refer to Chinese and Koreans.
When challenged by newspaper reporters, Ishihara declared that since the respected
dictionary Daijirin does not define the term as discriminatory or derogatory, it is not
a discriminatory term (Mainichi Shimbun, April 4, 2000). The dictionary served, thus,
as an authoritative text, reference to which invokes the speakers mastery of the
dominant system of legitimate language (Bourdieu 1991) and puts him beyond the
reach of reporters criticisms.
 WOMEN AND WORDS 173

I next analyze Japanese dictionaries treatment of sexist lexical items, using the
three sexist terms onnadatera ni, rooba, and oorudo misu. My corpus includes dic-
tionaries of modern Japanese published over nearly a 60-year-period, from Meekai
kokugo jiten (MKJ 1943) to Sanseedoo kokugo jiten dai5han (SKJ:5). It comprises
66 dictionaries, which range in size from the large (NKDJ:2, with 500,000 entries)
to the very small (CKJ, with about 40,000). Categorized by size, the corpus includes
2 large (more than 300,000 entries), 16 medium-sized (100,000300,000 entries),
and 48 small (fewer than 100,000 entries) dictionaries. See the appendix at the end
of this chapter for details.
My survey focuses on the following three questions: (1) is the word included (Q1),
(2) if included, are there examples of usage (Q2), and (3) does the entry include supple-
mentary explanations or guidelines with respect to meaning or usage (Q3)?

9.3.1. Onnadatera (ni)

This term is composed of onna woman + the suffix -datera (ni), so I consider the
treatment both of the word itself and of the suffix. The majority of the dictionaries
include an entry for each.

9.3.1.a. Onnadatera (ni)

Q1. Fifty-four of the 66 dictionaries in my corpus had an entry for onnadatera (ni).
Twelve, all small dictionaries, did not (IKJ:4, IKJ:5, KHK, GSK, SKJ:5, SGS, CKJ,
RSK:14). Reekai shinkokugo jiten (the RSK series, 19841994) has never included
onnadatera (ni). Perhaps the editors in 1984 found it obsolete already, or perhaps
they judged it of little use to their primary target users, middle school students.
Charenji kokugo jiten (CKJ), which similarly targets middle school students, also
does not include it. Then, too, these dictionaries are small, so the reason for non-
inclusion may be size.
The fifth editions of Iwanami kokugo jiten (IKJ:5 1994) and Sanseedoo kokugo
jiten (SKJ:5 2001) dropped onnadatera (ni), exhibiting a change of policy from
previous editions. IKJ:5 changed the example provided in the entry for -datera from
onna woman_____ni to kodomo child______ni, and the subsequent IKJ:6 con-
tinued with this substitution. In fact, the term onnadatera (ni) disappeared completely
from this dictionary series. Sanseedoo gendai shinkokugo jiten (SGS 1998) also
decided to drop this entry, a change from the dictionarys previous policy (SGK:2
1992). Gakken shinkokugo jiten (GSK) first appeared in 1994 and Kadokawa hikkee
kokugo jiten (KHK) in 1995; neither of these included the term, suggesting recogni-
tion of its sexist and obsolete character. One early dictionary, Meekai kokugo jiten
(MKJ 1943), also did not include this word, but in such an early publication one might
speculate that the reason was its vulgarity rather than its sexist nature.
In IKJ:5, SKJ:5, and SGS (dictionaries of the more-than-60,000-entry size),
onnadatera ni was omitted. When one contrasts this with the other dictionaries of
similar size that continue to include the same entries from previous editions, these
three dictionaries can be commended for their thoughtful treatment of discrimina-
tory words.
174 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

Q2. Recently dictionaries have given more weight to examples of usage than
formerly and have been marketed with promises of abundant examples of usage.
Examples of onnadatera (ni) in use, however, are extremely rare. Of the 54 dictio-
naries in my corpus that had an entry for onnadatera (ni), only 15 had examples of
its use: 2 large dictionaries (NKDJ, NKDJ:2), 8 medium-sized dictionaries (KKD,
GKD:1, GKD:2, GSN, KJE:4, KJE:5, DJR:1, DJR:2), and 5 small dictionaries (IKJ:3,
IKJ:4, OKJ:r, SCGK, RKJ).
And, looking at the examples given, with the exceptions of Nihon kokugo daijiten
(NKDJ) , Gakken kokugo daijiten (GKD:1 and 2), and Shinchoo gendai kokugo jiten
(SCGK), which offer concrete examples drawn from literary works, we find that only
constructed examples are provided. Further, these constructed examples center around
only three situations: [onnadatera ni] . . . oozake o nomu [Despite being a woman,
she] drinks heavily, . . . agura o kaku [. . . she] sits cross-legged, and . . . tanka
o kiru [. . . she] swears. In other words, only these three situations come to lexi-
cographers minds when constructing examples. The replication of these examples
in dictionaries produced separately and independently suggests just how very lim-
ited and biased this word is in contemporary Japanese.
Judgments on whether drinking heavily, swearing, and sitting cross-legged are
actions appropriate for a woman or not are formed in particular sociohistorical con-
texts. Today these actions are not uncommon among Japanese women and cannot
simply be regarded as unbecoming. Accordingly, such examples of usage can no
longer be considered apposite.
Q3. One might think that the more recent the dictionary, the more meticulous usage
labels and supplemental notes would be. However, in this study, we see that Reekai
kokugo jiten (RKJ 1955), one of the older dictionaries in the corpus, had usage labels,
while the newest, Shuueesha kokugo jiten: 2 (SEK:2 2000), had none. The presence of
usage labels or supplemental notes seems linked neither to the dictionarys publica-
tion date nor to its size. Rather, these decisions are related to the attitudes of the
dictionarys producers. Of those dictionaries with long publication histories, Shinsen
kokugo jiten: 7 (SSKJ:7), Shinmeekai kokugo jiten: 4 (SMKJ:4), and a few others may
be commended for adding notes on the discriminatory nature of sexist words.
When a dictionary includes a discriminatory word as an entry, it needs to specify
the history and circumstances of the words use. Some dictionaries have provided
such information in supplemental notes. Meeji shoin seesen kokugo jiten (MSS), for
example, incorporated a notes column and included the following supplemental in-
formation in its entry for onnadatera (ni),
Onnadatera ni Note: A term used to express the feeling that something is not ap-
propriate for a woman to do. Used to confine women within a narrow framework
[of suitable activities] and to ridicule or deride those women who attempt to move
beyond that framework. (1994: 170)

This is one of the most detailed and appropriate of descriptions of sexist language in
Japanese dictionaries currently being published. The Notes column expends three lines
to point out the words meaning, the speakers intentions when it is used, and the fact
that it was used in the past but not todaythat it is, in fact, an archaic word. MSSs
Notes demonstrate that even a small dictionary can provide supplemental guidelines,
 WOMEN AND WORDS 175

when seen as necessary, about the meaning or usage of sexist words. Other dictionar-
ies have supplemented their onnadatera ni entries with the following information:
(1) that previously one could say otoko man -datera despite being a man as well
(GKR:1 [first edition only]); (2) that onnadatera can be used critically (SSKJ:7, GSN,
DJS, DJR:1 [but not 2], RKJ, NKDJ, NKDJ:2); (3) that it can be used in censuring
(OHK, OHK:r, SGK:1, SGK:2); (4) that it can be used in ridicule or contempt (MSS);
and (5) that it can be mixed with criticism or amazement (SMKJ:4, SMKJ:5).
Relatively recent dictionaries such as Sanseedoo gendai kokugo jiten (first pub-
lished in 1988), Daijirin (1988), and Daijisen (1995) have included usage labels and
supplemental notes from their first editions. However, time has not improved the
picture in all cases. Gendai kokugo reekai jiten 1 (GKR:1 1985), for example, in-
cluded information about the counterpart word otokodatera despite being a man
and provided appropriate supplemental notes concerning the historical facts surround-
ing the word, but these were deleted in the revised edition (GKR:2 1993). Daijirins
notes concerning usage were also eliminated from its second edition, demonstrating
that some revisions represent changes for the worse. And, of course, it is regrettable
that few dictionaries followed Reekai kokugo jitens very early (1955) example of
including such guidelines.

9.3.1.b. Datera
Q1. All but three dictionaries (GKR:1, GKR:2, RSK:1) have entries for the suffix -
datera. Commendably, neither Gendai kokugo reekai jitens first nor second editions
(GKR:1, 2) carry an entry for this morpheme, reflecting its decline. However, although
Reekai shinkokugo jitens first edition (RSK:1) has entries neither for onnadatera
(ni) nor for -datera, in the second edition (RSK:2), an entry for -datera is added and
onnadatera ni is given as an example, resurrecting this discriminatory form.
Q2. Derivations that use the suffix -datera are extremely limited today. Ex-
amples drawn from classical texts include hooshidatera despite being a priest/
monk, oyadatera despite being a parent, keeseedatera despite being a beauty,
and choonindatera despite being a merchant; but examples given as contemporary
are restricted to two: kodomodatera despite being a child and onnadatera. Dictio-
naries that include an entry for-datera must, nonetheless, present examples of de-
rivative terms that use it, and of the 58 dictionaries that gave examples, 54 gave
onnadatera ni. Behind this frozen collocation of onna and the -datera suffix lurks a
discriminatory social stereotype, a stereotype that continues to be authorized by the
appearance of this form in dictionaries. Dictionaries failure to incorporate changes
in the word usage and their uncritical replication of material from earlier editions
contribute to reproducing past, discriminatory practices using the authority of lexi-
cographical expertise. While large and medium-sized dictionaries should include an
entry for this suffix, in order to acknowledge its historical existence they should also
provide adequate supplemental notes about the historical background of its use.
Q3. There were 23 small and 3 medium-sized dictionaries with some sort of usage
label or supplemental note for the suffix -datera. Small dictionaries such as Charenji
kokugo jiten (CKJ; see also RSK:14) are targeted at middle school users who are
likely to need some guidelines for these terms. These dictionaries do not have entries
176 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

for onnadatera (ni) and, thus, usage notes occur only with -datera. Examples of the
usage labels and supplementary notes for -datera include: (1) vulgar parlance/slang
(CKJ); (2) a lowbrow way of saying something (RSK:2, 3, 4); (3) a word used in
criticism, scorn (IKJ: 4, 5, 6; FKJ; OHK:r; GSK; SMKJ:1, 3, 3-3, 4, 5; GKD:1, 2;
KJE:5); (4) used in a critical sense (OKJ:9); (5) not used in a positive sense (SKJ:2
5); and (6) not used in a very positive sense (SGK 1, 2; SGS). The supplemental notes
for -datera differ from those for onnadatera (ni), which are restricted to women as
the object. For -datera, guidelines that focus on its generally pejorative nature (e.g.,
ii imi de wa tsukawanai not used in a positive sense) are more numerous.
We see, then, that both onnadatera (ni) and -datera are terms that have lost vital-
ity as contemporary forms to such an extent that no original examples can readily be
provided. Despite this, many dictionaries have entries for both. This may not be neces-
sary; in small dictionaries, an entry for one or, in fact, for neither would suffice. If entries
are to be included, however, it is imperative that dictionaries follow the example of
Meeji shoin seesen kokugo jiten (MSS 1994) and allocate space for usage labels and
supplemental notes that inform young users in particular about the discriminatory na-
ture of the words and draw attention to the problems attendant upon their use.

9.3.2. Rooba
Q1. Since this word is hardly used today, one might consider an entry unnecessary,
at least in small dictionaries. However, entries for rooba were found in all 66 dictio-
naries examined.
Q2. Dictionary makers apparently find examples of the use of this word unnec-
essary. Only two large (NKDJ, NKDJ:2) and one medium-sized (SCKJ) dictionary
provide examples of its useand those are examples from classical texts. There were
absolutely no examples of its use as a contemporary lexical item. This could be be-
cause: (1) the meaning of the word is so self-evident as to require no exemplifica-
tion; (2) the word is not actually in current use, so no appropriate examples exist; or
(3) the word is obsolete, so exemplification is pointless.
Q3. I found no supplemental notes that indicated this word was critical or de-
rogatory. Some dictionaries did provide notes concerning register: bun literary
(SKJ:24), bunshoogo written language (SGK:1, 2; SGS), and zokugo slang
(KKJ:1). Characterizing rooba as literary seems to be based on the fact that although
it is not used in everyday conversation, it isor, at least, has beenused in writing.
The label zokugo slang, however, can be interpreted as a claim that it is a vulgar usage
in spoken language and not usedor at least, no longer usedin ordinary written lan-
guage. Each of these characterizations undoubtedly reflects a partial truth.
An additional and interesting aspect of the definitions of rooba is that when they
include something like toshi o totta onna an old woman (lit., a woman who has aged),
one sees a shift over time from . . . onna woman , which has connotations of woman-
as-sexual, to other expressions, such as . . . josee woman or . . . onna no hito woman,
which convey much more positive images than onna, as Jugaku (1979) also notes.
Unmodified by such forms as toshi o totta old or by no hito person, onna stresses
the sexual aspects of womanhood; the relatively unrefined connotation of onna is miti-
gated by the addition of no hito in one common contemporary term for woman, onna
 WOMEN AND WORDS 177

no hito.3 The term fujin has slightly formal overtones, is used to denote women in pub-
lic discourse, and de-emphasizes womens sexuality. Josee is currently the most un-
marked term for women (Urushida 1993). Changes in dictionaries choices of words
for woman in definitions of rooba over time are displayed in table 9.1.
From Meekai kokugo jiten (MKJ 1943) through Koojien-2han (KJE:2 1969) only
onna was used, but in Shinmeekai kokugo jiten (SMKJ:1 1971) fujin was used and in
Reekai shinkokugo jiten (RSK:1 1984) onna no hito was used. Josee appears for the
first time in the first edition of Daijirin (DJR 1988) and in Sanseedoo gendai kokugo
jiten (SGK:1 1988). Most dictionaries in my corpus with first editions in 1990 or
later (e.g., KHK, SGS, SEK:1 and 2, SPK, CKJ, MSS, JR, DJS) used josee. Of newer
editions, only Gakken shinkokugo jiten (GSJ), with a first edition published in 1994,
has toshi o totta onna. The other six cases of onna occurring in definitions of rooba
in dictionaries published after 1990 are all in second or subsequent editions of a dic-
tionary and simply reproduce a first edition definition using onna. In the case of the
Iwanami kokugo jiten (IKJ) series, we see that IKJ:4, 5, and 6, all published after
1985, eliminated the entry onnadatera and changed the examples given in the entry
for -datera; despite this, for the definition for rooba, toshi o totta onna remains un-
changed from the first edition on. Koojien (KJE:15), too, uses onna from the first
through the fifth edition. However, usage has generally changed from . . . . no onna,
a form offensive to women, to . . . no josee or . . . no onna no hito, suggesting that
dictionary makers recognize the social changes in the attitudes toward and treatment
of women that have taken place in the latter part of the twentieth century.

9.3.3. Oorudo misu

Q1. Fifty-four of the dictionaries in my sample have entries for oorudo misu; 10 (IKJ:
56, OHK:r, KHK, CKJ, MSS, RSK: 14), all published after 1983, did not. Of these
latter 10, this represents a change from previous editions for 2 (IKJ:5, OHK:r). One
can speculate that publishers judged this word obsolete or sexist. Alternatively,
changes in Japanese views of marriagei.e., the growing number of women and men
who do not marry, the rise in average age of marriage, etc.may have had some
influence on publishers.
Q2. Only three dictionaries (RKJ, NKDJ, NKDJ:2) included examples of usage.
This is hardly surprising, since oorudo misu is widely recognized to be a sexist term.
Q3. This word is a collocation of English old and miss that was produced in
Japan. Where present, supplemental notes centered exclusively around that fact. There

9.1 Words used for woman in

definitions of rooba
Word Through 1989 1990

onna 25 8
josee 3 18
onna no hito 2 2
fujin 6 1
None 1 0
Total 37 29
178 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

was no mention whatsoever of its sexist nature. One does see, however, a shift in the
use of onna versus josee similar to that in the definitions of rooba in the definitions
of oorudo misu (as, for example, konki o sugite mo kekkon shinai de iru onna (josee)
a woman who remains unmarried even after the normal time for marriage) (see
table 9.2). Among those dictionaries first published in 1990 or later, only Gakken
shinkokugo jiten (GSK) used onna in the definitionas they did for roobaignor-
ing the tides of change. In the other dictionaries, josee was recognized as more ap-
propriate than onna.

9.3.4. Summary
In this chapter, I have examined the treatment of three words in terms of Japanese
dictionaries. These terms, which have almost completely fallen out of use in the media,
continuewith the exception of one group of improved dictionariesto appear
just as before even in small dictionaries. However, if dictionaries are to boast that
they are a mirror accurately reflecting the meaning of words (from the preface to
Sanseedoo kokugo jiten 3 [SKJ:3 1982:1]), many changes are needed before they
adequately reflect the changes in societys view of gender relations and the words
that encode them.
Medium-sized and large dictionaries can afford the space to record sexist and
other discriminatory words, but if they do, they have an obligation to describe the
history and discriminatory nature of the words and to make efforts to avoid lending
their authority to offensive language, as happened in the Ishihara declaration. The
majority of todays dictionaries are not useful mirrors that reflect society as it is
today. We have also seen, however, that some publishers produce dictionaries that
do reflect the changes in society. One strongly hopes that, in the future, the majority
of dictionary makers, who continue unreflectingly to reproduce outdated entries, will
learn from these latter volumes and make thorough revisions.

9.4. Changes in words for describing women

in public discourse and the media

In contrast to the conservative practices of most lexicographers and dictionary edi-

tors, Japanese public discourse and the Japanese media have come to avoid sexist

9.2 Words used for woman in

definitions of oorudo misu
Word Through 1989 1990

onna 23 4
josee 11 15
onna no hito 1 0
None 0 2
No entry 2 8
Total 37 29
 WOMEN AND WORDS 179

language. The 1970s saw the beginning of demands to eliminate the use of discrimi-
natory language in the mediaprimarily centered around buraku sabetsu discrimi-
nation against burakumin.4
Sexist language became another object of protest. The movement against sex-
ist language took on clearer focus in 1975, the International Womens Year. Fur-
ther, after the 1985 ratification of the United Nations Convention on the Elimination
of All Forms of Discrimination against Women, domestic laws had to be brought
into compliance, which resulted in the 1986 passage of the Equal Employment Op-
portunity Law. In order to implement this and other laws, womens liberation and
independence had to be promoted, and for that, the deep-rooted notion of danson
johi androcracy (lit., respect for men, contempt for women) had to be reformed.
Old attitudes inhere in old, biased words and impede reform. Hence, reevaluation
of sexist language was required.
Another stimulus to the reevaluation of sexist words was the changes in their
use by the media. Newspaper publishers and broadcasting companies produced hand-
books of iikae alternative wording for old, discriminatory words and tried to make
comprehensive changes. The medias reforms were in part voluntary and in part the
result of external forces. Takagi notes that the use of expressions which reflected
traditional views of women as inferior to men was protested by womens liberation
organizations, and the mass media came to avoid their use. . . . Roojo came to be
substituted for rooba, the use of which, it was feared, would hurt older women
(1988:605).
One of the womens liberation organizations was the Kokusai Fujinnen o Kikkake
ni shite Koodoo o Okosu Onnatachi no Kai The Womens Action Group Occasioned
by the International Womens Year,5 which criticized a food company 1975 com-
mercial Watashi tsukuru hito, boku taberu hito I [feminine first-person pronoun]
make it, I [masculine first-person pronoun] eat it. Then, in 1989, the Japan Womens
Studies Association presented a list of demands to each private television station and
NHK Japan Broadcasting Corporation, Japans public television station. The asso-
ciation demanded that these media outlets cease using expressions that derogated
women or fostered sex discrimination. Among such expressions were those that con-
noted that womens value depended primarily on their appearance.
It is unclear whether media shifts in terminology were caused by these protests
and demands by womens organizations or the media voluntarily added sexist lan-
guage to their revisions of discriminatory words in general in the wake of the Buraku
Liberation Leagues censure, but the activities of womens organizations were widely
publicized and clearly influential in changing media practices (see also chapter 1,
this volume). As seen from the newspaper and fictional examples presented earlier,
media use of sexist terms has changed to suit the changing social landscape in ways
that dictionary treatments of the same terms have not.

9.5. Conclusion

My analysis of media and dictionary materials has shown that, over the course of the
late twentieth century, media have progressed much further toward eliminating sexist
180 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

language than have dictionaries. Feminist representations to the media have been
influential in bringing attention to sexist linguistic practice. What has played an even
larger role, however, seems to be the governments reevaluation of the biased lexicon.
In this regard, a particularly important role was played by the Kanagawa Womens
Plan (August 1984), which called for reexamination of all publications produced under
prefectural auspices with regard to (1) language that discriminates against or dero-
gates women, (2) passages that implicitly valorize traditional, fixed gender roles, and
(3) words and phrases for family, married couples, women, and men that evoke the
old patriarchal family system. This movement influenced other regional and local
governments, and revision movements flourished.
Then, in 1987, the Headquarters for Womens Issues (headed by the prime min-
ister) agreed upon the New National Action Plan toward the Year 2000, which made
clear, from a long-term perspective, the fundamental direction in which policy to-
ward women should be geared. This plan had the integrated goals of eliminating ideas
about fixed sexual divisions of labor, enabling women to realize their full potential
by creating social conditions within which women can participate fully in society,
and working toward the formation of a society where women and men can equally
contribute to development and stability.
The governments Headquarters for Promoting Gender Equality (Danjo Kyoodoo
Sankaku Suishin Honbu) issued their Plan for Gender Equality 2000 in 1996.
Among its 11 major aims were reconsideration of existing social systems and
practices from the perspective of equal participation, and enhancement of public
awareness and respect for the human rights of women in the media. In terms of
concrete measures, this plan included language reform through their promotion
of the use of non-sexist phrases in all informational and other publications pro-
duced by public agencies. To that end, the plan encouraged all regional and local
administrative offices to produce guidelines. Following this, Hyogo Prefecture
produced a Hyoogen handobukku Phrase handbook (Hyogo-kenritsu Josei Sentaa,
1996), Tokyo produced Danjo byoodoo shakai e no michisujiGaidorain The
route to a gender equal societyguidelines (Tokyo-to Seikatsu Bunkakyoku,
1995), and Kanagawa Prefecture produced Media to josee no jinken Media and
the human rights of women (Kanagawa Josei Sentaa, 1997), among others, appeal-
ing to the public to rethink their everyday language practices. For example, Media
to josee no jinken points out that the stereotypical picture of a housewife often
portrayed by the media reproduces and reinforces the gendered division of labor,
and urges women to become speaking subjects to empower themselves (1428).
In 1999, the Basic Law for a Gender-equal Society was implemented. It stipulated
that the formation of a gender-equal society shall be promoted based on respect for
the human rights of men and women, including: no gender-based discriminatory treat-
ment of men or women; and the securing of opportunities for men and women to exer-
cise their abilities as individuals (Law No. 78, article 3).
In this way, with the government taking a leadership role, the concrete revi-
sion of language has proceeded. This, one may argue, has resulted in mass media
reducing their use of sexist language. Womens movements in Japan, however, have
been a significant force that causes the government to act. Even though they may
not themselves have taken language reform as a direct goal, they worked to push
 WOMEN AND WORDS 181

for governmental action. The fruits of these efforts are slowly emerging in media
practice. Activism seems not, however, to have influenced the production of dic-
tionaries in the same way. We have seen that changes in Japanese dictionaries are
rather roundabout and are too slow to be called reforms. The dictionary provides
speakers with a list of authorized meanings for words; it gives them a fixed and
timeless meaningtimeless, that is, at least until the dictionarys next edition.
Speakers can and do use dictionary entries to make and, at times, to justify their
lexical choices; it is incumbent upon lexicographers and dictionary makers, there-
fore, to take adequate care not to perpetuate past social injustices in the lexical realm.
Including sexist terms such as onnadatera ni, rooba, and oorudo misu without in-
dicating their sexist natureno matter how dictionary makers may claim they are
just providing socially neutral semantic descriptions of such wordsfails to cap-
ture the damaging consequences of their use in real interactional contexts. Further,
the very claim that dictionaries should be excused from reform efforts because they
simply describe word meanings rather than prescribe their uses ignores the fact of
their textual authority. And that textual authority is political rather than linguistic
(Cameron 1985). It is regrettable that dictionariesthe ultimate authoritative
source for words and their usesare not more help in the process of lexical re-
form. Still, we must keep firmly in mind that the strength of readers and society is
essential to the project of changing the attitudes of editors and lexicographers. We
must continue to call problems in the choice, definition, and exemplification of
words to their attention.

. List of dictionaries
Name Code Publisher Date # Words

A. Small Dictionaries
Iwanami kokugo jiten: 1 IKJ:1 Iwanami Shoten 1963 > 57,000
Iwanami kokugo jiten: 2 IKJ:2 Iwanami Shoten 1971 > 57,000
Iwanami kokugo jiten: 3 IKJ:3 Iwanami Shoten 1979 > 57,000
Iwanami kokugo jiten: 4 IKJ:4 Iwanami Shoten 1986 > 57,000
Iwanami kokugo jiten: 5 IKJ:5 Iwanami Shoten 1994 > 57,000
Iwanami kokugo jiten: 6 IKJ:6 Iwanami Shoten 2000 63,000
Oobunsha kokugo jiten, shinteiban OKJ:r Obunsha 1973 ~ 76,000
Oobunsha kokugo jiten: 9 OKJ:9 Obunsha 1998 81,500
Oobunsha hyoojun kokugo jiten OHK Obunsha 1965 ~ 41,000
Oobunsha hyoojun kokugo jiten, shinteeban OHK:r Obunsha 1991 43,000
Kadokawa shinkokugo jiten KSK Kadokawa Shoten 1981 ~ 70,000
Kadokawa hikkee kokugo jiten KHK Kadokawa Shoten 1995 52,000
Gakken shinkokugo jiten GSK Gakushu Kenkyusha 1994 ~ 65,000
Gendai kokugo reekai jiten GKR:1 Shogakkan 1985 ~ 65,000
Gendai kokugo reekai jiten: 2 GKR:2 Shogakkan 1993 ~ 65,000
Koodansha kokugo jiten, shinban KKJ:1 Kodansha 1981 > 73,000
Koodansha kokugo jiten: 2 KKJ:2 Kodansha 1991 > 76,000
Sanseedoo kokugo jiten: 2 SKJ:2 Sanseido 1974 62,000
Sanseedoo kokugo jiten: 3 SKJ:3 Sanseido 1982 62,000
Sanseedoo kokugo jiten: 4 SKJ:4 Sanseido 1992 73,000
Sanseedoo kokugo jiten: 5 SKJ: 5 Sanseido 2001 76,000
continued
182 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

. continued
Name Code Publisher Date # Words

Sanseedoo gendai kokugo jiten SGK:1 Sanseido 1988 ~ 60,000

Sanseedoo gendai kokugo jiten: 2 SGK:2 Sanseido 1992 ~ 62,000
Sanseedoo gendai shinkokugo jiten SGS Sanseido 1998 ~ 70,000
Shuueesha kokugo jiten SEK:1 Shueisha 1993 92,000
Shuueesha kokugo jiten: 2 SEK:2 Shueisha 2000 ~ 92,000
Shuueesha poketto kokugo jiten SPK Shueisha 1996 72,000
Shinsen kokugo jiten, shinban SSKJ:1 Shogakkan 1974 ~ 73,800
Shinsen kokugo jiten: 6 SSKJ:6 Shogakkan 1987 ~ 81,500
Shinsen kokugo jiten: 7 SSKJ:7 Shogakkan 1994 ~ 83,000
Shinchoo gendai kokugo jiten SCGK Shinchosha 1985 77,000
Shinmeekai kokugo jiten SMKJ:1 Sanseido 1971 68,385
Shinmeekai kokugo jiten: 2 SMKJ:2 Sanseido 1974 68,386
Shinmeekai kokugo jiten: 3 SMKJ:3 Sanseido 1981 ~ 70,000
Shinmeekai kokugo jiten: 3, 3rd printing SMKJ:33 Sanseido 1985 ~ 70,000
Shinmeekai kokugo jiten: 4 SMKJ:4 Sanseido 1989 ~ 73,000 +
Shinmeekai kokugo jiten: 5 SMKJ:5 Sanseido 1997 ~ 73,000
Charenji kokugo jiten CKJ Fukutake Shoten 1991 ~ 40,000
Deerii konsaisu kokugo jiten: 2 DKK:2 Sanseido 1995 70,000
Fukutake kokugo jiten FKJ Fukutake Shoten 1989 60,000
Meekai kokugo jiten MKJ Sanseido 1943 ~80,000
(Reproduction edition, 1997)
Meekai kokugo jiten, Kaiteeban MKJ:r Sanseido 1952 unknown
Meeji shoin seesen kokugo jiten MSS Meiji Shoin 1994 ~ 50,000
Reekai kokugo jiten RKJ Chukyo Shuppan 1955 > 40,000
Reekai shinkokugo jiten RSK:1 Sanseido 1984 > 40,000
Reekai shinkokugo jiten: 2 RSK:2 Sanseido 1987 45,000
Reekai shinkokugo jiten: 3 RSK:3 Sanseido 1990 46,000
Reekai shinkokugo jiten: 4 RSK:4 Sanseido 1994 48,000

B. Medium-size Dictionaries
Kadokawa kokugo daijiten KKD Kadokawa Shoten 1982 ~ 155,000
Gakken kokugo daijiten GKD:1 Gakushu Kenkyusha 1953 ~ 100,000
Gakken kokugo daijiten: 2 GKD:2 Gakushu Kenkyusha 1988 102,000
Gensen GSN Shogakkan 1986 150,000
Koojien KJE:1 Iwanami Shoten 1955 ~ 200,000
Koojien: 2 KJE:2 Iwanami Shoten 1969 ~ 200,000
Koojien: 3 KJE:3 Iwanami Shoten 1983 > 200,000
Koojien: 4 KJE:4 Iwanami Shoten 1991 ~ 220,000
Koojien: 5 KJE:5 Iwanami Shoten 1998 ~ 230,000
Shinchoo Kokugo JitenGendai, Kogo SCKJ:1 Shinchosha 1965 unknown
Shinchoo Kokugo Jiten -Gendai, Kogo: 2 SCKJ:2 Shinchosha 1995 > 130,000
Jirin JR Sanseido 1993 unknown
Daijisen DJS Shogakkan 1995 > 220,000
Daijirin DJR:1 Sanseido 1988 ~ 220,000
Daijirin: 2 DJR:2 Sanseido 1989 ~ 223,000
Nihongo daijiten NDJ Kodansha 1989 175,000

C. Large Dictionaries
Nihon kokugo daijiten NKDJ Shogakkan 1973 450,000
Nihon kokugo daijiten 2-han NKDJ:2 Shogakkan 2000 500,000

Key: r = revised.
 WOMEN AND WORDS 183

Notes
1. The ie household system describes a patrilineal stem family/corporate unit, contin-
ued through male primogeniture; until World War II, the household head enjoyed consider-
able power and authority over other members of the unit. The system was abolished upon the
enactment of the postWorld War II Civil Code, but the ie construct continues to have con-
siderable cultural force even today.
2. All definitions and examples are translated from the original Japanese. Each dictio-
nary examined is identified by a code provided in the appendix in this chapter.
3. The relation of the unrefined and sexual connotations of onna to connotations of class
position is interesting but falls outside the scope of this study.
4. In the Tokugawa period (16031868), a system that assigned persons in specific
professions to a status lower than the general population was institutionalized; these people
were forced to live in certain restricted areas (buraku) and were called burakumin people
of the buraku. This system was abolished in the Meiji period (18681912), but bias against
residents of the areas associated with this status and people who come from those areas
remains.
5. This group changed its name to Koodoo suru Onnatachi no Kai Womens Action
Group in April 1986, and in 1996, their original goals deemed accomplished, they disbanded.

References
Bourdieu, Pierre (1977). The economics of linguistic exchanges. Social Science Information
16(6): 645668.
(1991). Language and symbolic power. Cambridge, MA: Harvard University Press.
Cameron, Deborah (1985). Feminism and linguistic theory. New York: St. Martins Press.
CD-Mainichi Shimbun (CD-Mainichi Newspaper) (19931998). Tokyo: Nichigai Asoshieetsu.
CD-ROM-ban shinchoo bunko no hyakusatsu (A hundred volumes of Shincho-bunko on
CD-ROM) (1995). Tokyo: Shinchosha.
Endo, Orie (1988). Oi o arawasu kotoba (2)Shimbun ni miru sengo 40nen no suii (Words
describing old age (2)Transitions in 40 years of postwar newspapers). Kotoba 9:
4161.
(2000). Ninki dorama no hanashikotoba ni miru seesa: TV dorama Byuutifuru raifu
no mojika shiryoo kara (Sex differences in the dialogue of popular dramas: From the
script of the TV drama Beautiful life). Kotoba 21: 1323.
Gal, Susan (1989). Language and political economy. Annual Review of Anthropology 18: 345
367.
Gendai yoogo no kiso chishiki 1972 (A basic encyclopedia of current terms, 1972) (1972).
Tokyo: Jiyukokuminsha.
Gendai yoogo no kiso chishiki 1991 (A basic encyclopedia of current terms, 1991) (1991).
Tokyo: Jiyukokuminsha.
Hasegawa, Kunio (ed.) (1948). Gendai yoogo no kiso chishiki (A basic encyclopedia of cur-
rent terms). Tokyo: Jiyukokuminsha.
Hongo, Akemi (1985). Shoojo kara musume e no shoo (From little girl to young lady). In
Kotoba to Onna o Kangaeru Kai (eds.), Kokugo jiten ni miru josee sabetsu (Sexism in
Japanese dictionaries), 6788. Tokyo: San-ichi Shobo.
Jugaku, Akiko (1979). Nihongo to onna (The Japanese language and women). Tokyo: Iwanami
Shoten.
Kyodo Tsuushinsha (eds.) (2000). Kisha handobukku dai8han (Journalists handbook, 8th
edition). Tokyo: Kyodo Tsushinsha.
184 LINGUISTIC IDEOLOGIES AND CULTURAL MODELS

Muto, Yasuo (1997). Kaisetsu (Commentary). Meekai kokugo jiten: Fukkokuban (Meikai
Japanese dictionary: Reproduction). Tokyo: Sanseido.
Takagi, Masayuki (1988). Fukaigo (Offensive words). In H. Kindaiichi et al. (eds.), Nihongo
hyakka daijiten (The encyclopedia of Japanese), 605. Tokyo: Taishukan Shoten.
Urushida, Kazuyo (1993). Fujin, onna, josee: Onna no ippan koshookoo (Matron,
woman, lady: Thoughts on common terms for women). In K. [Reynolds] Reinoruzu-
Akiba (ed.), Onna to nihongo (Women and the Japanese Language), 123158. Tokyo:
Yushindo.
 PART III

REAL LANGUAGE, REAL PEOPLE

This page intentionally left blank
 10

YUKAKO SUNAOSHI

Farm Womens Professional

Discourse in Ibaraki

R ecent studies on language and gender in Japanese (e.g., Okamoto 1995, Inoue
2002, chapter 3, this volume) suggest that Japanese womens language as charac-
terized in terms of a set of linguistic features in Standard Japanese (SJ)1 is in fact a
prescribed norm and thus does not represent the way Japanese women actually use
language. While this realization is an important step in the study of womens lan-
guage use in Japanese, it must also be pointed out that we still know relatively little
about the real language practice of Japanese women and men, who are socially di-
verse and heterogeneous, and that its investigation entails a consideration of how
gender interacts with other social variables, such as region, class, and occupation.2
Although the image of the stereotypical Japanese woman is associated with the wife
of a businessman (the so-called saraiiman white-collar worker, lit. salaryman) or
with an educated urban middle-class woman, the lives of women who do not belong
to this category have been increasingly studied by scholars from anthropological and
sociohistorical perspectives (e.g., Dore 1978, Kondo 1990, Hunter 1993, Roberts
1994). These studies show that Japanese womens lives are truly diverse. Unfortu-
nately, however, none of these studies has investigated how such diversity manifests
itself in womens language use or how it is related to the notion of Japanese womens
language. Whether we say that Japanese womens language is an ideological norm
or an (at least partial) reality for some Japanese women, Japanese womens language
practice is still viewed from the perspective of highly educated women (and men)
who are living in or aspiring to be associated with Tokyo. However, for women who
are outside this category, and especially for those who reside in regional Japan, Japa-
nese womens language most likely plays a different role, if any at all. As Bucholtz

187
188 REAL LANGUAGE , REAL PEOPLE

(1999) pointed out, the field of language and gender has traditionally exoticized lin-
guistic others who are not part of the dominant group. The case of the Japanese
language is no exception. In fact, studies of linguistic others have barely started
with respect to Japanese womens language use. What needs to be done now is to
analyze and understand the speech of those linguistic others in their own right in-
stead of treating them as deviant examples. For example, women who live in regional
Japan and who are in working-class occupations may not use features of Japanese
womens language, not because they are unable to master them but because fea-
tures of Japanese womens language are simply not in their daily language reper-
toire. Nevertheless, the speech of these women could be regarded by those who closely
associate themselves with Japanese womens language as marginal, of lower quality,
or even not really part of the speech of real Japanese women.
This chapter investigates the language use of four women engaged in agricul-
ture in rural Ibaraki Prefecture in Japan. In order to understand the womens dialect-
dominant speech, the relationship between regional dialects and SJ is reviewed first
from a historical perspective. Then I analyze tape-recorded interactions of three
Ibaraki female farmers and an Agricultural Extension adviser. My analysis reveals
that these women used few, if any, Japanese womens language features to con-
struct their gender; instead, they used dialectal and SJ features to negotiate their re-
lationships and interactional outcomes. The findings of this chapter suggest that
Japanese womens language is not shared by all women as an ideology and hence
not practiced by all women.

10.1. Historicizing and diversifying the research

of language and gender

It is well documented that the process of selecting a standard variety for Japanese
speakers was a painstaking and controversial one about which a number of scholars
and policy makers had argued various possibilities since the late nineteenth century
(Shibata 1980, Sanada 1991, Lee 1996). Among those who discussed the selection
of SJ was the influential professor Ueda Kazutoshi. He suggested in 1895 that SJ
should be based on the Tokyo elites language. The very fact that a number of intel-
lectuals had to repeatedly state that SJ should be based on Tokyo speech shows how
prevalent regional dialects were in the lives of Japanese (Lee 1996).
While SJ was gradually implemented in elementary education across Japan since
the early twentieth century, the strong antidialect attitude called the hoogen bokumetsu
undoo dialect extermination movement created resentment on the part of dialect
speakers (Sanada 1991). In the last few decades SJ has become much more acces-
sible and familiar to Japanese all over the country. At the same time, the overall at-
titude toward dialects has softened and using ones own regional dialect remains an
essential part of a Japanese persons daily speech. In contemporary Japan, people
commonly use their own regional dialects and SJ in a diglossic manner (Sanada et al.
1992). However, as Okamotos (2000) data shows, SJ is not in complementary dis-
tribution with a regional dialect in any given persons speech. Rather, the two tend
to be mixed in a complex manner in one utterance, depending on the setting and the
 FARM WOMEN S PROFESSIONAL DISCOURSE IN IBARAKI 189

addressee.3 It is in this context that the dialect-dominant speech of the four women
in this chapter can be better understood.
Currently morphological and lexical features of Japanese womens language
(as well as those of Japanese mens language) are commonly treated as if they were
part of (spoken) SJ grammar, as can be seen in Japanese language textbooks for for-
eign learners (Siegal & Okamoto 1996). Extensive use of Japanese womens lan-
guage features may, however, give the impression that the speaker is a traditional
housewife from Yamanote. Yamanote is the western part of Tokyo, a middle- and
upper-class residential area, whereas the eastern part, Shitamachi, is associated with
lower-middle-class and blue-collar workers (Hibiya 1988). However, the Yamanote/
Shitamachi designation has now become symbolic of class divisions rather than a
strict geographic division (Lebra 1993), and a distinct speech style is associated with
each. Miyako Inoue (2002, chapter 3, this volume) regards Japanese womens lan-
guage as a product of Japans modernization, part of which started to emanate via
the kateeshoosetsu domestic novel in the late nineteenth century. Although more
investigation into this historical process is needed, it seems safe to say that SJ was
modeled after Yamanote speech and that the current Japanese womens language
evokes the image of a stereotypical Yamanote woman. It can also be said that Japa-
nese womens language is the feminine version of standard spoken Japanese.4 Ac-
cordingly, it is understandable that womens language use in regional Japan exhibits
a completely different pattern from Japanese womens language.
There are at least two possibilities regarding the relationship between Japanese
womens language and the language use of women in regional Japan. The first pos-
sibility is that such womens use of Japanese womens language is proportional to
the degree of their use of SJ. The second possibility is that their dialect has its own
system of gender marking. The reality may differ from community to community
and from individual to individual. M. Inoue (in preparation) mentioned a female
consultant who worked in Tokyo but was originally from regional Japan. For her,
the use of Japanese womens language features demonstrated her urbanization rather
than her femininity. Likewise, other women in regional Japan may not consider their
use of Japanese womens language an essential part of being a womanly woman.
As for the two possibilities mentioned earlier, neither was confirmed by the current
data. The women whose speech is analyzed here exhibited virtually no morphologi-
cal or lexical features of Japanese womens language, though there were features
of SJ in their speech. Their speech needs to be compared with that of female aspirers
in Ibaraki, who align with Tokyo-based hegemonic feminine identities and use rela-
tively more SJ features. The current data did not conclusively show whether or not
their dialect had its own gender-marking system. These results suggest that the role
of gender norms for speech should be reconsidered, since they may not be uniformly
relevant to all Japanese women.

10.2. Methodology

This chapter analyzes interactions between an Agricultural Extension adviser and

three farm women. The data was drawn from conversations that took place in a
190 REAL LANGUAGE , REAL PEOPLE

farming village in Ibaraki Prefecture during the summer of 1994, supplemented by

follow-up interviews with Hayashi-san, the Agricultural Extension adviser (the most
recent interview took place in late 2002).5 Ibaraki is my home prefecture, which was
helpful in finding study participants and understanding the dialect.6 I accompanied
Hayashi-san as she conducted her routine visits to the family farm of each of the three
women. Hayashi-san had mentioned to the three women earlier that someone inter-
ested in her work and their language would come with her. Each visit was 30 min-
utes to one hour in length, and the conversations were tape-recorded. Hayashi-san
talked mainly with the women, but their husbands were also present or joined them
later in the conversation.
Ibaraki is an agricultural prefecture located approximately two hours north of
Tokyo, with its own dialect called Ibaraki-ben Ibaraki vernacular or Ibaraki hoogen
Ibaraki dialect. All four women have lived in Ibaraki throughout their lives. Hayashi-
san is in her early fifties, and has worked as an adviser for 30 years in different local
offices in Ibaraki Prefecture. After high school she attended a specialized school to
study agriculture in order to become an adviser. The three women interviewed here,
Hada-san, Oki-san, and Nagano-san, are all in their forties, and grew up in or near
the community where they currently live. All of their parents are sengyoo nooka full-
time farmers, and these women and their husbands are full-time farmers as well.
These women and their husbands all completed high school, and all three couples
were married with the help of matchmakers. The main reason that the women mar-
ried into farm families was, according to Hayashi-san, that they (and their parents)
thought they would not suffer from poverty, since a farm family always has land and
a house. The Hadas grow rice, tomatoes, and lettuce; the Okis, roses and rice; and
the Naganos, maitake mushrooms and rice.
Although the three womens profiles are typical of those of traditional farmers
wives (nooka no yome), they also hold leading positions in their farming commu-
nity. Hada-san is president of Seekatsu Kaizen Guruupu, Local Life Improvement
Group, a womens group, which promotes the communitys products locally and
fosters communication with other members of the community. Oki-san and Nagano-
san are 2 of the 30 fujin noogyoo shi female agricultural specialists in the prefec-
ture, who play a role as community leaders. These positions, which show active
involvement in the farming community, are some of the recent outcomes of the ef-
forts of the adviser and the farm families to elevate womens status in their commu-
nity. Before examining these womens interactions, in the next section, I briefly
describe attitudes toward and features of Ibaraki dialect (henceforth, ID).

10.3. Language attitudes and dialect features

10.3.1. Attitudes toward ID
Ibaraki is one of the northernmost prefectures of the central Kanto region (to which
Tokyo belongs), located at the border with the Tohoku region, an agricultural area
located at the northern part of the main island. Linguistically, ID is considered much
closer to Tohoku dialects than to other varieties spoken in the Kanto region (Kindaichi
 FARM WOMEN S PROFESSIONAL DISCOURSE IN IBARAKI 191

1977[1957]). According to Fumio Inoues (1989) survey of Japanese attitudes

toward regional dialects, ID was rated close to Tohoku dialects, which were given
the most negative evaluation of all dialects examined with regard to emotion- and
intelligence-related associations, receiving evaluations such as zonzai rough,
wakai josee ni fusawashiku nai not suitable for young women, and kikitori nikui
difficult to comprehend.
The use of ID may strike Japanese hearers, whether or not they themselves are
speakers of ID, as tsuchi-kusai earthy and inakappoi countrified. As Labov (1972)
said, users of stigmatized features themselves can hold a negative attitude toward the
use of those features. Speakers of ID are no exception. At the same time, as the current
data shows, this does not necessarily mean that all speakers of ID attempt to speak in
SJ. In other words, at one end of the scale are aspirersthose in Ibaraki who are aware
of this negative image and aspire to the values and speech of the Center.7 At the other
end of the scale are those who have much more locally centered values and norms, like
the participants in this study. While they are capable of comprehending SJ in televi-
sion news and in newspapers and other print materials, they do not seem to be con-
cerned about talking like the people of Tokyo. That is, in formal settings they use what
they consider standard features in (at least some parts of their) speech (cf. Sanada
1996). Yet their speech overall, regardless of its level of formality, usually retains more
dialectal features than the speech of aspirers. It seems that those who are engaged in
agriculture tend to have this latter, locally centered mentality.

10.3.2 Features of ID
Hoogengaku dialectology is an established field in Japanese linguistics, and most
major regional dialects, including ID, have been documented and described using
data from representative speakers of these dialects. Table 10.1 below lists some rep-
resentative features of the dialect. Example (1) illustrates these features with excerpts
taken from my own data.

(1) (a) Voicing of stops

[Hayashi-san] . . . shigodo yaruttsuttatte igenai wake dakedo . . .
[SJ equivalent] . . . shigoto yarutte ittatte ikenai wake dakedo . . .
[English translation] . . . (we) cant go to work even if (we) were to work
(b) Long vowels
[Nagano-san] . . . mi:nna iretchaa . . .
[SJ equivalent] . . . minna irechau . . .
[English translation] . . . (Ill) go ahead and put everything . . .
(c) /i/ - /e/Vowel merger
Shokuin workers and shokuen salt as pronounced by ID speakers sound virtu-
ally identical.
(d) Deletion of question/confirmation marker
[Hayashi-san] . . . a soo na no ke (falling intonation)
[SJ equivalent] . . . a soo na no (falling intonation)
[English translation] . . . is that right?
192 REAL LANGUAGE , REAL PEOPLE

10.1 Representative features of Ibaraki dialect

Features of Ibaraki dialect Remarks

Phonological characteristics
(a) Voicing of /k/ and /t/ word-internally and word-
finally (Miyajima 1961, Kanesawa 1984)
(b) Diphthongs such as /au/, /ai/ tend to become long /ai//ee/ is mentioned in Ide (1990) as
vowels /aa/, /ee/ respectively (Kanesawa 1984) one of the vulgar expressions that
women cannot use in Standard Japanese/
Japanese womens language
(c) Vowels /i/ and /e/ lose their distinction and both
become centralized (Miyajima 1961, Kanesawa
1984)
(d) Sentences have final rising intonation while the
pitch-accent pattern of the rest of the sentence is
flat (ikkee akusento flat/single-pattern pitch-
accent) (cf. Kanesawa 1984 and Kindaichi 1995
[1948])

Morphological characteristics
(e) Question marker -ke with a rising intonation
Confirmation marker -ke with a falling intonation
(Kanesawa 1984)

Stylistic characteristics
(f) Absence of honorifics, except for a few addressee Ibaraki is known to be an honorific-less
honorifics (Miyajima 1961, Kanesawa 1984, area (mukeego chitai)
Yamada 1986, Taguchi 1995 [1944])

Among the features in table 10.1, (c) and (d) are not indicated in the transcripts in
the examples used in this chapter, although both features were constantly present
throughout the recording.8 In particular, as akusento pitch-accent pattern is said to
be the most difficult area to standardize (Kindaichi 1977), the speakers ikkee-akusento
flat/single-pattern pitch-accent speech pattern is evident, making their speech sound
more dialect-dominant than may sometimes appear in the transcript. The four
womens speech consisted of features of both ID and SJ, and it was sometimes dif-
ficult to separate them neatly (see Okamoto 2000 for similar examples). Neverthe-
less, overall their speech would be considered relatively heavy ID not only by outsiders
but also by Ibaraki people themselves, especially those who live in cities.
The absence of honorifics, as described in (f) in table 10.1, needs to be discussed
further. In general, the presence or absence of honorifics in a particular dialect de-
pends on how historically complex (that is, class- or caste-stratified) the region has
been (Yamada 1986; cf. Miyake 1995). For example, Kyoto dialect, the dialect of
the historical capital, is known to have an elaborate system of honorifics. The exis-
tence of higher and lower caste speakers in Kyoto contributed to the dialects devel-
opment of honorifics. As Ibaraki has mainly been an agricultural prefecture, the
simplicity of its system of honorifics makes sense according to this logic.9 In the
 FARM WOMEN S PROFESSIONAL DISCOURSE IN IBARAKI 193

grammar of ID, there are no referent honorifics and only a few occasional examples
of addressee honorifics (e.g., nansho please, as in asobi ni ki-nansho Please come
visit; however, none was observed in the current data).
This near absence of honorifics in the dialect leads one to think that gender dif-
ferences in honorific use are unlikely to be observed in ID. According to Ide (1990),
one characteristic of Japanese womens language is womens more frequent use
of honorifics than mens as a result of using politer forms to the same addressee. But
if honorifics are hardly available to any speaker in the Ibaraki repertoire, then this
characteristic of womens honorifics use would not hold in ID. However, nowadays
it is difficult to find speakers who use exclusively dialectal features in their speech,
due to their exposure to SJ through education and the mass media. Thus, speakers in
Ibaraki have access to honorifics in SJ, as seen later in example (4). Consequently, it
is possible that the women and men of Ibaraki may employ these honorifics in SJ to
varying degrees as a way of gender marking.
There is no mention of gender differences in ID in the literature. This is a curious
fact, considering how thoroughly dialectologists have examined other aspects
of the dialect. Compared to the abundant literature in the field of Japanese dialec-
tology in general, only a small number of studies have dealt with gender contrasts
(some of the few exceptions are Haig 1990 and J. Nakamura 1996 on the retention
of phonological features in Nagoya and Ina of Nagano respectively, and Yamada
1986 on gender differentiated honorific use in a community in Aichi).10 Kindaichi
(1977[1957]) asserted that gender contrast in speech is a recent, urban phenomenon
and is rarely observed in farming and fishing villages, though he did not elaborate
on exactly how it became an urban phenomenon.
When I inquired about this issue to several female and male farmers in Ibaraki
(not the speakers in this chapter), they answered that it is difficult to give a clear answer
as to how women and men speak differently in ID. The farmers thought that sen-
tence-ending forms such as dappe and dabe (SJ: daroo, expressing conjecture) are
uttered more strongly and with more confidence by men than by women. The other
tendency the farmers pointed out is that, among the younger generation, men seem
to use more dialectal features than women (cf. Labov 1972, Trudgill 1974, Nakamura
1996). Of course these intuitive observations require a careful analysis of actual data
in order to be confirmed. Nevertheless, it seems to be the case that gender difference
in ID is not nearly as observable or as established as in SJ, and is clearly less salient
at the ideological level. At least at the morphological and lexical level, there does
not seem to be a marked difference. Moreover, as (b) and (f) in table 10.1 show, there
are some features that are considered manly or vulgar (and thus available only to
men) in SJ but are used gender-neutrally in ID. Another example of this category is
the first-person pronoun ore, which is supposed to be available only to male speak-
ers. However, my observations show that a group of farm women frequently used
ore as a standard first-person pronoun at their product development meeting, as seen
in example (2), (cf. chapters 11 and 14, this volume).

(2) [ID] Ore ge sa kita ra . . .

[SJ equivalent] Watashi no uchi ni kita ra . . .
[English Translation] If you come over to my place . . .
194 REAL LANGUAGE , REAL PEOPLE

In short, such examples illustrate how important it is to observe language use by

women and men in regional Japan independently from the Japanese womens lan-
guage ideology.

10.4. The work of the Agricultural Extension adviser

in contemporary Japanese agriculture

Due to financial difficulties and insecurity in contemporary Japanese agriculture, it

is now common for one or more members of a farm family (most likely the husband)
to have another job outside farming. Consequently, in 1992 about 70% of the agri-
cultural laborers between 30 and 59 years of age were women (Soorifu 1993).
Though it is evident from these statistics that women are indispensable in Japa-
nese agriculture, the role of farm women has been underestimated. It is well known
that nooka no yome farmers wives have traditionally been treated as ushi ya uma
chattel (lit., cattle and horses). However, such an attitude has made it difficult for
farmers sons to find wives now that women have started having more options in
their lives (cf. Gal 1978).
In addition to the problem of finding wives, competition with domestic and for-
eign products is getting stiffer; now farms need to change and improve their meth-
ods of management so that they can attract the younger generation and expand or
even maintain their businesses. This desperate situation, along with the general de-
cline of agriculture in Japan, has led even some of the most conservative Japanese
farmers and government officials to reconsider their attitudes toward women and the
traditional ways of running family farming businesses.
The kairyoo-fukyuu-in Agricultural Extension adviser (hereafter adviser) is
a position in the regional offices of Noogyoo-kairyoo-fukyuu-sentaa The Agri-
cultural Extension Service, which is under the control of the Ministry of Agri-
culture, Forestry, and Fisheries. Even though many advisers mainly deal with
technical improvements such as teaching farmers how to grow products more ef-
fectively, another important goal has been to improve the quality of life of farm-
ing families. This effort was started to attract the younger generation, especially
women, to agriculture in order to stem the decline in this sector (Japan Agricul-
tural News, July 6, 1994). Thus, some of the advisers, including Hayashi-san, the
one in this study, specialize in dealing with issues relating to the improvement of
the lives of farming families.
Advisers work toward small goals and use these as stepping stones to the ulti-
mate goal. The ultimate or macrogoal in this case is to improve womens status in
the farms, and smaller or micro goals include dividing jobs clearly within a farm family
and providing wives with the knowledge and skills necessary for them to be able to
perform their tasks responsibly. The means by which advisers help farm households
include changing their management styles to enable women to have regular days off,
to employ a monthly payment system, to encourage wives to create personal bank
accounts, and to teach wives how to keep accounts using computers and to be ac-
tively involved in promoting their products.
 FARM WOMEN S PROFESSIONAL DISCOURSE IN IBARAKI 195

The county agent, then, educates, supports, and advises women (and men) in
these agricultural family businesses. All parties involved (i.e., adviser, wives and
husbands in farming families) are aware that raising womens status is a main goal
in the work of the adviser. However, in reality, the degree of acceptance of this idea
differs across gender and generation.

10.5. Dialect use, solidarity building, and

empowerment of farm women

In the current data, the word empowerment most accurately describes the womens
interactions, and it seems to be a two-way process. That is, as described earlier,
Hayashi-sans very job is to empower farm women, but her effort needs to be re-
ciprocated by their active involvement in the dialogue. In the interactions between
Hayashi-san and the three women, two major strategies are found. The first strat-
egy is the use of ID, or the avoidance of SJ, especially honorific features, as a means
of building solidarity. The use of this strategy also means that morphological and
lexical features of Japanese womens language are rarely used in single-sex
professional interactions. The second strategy is regarding the women as the repre-
sentatives of their households who manage family business matters, rather than treat-
ing them as farmers wives. In this discursive construction of the three womens
new position in their family businesses, both Hayashi-san and the farm women were
equally eager participants. Before, when visits were made by advisers who only
taught technical matters, extensive communication between the wife of a farmer
and an adviser was uncommon. Hada-san and Oki-san said that until Hayashi-san
started visiting them they had thought it was their husbands job to talk to the ad-
visers from the governmental office (i.e., to people with superior status).11
Oki-sans husband was present during the entire conversation, but he remained
a listener almost throughout and made comments only occasionally. Hada-sans and
Nagano-sans husbands joined their wives and Hayashi-san later in the conversations.
Hayashi-san paid attention to the husbands and actively included them in the con-
versations by, for instance, asking Mr. Hada how his sprained foot was. However,
the husbands seemed to be aware of the conversational dynamics; that is, they acted
as if Hayashi-san was talking to their wives, not to them, about whatever the wives
were in charge of in the agricultural business. This dynamic seemed to be created as
a result of Hayashi-sans clear intention to treat the women as professionals who were
as competent as their husbands. Her treatment of the women was reciprocated by
their articulating business matters and their needs.
In (3), both strategies enumerated earlier are observed. Hayashi-san and Hada-
san are talking.

(3)12 Hayashi-san = H, Hada-san = HW, Mr. Hada = HH, R = Researcher

(Hayashi-san tells Hada-san that the researcher has been studying in the United States.)
1 HW: hnnn kanshin shitchaa kanshin shitchimaa
2 (to R, who replies no, no) itteppe yoo honto da yoo naa
196 REAL LANGUAGE , REAL PEOPLE

3 H: nee mukashi wa watashira no koro ni wa(.)sonna toko e

4 iku nante yuu no wa kangae mo tsuka-nee
(Hada-san agrees and switches her topic to her recent noogyoo kenshuu agricultural
study trip to Europe sponsored by Hayashi-sans government office.)
5 HW: dagaa ii chansu da naa ado ni mo sagi ni mo kore wa nee naa do omotta
gara
6 toochan nantsuoo-to kantsuoo-to kama-nee yaa to omotte
(Mr. Hada comes home and enters through the door in front of where we are seated.)
7 H: konnichi wa(.)ojama shitemasu doomo
8 HH: iya doomo mada: mada(xxx)
9 H: mada(.) mada: mada
10 H: konda kageashi a deginai ne

English Translation
1 HW: Hmmm, Im so impressed! I am so impressed!
2 (to R, who replies no, no) Really, Im telling you, I mean it.
3 H: Right, in the old days, in our days, we couldnt even imagine
4 going to such a place [i.e., abroad].
(Hada-san agrees and switches her topic to her recent noogyoo kenshuu agricultural
study trip to Europe sponsored by Hayashi-sans government office.)
5 HW: So, because I thought it would be a good opportunity, there wouldnt be this
[kind of opportunity] later on,
6 I thought I wouldnt care no matter what my husband says [about my going
on the trip].
(Mr. Hada comes home and enters through the door in front of where we are seated.)
7 H: Hello, how are you?
8 HH: Hi, how are you? not yet not yet (xxx)
9 H: (looking at HHs sprained foot) not yet not yet not yet
10 H: you probably cant run with this (foot), can you?

In this excerpt, Hayashi-san and Hada-san use dialect-dominant language. In

fact, to those unfamiliar with ID, Hada-sans utterance kanshin shitchaa, kanshin
shitchimaa (SJ/Japanese womens language equivalent: kanshin shichau, kanshin
shichau wa) in line 1 may sound strong or even coarse (thus unfeminine) due to
the long vowel /aa/ instead of /au/, a vulgar feature according to ideologies of SJ /
Japanese womens language.
Likewise, Hada-sans utterance toochan nantsuoo-to kantsuoo-to kama-nee yaa
(SJ /Japanese womens language equivalent: otoosan ga nante ioo to kamawanai
wa) in line 6 must sound rather masculine and unsophisticated by the standard of SJ
/Japanese womens language. First, the term of reference toochan, literally fa-
ther, is a variant that can be used to refer to ones husband. Other variants include
otoosan and papa, where the former sounds most generic and the latter has a West-
ern flavor, indexing certain (desired) attributes of the family such as urban, edu-
cated, or Westernized. According to the standard of SJ /Japanese womens language,
toochan, the form Hada-san used, sounds less sophisticated compared to the other
two variants. Nantsuoo is a truncated form of nante ioo, and again, according to the
 FARM WOMEN S PROFESSIONAL DISCOURSE IN IBARAKI 197

rules of SJ and Japanese womens language, it indexes coarseness. Moreover, in

kama-nee, the long vowel /ee/ instead of the diphthong /ai/ was used, which suppos-
edly indexed coarseness like the other features.
However, these features are all acceptable and normal and do not index coarse-
ness in ID; and the dialectal form of the utterances shows Hada-sans active involve-
ment in the conversation and her open relationship with Hayashi-san. The content is
tied both to Hada-sans professional growth and to her personal feelings about it.
When Hada-sans husband suddenly comes home and joins the conversation, how-
ever, Hayashi-san uses a more formal expression in line 7 konnichi wa, ojama
shitemasu, greeting him not only in SJ but also with an addressee honorific form,
-masu. This utterance makes a sharp contrast with the rest of the utterances, which
are dialect-dominant and contain virtually no honorifics. Hayashi-sans remark to
Mr. Hada in line 10 quickly reverts to the dialect-dominant speech. After line 10,
once she showed her concern about his sprained foot and introduced me to him, Hada-
san became the main conversational partner once again and Mr. Hada became a by-
stander. In this interaction, the formal-informal contrast is expressed as the contrast
between SJ and ID. Hayashi-san switched her speech in a similar way when Mr.
Nagano joined her conversation with Nagano-san.
Hayashi-san was successful in relating to each womans situation during her
visits. Among the three visits recorded, this rapport making in progress was best
observed in her interaction with Oki-san, whom Hayashi-san had met only once before
(whereas she had known the other two women for several months). In her conversa-
tion with Oki-san, whom Hayashi-san knew the least, use of the SJ addressee honor-
ific affirmative form -desu and its variant -desho(o) was observed. It seems that both
of them used politer forms of speech to indicate the appropriate social distance be-
tween them. At the same time, in order to create rapport, they retained dialectal fea-
tures in their speech as well. Part of the interaction is shown in example (4).

(4) Hayashi-san = H, Oki-san = OW, Mr. Oki = OH

(Hayashi-san and I have just been invited to enter the room where the farmers keep cut
roses cool. We all sit down to talk. It is a very humid, hot day in mid-August, and this
room is the only cool place the family has.)
11 H: kondo soto ni dete igu to mata tsukareru desho
12 OH: soo soo(xxx)
13 H: nde kon nakaa suzushiku tatte hausu n nakaa mata(.)mata
14 mata sugoi atsui n da mon ne? nn
15 OW: (1)atsui nante mon ja nee kurui soo
16 H: so kurui soo desho hontoni ne?(.)nanni mo shinai de ita tte

17 kuruisoo nan da mon koa hausu n naka ja naku tatte

18 H: ijoo-daa-ne?
19 OW: nn kotoshi wa ijoo desu yo ne?

English Translation
11 H: And then when you go outside, you get tired again, I bet.
12 OH: yeah yeah
198 REAL LANGUAGE , REAL PEOPLE

13 H: Even though this room is cool, the greenhouse is again

14 extremely hot, isnt it? yeah
15 OW: Its far worse than hot I almost go crazy
16 H: Yeah, I bet you almost go crazy, dont you?, really. Even if you dont do
17 anything you would go crazy, even if you arent in a greenhouse . . . (so
inside the greenhouse must be terrible)
18 H: It is abnormal, I agree
19 OW: Its unusually (hot) this year, dont you think?

This excerpt is near the beginning of the recording at the Okis home. Hayashi-
san and I, having moved from the car to the room, felt relieved to be in a cool place.
We could tell that the Okis had been working in the heat for hours. Hayashi-sans
focus was on sympathizing with Oki-san and at the same time applauding her hard
work. In lines 11 and 16 Hayashi-san used the addressee honorific form desho, while
in line 19 Oki-san used desu. However, the desu ending was not used at every pos-
sible place; in fact, plain forms were also used frequently in (4), as well as through-
out the interaction. In SJ, the simple use of addressee honorifics is considered lower
on the scale of politeness than other combinations of addressee and referent honor-
ifics (Ogino 1986, Ide 1990). However, in the context here, where the use of the dia-
lect with almost no honorifics is the norm of local interactions, the womens use of
desu and desho stands out and indexes the distance between the speaker and addressee.
At the same time, other dialectal elements such as the voicing of /k/ (as in dete igu to
when [you] go outside by Hayashi-san in line 11) and /ee/ in place of /ai/ (as in
atsui nante mon ja nee its far worse than hot by Oki-san in line 15) contribute to
their solidarity building.
The next segment comes from Hayashi-sans interaction with Nagano-san, whose
household has just started growing maitake mushrooms. At this point, the Naganos
are still deciding the best way of promoting their mushrooms in the market. Hayashi-
san visited the Naganos to discuss the packaging, naming, and advertising of their
new product. In example (5), one can see that Hayashi-san encourages Nagano-san
by making positive comments as a consumer: younger people are interested in a
healthy diet these days, and mushrooms would attract these people since they are
low in calories and high in minerals. Maitake mushrooms are relatively new and not
as common as shiitake mushrooms, and Hayashi-san points this fact out while em-
phasizing the maitakes unique texture, which is harder than that of the shiitake.
The two womens speech exhibits IDs pitch-accent pattern, and without the formal
endings of SJ it sounds casual. As a result of overlaps (lines 21 and 22; lines 23 and
24), as well as quick chime-ins with each other throughout the excerpt (but expecially
lines 25 and 26), the interaction sounds lively and involved. It is as if Hayashi-san
and Nagano-san are co-constructing a profile of maitake mushrooms by energeti-
cally and rhythmically throwing out their positive characteristics. By the end of their
interaction, Nagano-san managed to come up with the copy she would use to market
her mushrooms. It emphasized the mushrooms unique texture and nutrition.

(5) Hayashi-san = H, Nagano-san = NW

20 H: katai toko ga oishii ne mata ne
 FARM WOMEN S PROFESSIONAL DISCOURSE IN IBARAKI 199

21 oishii ne nn
22 NW: katai toko oishii ne anooo mazegohan mo
23 H: shii kedomo (xxx) (xx ra)uken ne aa
24 NW: oi shii kedomoo(.)katai tokoo ga wakai hito dakara ne
25 H: nn hagire ga ii n da=
26 NW: (1) aa koshi ga aru n da yo ne anoo =soo

English Translation
20 H: The hard parts are delicious, especially
21 delicious, yeah
22 NW: the hard parts are deli- cious weeell, mixed rice [with maitake] is
23 H: -ty, but (xxx)
24 NW: tasty, but the tough parts are [good] since [the targeted consumers are]
young people [who like crisp food more than old people do]
25 H: yeah theyre fresh and crisp
26 NW: (1) well, they (maitake) have good texture well right

In line 25, Hayashi-san uses the affirmative ending hagire ga ii n da its fresh and
crisp, which is considered masculine in SJ (Reynolds 1990) but not in ID.
To summarize, Hayashi-san employed two strategies in order to establish soli-
darity and empower the farm women, Hada-san, Oki-san, and Nagano-san. First, all
the womens use of ID essentially without honorifics contributed to their establish-
ing rapport and engaging in intimate yet task-oriented conversations. In their inter-
actions, the adviser, Hada-san, and Nagano-san used dialect-dominant speech. Their
speech contained virtually no Japanese womens language features but included a
number of Ibaraki dialectal features, some of which are considered vulgar according
to the rules of Japanese womens language. Even in the conversation between the
Advisor and Oki-san, who were more recent acquaintances, both utilized a mixture
of SJ addressee honorifics and dialectal features simultaneously to maintain the so-
cially appropriate distance (as relative strangers) and to establish rapport. Second,
Hayashi-san made sure that it was the wife, instead of the husband, who was the rep-
resentative to whom she came to talk, even though she included the husband in the
conversation. By doing so, she successfully placed the wives in a more responsible,
respected position than they have been traditionally accorded and she interacted with
them as professionals. At the same time, the three farm women were by no means
passive receivers of Hayashi-sans treatment. Instead, they were simultaneously
empowering themselves, as well as being empowered, by actively voicing their pro-
fessional selves in their discourse with her.

10.6. Conclusion and implications

In this chapter, the discourse of an Agricultural Extension adviser and three women in
a farming community in Ibaraki was analyzed. From the womens extensive use of dia-
lectal features as well as the lack of morphological, phonological, and lexical features
of Japanese womens language in their speech it is evident that we need to go beyond
200 REAL LANGUAGE , REAL PEOPLE

analyzing the presence or absence of Japanese womens language in Japanese womens

language use. The use of Japanese womens language was not part of the speech rep-
ertoire of these four women. This finding is better understood when the relationship
between regional dialects and SJ in Japan is historically reviewed. These women have
always spoken in ID, with some elements of SJ mixed in depending on the context.
Though a small-scale study, this chapter suggests a crucial point in the field of
language and gender. That is, gender may not always be the most salient category
that can affect speakers. In order to understand how women speak, it is important to
consider their regionality and occupation and other aspects of their identities, as well
as the sociohistorical background of their linguistic resources. Regarding the four
Ibaraki womens speech and lives, it is likely that they have more in common with
other members of the rural population of Ibaraki or with other people engaged in
agriculture elsewhere in Japan than with traditional Yamanote housewives. Indeed,
the use of Japanese womens language is itself a regional phenomenon; Japanese
womens language is the ideal shared by women of a certain socioeconomic back-
ground who reside in Tokyo or aspire to Tokyo values, including language use. Japa-
nese women who have internalized the ideology of Japanese womens language
may aspire to, negotiate with, or even reject Japanese womens language. How-
ever, the speech of other women, for whom Japanese womens language is more
fictive than real (cf. Inoue 2002 and chapter 3, this volume), may not even include
Japanese womens language. Nevertheless, thus far Japanese womens language
has been treated as if it were the only speech style Japanese women should follow or
resist. Part of the reason for this may be that even we scholars of language and gen-
der have been deeply influenced by the social norm and gender identity associated
with Japanese womens language.
Of course, in order to confirm my observations more research needs to be con-
ducted about womens and mens speech in various regions and occupations in Japan.
There are still many questions to be answered regarding the relationship between
dialects, SJ, and Japanese womens language and how or if gender difference is
marked in various dialects.
I suspect that the gender ideology of Japanese womens language has not quite
reached or, rather, has not been internalized by, locally centered people, including
my study participants. At the same time, women and men of rural Ibaraki do under-
stand what Japanese womens language features are supposed to index when they
encounter these features on television or in novels. What kind of position, then, does
Japanese womens language occupy in the language ideology and language use of
these rural women and men of Ibaraki? Also, if there is a local gender-marking sys-
tem in ID, how do female aspirers negotiate uses of the dialects gender-preferred
elements and Japanese womens language? To answer these questions requires a
larger scale study than is possible in this chapter.

Notes
I thank the women and men in Ibaraki who helped me gather data for this research. My spe-
cial appreciation also goes to Mary Bucholtz and the editors of this volume for their valuable
comments.
 FARM WOMEN S PROFESSIONAL DISCOURSE IN IBARAKI 201

1. The term Japanese womens language is ambiguous. For example, it may refer to
any speech Japanese women use or to vaguely defined feminine speech. But throughout
this essay, I use Japanese womens language to specifically refer to a set of morphological,
phonological, and stylistic features that remind a Japanese native speaker of a stereotypical
Tokyo woman.
2. But see chapter 15 for regional differences in mens use of sentence-final particles.
Ogawa and [Shibamoto] Smith (1996) also deal with regionality in gay speech.
3. The degree of retention of a dialect differs from region to region, as well as from
person to person. Kansai (e.g., Osaka, Kyoto) dialects preserve high ethnolinguistic vitality
next to SJ in current Japanese society perhaps due to the regions history; Kansai dialects
speakers tend to be proud of their dialect and use them even when they are outside their re-
gion. However, other speakers, including users of ID, tend to restrict their dialect use to within
their own communities. See Inoue (1989) for surveys of various dialects.
4. However, Momoko Nakamura (2001) argues that users of Japanese womens lan-
guage, while being praised for their femininity, have also been subject to criticism that their
speech is deviant from the standard (that is, standard male speech, which is often automati-
cally subsumed under SJ).
5. All speakers names that appear in this chapter are pseudonyms.
6. Despite Ibarakis proximity to Tokyo, its dialect can be so distinct from SJ as to be
unintelligible to SJ speakers, especially when the ID speaker is older, locally based, and
working-class.
7. Strictly speaking, there is a difference between SJ and the Tokyo dialect, the former
being the standard established for the mass media and education and the latter being the daily
spoken variety of Tokyo residents. Contemporary Japanese living in regional Japan seem to
be influenced by the central language at these two different levels (see Inoue 1994 and Sanada
1996, for example). However, it is beyond the scope of this chapter to differentiate these two
levels or to develop a discussion on this matter.
8. Feature (c), vowel merger, is not indicated in the transcript for two reasons: the merger
was not necessarily observed with all tokens of /i/ and /e/, and the degree of centralization
was not uniform across occurrences. These variations are most likely due to the influence
from SJ.
9. However, Irvine (1998) notes that class differences alone cannot sufficiently explain
the existence of honorifics. In addition, this explanation leaves a question regarding ID. Mito,
the Ibaraki prefectural capital, was where one of the three Tokugawa family branches was
located during the Tokugawa Shogunate. In that sense, at least part of Ibaraki was histori-
cally complex between the seventeenth and the nineteenth century, though the dialect is said
to be honorific-less. Further research is needed on this issue.
10. At the anecdotal level, I have heard people from different parts of the Kansai region
(e.g. Osaka, Kyoto, and Nara) describing gender-differentiated features in their own dialects.
11. In fact, until they became amalgamated into one position a few years ago, advisers
had two tracks: noogyoo kairyoo-fukyuu-in agricultural adviser and seekatsu kairyoo-fukyuu-
in life adviser, at the ratio of approximately 10:1. Hayashi-san was employed as the latter.
On the one hand, the former position dealt with technology-related issues and was mostly
occupied by males, who interacted with male members of farm families. On the other hand,
the latter position dealt with issues to improve the quality of lives of these families, as dis-
cussed in this chapter. Like Hayashi-san, women mostly occupied the latter position, and its
job was traditionally considered secondary to the formers.
12. Transcription conventions are:
? rising intonation
(xxx) unclear section that could not be transcribed
202 REAL LANGUAGE , REAL PEOPLE

= latching
(1) number of seconds of pause
(.) micropause of less than one second
boldface analyzed features
underline dialectal features
5 H: aaaaa ccccc temporal overlap. In this example, H utters aaaaa, followed
6 OH: bbbbb by OHs bbbbb, then Hs ccccc

References
Bucholtz, Mary (1999). Bad examples: Transgression and progress in language and gender
studies. In M. Bucholtz, A. C. Liang, and L. A. Sutton (eds.), Reinventing identities:
The gendered self in discourse, 324. New York: Oxford University Press.
Dore, Ronald P. (1978). Shinohata: A portrait of a Japanese village. Berkeley: University of
California Press.
Gal, Susan. (1978). Peasant men cant get wives: Language change and sex roles in a bilin-
gual community. Language in Society 7:116.
Haig, John H. (1990). A phonological difference in male-female speech among teenagers in
Nagoya. In S. Ide and N. H. McGloin (eds.), Aspects of Japanese womens speech, 5
22. Tokyo: Kuroshio.
Hibiya, Junko (1988). A quantitative study of Tokyo Japanese. Ph.D. dissertation, Univer-
sity of Pennsylvania. Ann Arbor: University of Michigan.
Hunter, Janet (1993). Japanese women working. London: Routledge.
Ide, Sachiko (1990). How and why do women speak more politely in Japanese? In S. Ide and
N. H. McGloin (eds.), Aspects of Japanese womens language, 6379. Tokyo: Kuroshio
Inoue, Fumio (1989). Kotobazukai shinfuukee: Keego to hoogen (New views on ways of speak-
ing: Honorifics and dialects). Tokyo: Akiyama Shoten.
(1994). Hoogen-gaku no shinchihee (New horizon of dialectology). Tokyo: Meiji Shoin.
Inoue, Miyako (2002). Gender, language, and modernity: Toward an effective history of Japa-
nese womens language. American Ethnologist 29(2): 392422.
(in preparation). Vicarious language: The political economy of gender and speech in
Japan. Book manuscript.
Irvine, Judith (1998). Ideologies of honorific language. In B. B. Schieffelin, K. A. Woolard,
and P. V. Kroskrity (eds.), Language ideologies: Practice and theory, 5167. New York:
Oxford University Press.
Kanesawa, Naoto (1984). Ibaraki ken no hoogen (ID). In K. Iitoyo, S. Hino, and R. Sato (eds.),
Kanto chihoo no hoogen (Dialects of Kanto region). Tokyo: Kokusho Kankokai.
Kindaichi, Haruhiko (1977 [1957]). Nihongo (Japanese language). Tokyo: Iwanami Shoten.
(1977). Nihongo hoogen no kenkyuu (Study of Japanese dialects). Tokyo: Tokyo-do
Shuppan.
(1995 [1948]). Ikkee akusento ni tsuite no ichi koosatsu (A study of one-pattern accent).
In F. Inoue, K. Shinozaki, T. Kobayashi, and T. Oonishi (eds.) (1995), Kanto hoogen
koo 1: Kanto ippan, Ibaraki ken, Tochigi ken (Dialects of Kanto region 1: Kanto region
in general, Ibaraki prefecture, Tochigi prefecture), 257262. Tokyo: Yumani Shobo.
([=1948] Kotoba 7.)
Kondo, Dorinne (1990). Crafting selves: Power, gender, and discourse of identity in a Japa-
nese workplace. Chicago: University of Chicago Press.
Labov, William (1972). Sociolinguistic patterns. Philadelphia: University of Pennsylvania
Press.
 FARM WOMEN S PROFESSIONAL DISCOURSE IN IBARAKI 203

Lebra, Takie Sugiyama (1993). Above the clouds: Status culture of the modern Japanese
nobility. Berkeley: University of California Press.
Lee, Yeounsuk (1996). Kokugo to yuu shisoo (The ideology of the national language).
Tokyo: Iwanami Shoten.
Miyajima, Tatsuo (1961). Fukushima, Ibaraki, Tochigi. In Y. Endo, T. Hirayama, T. Okubo,
and T. Shibata (eds.), Toobu hoogen (Eastern dialects), 236263. Tokyo: Tokyo-do
Shuppan.
Miyake, Yoshimi (1995). A dialect in the face of the standard: A Japanese case study. In
J. Ahlers, L. Bilmes, J. S. Guenter, B. A. Kaiser, and J. Namkung (eds.), Proceedings of
the Twenty-first Annual Meeting of the Berkeley Linguistics Society, 217225. Berke-
ley: Berkeley Linguistics Society.
Nakamura, Junko (1996). Ina hoogen ni okeru hoogen hoji no danjo-sa (Gender contrast in
dialect retention in Ina dialect). Nihongo Kenkyuu 16: 2538.
Nakamura, Momoko (2001). Kotoba to jendaa (Language and gender). Tokyo: Keiso Shobo.
Ogawa, Naoko, and Janet S. [Shibamoto] Smith (1996). The linguistic gendering of an alter-
native Japanese lifestyle: Speech variation in the gay communities of urban Japan. In
R. Ide, R. Parker, and Y. Sunaoshi (eds.), SALSA: Proceedings of the Third Annual Sym-
posium about Language and Society, 36, 2840. Austin: University of Texas, Depart-
ment of Linguistics.
Ogino, Tsunao (1986). Quantification of politeness based on the usage patterns of honorific
expressions. International Journal of the Sociology of Language 58: 3758.
Okamoto, Shigeko (1995). Tasteless Japanese: Less feminine speech among young Japa-
nese women. In K. Hall and M. Bucholtz (eds.), Gender articulated: Language and the
socially constructed self, 297325. New York: Routledge.
(2000). The role of standard and non-SJ in conversations. Paper presented at the 2000
Annual Meeting of the American Association for Applied Linguistics, Vancouver, Brit-
ish Columbia.
Reynolds, Katsue Akiba (1990). Female speakers of Japanese in transition. In S. Ide and N.
H. McGloin (eds.), Aspects of Japanese womens language, 129146. Tokyo: Kuroshio.
Roberts, Glenda S. (1994). Staying on the line: Blue-collar women in contemporary Japan.
Honolulu: University of Hawaii Press.
Sanada, Shinji (1991). Hyoojungo wa ika ni seeritsu shita ka: Kindai nihongo no hatten no
rekishi (How Standard Japanese has developed: History of the development of contem-
porary Japanese). Tokyo: Sotakusha.
(1996). Chiiki-go no seetai shiriizu: Chiiki-go no dainamizumuKansai (Regional
variety series: Dynamism of regional varietiesKansai region). Tokyo: Ofu.
Sanada, Shinji, Katsumi Shibuya, Masataka Jinnouchi, and Seiju Sugito (1992). Shakaigengo-
gaku (Sociolinguistics). Tokyo: Ofu.
Shibata, Takeshi (1980). Kyootsuugo (Common language); Hyoojungo (Standard language).
In Kokugo-gaku daijiten (Dictionary of the national language studies), 219220, 729
730. Tokyo: Kokugo Gakkai.
Siegal, Meryl, and Shigeko Okamoto (1996). Imagined worlds: Language, gender and socio-
cultural norms in Japanese language textbooks. In Proceedings of the Fourth Ber-
keley Women and Language Conference, 667678. Berkeley: Berkeley Women and
Language Group.
Soorifu (ed.) (1993). Josee no genjoo to shisaku: Kawaru kazoku to josee no seekatsu
(Womens current situations and trials: Changing families and womens lives). Tokyo:
Okurasho.
Taguchi, Yoshio (1995[1944]). Ibaraki hoogen gohoo ni-san no koosatsu (A study of Ibaraki
hoogen usage). In F. Inoue, K. Shinozaki, T. Kobayashi, and T. Oonishi (eds.) Kanto
204 REAL LANGUAGE , REAL PEOPLE

hoogen koo 1: Kanto ippan, Ibaraki ken, Tochigi ken (Dialects of Kanto region 1: Kanto
region in general, Ibaraki Prefecture, Tochigi Prefecture), 295306. Tokyo: Yumani
Shobo. ([=1944] Hoogen kenkyuu 10.)
Trudgill, Peter (1974). Sociolinguistics: An introduction to language and society. Har-
mondsworth: Penguin.
Yamada, Tatsuya (1986). Hoogen to keego (Dialects and honorifics). In K. Iitoyo, S. Hino,
and R. Sato (eds.), Hoogen gaisetsu (Overview of dialects), 181204. Tokyo: Kokusho
Kankokai.
 11

HIDEKO ABE

Lesbian Bar Talk in Shinjuku, Tokyo

T wo decades ago, the difference between sex (as biological, natural, physical, real,
actual, unrefined) and gender (as social, fictitious, refined, forced) was presumed to
be clear and obvious for language and gender researchers. However, the distinction
between the two was questioned and challenged most powerfully by Butler, who
claims that sex by definition has been gender all along (1990:8) and that gender is
a discursive practice unfolding in an ongoing interaction and, thus, open to interven-
tion and resignification. An important aspect of Butlers work is her claim that gen-
der only exists in the service of heterosexism. Butler restructures gender as the
performative effect of repeated acts, a cultural fiction, which can as such no longer
privilege heterosexuality as the real or original sexuality at the expense of homo-
sexuality (Jagose 1996). Ueno (1999:48) argues that one of queer theorists contri-
butions is to demolish the distinction between heterosexuality as norm/origin and
homosexuality as deviation/imitation. Jagose defines queer as follows.

Queer describes analytical models which dramatise incoherence in the allegedly

stable relations between chromosomal sex, gender and sexual desire. Resisting that
model of stabilitywhich claims heterosexuality as its origin, when it is more prop-
erly its effectqueer focuses on mismatches between sex, gender and desires. In-
stitutionally, queer has been associated most prominently with lesbian and gay
subjects, but its analytic framework also includes such topics as cross-dressing, her-
maphroditism, gender ambiguity and gender-corrective surgery. Demonstrating the
impossibility of any natural sexuality, it calls into question even such apparently
unproblematic terms as man and woman. (1996:3)

205
206 REAL LANGUAGE , REAL PEOPLE

In contrast to language and gender studies that have a relatively substantial his-
tory of methodological and theoretical development, language and sexuality studies
are a new area of inquiry, provisionally called queer linguistics. Barrett argues that
queer linguistics provides a means for beginning to understand the ways in which
people actually construct and produce markers of queer identities and deal with the
ambiguity of identity categories and communities that are imagined differently by
different community members (1997:198).
In this chapter I look at how linguistic practice observed at lesbian bars in Tokyo
is related to the construction of queer identities. I have chosen the lesbian bar com-
munity because it is the most concentrated place where lesbians gather and interact
with one another on a personal level. My study does not attempt to describe how
Japanese lesbians speak in general, but rather it discusses how lesbians construct their
identities through a specific linguistic practice. This practice may be labeled mascu-
line speech with noticeable linguistic variation by individual speakers in the same
speech context.

11.1. Background

Japanese language and sexuality has been largely unexplored. Further, the limited
number of studies (e.g., Maree 1997, Ogawa & [Shibamoto] Smith 1997, Valentine
1997) that have appeared are mainly focused on the speech of gay men (cf. chapter
5, this volume). There have been more studies on European languages, mainly En-
glish, and some of these focus on lesbian speech. Topics of lesbian-centered research
cover (1) lexical variation (Kleinfeld & Warner 1997), (2) voice characteristics
(Moonwomon-Baird 1997), (3) the language use of comic-book characters (Queen
1997), (4) coming-out stories (Wood 1997), (5) identity construction through lan-
guage (Liang 1997, 1999; Neumann 1997), (6) language use in film (Queen 2000),
(7) grammatical gender (Livia 1997, Pastre 1997), (8) language use in life stories
(Moonwomon-Baird 2000), and (9) the textual analysis of personal ads (Livia 2000).
These studies attempt to uncover different aspects of lesbian lives and their identi-
ties and promise an exciting future for research on language and sexualities.
While the field of language and sexuality has started producing many studies of
private speech and of language in the media in recent years, we have seen a very
limited number of studies on public speech, especially in gay/lesbian bars. Both
Achilles (1993 [1967]) and Reads (1980) studies of non-Japanese gay bars discuss
the role and the significance of gay bars in providing a space for social interaction.
So far, there are only a few studies on non-Japanese lesbian bars, all of which focus
on how class-related differences play a role in bar-going habits as well as in ways of
articulating lesbian identity (e.g., Chamberland 1993, Gilmartin 1996). The most
recent study (Hankin 2002) analyzes how lesbian bars are represented in popular
culture such as films. Barretts (1997) study of gay identities and language use in
bar settings discusses the speech of African-American drag queens and analyzes
how they negotiate their overlapping identities by using code-switching strategies.
As for Japanese gay bars, there are a few works (e.g., Ootsuka 1995, Sunagawa
1998), which emphasize the importance of gay bars as a crucial space for developing
 LESBIAN BAR TALK IN SHINJUKU , TOKYO 207

and maintaining ones gay identity. There are no studies of Japanese lesbian bars,
although some popular womens magazines have published reports on a few, along
with interviews with the owners. Anise, a lesbian and bisexual information magazine
(joohooshi) that started in 1996 and ended in 1997, also discusses some lesbian bars.
Another joohooshi, Love Revolution (AKIKO 2000), lists 10 lesbian bars in the Tokyo
area. Few studies have focused on language use in either gay or lesbian bars.

11.2. The goals of this study

This chapter focuses on the relationship between identity and language use observed
among women at lesbian bars in Shinjuku, Tokyo. By examining specific linguistic
behaviors and interactions at bars it attempts to demonstrate how language effects
gender (Bucholtz 1999:6) and, in particular, how masculine power normatively
associated with specific linguistic forms is negotiated differently by individual speak-
ers. Here I use the word power not as an absolute or static term but as a relational and
processual term. In other words, power should be understood as interactionally ne-
gotiated through submission to or resistance against force among participants (Fou-
cault 1990). As Kramarae, Schultz, and OBarr (1984:11) argue, regardless of the
definition of power the resources available to exert or resist influence are recurrent,
similar, andin societies at peacechiefly verbal.
I chose a lesbian bar as a setting for ethnographic linguistic study because les-
bians, in contrast to gays, have been generally underrepresented in Japanese gender
studies. Watanabe (1990) argues that lesbianism in Japan has historically been
marginalized and is much less well documented than male homosexuality. As
Moonwomon-Baird puts it, Lesbian practice is regarded as marked behavior, but
goes unremarked much more than is true of gay male practice, even in this era of
both friendly and hostile societal discourse on queers (1997:202).
Specifically, this study discusses two issues: (1) naming and identity construc-
tion in discourse and (2) linguistic behavior and interaction at lesbian bars. In the
first part of this chapter, I examine how certain terms of social categorization are
used to differentiate one individuals identity from others and how the social spaces
of lesbian bars help individuals construct, renew, or vitalize their identities. Then I
discuss how lesbian speech at bars tends to contain extensive use of masculine forms
with wide contextual and individual variation. As I will discuss later, these two phe-
nomena are crucial to understanding the relationship between Japanese lesbian iden-
tities and linguistic practice.

11.3. Method and data collection

Tokyos Shinjuku Ni-choome the Second Block, an area as small as 300 meters by
400 meters, is often referred to as the worlds largest gay town. It was formerly a
district inhabited by Japanese female prostitutes who served American personnel
during the Occupation after World War II (Treat 1999). According to a recent gay
travel guide (Uminarikan 1998, cited by Sunagawa 1998), there are 217 gay bars in
208 REAL LANGUAGE , REAL PEOPLE

Ni-choome.1 By contrast, there are 12 lesbian bars in Ni-choome and one in San-
choome the Third Block in Tokyo, four in Osaka, and one each in Kyuushuu and
Hokkaido.
The fieldwork for this study was conducted between September of 1999 and June
of 2000 in the Shinjuku area of Tokyo. I visited lesbian bars almost every week, spending
at least two to three hours as a customer and researcher at each bar. I also corresponded
with employees via e-mail. In total, there are 13 lesbian bars in Shinjuku, and I visited
12 of them; I spent considerable time at some, and others I visited only once. Not all
lesbian bars are the same; some target older middle-class professional women while
others target younger nonprofessional women, both middle- and working-class. How-
ever, some bars have both types of women. All employees of the bars are lesbians.
Whenever I visited a bar for the first time, I arrived at opening time so that I
could be the first customer and spend time alone with the owner or the manager of
the bar to explain my research. This is when I asked for permission to take field notes.
I was allowed to have a notebook in front of me and to make notes at any time.
Oftentimes employees jokingly said to me, a mata nan ka kaiteru shes writing
something again while they were serving other customers. I conducted interviews
with owners and employees but also interacted with other customers, all of whom
knew of my research project. I did not tape-record interactions at bars, nor did I con-
duct interviews outside the bars. There are two reasons for not recording the interac-
tions. The first is that there was always loud music playing at the bars, which would
have made recording almost impossible. Second, and more important, the bar own-
ers and employees did not agree to recording. The reason was clear. In most bars,
women are anonymous and it was important that they be allowed to remain so. Tape-
recording is one of the most important methods of collecting sociolinguistic data; in
particular, it is crucial for analyzing phonological aspects of linguistic structure. Yet
careful field notes can provide important insights into the social use and signifi-
cance of lexical, grammatical and rhetorical forms (Gaudio 2001:40). Other means
of text (such as e-mail) are also useful for examining language use, and I drew on
these as well. My sources of data thus are (1) interviews, (2) field notes, (3) articles
from magazines, and (4) e-mail messages.

11.4. Data analysis and discussion

11.4.1. Naming and identity construction in discourses
at lesbian bars
The definition of the community served by lesbian bars turned out to be broader than
I had expected. While some owners and employees described their businesses as places
that catered only to a female clientele, some welcome onabe and nyuu haafu trans-
sexual/transgendered people and even gay male friends of customers or employees.
Owners and employees generally assume that female clientele are homosexual or
bisexual but welcome others. During my visit to one bar, I encountered a situation in
which a jooren regular customer brought her friend and introduced her to the owner
and employees as a nonke straight. It appears that a customers sexuality has to be
stated explicitly if she or he is straight.
 LESBIAN BAR TALK IN SHINJUKU , TOKYO 209

Just as Sunagawa (1998) emphasizes the difference between a gay bar and a show
pub in which there is an entertainment show by nyuu haafu, I found it crucial to dif-
ferentiate between a lesbian bar and an onabe bar. The word onabe pan, which is
parallel to okama pot, meaning male homosexuals, is often used to refer to lesbi-
ans in a broad sense, but the two words, lesbian and onabe, are not the same (cf.
Valentine 1997), as is clearly demonstrated in example (1). In this example,2 the
manager (A) of a lesbian bar is talking about her sexuality.

(1) Yoosuru ni jishoteki ni wa dooseeaisha tte koto ni narimasu ne. Onabe to chigau no wa,
jibun no josee to yuu see o mitometa ue de, josee ga suki tte yuu.

Anyway, I am a homosexual by the dictionary definition. How I am different from onabe

is that I accept my sex as female and like women.

Another employee (B) at a different lesbian bar3 contends that a lesbian is a woman
who feels comfortable with her female body (in other words, with her biologically
female sex) and who chooses a woman as a partner. Unlike straight women, how-
ever, her identity as a woman is constructed through a relationship with another
woman. However, according to this employee, an onabe loves women and chooses
a woman as a partner, but an onabes social and emotional identity is male. Thus,
both the concept of onabe and that of lesbian challenge the conventional hetero-
sexual gender arrangement, yet her explanation remains restricted within the bi-
nary woman/man.
The same employee (A) explains the difference between onabe and rezu (or
rezubian) lesbians in example (2).

(2) Chigaimasu ne. Ishikiteki ni mo. Kore wa yoku iwareru n desu kedo, onabe san wa,
rezu no ko to wa tsukiattari, anmari shinai n desu yo. Sore wa, onabe san wa jibun ga
otoko to shite miraretai tte yuu no ga aru kara, onna o suki na rezu no ko to wa,
tsukiaenai.
Yes, we are different. Consciously as well. People often say thisonabe do not date
lesbians. Because they want to be seen as men, they cannot date a lesbian who likes
women.

It is significant that the speaker added -san Ms./Mr. to onabe and ko child to rezu.
The former emphasizes the distance between the speaker and the referent, while the
latter constructs an intimate or friendly relationship. This kind of identity construc-
tion through naming is also found in other interviews.
Onabe bars, then, are butch-type bars where staff members usually wear mens
clothes and work as hosts. There are three onabe bars in Tokyo (Valentine 1997).
Customers at onabe bars are usually half lesbians and half nonke straight women.
According to many employees of rezubian bars, these bars are places where custom-
ers, who may be onabe or lesbians, and employees are on equal terms, whereas in
onabe bars employees are there to serve customers. They also suggest that custom-
ers who are tired of onabe bars may move on to host bars, where young straight men
serve women.
210 REAL LANGUAGE , REAL PEOPLE

Nine of the 13 lesbian bars accept only women, including heterosexuals, onabe,
and nyuu haafu. In example 3, a manager (C) of another lesbian bar talks about her
clientele.

(3) Toriaezu, rezu no okyaku sama ga taihan nan desu kedo, rezu ja nai ko mo iru n desu
ne. Rezu wa rezu de kakusazu ni, soo ja nai ko mo tanoshiku. Betsu ni itsu mo rezu no
wadai bakkari tte wake nai kara. Dansee wa ne, watashi no hontoo ni goku shitashii
homo no kata to ka, soo yuu hito dake desu ne. Yappari onna no ko ga dansee no me o
ki ni shinai de kiraku ni nomeru tte koto de, kihonteki ni dansee wa dame desu. Onabe
no ko to ka nyuu haafu no ko to ka kiraku ni asobi ni kite kureru shi, futsuu no onna no
ko datte ii shi. Mattaku futsuu no otoko wa dame. Dono janru ga doo ka ja nakute,
mainoritii de zenzen oorai. Soo soo, guroobaru na imi de nan de mo oorai.
Anyway, the majority of our customers are lesbians, but there are some who arent.
For lesbians, they dont have to hide their identity. Nonlesbians can have a good time
here too. We dont talk only about lesbianism, you know. As for male customers, only
a few close homosexual friends of mine can come here. As a rule, we dont allow male
customers because we want to provide a place where women feel comfortable drinking
without worrying about mens eyes. People who are onabe or nyuu haafu come here as
well. Ordinary [meaning heterosexual] women are also welcome. But ordinary men are
not welcome here. Im not categorizing people, but as long as they are a minority, theyre
totally welcome. Yes, anybody in a global sense is welcome.

Whereas speaker (A) in example (2) added -san to onabe, speaker (C) uses the term
ko child/young person (instead of hito or kata, two terms for person), as in onabe
no ko a young onabe and nyuu-haafu no ko a young nyuuhaafu. This shift indi-
cates that onabe are here treated as in-group members (customers). This use of ko is
a reflection of the fact that the majority of customers are in their twenties (younger
than the owners or employees) and also that employees want to sound more inclu-
sive and intimate. The alternative words, hito or kata, are more formal and distant.
Interestingly, the women I interviewed never use ko to refer to men; instead they use
dansei men or otoko men or even homo no kata homosexuals. Otherness is
emphasized by nonuse of ko.
Another interesting usage is found in the term futsuu ordinary. In example (3),
speaker (C) uses the term futsuu to refer to heterosexuals. I was initially shocked to
encounter this word, because I thought that the speaker meant that she considered
herself and other lesbians not ordinary, although it is quite possible that by futsuu
she meant run-of-the-mill or even dull, and viewed being not futsuu, or being
lesbian, in a positive light. However, the word futsuu is not always used for distin-
guishing lesbians from heterosexuals, as example (4) demonstrates. Here the man-
ager of a lesbian bar, (A), uses the term futsuu when she is asked about the role and
the function of such a bar.

(4) Minna hontoo ni futsuu no renai o shite iru n desu yo ne. Naimenteki na bubun ga chigau
tte yuu ka. Futokutei tasuu no hitobito ni yotte tsukuridasareta imeeji to yuu mono ga,
henken o umidashita tte yuu no wa aru to omou. Shinjitsu o wakatte nai to yuu ka. Futsuu
 LESBIAN BAR TALK IN SHINJUKU , TOKYO 211

no onna no ko demo, rezu baa tte donna tokoro na no ka na, mitai na kanji de kuru shi,
kite mireba, a, nan da futsuu no mono nan da mitai na. Onna no ko hitori de mo anshin
shite nomi ni kite kuremasu yo. Futsuu no onna no ko mo ippai kimasu. Shufu no hito
mo iru shi, kareshi ga iru kedo kuru ko mo imasu shi ne.
Lesbians have ordinary love relationships, you know. Internally, we are different. Some
people created the image of lesbians as different, which created prejudice, I think. They
dont know the real truth. Ordinary women come here because theyre curious. Once
they come, they realize how ordinary we are. Girls can feel comfortable coming here on
their own to drink. Lots of ordinary women come here, including housewives and women
who have boyfriends.

This manager emphasizes that lesbians have futsuu ordinary love relationships. I
argue that the speaker characterizes lesbians love relationships as futsuu because
she wants heterosexuals to be inclusive of her by thinking of her as ordinary. In other
words, she is not using this word to construct differences among customers but to
include people of both sexual orientations.
Example (5), from a conversation with the manager (D) of a different bar, un-
derscores this point.4

(5) Intaanetto mite kuru onna no ko to ka tte wari to booisshu na onna no ko to ka kitai
shite kuru ko mo iru kedo, soo yuu imi de wa, watashi ga kekkoo futsuu to yuu ka, gaiken
ga booishhu ja nai kara, sore o meate ni kita okyaku san ni wa warui ka na tte.
There are those who come here after seeing us on the internet, expecting that Ill be
boyish looking. In that sense, how should I put it? Im pretty ordinary looking. I dont
have a boyish look. I feel kind of bad for customers who come expecting that.

Here the word futsuu is used to contrast with boyish looks. By using futsuu, the speaker
D means to say that she is feminine looking. In other words, she wants to emphasize
that feminine as well as boyish types of lesbians are welcomed at her bar.
Identity thus is demonstrated through naming and lexical contrast. While in
example (3) the word futsuu contrasts with lesbian, example (4) treats the same word
as part of what it means to be a lesbian. In example (5), it distinguishes feminine
appearance from boyish looks. These three examples suggest the complexity and
situatedness of peoples identities, as well as the rhetorical flexibility they have to
position themselves in different ways at different moments for different purposes.

11.4.2. Linguistic behavior and interactions at bars

In the first part of this chapter, I examined how identities can be examined through
naming and lexical contrast. In this section, I discuss several linguistic features found
in interactions at lesbian bars among employees as well as between employees and
customers: (1) formal versus informal styles, (2) first-person pronouns, (3) second-
person pronouns, (4) commands and requests, and (5) sentence-final particles. These
features are chosen because these are the areas where the construction of gender dif-
ferences is often observed and recognized.
212 REAL LANGUAGE , REAL PEOPLE

11.4.2.1. Formal versus informal styles

The first feature, formal versus informal style of speech, is mainly concerned with
owners and employees approaches to language use at their own bars. At one end
of the spectrum, I observed a striking difference between an owner and her employ-
ees whereby the former stayed in an informal style (e.g., using direct sentence end-
ings) and the latter used more formal speech (e.g., distal sentence endings). How-
ever, both of them shifted their level of speech when talking to customers. Although
it is often said that women use a more formal/polite style (Ide 1979), more than a
few times I observed the exclusive use of direct style by employees while custom-
ers continued to use distal endings. I argue that the use of direct style by employ-
ees comes from their goal to create an atmosphere in which customers are brought
into an inclusive groupa kind of performed friendliness that may nurture a strong
sense of solidarity. For example, I observed one interaction between an employee
in her early twenties and a customer in her thirties talking about relationships. It
sounded as if a rude young brother was talking to his older sister, if judged by the
protocols assumed to govern interpersonal communication in heterosexual-centered
contexts. This employee stayed in direct style even when talking to her supervisor.
It seems that she has made a conscious choice regarding her linguistic style at her
workplace.
At some bars where owners hardly ever show up, the linguistic atmosphere feels
much freer. In fact, one owner complained to me about the use of bad language at
her bar on her day off (when her employee manages the bar). Thus, one may expe-
rience a completely different linguistic situation at the same bar from day to day.

11.4.2.2. First-person pronouns

As with the contrast between formal and informal speech, the use of first- and second-
person pronouns contributes to defining how speakers relate to one another at les-
bian bars. I discuss first-person pronouns in this section and second-person pronouns
in the following section.
Previous studies have claimed that the use of first- and second-person pronouns
exhibits the most gender-differentiated characteristics of the Japanese language. Ide
(1979) asserts that the first-person pronouns watakushi and watashi are used by both
sexes, the forms boku and ore (standard) as well as wagahai and washi (nonstand-
ard) are used exclusively by men, and the forms atakushi and atashi (standard) as
well as atai and uchi (non-standard) are used by women. Kanamaru (1997) claims
that jibun is also a masculine form. Jibun is a reflexive pronoun; its use as a personal
pronoun is relatively old-fashioned and is often associated with men in sports or in
militaristic groups such as the jieetai self-defense army or the police force. It is clear
that these options for first-person pronouns reflect the strong sense of an idealized
form of womens and mens speech. However, the choice of personal pronouns is
not as categorical as was once believed. Women in different parts of Japan may have
different options and women in different parts of the age spectrum may use different
forms (see chapter 10, this volume, for farm womens use of ore and chapter 14, this
volume, for junior high school girls use of boku and ore).
 LESBIAN BAR TALK IN SHINJUKU , TOKYO 213

An example from the bisexual and lesbian magazine Anise (Hagiwara 1997) gives
some insight into the use of first-person pronouns among lesbians, onabe, and trans-
sexuals. Examples (6) through (9) are drawn from the magazine transcript of a panel
discussion among six people, two self-identified female-to-male transsexuals (T1 and
T2), two self-identified onabe (O1 and O2), and two self-identified lesbians (L1 and
L2). They are talking about their gender.

(6) T1: Boku wa rezubian ga kirai nan ja nakute, rezu ikooru onna, jibun5 ga onna ni
mirareru no ga iya datta.
It is not that I dislike lesbians, but lesbian means a woman. I didnt want to be per-
ceived as a woman.
O2: Jibun mo yoku rezu tte iwarete jibun wa otoko nan da tte tomodachi to kenka
shimashita.
I was also told that Im a lesbian. I used to fight with my friends for saying that Im
a man.

(7) T2: Boku wa nenrei ga agareba penisu ga haete kuru mon da to omotte ita n desu yo.
I believed that once I got older, I would grow a penis.

(8) L1: Watashi wa monogokoro tsuita toki kara, zutto onna no ko ga suki datta. Otoko to
ka onabe ni naritai tte kimochi mo atta kedo, seken no hito wa otoko na no ka onna na
no ka waketagaru n da yo ne. Ippan shakai de wa joshi toire ni wa hairenai shi.
Ever since I was a child, I always liked girls. There was a time when I wanted to be a
man or onabe. People want to categorize themselves into men or women. In this soci-
ety, I cant use a public womens bathroom.
L2: Watashi wa joshi toire ni hairu yo.
I use a womens bathroom.

(9) O2: Jibun mo danjo to ka kangaenai hoo nan desu yo.

I tend not to think about if someone is a man or woman.

The three groups of speakers use first-person pronouns distinctively; the trans-
sexuals use boku, the onabe use jibun, and the lesbians use watashi almost uniformly.
It is as if there is a rule for them to use a different personal pronoun depending on
their own identity (at least in this context). Gender crossing reflected in this use of
first-person pronouns has nothing to do with these women wanting to be men; in-
stead, it is the appropriation of masculine resources.
The use of first-person pronouns shown in examples (6) through (9), however,
does not mean that each speaker uses the same pronoun no matter what context s/he is
in. Further analysis suggests that speakers negotiate gendered speech norms in each
context. This negotiation becomes more apparent if we examine examples in which a
speaker shifts the use of pronoun forms, which is frequently observed at lesbian bars.
In fact, one of the participants in the panel discussion (L1) is an employee at a lesbian
bar I often visited, where I observed her use of jibun. Moreover, I observed the use of
several first-person pronouns at lesbian bars: watashi, atashi, ore, washi, and jibun,
214 REAL LANGUAGE , REAL PEOPLE

with the same speaker using multiple first-person pronouns depending on the context.
One of the first questions I often asked at bars concerned the use of personal pronouns.
Younger employees in their early twenties exclusively listed jibun oneself (a form
commonly considered masculine) as their favorite first-person pronoun. One employee
explained that she uses the term because watashi and atashi exhibit too much feminin-
ity, but she does not want to use boku or other masculine first-person pronouns be-
cause she is not a man. She said that she refuses to dooitsuka-suru merge and identify
with men. An employee at a different bar argued, Why do I need to use boku? I dont
even like men. Her use of jibun is illustrated in example (10).

(10) Kutte nee, jibun wa. Nan da, nan ni mo nee jan.
I havent eaten. What! Theres nothing left.

This employee was 21 years old and had been working at the bar for several
months. Asked by her female supervisor if she had eaten something before coming
to work, she replied, using jibun. During the course of the evening, this young em-
ployee maintained the use of jibun irrespective of whether she was talking to her
coworkers or to customers. As she puts it, for her, jibun is the most neutral personal
pronoun available.6 The use of jibun was observed at all the bars in my study.
However, at a bar down the street I observed a quite different interaction. The
employee in example (11) was 20 years old and had been working almost a year.
When I entered, there was nobody but her in the bar. We introduced ourselves and
started talking about various things such as why she started working there and whether
she enjoyed working at the bar. When I noticed her using jibun in our interaction, I
asked the reason for her choice. Her answer was that she did not want to sound too
feminine by using watashi or atashi. She added that jibun was her favorite term and
that she did not use other masculine personal pronouns, unlike her boss, who uses
washi. It is interesting to note that washi presents a more masculine nuance to her
than jibun. However, her answer does not reflect her actual speaking practices. She
and I had been talking alone for more than two hours when the telephone rang. The
caller was a regular customer, whom this employee knew very well. The employee
explained to me that there had been a party at the bar the night before during which
this customer got drunk and did some crazy things that upset the employees and other
customers, but the customer claimed that she did not remember anything. The em-
ployee did not believe she had forgotten. The employees side of the conversation
was as follows.

(11) Omee na, fuzaken na yo, ore oko-ru yo.

Are you kidding? Ill get mad at you.
Jaa, A-san ni kiite mi na yo.
Then, ask A.
Anne, soo da yo, anta, minna ni meewaku kaketa n da kara.
You bothered everyone.
Ore sugee koshi itai mon. Koshi ni kita yo.
As for me, my lower back hurts. It really hurts.
 LESBIAN BAR TALK IN SHINJUKU , TOKYO 215

Here we see the employees shift from jibun in her conversation with me to ore
in her interaction with her customer. In the first instance of ore, the speaker is ex-
pressing her anger or frustration and thus is very emotionally involved. In the sec-
ond instance, she is describing how much pain she has now thanks to the customer.
The use of ore in this example as opposed to her previous use of jibun exhibits nego-
tiation of multiple identity positions in relation to different contexts. For her, jibun is
used in more formal settings of interaction, such as an interaction with a relatively
new customer who is also a researcher, but ore is her preferred choice in more inti-
mate and emotional contexts. It seems that since all the first-person pronouns (ex-
cept the relatively formal ones such as watakushi and watashi) are normatively
associated with either women or men, lesbians consider jibun the least gendered if
one wants to be, literally, simply oneself (see also Maree 1997 for discussion of the
use of jibun by onabe).
The use of boku is quite different. The speaker in example (12) is a lesbian cus-
tomer in her mid thirties who works at a computer graphics company. She and I had
been discussing the use of first-person pronouns for a while and I had noticed that
she used atashi in our conversation. When I asked if she ever used different first-
person pronouns, she answered by saying:

(12) Atashi wa kyosee o haru toki, boku o tsuka-u.

I (atashi) use I (boku) when [I] make a false show of power.

She added that she uses boku in arguments at her workplace with her male boss, who
may suspect that she is lesbian, and claims that this helps her situate herself at the
bosss level. Here the speaker explicitly recognizes the forcefulness attached to
masculine forms. She also expressed the belief that the feminine first-person
pronoun, atashi, did not make her strong. However, in explaining the use of boku
the speaker used the term kyosee false show of power, literally, kyo emptiness +
see power, force, which implies merely a superficial or even empty power.
These examples demonstrate that first-person pronouns are carefully and con-
sciously chosen by each speaker at lesbian bars since they are one of the most obvi-
ous linguistic features generally associated with sex/gender.

11.4.2.3. Second-person pronouns

Second-person pronouns are also considered highly gendered, but to a lesser degree
than first-person pronouns. Many researchers identify two forms (anata and anta)
for women and five or more forms (anata, anta, kimi, omae, kisama, otaku, and temee)
for men (Ide 1979, Kanamaru 1997). However, actual uses of second-person pro-
nouns do not necessarily conform to these gender classifications.7
I found two types of second-person pronouns at the bars I studied: anta and omee.
The frequency of these two terms is not high, since speakers generally prefer to use
nicknames or first names. Anata you, a more formal second-person pronoun, is
almost never used. The use of omee, a very casual masculine second-person pro-
noun, is found among close friends at bars, between employees and customers,
and between customers. In example (11), the speaker used both anta and omee in
216 REAL LANGUAGE , REAL PEOPLE

criticizing one of her customers. Omee is used to express the employees extreme
rage toward the customer (omee na, fuzaken na yo), whereas anta accompanies an
attempt at persuasion (soo da yo, anta, minna ni meewaku kaketa n da kara). Thus,
the shift between the two pronouns reflects the change in her emotional state.
The customer (in ex. 12) who uses boku in arguments with her boss also told
me that she uses omee when she argues with close male friends, adding that omee
has a certain forcefulness that cannot be expressed by anta or omae, a term that is
stereotypically associated with male speakers. The speakers in examples (11) and
(12) both assert that omee helps them argue more persuasively. Both speakers are
manipulating the pragmatic meaning (forcefulness) attached to the term omee.

11.4.2.4. Commands and requests

Other linguistic forms that are often associated with gender roles are found in com-
mands and requests. For instance, the use of bald imperative verb forms, such as
tabero eat, is generally considered to indicate that the interlocutor is not in a su-
perior position and also that the speaker is a man. This type of imperative is tradi-
tionally categorized as a masculine form (Martin 1975), and women are advised
to avoid using it. For example, Tohsaku (1999:43) notes that this type of impera-
tive form sounds very blunt and harsh and says that female speakers should not
use it at all. In other words, women are culturally or ideally discouraged from us-
ing bald imperatives. Furthermore, Japanese language textbooks such as Tohsakus
reinforce this notion.
Despite this prescription, example (11) includes strong imperative forms. The
first negative imperative, fuzaken na dont mess with me, can be articulated differ-
ently in different contexts, from the most formal fuzakenai de kudasai, to fuzakenai
de, to the very informal fuzakeru na. The assimilation and reduction of ru (in fuzakeru
na) to n (in fuzaken na) indexes even stronger roughness or toughness. As noted
earlier, the speaker in this example is extremely angry and emotionally involved. The
second imperative form, kiite mi na yo try to ask, is less strong than kiite miro.
Example (13) immediately continues the conversation in example (11).

(13) Chigau yo. Pantsu to bura wa tsukete ta yo.

No, you were wearing your underpants and a bra.
Un, itte ta. no yaroo tte. Dakara koi tte.
Yes, she was saying damn her. So, come here.
Mada A san wa inai yo. Iru yo. Maa soo da to omou yo.
A isnt here yet. Yes, shes here. Well, I think so.
Soo yuu koto ja nai kedo ne. Okorareru yo, mata A san ni.
I dont think so. Youll be chewed out by A again.

After the phraseno yaroo damn her, which is an extremely strong way to
address someone, another strong imperative form, koi, is used.8 This direct affirma-
tive command form koi is normally used by men, as noted earlier, but it is used here
to make a strong command.
 LESBIAN BAR TALK IN SHINJUKU , TOKYO 217

11.4.2.5. Sentence-final particles

The last linguistic feature I discuss here is sentence-final particles, which are often
studied in Japanese language and gender research (Uyeno 1971, McGloin 1997,
Okamoto & Sato 1992, Okamoto 1995, Mastsumoto 1996, Abe 1999, 2000). The
functions of such sentence-final forms have been identified as (1) indicating the
speakers emotions and attitudes such as doubt, caution, and confirmation (Martin
1975:914, Makino & Tsutsui 1986:45); (2) encouraging rapport between speech
participants (Makino & Tsutsui 1986:45); (3) achieving a close monitoring of the
feelings between speech participants (Maynard l987:28); and (4) expressing ones
own masculinity and/or femininity (Makino & Tsutsui l986:49). Traditional gender
classification of sentence-final particles suggests three types: (1) feminine forms,
which are said to be predominantly used by female speakers (e.g., wa with a rising
intonation, na no/na no ne, kashira); (2) masculine forms, used by male speakers
(e.g., zo/ze, da, da yo); and (3) neutral forms, used by both women and men (e.g., yo
ne, ka na). As Okamoto (1995) argues, the traditional classification should be un-
derstood as a reference point rather than a description of actual use.
The actual use of sentence-final particles at the lesbian bars bears this point out.
The most frequently used sentence-final particles are yo and its variations such as da
yo and da yo ne. For example, in (14) an employee and a few customers are chatting
while looking at a magazine article, which describes an onabe bar.

(14) Customer: Kore otoko da yo, hakkiri itte.

This is a man, frankly speaking.
Employee: Kurabete moraitaku nai ne.
I dont want you to compare me with this guy.

The customer points out that the figure in the picture in the magazine looks like a
man. She uses da yo in this utterance. As mentioned earlier, in the conventional classi-
fication, da + yo is regarded as a masculine form. Yet it seems that this form is used
more commonly used by many women than it used to be (Abe 2000, chapter 13, this
volume). Thus, it is possible that this form is neutral for this speaker.
Other conventional masculine forms are found in example (15). Here younger
employees, who typically refer to themselves as jibun, are talking with some cus-
tomers in their thirties. Two of them are trying to remove a spot from the surface of
a toy.

(15) Employee: Kosutte mo torenee n da.

We cant get rid of it even by rubbing.
Customer : Nani ka iro ga chiga-u zo.
Hey, the color is somehow different!

The employee uses da, a (moderately) masculine form, and the customer uses zo,
which is often classified as strongly masculine. In addition, the rough negative form,
nee, is used instead of nai. This use of nee was also observed at a different bar, as
shown in example (10) earlier. There are two instances of nee in (10). Moreover, the
218 REAL LANGUAGE , REAL PEOPLE

speaker uses kuu to eat instead of taberu to eat. The latter is a standard form of the
verb, while kuu suggests masculine speech. This speaker elsewhere uses the plain
interrogative form ka (verb + ka) in Ureshii ka are you happy? and Nomu ka do you
want to drink? another construction associated with masculine speech. This type of
question form is usually considered too strong and rude for women to use.
In sum, these speakers are rejecting the forms that they feel are too feminine, while
adopting masculine or neutral forms. Their use of masculine forms, however, does not
mean that lesbians want to be identified as men, as the discussion on the difference
between lesbian and onabe clearly demonstrates, nor is it simply the case of butch les-
bians speaking like men. On the other hand, even though they may consider themselves
women (also indicated earlier in their comparison of lesbians and onabe), they are not
using stereotypical feminine forms. I argue, therefore, that their linguistic choices in-
dicate that they are marking their difference in speech, which, they believe, supports
their identities as lesbians. For instance, the use of direct/informal style and masculine
sentence-final particles may be a way to create an intimate relationship among lesbi-
ans at the bars they frequent or to treat them as in-group members. Further, their use of
masculine first-person pronouns may be an appropriation of mens resources, which
enable them to express (real or imagined) powerfulness. In other words, some lesbians
at bars recognize the assumed dominant power associated with masculine speech. It is
also important to note that the same speaker may shift her speech style depending on
the context. Recall, for example, that one speaker who used jibun I with me switched
to ore I when talking to someone else. Shifts such as this indicate the speakers ne-
gotiation of gendered identity in interaction.

11.6. Conclusion

In this chapter, I have examined language use at lesbian bars in Tokyo. In the first
part of the chapter, I have shown how lesbians use a variety of category names to
identify themselves or distinguish themselves from others, from which I conclude
that gender identities for these lesbians are not fixed, certain, or natural. Rather, they
are constantly negotiated among themselves. In the latter part of the chapter, I ana-
lyzed characteristic speech styles observed at lesbian bars, in particular, the com-
mon use of linguistic features stereotypically associated with male speakers. I have
argued that speakers use masculine speech styles in a complex way in order to ex-
press a variety of context-dependent meanings related to their lesbian identities and
relationships. In sum, this study demonstrates that gender identities are not fixed,
but negotiated and constructed through language use in interactions.

Notes
1. There are also quite a few gay bars in other parts of Tokyo, such as Shinbashi, where
the first gay bar, Yanagi, was opened in 1950; there were 50 to 60 in other parts of Tokyo by
1960 (Sunagawa 1998).
2. This interview was taken from an article in a magazine called Marumaru, which spe-
cializes in introducing bars and men who work for the bars. (These bars are mostly in Shinjuku.)
 LESBIAN BAR TALK IN SHINJUKU , TOKYO 219

A male reporter interviews many owners and managers of bars, including host, onabe, and
lesbian bars. Four lesbian bar owners/managers are interviewed, which is shown in examples
(1) through (5) (Marumaru 2000: 2426).
3. She has been working for two years at the oldest lesbian bar. She is 21 years old.
4. Although this example is taken from a magazine interview, this manager and I vis-
ited at her bar quite often.
5. Here jibun is a reflexive pronoun, not a personal pronoun.
6. A similar explanation was given when I interviewed a cross-dresser who prefers the
use of watashi. S/he told me that it was her/his conscious attempt to be neutral. S/he argues
that many gay men use atashi or ataiwhich is often associated with womenbut that s/he
intentionally rejects these words because s/he does not want to be slotted into one of the bi-
nary categories of women or men. S/he contends that watashi is a neutral first-person
pronoun and is not as formal as watakushi.
7. Kimi is often used by female schoolteachers. I myself use it when talking to my former
students or close male and female friends. The use of omae is much more common by male
speakers, but it seems to be used by some women, particularly by speakers of some regional
dialects.
8. This type of imperative is also found in Abes (2000) study of womens speech, in
which the speaker, a film directors wife, uses it 1.86% of the time.

References
Abe, Hideko (1999). Power negotiation between Japanese females and males. Intercultural
Communication Studies: 8(1): 115.
(2000). Speaking of power: Japanese women and their language. Munich: Lincom
Europa.
Achilles, Nancy (1993[1967]). The development of the homosexual bar as an institution. In
H. Abelove, M. A. Barale, and D. M. Halperin (eds.), Lesbian and gay studies reader,
175182. New York: Routledge.
AKIKO (2000). Love Revolution 2.
Barrett, Rusty (1997). The homo-genius speech community. In A. Livia and K. Hall (eds.),
Queerly phrased: Language, gender, and sexuality, 181201. New York: Oxford Uni-
versity Press.
Bucholtz, Mary (1999). Bad examples: Transgression and progress in language and gender
studies. In M. Bucholtz, A. C. Liang and L. A. Sutton (eds.), Reinventing identities: The
gendered self in discourse, 324. New York: Oxford University Press.
Butler, Judith (1990). Gender trouble: Feminism and the subversion of identity. New York:
Routledge.
Chamberland, Line (1993). Remembering lesbian bars: Montreal, 19551975. The Journal
of Homosexuality 25(3): 231269.
Foucault, Michel (1990). The history of sexuality: An introduction. New York: Random House.
Gaudio, Rudolf P. (2001). White men do it too: Racialized (homo)sexualities in postcolonial
Hausaland. Journal of Linguistic Anthropology 11(1): 3651.
Gilmartin, Katie (1996). We are not bar people: Middle-class lesbian identities and cul-
tural spaces. Journal of Lesbian and Gay Studies 3: 151.
Hagiwara, Mami (ed.) (1997). Sexuality. Anise 3: 2427.
Hankin, Kelly (2002). The girls in the back room. Minneapolis: University of Minnesota.
Ide, Sachiko (1979). Onna no kotoba, otoko no kotoba (Womens language, mens language).
Tokyo: Nihon Keizai Tsushinsha.
220 REAL LANGUAGE , REAL PEOPLE

Jagose, Annamaire (1996). Queer theory: An introduction. New York: New York University
Press.
Kanamaru, Fumi (1997). Ninshoo daimeeshi/koshoo (Person pronouns and address terms). In
S. Ide (ed.), Joseigo no sekai (The world of womens language), 1532. Tokyo: Meiji Shoin.
Kleinfeld, Mala S., and Noni Warner (1997). Lexical variation in the deaf community relat-
ing to gay, lesbian, and bisexual signs. In A. Livia and K. Hall (eds.), Queerly phrased:
Language, gender, and sexuality, 5884. New York: Oxford University Press.
Kramarae, Cheris, Muriel Schultz, and William OBarr (eds.) (1984). Language and power.
Beverly Hills: Sage.
Liang, A. C. (1997). The creation of coherence in coming-out stories. In A. Livia and K. Hall
(eds.), Queerly phrased: Language, gender, and sexuality, 287309. New York: Ox-
ford University Press.
(1999). Conversationally implicating lesbian and gay identity. In M. Bucholtz A. C.
Liang, and L. A. Sutton (eds.), Reinventing identities: The gendered self in discourse,
293312. New York: Oxford University Press.
Livia, Anna (1997). Disloyal to masculinity: Linguistic gender and liminal identity in French.
In A. Livia and K. Hall (eds.), Queerly phrased: Language, gender, and sexuality, 349
368. New York: Oxford University Press.
(2000) Bi, androgyne Sabstenir: Lesbian community creation through the personals.
Paper presented at the First International Gender and Language Conference, Stanford, CA.
Makino, Seiichi, and Michio Tsutsui (1986). A dictionary of basic Japanese grammar. Tokyo:
Japan Times.
Maree, Claire (1997). Jendaa shihyoo to jendaa no imisee no henka: Eega Shinjuku Boys ni
okero onabe no baai (Gender indexicality and semantic shifts in meanings: the case of
an onabe in the documentary film Shinjuku Boys). Gendai Shisoo 25(13): 263278.
Martin, Samuel (1975). A reference grammar of Japanese. New Haven: Yale University.
Marumaru (2000). Best spot. Marumaru: 2426.
Matsumoto, Yoshiko (1996). Does less feminine speech in Japanese mean less femininity?
In N. Warner, J. Ahlers, L. Bilmes, M. Oliver, S. Wertheim, and M. Chen (eds.), Gen-
der and belief systems: Proceedings of the Fourth Berkeley Women and Language Con-
ference, 455467. Berkeley: Berkeley Women and Language Group.
Maynard, Senko Kumiya (1987). Japanese conversation: Self-contextualization through struc-
ture and interactional management. Norwood, NJ: Ablex.
McGloin, Naomi Hanaoka (1997). Shuujoshi (Sentence-final particles). In S. Ide (ed.), Joseego
no sekai (World of womens language). Tokyo: Meijishoin.
Moonwomon-Baird, Birch (1997). Toward the study of lesbian speech. In A. Livia and
K. Hall (eds.), Queerly phrased: Language, gender, and sexuality, 202213. New York:
Oxford University Press.
(2000). Tales of lesbian times: A temporal template for the life story present. Paper
presented at the First International Gender and Language Conference, Stanford, CA.
Neumann, Tina M. (1997). Deaf identity, lesbian identity: Intersections in a life narrative. In
A. Livia and K. Hall (eds.), Queerly phrased: Language, gender, and sexuality, 274
286. New York: Oxford University Press.
Ogawa, Naoko, and Janet S. [Shibamoto] Smith (1997). The gendering of the gay male sex
class in Japan: A preliminary case study based on Rasen no Sobyoo. In A. Livia and
K. Hall (eds.), Queerly phrased: Language, gender, and sexuality, 402415. New York:
Oxford University Press.
Okamoto, Shigeko (1995). Tasteless Japanese: Less feminine speech among young Japa-
nese women. In K. Hall and M. Bucholtz (eds.), Gender articulated: Language and the
socially constructed self, 297325. New York: Routledge.
 LESBIAN BAR TALK IN SHINJUKU , TOKYO 221

Okamoto, Shigeko, and Shie Sato (1992). Less feminine speech among young Japanese fe-
males. In K. Hall, M. Bucholtz, and B. Moonwomon (eds.), Locating power: Proceed-
ings of the Second Berkeley Women and Language Conference, 478488. Berkeley:
Berkeley Women and Language Group.
Ootsuka, Takashi (1995). Ni-choome kara uroko: Shinjuku gee sutoriito zakki-choo (Shinjuku
2-chome gay street note). Tokyo: Shooeisha.
Pastre, Genevieve (1997). Linguistic gender play among French gays and lesbians. In A. Livia
and K. Hall (eds.), Queerly phrased: Language, gender, and sexuality, 369379. New
York: Oxford University Press.
Queen, Robin M. (1997) I dont speak spritch: Locating lesbian language. In A. Livia and
K. Hall (eds.), Queerly phrased: Language, gender, and sexuality, 233256. New York
and Oxford: Oxford University Press.
(2000). Language use in mainstream lesbian films. Paper presented at the First Inter-
national Gender and Language Conference, Stanford, CA.
Read, Kenneth E. (1980). Other voices: The style of a male homosexual tavern. Novato, CA:
Chandler and Sharp.
Sunagawa, Hideki (1998). Sekushuaritii no saiteegi ni mukete (Toward the redefinition of
sexuality). Unpublished masters thesis, University of Tokyo.
Tohsaku, Yasu-Hiko (1999). Yookoso: Continuing with contemporary Japanese. Boston:
McGraw-Hill College.
Treat, John Whittier (1999). Great mirrors shattered: Homosexuality, orientalism, and Japan.
New York: Oxford University Press.
Ueno, Chizuko (1999). Gender studies: Gender trouble. Gendaishisoo 27(1): 4477.
Uminarikan (ed.) (1998). Otoko-gai mappu 98 nenban ([Gay] mens map of 1998). Tokyo:
Uminarikan.
Uyeno, Tazuko Y. (1971). A study of Japanese modality: A performative analysis of sen-
tence particles. Unpublished Ph.D. dissertation, University of Michigan.
Valentine, James (1997). Pots and pans: Identification of queer Japanese in terms of discrimi-
nation. In A. Livia and K. Hall (eds.), Queerly phrased: Language, gender, and sexual-
ity, 95114. New York: Oxford University Press.
Watanabe, Mieko (1990). Uuman rabingu (Women loving). Tokyo: Gendai Shokan.
Wood, Kathleen M. (1997). Narrative iconicity in electronic-mail lesbian coming-out sto-
ries. In A. Livia and K. Hall (eds.), Queerly phrased: Language, gender, and sexuality,
257273. New York: Oxford University Press.
222 REAL LANGUAGE , REAL PEOPLE

12

YUMIKO OHARA

Prosody and Gender

in Workplace Interaction
Exploring Constraints and Resources
in the Use of Japanese

12.1. Constraints and resources

Many previous studies of the Japanese language have made a distinction between
the speech used by women and men (e.g., Ide 1979, Jugaku 1979, Shibamoto 1985,
Shibatani 1990). One of the primary bases for this distinction has been the observa-
tion that the Japanese language includes certain gentler linguistic items that are used
by women (e.g., the first-person pronoun atashi and the sentence-final forms wa and
no yo) and certain rougher items that are used by men (e.g., the first-person pronoun
ore and the sentence-final particles zo and ze). It is frequently assumed that it is so-
cially unacceptable for women and men to cross these boundaries and use linguistic
forms inappropriate for their gender. In other words, social expectations about lan-
guage in Japan work as constraints that control, or at least influence, how women
and men actually use language.
There is, however, a set of emerging studies that, primarily through empirical
studies of actual language use, indicates that especially female speakers of Japanese
are not actually as constrained in their use of language as previously thought (Inoue
1994, Okamoto 1995, Matsumoto 1996). Okamoto, for example, found that young
women in conversations among friends were not averse to making use of the sen-
tence-final particles zo and ze as well as vulgar, strongly masculine, and very
informal lexical expressions (1995, 1996, 1997). These recent findings have, on the
one hand, been used by researchers to challenge the female speechmale speech di-
chotomy in the Japanese language. Inoue (1994:322), for one, acknowledging both
tremendous social and contextual diversity of language use, points out that there is

222
 PROSODY AND GENDER IN WORKPLACE INTERACTION 223

actually no empirically tested single set of speech patterns for all Japanese women
and suggests that Japanese womens speech be treated as merely one particular
representation of how Japanese women speak (see also chapter 1, this volume). On
the other hand, these findings have led researchers to underscore the degree to which
language serves as a set of resources that allows women (as well as all speakers of a
language) to construct a variety of gendered identities according to different situa-
tions. In other words, instead of always and obligatorily using language according to
social constraints, women are much freer than previously depicted to use various
linguistic resources to project different images, which themselves may not necessar-
ily conform to the cultural models of normative Japanese womanhood.
To be sure, research that has emphasized Japanese womens access to a wide
range of linguistic resources, including those once deemed to belong strictly to men,
does not deny the existence of gender-based constraints on language use. Still, rec-
ognition of the diverse speech styles adopted by Japanese women has prompted a
significant shift in research focus. Rather than being viewed as strict followers of a
predetermined code, Japanese womens language, female speakers of Japanese have
been recast as agents capable of using a variety of linguistic forms in innovative and
resourceful ways that challenge or even resist tradition. As Matsumoto (1996:464)
concludes, based on her study of womens speech both in contemporary magazine
advertisements and in late nineteenth- and early twentieth-century Japan, young
womens use of unconventional speech styles indicates not only a greater role for
speakers agency than previously recognized but also young womens ability to
produce through their choice of language resistance to the dominant ideology that
frames the normative concept of femininity.
In this chapter, I attempt to shed further light on resources and constraints in the
use of language by Japanese women and men by focusing on one particular aspect
of language, namely, prosody. More specifically, I examine the voice pitch levels
used by speakers of Japanese in their places of employment. Based on this chapters
examination of naturally produced speech by women and men, I contend that Japa-
nese women are constrained in their pitch behavior, but at the same time I describe
some of the ways that voice pitch functions as a resource that enables both the fe-
male and male employees to accomplish the speech activities that constitute their
roles and relations at their workplaces. In particular, I show that a high-pitched voice
can be used strategically by both female and male speakers of Japanese to empha-
size specific aspects of their utterances. I describe the procedures used in this study
in more detail in section 12.3, but I first explain in the next section the basis for using
prosody to study gender in a specific linguacultural environment such as Japan.

12.2. Voice pitch and cultural assumptions:

Making the connection

Traditionally, scholars in the fields of phonetics, communication, and speech have

most frequently attempted to explain voice pitch variation in terms of human anatomy,
physiology, physical and emotional states, age, or the structure of the language
being spoken. However, some recent research has suggested that culture plays an
224 REAL LANGUAGE , REAL PEOPLE

important role in the production of voice pitch. For example, in some of my previ-
ous research (Ohara 1992, 1997, 1999), I measured the voice pitch levels of bilin-
gual speakers of Japanese and English and discovered that the results of differences
across language and gender could best be explained in terms of culture. Finding that
the bilingual women produced higher fundamental frequency levels when speaking
in Japanese than in English while the bilingual men did not vary their pitch across
languages, I suggested that this difference could not be explained in terms of the struc-
tures of the languages, physiological differences, or variances in emotional or health
states. Instead, I posited that the results were best understood in terms of sociocul-
tural factors. Referring to previous research that has linked a high voice pitch to a
womans role in Japanese society (Loveday 1986), I argued that female speakers of
Japanese were under cultural constraints that did not necessarily apply to men. More
specifically, the claim was that these cultural constraints centered on expectations
about femininity in Japanese society; one way Japanese women are expected to exert
a feminine persona is through the use of a high-pitched voice.
These claims were supported by research that focused on the perception of voice
pitch in Japanese (Ohara 1993, 1997; van Bezooijen 1995, 1996). Van Bezooijen
(1995), for instance, in a comparative study of sociocultural aspects of pitch in Japan
and the Netherlands, found a preference in Japanese culture for womens use of a
high-pitched voice, while the Dutch culture preferred a low-pitched voice. In the same
study she also observed a strong association between physical and psychological
powerlessness (small, weak, indirect, modest, etc.) and a high-pitched voice. Her
results, in other words, strengthen the argument for the existence of cultural expec-
tations directly concerning female pitch behavior: If a higher-pitched voice is deemed
more socially desirable, then many women may want to project socially desired
images by using a high-pitched voice. That is, they may elevate their ptich in order
to be heard as smaller, weaker, more indirect, and more modest.
Yet while this previous research has indicated imbalances in the voice pitch
behavior of women and men in Japanese and, at the same time, suggested that gen-
der norms and expectations in Japanese culture lead women to employ a high-pitched
voice in order to explicitly express their femininity, there are two areas that require
methodological refinement. First, most of my previous research (Ohara 1992, 1997,
1999), as well as the majority of research on voice pitch, has been based on data
produced under laboratory conditions, with speakers reading isolated sentences into
a tape recorder. In this study, by contrast, I move the research out of the laboratory
and work with data that was tape-recorded in naturalistic settings. Because voice pitch,
like other aspects of language such as syntax, phonology, and morphology, is some-
thing people use to express ideas, intentions, identities, and so on in real interactional
encounters, this was deemed a necessary methodological move. Second, a method
often used in previous research is to examine fundamental frequencies of speech
samples. In the research reported here, in addition to this method I base my analysis
of the naturalistic data on pitch extractions that actually show the flow of the pitch
contours of the speakers utterances. Only by attempting to place voice pitch within
actual interactions and track how it changes and fluctuates as speakers engage in
different speech acts will it be possible to gain a deeper understanding of what voice
pitch means in human communication.
 PROSODY AND GENDER IN WORKPLACE INTERACTION 225

12.3. Data and procedures

The analysis focuses on the speech of four native speakers of Japanese, two women
and two men, who at the time of the study were working in businesses located in
Honolulu, Hawaii. All four of the speakers were middle-class and in their mid-to-
late thirties at the time of the study. The recordings took place over two two-week
periods, one at the end of 1999 and the other at the beginning of 2000. During these
intervals, the speakers were given a tape recorder and a microphone and asked to
record their own interactions during their workday. For all of the recordings, the
researcher was not present. A total of 15 hours of conversations was recorded, ap-
proximately 7 hours and 2 hours respectively for the two women and 4 hours and
2 hours respectively for the two men.
As a part of their interactions with customers, coworkers, business associates,
and acquaintances, the four speakers typically performed and engaged in a number
of different speech acts, including greetings, requests, apologies, posing and answer-
ing questions, negating prior utterances, and closings. In this chapter, in order to
compare and contrast the pitch levels of the speakers across speech acts, I focus the
analysis on two speech acts, requests and negation. These two speech acts were cho-
sen because of the frequency with which they appear in the data. If gender differen-
tiation in voice pitch levels is indeed a pervasive and fundamental feature of the
Japanese language, then we would expect to see this differentiation in the accom-
plishment of very common actions, such as negation and requests, in Japanese social
interaction.
With regard to requests, the analysis focuses specifically on those utterances
in which the speakers told their interlocutors to wait for a moment (in Japanese,
some variant of chotto matte kudasai Please wait for a moment).1 Based on the
perlocutionary force of the utterances, I categorized them together as one group, which
I refer to as requests. Although negation is a category constituted by a wide variety
of utterances, for example, negation of ones own previous statements (i.e., I said X
previously, but I was wrong. Its actually Y) and negation of statements produced
by others (i.e., What you said isnt true), I consider in the analysis only one type,
namely, negation of ones own previous statements. The reason for this choice is that
these kinds of utterances are common in the data. While the speakers rarely directly
negated statements made by customers, they often had to negate their own prior state-
ments as a part of explaining things to customers (i.e., I told you we could do X
earlier, but actually it wont be possible to do X). In addition, this kind of negation
was not uncommon in interaction with acquaintances.
In order to consider the effects of context, I examine these two speech acts as
they were produced to two different types of interlocutors, customers and acquain-
tances. The term customer is used to refer to buyers or potential buyers of the service
provided by the company. Acquaintance refers to people whom the speakers them-
selves considered close friends, including some co-workers who were equal in rank
to the speakers. For the purposes of the analysis, a total of 64 utterances, 32 requests
and 32 negations, were first located within the 15 hours of recorded data. For each of
the four speakers, 8 utterances per speech act were chosen, 4 that were made to
customers and 4 made to acquaintances. The 64 utterances were first subjected to a
226 REAL LANGUAGE , REAL PEOPLE

fundamental frequency analysis that derived numerical values for the average, maxi-
mum, minimum, and standard deviation. Next, from the 64 utterances, the 16 most
typical with respect to fundamental frequency values were chosen for a more detailed
analysis that included examination of pitch extractions. This time, for each of the
four speakers 4 utterances were chosen, 2 requests (1 to a customer and 1 to an ac-
quaintance) and 2 negations (1 to a customer and 1 to an acquaintance).

12.4. Results

The results are divided into three subsections. The first presents the findings of the
fundamental frequency analysis for each of the 16 utterances chosen for further analy-
sis. The second displays pitch extractions of some of the 16 utterances in order to
point out similarities as well as differences across addressees, contexts, and the gen-
ders of the speakers. The third provides further details about the pitch extractions as
well as additional extractions in order to highlight the relationship between prosodic
patterns and the notion of emphasis.
For the sake of clarity, I have labeled the speakers A, B, C, and D, with A and B
being the female speakers and C and D the males. In addition, I have used a code to
indicate the interlocutor. The small letter a indicates acquaintance, and c signifies
customer. Thus, the two symbols together, Aa, signify Speaker As utterance directed
to an acquaintance.

12.4.1. Fundamental frequency analysis

This subsection presents the results of the fundamental frequency analysis of the 16
utterances that were most representative in terms of their fundamental frequency
levels. Table 12.1 shows the results of the analysis for the eight requests. The utter-
ance itself is shown in the last column on the right.
The numerical values show some noteworthy tendencies regarding the partici-
pants gender. First of all, although the average pitch levels for the two female speak-
ers, A and B, are considerably higher than those of speaker C, they are not necessarily
higher than those of speaker D. This point would seem to contradict the previous
findings that women use a higher pitched voice than those of men when speaking
Japanese. In addition, a female speaker, B, used minimum pitch levels that were lower
than those of male speaker D, another point where the numerical values seem to con-
tradict previous findings. The only place where the fundamental frequency values
display recognizable gender differentiation is across interlocutors. The female speak-
ers used higher fundamental frequency levels when speaking to customers than they
did when speaking to acquaintances, but the male speakers C and D did not show
much variation across interlocutors. Moreover, the variation they did exhibit was not
patterned, as the womens was. That is, both women had higher fundamental fre-
quencies when addressing customers than when addressing acquaintances, but one
of the male participants raised his voice pitch slightly when talking to customers,
while the other participant employed a somewhat lower pitch when interacting with
customers. The standard deviations show that for the female speakers fundamental
 PROSODY AND GENDER IN WORKPLACE INTERACTION 227

12.1 Speakers fundamental frequency levels when producing requests

Gender Speaker Average Max. Min. sd* Utterance

Female Aa 217 277 167 35.39 a chotto mattee

oh, please wait
Ac 260 352 174 49.84 shoo shoo omachi
kudasaimasee
please wait for a moment
Ba 166 221 116 30.38 chotto mate chotto mate
wait a little; wait a little
Bc 202 365 133 61.91 shoo shoo omachi itadakemasu ka
would you please wait for a moment?
Male Ca 130 166 115 19.78 mateyooo
wait
Cc 139 169 106 17.24 haai chotto matte kudasaai
yes, please wait for a moment
Da 218 260 160 39.36 aa chotto matte
um, please wait
Dc 211 243 135 29.08 hai ja sochira de matte kudasai
yes, please wait over there
*The abbreviation sd stands for standard deviation, which indicates the degree to which fundamental frequency levels
varied throughout the utterance.

frequencies varied more when they interacted with customers than when they inter-
acted with acquaintances. Furthermore, Speaker B shows a greater variation across
interlocutors than Speaker A. For the male speakers, the standard deviation figures
do not vary much across interlocutors, especially for Speaker C.
Table 12.2 shows the fundamental frequency levels for each speaker when pro-
ducing negations.
While the female speakers did show a pattern of elevated pitch when interacting
with customers, some tendencies seen in negation, like those in requests, appear to
contradict the results of previous studies. Especially for Speaker B, the average pitch
level used when talking to an acquaintance is lower than the levels used by the two
male speakers when they spoke with acquaintances. In addition, not only is Speaker
Bs maximum pitch level when speaking to an acquaintance lower than the two male
speakers maximum pitch, but also Bs minimum pitch level with an acquaintance is
nearly identical to those of the two male speakers.
Thus, speaking at a general level, the results would seem to be quite equivocal.
On the one hand, it is difficult to conclude that Japanese women are under a general
constraint to use high pitch levels if, under some conditions (such as speaking to
acquaintances), at least some speakers (such as Speaker B in this study) are shown to
use pitch levels equivalent to or even lower than those used by Japanese men. On the
other hand, the fact that only the female speakers raised their voice pitch levels when
speaking to customers would seem to suggest that their choice of pitch levels is not
random. It would suggest that there was, at least when speaking to customers, some
underlying motivation for employing a high-pitched voice. In order to enhance un-
derstanding of these results, I want to examine in more detail some of the requests
and negations made by the speakers. To do so, I will present some of the pitch
228 REAL LANGUAGE , REAL PEOPLE

12.2 Speakers fundamental frequency level when producing negation

Gender Speaker Average Max. Min. sd Utterance

Female Aa 209 326 156 38.98 anoo kooraru no hoo ano moobu ja nai
hoo nee
um, the one thats coral, not the one
thats mauve
Ac 241 405 169 50.98 de ano tsuu ga anoo seezoo chuushi
natte surii subete kawatteru n desu
kedo
um, those labeled 2 were discontinued
and have been all changed to 3
Ba 172 208 104 32.28 namamono ja nai nihon no pantsu
its not a raw food; its a Japanese
underwear
Bc 205 462 134 77.29 aite ita n desu keredomo nihon no hoo
kara yoyaku haitte shimaimashitee
it was vacant, but a reservation from
Japan came in
Male Ca 185 235 105 24.13 koko ni wa nai to omoun da kedo
I think we dont have it here
Cc 171 221 107 31.29 are ima muryoo ja nai mitai desu nee
it seems that it isnt free now
Da 184 228 106 35.53 chigau kapiolani ja nakute kapahulu
datta yoo
its wrong; it wasnt kapiolani but
kapahulu
Dc 186 237 112 38.24 igai nan desu kedo renzoku ja torenai
desu nee
although its unexpected, it cannot be
reserved for consecutive days

extractions of these 16 utterances. Examination of the pitch ranges and contours shown
in the extractions makes it clear that the notions of constraints and resources are both
applicable to the ways that the speakers employed voice pitch.

12.4.2. Pitch extractions

The advantage of showing pitch extractions is that they make it possible to view the
pitch contour and pitch range with which the utterances were produced. By showing
in this subsection the pitch extractions for some of the utterances I hope to highlight
differencesnot easily seen in a purely numerical analysisin the way that the fe-
male and male speakers produced their requests and negations.
Figure 12.1 shows the pitch extraction for As request to an acquaintance. As
the extraction shows, this utterance starts around 240 hertz, continues at that level a
short while, and ends around 180 hertz.
Figure 12.2 shows the pitch extraction for As request to a customer. As the ex-
traction shows, the first segment of this utterance, shoo shoo a little, starts around
 PROSODY AND GENDER IN WORKPLACE INTERACTION 229

ARa

a chottoma ttee

500
450
400
350
300
250
200
150
100
Hz
ms 600 1200 1800 2400 3000

12.1 Pitch extraction of Speaker As request to an acquaintance

280 hertz, rises to around 350 hertz and then ends at around 170 hertz. The pitch
range used for this utterance is much larger than that for the utterance shown earlier
in figure 12.1. This pattern, whereby a speaker used a much larger range of voice
pitch levels to a customer than to an acquaintance, was also apparent in the pitch
extractions for Speaker Bs requests. It was not, however, observed in the extrac-
tions for the requests made by the two male speakers.
Figure 12.3 shows the pitch extraction for Cs request to an acquaintance. As
the extraction shows, this utterance starts around 110 hertz and then rises and fin-
ishes at around 160 hertz. Figure 12.4 shows the pitch extraction for Cs request to a
customer. The first part of this utterance, haai yes, is produced at around 160 hertz

ARc

shooshooomachikudasaimasee

500
450
400
350
300
250
200
150
100
Hz
ms 600 1200 1800 2400 3000

12.2 Pitch extraction of Speaker As request to a customer

230 REAL LANGUAGE , REAL PEOPLE

CRa

mateyoo

500
450
400
350
300
250
200
150
100
Hz
ms 600 1200 1800 2400 3000

12.3 Pitch extraction of Speaker Cs request to an acquaintance

and the latter part starts around 160 hertz and ends around 100 hertz. Unlike Speak-
ers A and B, there is no considerable difference between the extractions of requests
that are directed toward acquaintances and those directed toward customers. Although
I will not take the space to show the extractions for Speaker Ds requests, it can be
noted that like Speaker C, there was little variance in the shape of the extractions of
requests to acquaintances and customers.
Having presented some of the pitch extractions for the eight requests produced
by the speakers, I now turn to the pitch extractions of the negations they uttered. Fig-
ure 12.5 shows the pitch extraction for Bs negation to an acquaintance. As can be
seen, the first part of this utterance, namamono ja nai it isnt a raw food, starts around

CRc

haai chotto matte kudasaai

500
450
400
350
300
250
200
150
100
Hz
ms 600 1200 1800 2400 3000

12.4 Pitch extraction of Speaker Cs request to a customer

 PROSODY AND GENDER IN WORKPLACE INTERACTION 231

BNa

namamono janai nihonno pantsu

500
450
400
350
300
250
200
150
100
Hz
ms 600 1200 1800 2400 3000

12.5 Pitch extraction of Speaker Bs negation to an acquaintance

200 hertz and ends at about 130 hertz. The latter part begins at about 180 hertz, goes
up until it reaches around 200 hertz, and then drops and ends around 160 hertz. Fig-
ure 12.6 shows the pitch extraction for Bs negation to a customer.
The most striking feature of the utterance is the significantly high peak, over
450 hertz, within this segment. As the extraction shows, the utterance starts around
200 hertz, goes up to about 250 hertz, then goes down to around 150 hertz, next
rises slightly to around 180 hertz, then rises dramatically to the highest peak of the
utterance, 450 hertz, next falls to around 160 hertz, goes up again to about 260
hertz, and finally drops and ends around 150 hertz. Of the two female speakers,
Bs negation to a customer shows the larger range of voice pitch levels. However,

BNc

aitetandesukeredomo nihonnohoo karayoyaku hai tte shimaimashitee

500
450
400
350
300
250
200
150
100
Hz
ms 600 1200 1800 2400 3000

12.6 Pitch extraction of Speaker Bs negation to a customer

232 REAL LANGUAGE , REAL PEOPLE

like B, As negation to a customer showed a larger range than did her negation to
an acquaintance.
One of the most interesting aspects of the negation extractions is the pitch ranges
used by men compared to women. Very similar to their use of pitch in requests, nei-
ther Speaker C nor D showed much variation in his pitch levels when producing
negation to an acquaintance or to a customer. Here I show Ds negation to an ac-
quaintance (figure 12.7). This utterance starts around 180 hertz, rises to about 220
hertz, falls to around 120 hertz, then goes up again to about 200 hertz, and then falls
and ends around 100 hertz. Figure 12.8 shows Speaker Ds negation to a customer.
As the extraction shows, this utterance starts at around 220 hertz, goes up to about
250 hertz, then declines to about 130 hertz, rises again to around 250 hertz, falls to
approximately 140 hertz, and ends at around 100 hertz. The range of pitch levels,
from 100 hertz to 250 hertz, is slightly more than the other three negations produced
by the two male speakers but not nearly as large as the ranges produced by female
speakers in the same situation. In the next subsection, I examine this and other pitch
extractions in more detail, focusing especially on the connection between voice pitch
and emphasis.

12.4.3. Pitch contours and the concept of emphasis

As noted earlier, the results of the fundamental frequency analysis shown in section
12.4.1 are particularly striking because they seemingly make it difficult to argue that
only the female speakers are subject to cultural constraints that lead them to employ
high-pitched voices in order to explicitly express their gendered identity. Contrary
to previous research which found that Japanese men basically do not alter their pitch
across language or context, (Ohara 1992, 1993, 1997, 1999) the findings here, all
based on naturally occurring conversations, demonstrate that the male speakers in
this data did use different levels of voice pitch in constructing their utterances. In

DNa

chigau k a p i o l a n i j a n a k u t ee k a p a h u l u d a tt a y o o

500
450
400
350
300
250
200
150
100
Hz
ms 600 1200 1800 2400 3000

12.7 Pitch extraction of Speaker Ds negation to an acquaintance

 PROSODY AND GENDER IN WORKPLACE INTERACTION 233

DNc

igainandesukedorenzokujaatorenai desunee

500
450
400
350
300
250
200
150
100
Hz
ms 600 1200 1800 2400 3000

12.8 Pitch extraction of Speaker Ds negation to a customer

fact, as discussed, it was sometimes the case, especially in interaction with acquain-
tances, that the pitch levels used by the male speakers were higher than those used
by one of the female speakers, Speaker B. Such results apparently make it impos-
sible to say that there exist underlying social constraints that apply only to the women
and not the men.
Yet before we begin to suggest that Japanese women and men use voice pitch in
whatever way they like and that there are no apparent gender-differentiated cultural
constraints concerning pitch behavior, it should be noted that the pitch extractions
highlight an important point concerning the use of voice pitch in natural interaction.
To demonstrate this point, it is necessary to observe some pitch extractions and ex-
amine at what points rises in pitch occurred. Figure 12.8 offers a good illustration of
this point, and therefore, I discuss it here in further detail. It should first be noted that
in making this negation the male speaker is responding to a customer about hotel
room reservations. Since it is at the point where the word renzoku consecutively is
produced that the highest level of fundamental frequency occurs and since the word
itself is constituted by a large range of voice pitch levels (from 150 to 250 hertz), it
appears that the emphasis of the utterance igai nan desu kedo renzoku ja torenai desu
ne although it is unexpected, it cannot be reserved consecutively is on the term
renzoku consecutively. Thus, what this extraction seems to suggest is that sharp
rises and falls in voice pitch can be used for emphasis; in other words, by raising
voice pitch momentarily, a speaker can add emphasis to that particular part of an
utterance. This phenomenon was also evident in the female speakers speech. For
example, in the following pitch extraction (figure 12.9), which shows a question posed
by Speaker B to a customer, it can be seen that the speaker uses a rise and fall in
pitch to repeat a term that is an especially important aspect of the utterance.
As the pitch extraction shows, Speaker B marks the term pin nanbaa PIN num-
ber by raising its pitch level as well as by repeating it. Prior to this utterance, Speaker
B had been having difficulty communicating with this particular customer. Despite
234 REAL LANGUAGE , REAL PEOPLE

BQc

pinnanbaa tt e i u anoo pinnanbaatteyuubangooarimasu ka

500
450
400
350
300
250
200
150
100
Hz
ms 600 1200 1800 2400 3000

12.9 Pitch extraction of Speaker Bs question to an acquaintance

Speaker Bs attempt to explain how to make a phone call from the hotel room, the
customer had failed to understand. Thus, since use of a telephone card with a PIN
number is crucial to making a direct call, it is understandable that Speaker B empha-
sizes the term pin nanbaa PIN number in this utterance.
A similar phenomenon can be seen in figure 12.6. This extraction, in which the
utterance nihon no hoo kara from the Japanese side receives an extremely high pitch,
is another good example of high pitch as well as sharp rise and fall being used for
emphasis. This utterance was made by Speaker B in response to a customers ques-
tion concerning the availability of a room. This is apparently a difficult task, since
Speaker B in her earlier conversation with the customer had said that a room was
available and in this particular utterance had to negate her earlier comment. The
utterance of the segment nihon no hoo kara from the Japanese side with an ex-
tremely high pitch and a sharp rise and fall, exceeding 450 hertz and with a range
of over 300 hertz, emphasizes that it was a reservation from Japan that caused this
misunderstanding.
The examples presented and discussed in this subsection indicate that at least
part of the pitch variances seen in the data can be explained by recognizing that voice
pitch can be used to emphasize certain aspects of speech. This point is further sup-
ported by the observation of a few instances in the data in which one of the male
speakers, Speaker C, employed an extremely high pitch level that reached around or
over 400 hertz. Figure 12.10 provides an illustration.
In this extraction, in which Speaker C was negating something said to him by
his coworker, he uses a high pitch level, around 400 hertz, three times within this
short segment. In particular, the word kyaku customer starts around 240 hertz and
ends at around 440 hertz, giving it a pitch range of 200 hertz. Similarly, it can be
seen that another part of the utterance, kureru ka yo anna no would they give that
kind of thing, starts around 240 hertz, then goes above 400 HZ, and ends up around
 PROSODY AND GENDER IN WORKPLACE INTERACTION 235

CNa2

kyakuni kureru kayoannanojibun no dayoare

500
450
400
350
300
250
200
150
100
Hz
ms 600 1200 1800 2400 3000

12.10 Pitch extraction of Speaker Cs second negation to an acquaintance

130 hertz, showing a pitch range of close to 300 hertz within this rather short seg-
ment. The other part of this utterance, jibun no da yo are theyre mine, hits 440
hertz with a range of over 300 hertz, almost as high as any utterance observed in this
analysis, regardless of speaker gender. When we consider the fact that these utter-
ances were made in response to a comment from the speakers coworker that accused
Speaker C of illicitly receiving special gifts from guests and trying to sell these items
to her, we can understand how he, in his attempt to strongly deny the accusation,
would want to emphasize certain parts of his negation by making use of the avail-
able resources, including a high-pitched voice. Indeed, this finding that voice pitch
can be used for emphasis is in agreement with previous research on stress patterns
in which the stressed or emphasized segment is found to be marked by high funda-
mental frequency or a rise or sharp fall (OShaughnessy 1987, Sugito, Inukai, &
Sadanobu 1997).

12.5. Discussion

The idea that a high pitch level can be used by both women and men to accomplish
an action such as emphasis is consistent with recent work on gender and language in
Japanese that has stressed womens ability to use a variety of linguistic forms as re-
sources for the construction of interaction and the projection of different gendered
images. Just as this emergent set of research has stressed that the strict boundaries
drawn between the language of women and men is not as rigid as once thought
(Kobayashi 1993; Inoue 1994; Okamoto 1995, 1996, 1997; chapters 1011, 1315
this volume), the point made clear in the analysisthat high-pitched voice is not
something that solely belongs in the predetermined domain known as Japanese
womens languageprovides further evidence for the claim that, because of the
multiplicity of their functions, normatively gendered linguistic forms may serve as
236 REAL LANGUAGE , REAL PEOPLE

resources for speakers in social interaction (Ochs 1992, Trechter 1999). As the re-
sults of the analysis have demonstrated, high-pitched voice is a device that can be
used by both women and men to create a certain effect, namely, emphasis.
The result that men also are shown at times to use a high-pitched voice as a re-
source is especially striking given the relative lack of attention accorded in previous
research on Japanese to the language used by men. Although previous research that
focused on the reading of isolated sentences in a laboratory has tended to describe
the voice pitch levels used by Japanese men as being very narrow in range, the re-
sults of this study suggest that the men in this study, like the women, were dynamic
users of language. In other words, as social actors who closely monitor the details of
social interaction they were very capable of employing a high pitch at strategic points
in the discourse, for example, when they needed to emphasize certain parts of their
utterances.
And just as it is necessary to underscore how great a resource a high-pitched voice
can be in interaction, by the same token, we would have to say the same for a low-
pitched voice as well. Although previous research has said very little about womens
capacity to use lower pitch levels, this study has not only shown that men sometimes
use high-pitched voice, but it has also demonstrated that Japanese women utilize lower-
pitched voices in some situations. In particular, Speaker B used voice pitch levels that
were as low as levels used by the men when speaking to acquaintances.
However, while voice pitch can be seen as a resource available for pragmatic
purposes to speakers of Japaneseboth women and menengaged in everyday
conversational work, my findings do not by any means suggest that there are no
gender-associated patterns of pitch usage. The results clearly indicate that the female
speakers but not the male speakers varied their pitch levels considerably and sys-
tematically across interlocutors. More specifically, the results showed that the female
speakers employed a considerably higher pitch when speaking to customers than when
talking to acquaintances. This finding suggests, on the one hand, that high voice pitch
was being used for the purpose of politeness. This suggestion is in accord with the
observation made by Loveday (1986) that Japanese women are expected to use a
high-pitched voice to make their utterances more polite. On the other hand, espe-
cially given the lack of variation in mens voice pitch according to interlocutor, this
finding also suggests that the cultural constraint mentioned earlier might be at work:
women are expected to use a high pitched voice to express femininity. In fact, we
might say, based on this finding, that these cultural constraints are related not only to
femininity but also to politeness (see Ohara 1999 for further discussion on pitch and
politeness). In other words, even though in some circumstances, such as when talk-
ing to a close acquaintance, Japanese women might be culturally permitted to use a
low-pitched voice, there are certain situations, for example, interactions with a cus-
tomer, in which they are expected to use higher pitch levels to enhance the polite
nature of their linguistic expressions.
To be sure, it needs to be emphasized, as Okamoto (chapter 2, this volume) has
done, that politeness is a very situated phenomenon. Although some early research
has suggested as a general principle that women are expected to be more polite than
men when speaking in Japanese, more recent research has shown there to be many
 PROSODY AND GENDER IN WORKPLACE INTERACTION 237

situations in which this is not the case (Okamoto 1996; chapters 2, 10, and 11, this
volume). Accordingly, I do not want to suggest that women in all workplaces are
expected to speak more politely to customers. Likewise, the results of this study
do not necessarily mean that we should generalize too much and assume that a
high-pitched voice in employeecustomer interaction will always be utilized for
politeness. Only further research can make it possible to understand how deep the
relationship between voice pitch and politeness runs. Nonetheless, the results of this
study, especially the fact that the women used higher-pitched voices when speaking
to customers while men did not, would strongly suggest that in this particular situa-
tiondealing with customers who were apparently very important to a particular
businesswomen and not men face constraints such that they are expected to use a
high-pitched voice to express politeness.

12.6. Conclusion

Using a phonetic analysis of naturalistic data, this chapter has paid tribute, in many
ways, to the dynamic quality of voice pitch as it is actually used in social interaction.
Like other aspects of language, it is a resource that is used by speakers, both women
and men, to create specific effects and project certain images that are crucial to the
construction and accomplishment of social interaction. At the same time, however,
by focusing specifically on how voice pitch is connected to both gender and culture
this chapter has supported the claim in previous research that women in Japan face
cultural constraints that lead them to raise the pitch of their voice in order to project
a feminine image. In particular, the results of the analysis in this chapter suggest that
the link between voice pitch and femininity is mediated by a third concept, namely,
politeness.2 As the results showed, it was the female speakers and not the male speak-
ers who consistently used a higher-pitched voice when speaking to customers than
when speaking with acquaintances.
Because of my desire both to focus on speech activities actually performed by
speakers in social interaction and to present pitch extractions of all the utterances
examined, I was only able to examine in this chapter the speech of four speakers
as they produced two speech acts. Still, the study presented here represents an im-
provement on previous laboratory results. By attempting to situate voice pitch within
actual interactional encounters I was able to observe how some prosodic aspects
were used by real people in real contexts. Moreover, by paying attention to con-
text and observing how the employees used voice pitch to different interlocutors,
customers, and acquaintances, I was able to describe the complex relationship that
exists among gender, voice pitch, and the identities of participants in interaction.
Especially when we attempt to understand how an aspect of language such as
prosody interacts with gender in a specific cultural environment, the context of the
interaction cannot be ignored. In addition, the methodological adjustments made
in this study may facilitate our understanding of the speech not only of women but
also that of men. Further analyses in a similar vein will make it possible to gain
insight into the ways that both women and men use various pitch levels in various
238 REAL LANGUAGE , REAL PEOPLE

interactions. In the future, by continuing to examine naturally produced speech and

considering social, cultural, and textual contexts it will be possible to gain a much
deeper appreciation of what voice pitch means to human communication, includ-
ing both how it serves people as a resource and how its production is subject to
cultural constraints.

Notes
1. Although I call this category request, the actual utterances seem to range from what
might be considered a request to an order. For instance, an utterance such as hai chotto matte
kudasai yes, please wait for a second that is produced to a customer can be classified as a
request without much dispute while a similar utterance to an acquaintance, chotto mate chotto
mate wait for a little; wait for a little, can be said to be more of an order.
2. While I do not have space in this chapter to discuss politeness in more detail, it should
be noted that the relationship among language, gender, and politeness has been a topic of
extensive study (see, e.g., Lakoff 1975, Smith-Hefner 1988, Ide 1990, Holmes 1995).

References
Holmes, Janet (1995). Women, men, and politeness. London: Longman.
Ide, Sachiko (1979). Onna no kotoba, otoko no kotoba (Womens language, mens language).
Tokyo: Nihon Keizai Tsushinsha.
(1990) How and why do women speak more politely in Japanese. In S. Ide and N. H.
McGloin (eds.), Aspects of Japanese womens language, 6379. Tokyo: Kuroshio.
Inoue, Miyako (1994). Gender and linguistic modernization: Historicizing Japanese womens
language. In M. Bucholtz, A. C. Liang, L. Sutton, and C. Hines, (eds.), Cultural perfor-
mances: Proceedings of the Third Berkeley Women and Language Conference, 322343.
Berkeley: Berkeley Women and Language Group.
Jugaku, Akiko (1979). Nihongo to onna (The Japanese language and women). Tokyo: Iwanami
Shoten.
Kobayashi, Mieko (1993). Sedai to joseego: Wakai sedai no kotoba no chuuseeka ni tsuite
(Generation and womens language: On neutralization on the speech of the young
generation). Nihongogaku 12: 181192.
Lakoff, Robin (1975). Language and womans place. New York: Harper and Row.
Loveday, Leo (1986). Explorations in Japanese sociolinguistics. Amsterdam: John Benjamins.
Matsumoto, Yoshiko (1996). Does less feminine speech in Japanese mean less femininity?
In N. Warner, J. Ahlers, L. Bilmes, M. Oliver, S. Wertheim, and M. Chen (eds.), Gen-
der and belief systems: Proceedings of the Fourth Berkeley Women and Language Con-
ference, 455467. Berkeley: Berkeley Women and Language Group.
Ochs, Elinor (1992). Indexing gender. In A. Dutanti and C. Goodwin (eds.), Rethinking con-
text: Language as an interactive phenomenon, 344358. Cambridge, MA: Cambridge
University Press.
Ohara, Yumiko (1992). Gender-dependent pitch levels: A comparative study in Japanese and
English. In K. Hall, M. Bucholtz, and B. Moonwomon (eds.), Locating power: Proceed-
ings of the Second Berkeley Women and Language Conference, 469477. Berkeley:
Berkeley Women and Language Group.
(1993). Koe no takasa kara ukeru inshoo ni tsuite (Images of voice pitch). Kotoba
14: 1419.
 PROSODY AND GENDER IN WORKPLACE INTERACTION 239

(1997). Shakaionseegaku no kanten kara mita nihonjin no koe no takasa (Japanese

pitch from a sociophonetic perspective). In S. Ide (ed.), Sekai no joseego (Womens lan-
guages in the world), 4258. Tokyo: Meiji Shoin.
(1999). Performing gender through voice pitch: A cross-cultural analysis of Japa-
nese and American English. In U. Pasero and F. Braun (eds.), Performing and perceiv-
ing gender, 105116. Opladen: Westdeutscher Verlag.
Okamoto, Shigeko (1995). Tasteless Japanese: Less feminine speech among young Japa-
nese women. In K. Hall and M. Bucholtz (eds.), Gender articulated: Language and the
socially constructed self, 297325. New York: Routledge.
(1996). Representation of diverse female speech styles in Japanese popular culture.
In N. Warner, J. Ahlers, L. Bilmes, M. Oliver, S. Wertheim, and M. Chen (eds.), Gen-
der and belief systems: Proceedings of the Fourth Berkeley Women and Language Con-
ference, 575588. Berkeley: Berkeley Women and Language Group.
(1997). Social context, linguistic ideology, and indexical expressions in Japanese.
Journal of Pragmatics 28: 795817.
OShaughnessy, Douglas (1987). Speech communication: Human and machine. Reading, MA:
Addison-Wesley.
Shibamoto, Janet S. (1985). Japanese womens language. New York FL: Academic Press.
Shibatani, Masayoshi (1990). The languages of Japan. Cambridge, MA: Cambridge Univer-
sity Press.
Smith-Hefner, Nancy (1988). Women and politeness: The Javanese example. Language in
Society 17(4): 535554.
Sugito, Miyoko, Takashi Inukai, and Toshiyuki Sadanobu (1997). Bun no koozoo to purosodi
(Sentence structure and prosody). In Onsee bunpoo kenkyuukai (Spoken Language Working
Group) (ed.), Bumpoo to onsee (Grammar and prosody), 320. Tokyo: Kuroshio.
Trechter, Sara. (1999). Contexualizing the exotic few: Gender dichotomies in Lakhota. In
M. Bucholtz, A. C. Liang, and L. A. Sutton (eds.), Reinventing identities: The gendered
self in discourse, 10122. Oxford: Oxford University Press.
van Bezooijen, Rene (1995). Sociocultural aspects of pitch differences between Japanese
and Dutch women. Language and Speech 38(3): 253266.
(1996). The effect of pitch on the attribution of gender-related personality traits. In
N. Warner, J. Ahlers, L. Bilmes, M. Oliver, S. Wertheim, and M. Chen (eds.), Gender
and belief systems: Proceedings of the Fourth Berkeley Women and Language Confer-
ence, 755766. Berkeley: Berkeley Women and Language Group.
240 REAL LANGUAGE , REAL PEOPLE

13

YOSHIKO MATSUMOTO

Alternative Femininity
Personae of Middle-aged Mothers

T he origins of the common perceptions of femininity and of female roles within

Japanese society have been illuminated in sociological and anthropological studies
of Japanese women. For example, Lebra, describing the traditional socialization and
discipline of young people in Japan, discusses an aspect of discipline that she labels
femininity training: training in manners and comportment according to the cultur-
ally standardized codes of propriety that were imposed upon young women (1984:42).
According to her description, which is largely based on the prewar and wartime ex-
perience of women, the three dimensions of femininity training were modesty,
elegance, and tidiness. Women were trained to efface themselves in behavior
that included posture, movement, and speech, and to portray elegance and tidiness
in their handling of objects. With regard to speech according to Lebra, the prescrip-
tion of modesty demanded overall reticence, a soft voice, the polite and feminine
style of speech, and avoidance of exposing the oral cavity (1984:42).
There have been numerous studies with regard to the feminine style of speech,
as discussed also in other chapters of this book (e.g., chapters 1 and 2). Such speech
is purportedly manifested morphologically in the choice of sentence-final expres-
sions, referential terms, and honorifics; these forms as used by women are seen as
indicating softness, nonassertiveness, and politeness (e.g., Ide 1982, Shibamoto
1985, Mizutani & Mizutani 1987, Reynolds 1990, Shibatani 1990, Horii 1993).
One common point in these previous considerations of gender indexing in Japa-
nese is a focus on the conventional and stereotypical association between certain
linguistic forms and the gender identity of the speaker. Specifically, such studies
examined which forms are used in womens language and why, in light of Japa-

240
 ALTERNATIVE FEMININITY 241

nese cultural and historical background, they are regarded as characteristic of

women.
The aspects of femininity that Lebra points out as the goals of femininity train-
ing evoke an image of traditional womanhood that is most associated with the Con-
fucian ideology of ryoosai kenbo good wife, wise mother. The concept of ryoosai
kenbo has been described as an amalgamation of Confucian teaching on womanhood
and the Western cult of domesticity, which developed in the mid-nineteenth to early
twentieth centuries (e.g., Tamanoi 1990, Uno 1993). These ideologies assign to
women the duties of subservience to men and of nurturance of family.
However, researchers have also noted signs of womens resistance to the domi-
nant ideology (e.g., Tamanoi 1990; chapters 14, this volume), although such signs
should be understood as symptomatic of incremental changes rather than as the start
of an attitudinal revolution. Yet while such resistance may be incremental and frag-
ile, it can gradually affect society, as noted by Rosenberger (1996). Other studies
have shown that Japanese women in the last few decades have been given more op-
portunities to join the labor force (Fujimura-Fanselow & Kameda 1995) and have
the potential to increase their status in the society (Iwao 1993). These studies sug-
gest that women have a chance to choose not to adhere to the concept of femininity
that has been the target of socialization training as described by Lebra.
Against the background of this social context, I discuss the language use of Japa-
nese middle-aged women, especially mothers who do not work outside their homes.
Among women, such stay-at-home mothers would appear to play the most traditional
role and therefore could be expected to reflect more than others a traditional concept
of femininity. In spite of their central position in forming the image of traditional
femininity, the language of these middle-aged housewives and mothers has not been
studied as much as that of professional women (e.g., [Shibamoto] Smith 1992). In
this chapter, I analyze the real voices of such women, discuss the female personae
conveyed in their speech, and consider whether and how these personae differ from
(or are similar to) the stereotypical construction of femininity.
The examples discussed in this chapter are drawn from recordings collected
during 19961997 that contain natural conversations of speakers, especially middle-
aged (35-to-60-year-old) women. The analysis focuses particularly on one Japa-
nese woman in her forties who is a mother of three teenagers. The data analyzed
here is drawn from (1) a 120-minute recording of conversations of this woman with
her fellow Parent Teacher Association committee members and (2) a 90-minute
recording of conversations of the same speaker with fellow members of a womens
amateur chanson group.
An analysis of the pragmatic and social meanings of the linguistic forms that
the speaker and her interlocutors chose demonstrates that, contrary to the common
and traditional assumption that middle-class middle-aged womens (especially moth-
ers) speech is stable, gentle, and normatively feminine, expressions that are diverse
in force and style are widely used throughout their conversations. This nonuniformity
in style is observed even within an individuals utterances in the same speech situa-
tion. I suggest, based on this analysis, that the female personae that these middle-
aged speakers portray diverge from the traditional norms of femininity that were
inculcated in the femininity training described in Lebra (1984). As a result of such
242 REAL LANGUAGE , REAL PEOPLE

divergences, the femininity of Japanese women is seen to be not singular but mul-
tiple, encompassing alternative female personae that women can represent. The di-
vergences also highlight the interaction between individual agentive choice and social
constraints in constructing ones persona. (See also chapter 12, this volume.)
Two concepts useful in the following discussion are those of forceful and deli-
cate stance. These stances can be conveyed linguistically or extralinguistically. Studies
in linguistic pragmatics have shown that the same assertion, question, order, and so
on, can be expressed forcefully or delicately depending on the speakers intention or
stance. In relation to the gender of the speaker, however, as exemplified by the femi-
ninity training mentioned earlier, women in Japan have been normatively expected
to adopt a delicate stance, while men have been expected to be forceful. For those
reasons, linguistic forms that pragmatically convey forceful stance, such as the sen-
tence-final particle zo, which conveys ones determination and force, and the copula
da, which conveys definiteness, have stereotypically been regarded as mens expres-
sions, while expressions that convey a delicate stance, such as the sentence-final
particle wa, which expresses a mild self-confirmation, and the use of an abbreviated
request rather than the imperative (e.g., yamete . . . please stop rather than the im-
perative yamero stop!), have been labeled as womens expressions. In the follow-
ing, I consider in more detail how these stances figure in conversations among
middle-aged mothers.

13.1. Mixture of styles: The speech

of middle-aged women

Example (1) is an excerpt of a 120minute conversation among four women whose

children attended the same public middle school in a middle-to-upper-middle-class
residential neighborhood in Tokyo. The women were members of the local PTA
Public Relations Committee and ranged in age from late thirties to early forties. They
know one another relatively well, but more through their children and the PTA
activities than privately. The main speaker, whom I refer to as Minako (M), was the
oldest (43 years old at the time of the recording) in this group. The other speakers
are referred to as Chie (C), Eri (E), and Fumiko (F).1 The conversation took place in
the course of a car trip to check a park on the outskirts of Tokyo where the school
was planning to hold a sporting event. In the first excerpt, which was recorded while
they were still on city streets, Minako was the driver and was unsure of the correct
route. Single underlines indicate expressions that convey a forceful stance, while
double underlines indicate speech that communicates a delicate stance.2

(1) 1 M: chotto kotchi iketakke

:
2 C: . . . ki no oto ga sugoo(ku)
3 M: a, yabbee, kore migi shika ikenai wa.
4 E: a, honto da
5 M: arararara
6 E: minna ippoo-tsuukoo ni natchatte
 ALTERNATIVE FEMININITY 243

7 M: kotchi hairitakunai n da yo na
8 E: ara, zutto da wa
9 C: koko n toko zutto soo nan desu yo, michi ga
10 E: a soo nan da yo ne
11 M: ura ikitai no yo ne, ura o
1 M: I wonder if we can go this way
:
2 C: . . . the sound of trees was really terri(ble)
3 M: Oh, damn! I can only turn right here.
4 E: Oh, yeah, thats true
5 M: Dear, dear, dear
6 E: They became one-way
7 M: I dont want to go in here
8 E: Oh, all of them are
9 C: Its all like that around here; the roads are
10 E: Yeah, thats right, isnt it?
11 M: I want to take a back road, back road

One notable feature in the conversation in example (1) is that expressions that
convey a forceful stance alternate with expressions that are delicate in force even
within the same setting and topic of speech. The singly and doubly underlined sen-
tence-final expressions and interjections in the example are salient in this regard.
Before we start examining the conversation in example (1) in detail, it is helpful
to first consider elements whose pragmatic effect is to convey forceful or delicate
stances. For example, one of the typical elements that bring about a delicate and at-
tenuated force of speech is the pragmatic particle wa, which is used after a predicate,
as can be seen in lines 3 and 8 in example (1). Wa was used historically as an exclama-
tion, and it expresses as its core meaning the speakers mild self-addressed confirma-
tion about the propositional content of the sentence (Uyeno 1971, 1972; Nihon kokugo
dai-jiten 1978).3 Since it marks an utterance as addressed to oneself, the pragmatic force
of an assertion with wain particular, its claim to the addressees beliefis less in-
tense than that which is conveyed by the absence of a particle or by the presence of
more assertive particles such as yo, ze, and zo. When wa is uttered with a rising or sus-
taining intonation, as in all of the instances in example (1), the combination of intona-
tion contour and core meaning of wa conveys an open-endedness that attenuates the
speakers assertion. For example, in the utterance in line 8, ara, zutto da wa Oh, all of
them are, the sentence-final wa makes the utterance delicate and less intense.4 Simi-
larly, the interjection ararara dear, dear, dear in line 5 and its core form ara oh or
oh, dear in line 8 are relatively delicate or mild expressions of surprise.
In contrast to the delicate expressions discussed earlier, the interjection yabbee
dangerous, in line 3 (an intensified version of a more common expression, yabee),
although it has gained currency among young speakers, originated in gangsters slang
and expresses the strong feeling of the speaker about an unwanted situation. Another
salient element that makes an utterance more forceful is the use of the copula da,
which follows a noun or an adjective, as in lines 4, 7, and 10 of example (1). From a
244 REAL LANGUAGE , REAL PEOPLE

functional-linguistic (and pragmatic) point of view, the plain form copula da conveys
as its modal function the speakers certainty and definiteness toward the propositional
content of the sentence (Teramura 1984, Konomi 1994). When the plain form da is
used, the sentence conveys a stronger sense of assertiveness. For example, the utter-
ance in line 4 honto da thats true is more forceful than its counterpart without da as
in honto true. However, this forcefulness can be overridden when the copula is fol-
lowed and attenuated by the particle wa, as in line 8, zutto da wa all of them are.5
An alternating distribution of forceful and delicate expressions within a single
context as observed in example (1) is interactionally interesting, but it deserves spe-
cial attention in light of the normative and conventional association of the use of
specific forms such as wa and da with the gender identity of the speaker. This con-
ventional association is exemplified in example (2), which is an excerpt from a well-
known textbook of Japanese for nonnative speakers that contrasts the differences
between sentence-ending expressions in the informal or familiar speech of women
and men. In this description, as shown in boldface, the use of the copula da is asso-
ciated with mens speech while the pragmatic particle wa is associated with womens
speech.

(2) In polite or formal speech, there is very little difference between men and women, but
in familiar speech, there are some differences between the two.
Sentence Ending
Men usually say Women usually say [English glosses mine]
yokatta yo, yokatta ne yokatta wa yo, yokatta wa ne (it) was good
iku yo, iku ne iku wa yo, iku wa ne will go
ikanai yo, ikanai ne ikanai wa yo, ikanai wa ne wont go
soo da yo, soo da ne soo yo, soo ne thats right
soo da, soo datta soo da wa, so datta wa (it) is so, (it) was so
ashita da yo, ashita da ne ashita yo, ashita ne (it) is tomorrow
(Mizutani & Mizutani 1977:150151)

The association between the expressions illustrated here and gender identity,
however, as argued by anthropological linguists and others, is not intrinsic to the
language but is culturally mediated. For example, Ochs (1993), in discussing the
indexicality of linguistic expressions, explains the conventional association between
female voice and the Japanese pragmatic particle wa, which directly indexes deli-
cate intensity. As explained by Ochs, this association is rooted in a complex of be-
liefs in Japanese culture about femininity and delicacy of expression. In other words,
an affective disposition of delicacy is part of the preferred image of women, which
both defines female voice and may motivate womens differential use of the form.
For this reason, a delicate linguistic style has been normatively associated with fe-
male speakers and a forceful style with male speakers.
As exemplified in example (1), the extent of usage of forceful expressions by
middle-class mothers raises questions about the functions and indexical meanings
of such forms in addition to, and in contrast with, the well-studied association of
delicate expressions with conventional female identity. It is not particularly plausible
to consider, in the speech context of (1), that the variation in the expressions found
 ALTERNATIVE FEMININITY 245

there is motivated by the speakers desire to indicate their alternating female and male
identities. Further, although several studies and newspaper columns reported in the
1990s the increasingly frequent use by teenage girls and young women of forceful
speech forms that were conventionally regarded as belonging to the realm of mens
language (e.g., Asahi Shimbun 19921994; Uchida 1993; Okamoto 1995; Matsumoto
1996, 1999), it is not quite convincing to postulate that the speakers identities are
simply oscillating between youthful and adult.
I return to example (1) to consider this question by analyzing the stretch of con-
versation given there. I further discuss the issues by examining another excerpt of
the same 120-minute conversation given in example (3) and an excerpt given in (7)
from a conversation of the same main speaker, Minako, with another group of women.
Returning to example (1), Minakos (Ms) use of the slang interjection yabbee
damn (lit., dangerous) in line 3 can be interpreted as a clear indication of her
frustration at not being able to keep the situation under control. On the one hand, her
utterance in line 7 with da (kotchi hairitakunai n da yo na I dont want to go in here)
can be understood similarly, that is, as a forceful expression used to express frustra-
tion. On the other hand, Eris (Es) choice of da in lines 4 (honto da oh, yeah, thats
true) and 10 (soo nan da yo ne thats right, isnt it?) does not seem to indicate such
frustration. On the contrary, both utterances seem to express positive attitudes of
agreement with or support for the other speakers. With the use of da, Eris assertions
are more definitive and convey a more forceful stance toward her expressions of
agreement.
Interestingly, Eri uses an attenuated expression in line 8zutto da wa all of
them areobserving and describing (seemingly to herself) the fact that the streets
they are crossing are all one-way. Eris choice between forceful and attenuated ex-
pressions may be explained according to the purposes of her speech act: forceful
expressions are used when agreeing with others opinions, while attenuated expres-
sions are used to express her own view. We may further explain this by saying that
Eri uses (at least) two strategies to convey her friendliness and thoughtfulness to
others. One strategy is to give strong support to her interlocutor by making straight-
forward and forceful assertions. Another strategy is to show her concern for others
by not pressing her own views and to convey friendliness in a more reserved man-
ner. The first strategy, on the one hand, has been commonly associated with the com-
munication style of men in informal settings and of some young women. The second
strategy, on the other hand, has been associated with adult women as a choice or as
a social constraint on their linguistic behavior (e.g., McGloin 1990, Mizutani &
Mizutani 1987, Okamoto & Sato 1992). Eri seems to use these two strategies of in-
dexing friendship effectively and to successfully create the persona of a supportive,
friendly person who is versed both in recent trends and in tradition. She is appar-
ently well liked among her peers and is seen as a pleasant person to be with.
In comparison to Eri, Minakos use of forceful and delicate expressions is not as
clearly demarcated as Eris. To express her feelings about the unwanted situations
that she has been put in, Minako uses forceful expressions in lines 3 (yabbee damn)
and 7 (hairitakunai n da yo na I dont want to go in here) but more delicate expres-
sions in lines 5 (arararara dear, dear, dear) and 11 (ikitai no yo ne I want to take
a back road). Therefore, we cannot explain Minakos choices of expressions simply
246 REAL LANGUAGE , REAL PEOPLE

by linking them to different speech acts. A more plausible explanation of her choice
is that Minakos speech pattern represents a more complex personaone that is
straightforward but delicate and recognizes the social constraints of being a tradi-
tional adult woman while attempting to avoid the covert stigma (Matsumoto
2002:345, 352) attached to traditional gender images. In this way, she balances and
negotiates between the two (or more) different types of personae that she presents.
What is common between these two mothers, however, is that neither one instanti-
ates the traditional femininity associated with the stereotypical image of Japanese
woman. Both represent (modern) alternatives to such beliefs.
We see more instances of variation in individual speakers choices in example
(3), an excerpt from the last part of the same 120-minute conversation that we saw in
example (1). The broken underline indicates a trendy expression.

(3) 1 M: are o hidari ni iku no?

2 E: un, 412, un
3 F: a! pikunikku rando tte kaite aru
4 E: sono saki o sasetsu
5 F: migi ga sagamiko eki da ne
6 M: oi oi
:
7 M: nani ka o-mise mo roku ni aitemasen koto yo (laughter)
8 E: (laughter)
9 F: nee
10 M: nanka shin to hora atchi-kotchi shimatteru yo. hora mada (laughter)
11 koohii nomitakute mo nomenai ja nai desu ka, kore ja.
12 E, F (laughter)
13 M: nani ka attara nomoo ne, toriaezu, sagami p pikunikku rando ni hairu mae
ni
14 E: mada mada
15 F: mada 10-ji dakara aite nai ne
16 E: aite nai ne. makudonarudo mo nai shi tomerareru toko mo nai
17 M: nai yo
18 aa waratchau yo ne
19 E: resuto hausu atte, yasumu toko nai no kashira. komatchau wa ne.
20 M: demo yokatta ne, hayaku tsuite ne.
21 F: nee. suiteta ne.
22 M: suiteta suiteta.
23 F: suitetara densha yori mo zutto hayai n da ne.
24 M: hayai wa yoo.
1 M: Do we go left over there?
2 E: Uh-huh, 412, uh-huh
3 F: Ah! It says Picnic Land.
4 E: Make a left turn after that
5 F: Lake Sagami Station is on our right, isnt it?
6 M: Hey, hey!
:
 ALTERNATIVE FEMININITY 247

7 M: It seems that the stores are hardly open. (laughter)

8 E: (laughter)
9 F: Yeah, right
10 M: Its kind of quiet. Look, stores are closed here and there. Look, theyre not
yet open.
(laughter)
11 M: Even if we want to, we cant have coffee, can we? If its like this
12 E, F : (laughter)
13 M: Lets have coffee, if theres any, one way or the other, before we get into
Sagami p-, Picnic Land
14 E: Theyre not open yet, not yet
15 F: Since its still 10 oclock, they arent open, are they?
16 E: They arent open, are they?. There isnt even a McDonalds, and theres no
place to stop, either.
17 M: No.
18 Oh, this is something to laugh at, isnt it?
19 E: Theres a rest area; I wonder if theres any place to rest.
I dont know what to do.
20 M: But its good that we got here quickly, isnt it?
21 F: Yeah; the traffic was smooth, wasnt it?
22 M: It was; it was.
23 F: When the traffic is smooth, driving a car is much faster than a train, isnt
it?
24 M: It is faster.

In this stretch of conversation, which was recorded close to their destination, the
women were looking for a place to rest and have something to drink before their
expedition at the park, Lake Sagami Picnic Land. As before, each speaker uses dif-
ferent linguistic strategies to convey stance and persona. Table 13.1 shows a com-
parison of the strategies used in this excerpt across the three speakers, Eri, Fumiko
and Minako.
Eri, who is looking at a map and navigating, uses the same pattern of alternating
expressions that we saw in example (1): when she agrees with someone, she uses
more forceful expressions, as in line 16, aite nai ne (they arent open, are they?), a
repetition of what Fumiko (Speaker F) has just said, while less forceful expressions
are chosen when she expresses her own thoughts and concern, as in line 19, yasumu
toko nai no kashira. komatchau wa ne (I wonder if theres any place to rest. I dont
know what to do).

13.1 Types of expressions used by speakers

Eri Fumiko Minako

Forceful expressions 1 5 6
Delicate expressions 1 0 2
Trendy expressions 0 0 1
248 REAL LANGUAGE , REAL PEOPLE

Fumiko, however, uses no attenuated expressions regardless of the content of

her utterances but most often compensates by adding the pragmatic sentence-final
particle ne(e), which gently invites the addressees involvement. The persona that
these expressions project may vary depending on the social background of the inter-
preter, but they generally give an image of someone who is wholesome and straight-
forward.
Minakos utterances again display a complex distribution of forceful and deli-
cate expressions. Minakos utterances from lines 6 to 13 are of particular interest. Oi
oi hey, hey! in line 6 is a more forceful exclamatory attention getter than, for ex-
ample, chotto (lit., a little) or ara (oh or oh, dear), which can be directed to
oneself as well as to others. After a pause, Minako introduces in line 7 the new topic
that the shops are hardly open, which is most likely to have prompted the exclama-
tion oi oi in the previous line. She repeats this observation in line 10, using expres-
sions that indicate a different attitude. The linguistic form in line 10 shimatteru yo
the stores are closed lacks the particle wa between the predicate and the assertive
particle yo, and is therefore relatively forceful. In contrast, the main predicate
aitemasen are hardly open (rather than aitenai) in line 7 is in what is commonly
termed the polite form, unlike the majority of the utterances in this recording. What
follows the polite-form predicate is the formal noun, or nominalizer, koto (lit., fact),
which functions to make the content of the preceding clause an established fact and
has a distancing effect. In this sense, the emotional impact of the statement is lessened,
and therefore its affective force is attenuated. Consequently, this form [predicate +
koto + yo] can fit the traditional cultural norm of femininity and is known as one of
the stereotypical old-fashioned forms of upper-middle-class female speech. It is likely
that the speaker is aware of the connotations of the form and is using it ironically for
humorous effect, as indicated by the laughter that follows the utterance and the overly
clear pronunciation of the words.
Another contrastive set of expressions may be observed in lines 11 and 13. The
contents of the two utterances are not exactly identical, but both concern the possi-
bility of having coffee. The example in line 13, nomoo ne lets have coffee, indi-
cates straightforward informal friendliness. The phrase used in line 11, ja nai desu
ka can we?: literally, isnt it the case that . . . ? with a polite copula desu has re-
cently caught the attention both of linguists and of speakers of Japanese (McGloin
1999, Asahi Shimbun 19921994) and is generally considered a trendy expression
typically used by young people, probably as a polite expression of involvement. The
use of such a trendy youthful expression also forms an interesting contrast with the
old-fashioned expression aitemasen koto yo in line 7.
Except for this ironically used old-fashioned expression, which brings about a
comical effect, none of the expressions used by the three speakers seems designed to
create a special effect as an unexpected expression, although in their own ways all of
the speakers use expressions that are different from those assumed to be used by
middle-class mothers. Minako tends to take the lead in conversations more than others,
but all the participants produce repetitions and agreements that indicate cooperation,
which is generally associated in the literature with female speech styles (Coates 1996).
The activities that the speakers are engaged indriving to a place as PTA members,
looking for a place to rest, and talking about these activitiesdo not seem peculiar
 ALTERNATIVE FEMININITY 249

in comparison to what mothers may normally do. We can thus conclude that the
examples give us a glimpsea slice of reality, as it wereof how middle-aged
women in Tokyo talk among themselves, in contrast to the common cultural images
of their speech.6
Minako, the main speaker in examples (1) and (3), has other groups of female
friends, including a group of amateur chanson singers who are alumnae of a univer-
sity chanson club and who perform concerts for themselves and with other similar
alumnae /alumni groups. Where in the earlier conversation Minako, at age 43, was
the oldest participant, here she is the youngest. Minako is one of the most prominent
members of the group. Example (4) is an excerpt from a conversation recorded in a
gathering of the group after a concert, in which Minako sang a solo. O (61 years old),
P (52 years old), and Q (exact age unknown) are among the other women in the con-
versation. In this conversation we can observe a different choice of copula forms by
Minako than in the earlier example, as well as contrastive uses of sentence-final forms
between speakers O and P. The distancing copula form desu, rather than the more
familiar form da, is indicated with wavy lines.

(4) 1 O: Okaasama suteki datta deshoo tte ittara, anoo.

2 P: soo itteta yo
3 O: soo ittemashita yo, soo itte irashita wa ee. suteki datta tte nee.
4 Q: aa soo.
5 M: demo chotto utai-kata ga chigatta ne tte iwarechatta
6 O: aa soo
7 P: waaa
8 Q: heee
9 R: hontoo
10 O: yokuu ja kiite ite kekkoo ne kiiterashita no nee
11 P: yoku kiiteru n daa
12 M: kashi machigaeru na tte iwareru n desu yo, oboeteru n desu yo, kekkoo
13 P: sugoi nee

1 O: When I said, Your mother was fabulous, wasnt she? umm.

2 P: They said so.
3 O: They said so, they said (honorific) so, yes. It was fabulous, right.
4 Q: Is that so.
5 M: But they [my children] told me [lit., I was told] that the way I sang was a
bit different
6 O: Is that so?
7 P: Wow.
8 Q: Gee.
9 R: Really?
10 O: Well then they were listening fairly well, they were listening (honorific).
11 P: They were listening well
12 M: Dont make mistakes in the lyrics, they [my children] told me. They
remember the lyrics fairly well.
13 P: Im amazed.
250 REAL LANGUAGE , REAL PEOPLE

O and P have been friends with Minako for some time, but Minako notably uses the
more polite and distancing copula desu in line 12. This may be on account of the age
difference between M and the other participants and Ms following or being con-
strained by the social expectations of linguistic behavior toward older individuals,
especially O, who uses honorific forms, even in this informal context when talking
to younger interlocutors.7
Another interesting point to note in this conversation is the contrast between Os
and Ps speech. O and P are both friendly to Minako, but their choices of expressions
are divergent. Although both O and P are older (61 and 52, respectively), P uses more
straightforward expressions that show camaraderie, similarly to E and F in earlier ex-
amples, while O uses more delicate expressions and honorific forms as well as the
prosody characteristic of traditional womens speech, thereby creating the image of a
stereotypical upper-middle-class female speaker. It is noteworthy that O uses such
delicate and honorific expressions very consciously. In both lines 3 and 10, she first
uses polite but nonhonorific form of expressions, ittemashita they said and kiite ite
were listening to the utterance but repairs (repeating the same content while revising
the form) to be more refined with the use of attenuated forms and honorifics. Interest-
ingly, O, the speaker of this traditional style commonly associated with (upper-)middle-
class housewives and not with working women, is the only participant who also works
outside the home, albeit part-time: she is a home economics teacher at a private high
school for girls. This fact confirms that style choice and job status do not directly cor-
relate. Os deliberate use of a traditional style suggests a few things: She may be trying
to prevent others from seeing her as less than middle-classsomeone who works out
of financial necessity. She may be trying to maintain the ideal style that she believes
to be expected of someone of her background. As the only participant who was ex-
posed to the prewar value system and education, she may still uphold in principle the
traditional good wife and wise mother ideology, the aforementioned femininity train-
ing, and their associated linguistic behavior. Alternatively, O, who is also the head of
the group, may have felt that, in order to successfully convey her compliments to
Minako, she could indicate her respect by using honorifics to describe Minakos
childrens behavior. These do not exhaust all possible explanations. The excerpt in
example (4) is therefore useful to illustrate again that women who share a similar back-
ground (e.g., age, class, education, and hobbies) do not necessarily choose the same
strategies for presenting themselves.

13.2. Conclusion: Alternative femininity and

the presentation of a complex persona

We have observed multiple functions of linguistic expressions and the variability of

female personae that are represented in the speech of middle-aged middle-class Japa-
nese women, focusing primarily on mothers who do not work outside the home. In
the data examined here, expressions of a forceful stance were used sometimes to
display frustration but at other times to accommodate others views and to be friendly.
Expressions of a weaker and delicate stance were used to present the speakers views
in a friendly and deferential style and also perhaps to convey the speakers delicacy.
 ALTERNATIVE FEMININITY 251

Because forceful expressions have been associated with youth (both female and male)
and delicate expressions with traditional women in Japan, it is likely that there were
also times when the mothers choices of expressions were motivated by the desire to
portray themselves as current and anticonformist, or traditional and classy (see
chapter 10, this volume, on class associations of Japanese womens language).
What these observations imply is that statements that present similar proposi-
tional, or semantic, content can be expressed with a variety of speech styles and forces,
even in the same speech situation, to the same addressee, and sometimes by the same
speaker. The only speaker in our data who could be characterized as upholding the
traditional image of femininity in her language use and in her speech was the 61-
year-old speaker O in example (4).
I should add in passing that the alternation between forceful and delicate stances
is not necessarily limited to middle-aged mothers or other female speakers. Male
Tokyo speakers have been known to use expressions perceived to be delicate such
as those given in the column labeled as Women usually say in example (2). For
example, I have observed the nonuse of the copula da, as in kore niku yo this is meat
and soo yo thats right (but not the combination of particles wa yo or nonfalling
intonation wa or wa ne).8 Further study is necessary to delineate the distribution of
such expressions and to discuss their implications, but it is important to note that the
range of expressions that are in fact used by both female and male speakers is broader
than in normative accounts.
The linguistically exhibited variability of female personae seems to generally
reflect the societal situation since the 1980s and 1990s when, as briefly mentioned
earlier, women in Japan were afforded more choices for their careers and lives. It is
not surprising that an expansion in ones social (and career) domain should have as
a corollary an expanded selection of available personae. What is particularly inter-
esting here, however, is that most of the women in my data are mothers without even
part-time jobs. Various interpretations can be given as to why mothers without out-
side careers should exploit linguistic choices to represent complex personae that dif-
fer from the traditional presentation of femininity. One possibility is that the traditional
cultural belief about femininity is an ideological construct not reflected in reality.
Another is that the social changes in recent years have affected even the paragons of
the nurturing role modelmothers. Perhaps to some extent both explanations are
true. Sociological studies on urban housewives show that they are becoming more
regionally active with increased social participation outside the home through part-
time jobs and volunteer activities (e.g., Imamura 1987, 1996; Fujimura-Fanselow &
Kameda 1995). Such changes may have enabled the construction of the female per-
sonae that are seen linguistically in this study. At the same time, speakers have always
been endowed with the ability to choose and construct their own personae based on
aspects of their lives, including their past and current experiences that they wish to
foreground, as well as on their awareness of and responses to cultural beliefs about
femininity. As Eckert and McConnell-Ginet (1992) have argued, language use is a
showcase of speakers individual histories of previous experiences and engagements
in various groups (or communities of practice) whose linguistic influences persist in
their engagement in current activities and in the negotiation of their individual
presentations of self.
252 REAL LANGUAGE , REAL PEOPLE

Even within the limited selection of conversational samples that I examined here,
the utterances of Minako exhibit a notable variety of expressions and stances. This
variety may be explained by the way she has lived her life. Minakos past and cur-
rent activities cut across several areas. She is active in the PTA at her sons school as
well as in an amateur (or even half-professional) singing group. In college, she pur-
sued a number of modern and traditional artistic as well as athletic interests. She
married an athlete and educator and became the mother of one daughter and two sons,
who are now in their teens. Minakos relatively broad current and past social net-
workalthough limited to a middle-class and mainly nonworking environment
seems congruent with her broad range of linguistic expressions.
The seemingly ambivalent style that embraces a range of stances provides the
speakers in these examples with the opportunity to present themselves in a complex
yet flexible fashion and allows them to negotiate their relationships with addressees
in more varied ways. According to my data and previous studies, such a style is not
prominent in younger womens speech9 but may be a characteristic of at least some
middle-aged speakers. It is not surprising that such a style is found among middle-
aged women, who are old enough to have had a variety of experiences and social
roles but young enough to feel affinity with the directness fostered in youth.
The examples cited in this chapter suggest that variations in linguistic style are
part of the speakers own linguistic repertoire rather than signaling the use of some-
one elses voice, such as a male voice. When, Minako, for example, switches from
one form to another, although the switch has a pragmatic and social effect, most of
the time it does not seem to invite some sort of Gricean implicature (1975), which
would result from a contextually unexpected choice of expressions.10 The alterna-
tion among types of forms may be viewed as presenting a broad but coherent image
of an individual speaker.
It should be noted here that not all middle-aged women draw from as wide a
stylistic palette as Minako and Eri in the preceding examples. For example, Fumiko
(ex. 3) and O and P (ex. 4) exhibit a consistent preference for one stance over an-
other. Depending on their social backgrounds and interactional goals, some speak-
ers may use predominantly delicate and attenuated expressions, some may favor more
forceful expressions, some may choose to use conventionally gendered expressions,
and some may employ a mixture of these. The examples presented here highlight
this linguistic individuality among speakers who could all too often be categorized
in terms of a single gender, age, class, or regional grouping.
I do not expect that the range of expressions shown in this chapter would in any
way exhaust the possible personae that women can choose to present. It is conceiv-
able that women of different age-groups and social backgrounds would exhibit other
expressions and stances and therefore other types of femininity not discussed here.11
For example, I considered in Matsumoto (1996) a cute, infantile feminine persona
associated with younger married women as another alternative to the traditional Japa-
nese expression of femininity. Echoing that earlier work, this study illustrates the
multiplicity and complexity of female personae that language can represent. These
works argue for the acknowledgment of the varieties of femininity at the outset of
language and gender studies rather than consider femininity as a singular concept.
While speakers recognize the cultural and social expectations and the constraints
 ALTERNATIVE FEMININITY 253

regarding their behavior, linguistic and otherwise, they still have room to negotiate
among these conditions to represent themselves in multidimensional ways. Acknowl-
edgment of the varieties of femininity and recognition of agentivity within the chang-
ing and not-very-clearly demarcated boundary of cultural and gender constraints
should be a touchstone for our further studies in language and gender ideology and
practice.

Notes
I would like to express my gratitude to the editors of this volume, Janet Shibamoto Smith and
Shigeko Okamoto, and to Mary Bucholtz, the series editor, for their useful comments and
suggestions. I am also indebted to the speakers who provided my data, especially Minako,
who kindly recorded her conversations. This chapter is a substantially revised version of a
paper that I presented at the First International Gender and Language Conference in 2000
(Matsumoto 2002).

1. All the participants names in this and other examples are pseudonyms.
2. Expressions that are labeled as forceful expressions and delicate expressions here
were referred to as so-called nonfeminine forms and feminine forms (respectively) in my
earlier paper (Matsumoto 1999) in accordance with the terms conventionally used in some of
the recent studies in the area. I use more pragmatically oriented terms in this chapter because
the focus of the discussion here is less to point out the discrepancy between the reality and
the normative ideology of womens speech (as it was in my earlier work) and more to under-
stand how women use linguistic devices to express a persona that transcends traditional con-
ceptions of femininity.
3. As a consequence, wa cannot be used with a sentence that conveys meanings incom-
patible with the concept of confirmation, such as a sentence that conveys a question. The sen-
tence ii wa (good/OK Sentence-Final Particle) cannot mean Is it OK? but means Its OK,
and *ii ka wa (good/OK Question Particle Sentence-Final Particle) is ungrammatical.
4. My position here is that the same core pragmatic meaning is conveyed regardless of
the intonation with which the sentence with wa is uttered, although I recognize that open-
endedness is more clearly conveyed with a rising intonation and the combination of rising
intonation with wa makes the utterance even less intense in force. There is a controversy over
the function of intonation contour in relation to the meaning of wa, in particular as to which
intonation contour indicates stronger emotion (Kitagawa 1977, McGloin 1990). My analysis
here is based more on the core meanings of the particle proposed by Uyeno (1971) and Nihon
kokugo daijiten (1978).
5. The particles yo, na, and ne, which are found in lines 7 and 10 (hairitakunai n da yo
na I dont want to go in and soo nan da yo ne thats right), express assertiveness and in-
volvement rather than delicacy.
6. A similar variety of forms is also found among middle-class middle-aged career women
(see Matsumoto 1996), but I do not go into that issue in this chapter, as it is not the main
concern here.
7. For further discussions and implications of mixed use of the masu form and plain
forms, see Cook (1998). Minako here can be described as more aware of her self-presenta-
tion and addressee deference than in her other utterances earlier (Cook 1998:87).
8. Speakers of such forms of delicate stance can be perceived by speakers of non-Tokyo
dialects as effeminate or as users of womens language (S. Iwasaki, S. Okamoto, personal
communication).
254 REAL LANGUAGE , REAL PEOPLE

9. This does not mean that younger womens speech totally lacks delicate expressions.
In the data that I observed, such expressions are found in younger womens speech but are
not as prevalent as in middle-aged womens speech.
10. I call this outcome an interactional implicature elsewhere (Matsumoto 1988).
11. See also other chapters in part II of this volume.

References
Asahi Shimbun (19924). Ima tookyoo-go wa (Tokyo-language now). Special weekly columns.
Coates, Jennifer (1996). Women talk: Conversation between women women friends. Oxford:
Blackwell.
Cook, Haruko M. (1998). Situational meanings of the Japanese social deixis: The mixed use
of the masu and plain forms. Journal of Linguistic Anthropology 8(1): 87110.
Eckert, Penelope, and Sally McConnell-Ginet (1992). Communities of practice: Where lan-
guage, gender, and power all live. In K. Hall, M. Bucholtz, and B. Moonwomon (eds.),
Locating power: Proceedings of the Second Berkeley Women and Language Conference,
8999. Berkeley: Berkeley Women and Language Group.
Fujimura-Fanselow, Kumiko, and Atsuko Kameda (eds.) (1995). Japanese women: New femi-
nist perspectives on the past, present, and future. New York: Feminist Press.
Grice, H. Paul (1975). Logic and conversation. In P. Cole and J. L. Morgan (eds.), Syntax
and semantics 3: Speech acts, 4158. New York: Academic Press.
Horii, Reiichi (1993). Joseego no seeritsu (Establishment of womens language). Nihongogaku
12(5): 100108.
Ide, Sachiko (1982). Japanese sociolinguistics: Politeness and womens language. Lingua 57:
357385.
Imamura, Anne E. (1987). Urban Japanese housewives. Honolulu: University of Hawaii Press.
. (ed.) (1996). Re-imaging Japanese women. Berkeley: University of California Press.
Iwao, Sumiko (1993). The Japanese woman: Traditional image and changing reality. New
York: Free Press.
Kitagawa, Chisato (1977). A source of femininity in Japanese: In defense of Robin Lakoffs
Language and womens place. Papers in Linguistics 10: 275298.
Konomi, Emiko (1994). The structure of the nominal predication in Japanese. Unpublished
Ph.D. dissertation, Cornell University.
Lebra, Takie Sugiyama (1984). Japanese women: Constraint and fulfillment. Honolulu:
University of Hawaii Press.
Matsumoto, Yoshiko (1988). Reexamination of the universality of face: Politeness phenom-
ena in Japanese. Journal of Pragmatics 12: 403426.
(1996). Does less feminine speech in Japanese mean less femininity? In N. Warner,
J. Ahlers, L. Bilmes, M. Oliver, S. Wertheim, and M. Chen (eds.), Gender and belief
systems: Proceedings of the Fourth Berkeley Women and Language Conference, 455
467. Berkeley: Berkeley Women and Language Group.
(1999). Japanese stylistic choices and ideologies across generations. In J. Verschueren
(ed.), Selected papers from the 6th International Pragmatics Conference, vol. 1: Lan-
guage and ideology, 352364. Antwerp: International Pragmatics Association.
(2002). Gender identity and the presentation of self in Japanese. In S. Benor, M. Rose,
D. Sharma, J. Sweetland, and Q. Zhang (eds.), Gendered practices in language. Stanford,
CA: CSLI Publications.
McGloin, Naomi Hanaoka (1990). Sex difference and sentence-final particles. In S. Ide and
N. H. McGloin (eds.), Aspects of Japanese womens language, 2341. Tokyo: Kuroshio.
 ALTERNATIVE FEMININITY 255

(1999). Pragmatic and discourse functions of the rhetorical negative question form,
zyanai desu ka. In M. Muraki and E. Iwamoto (eds.), Linguistics: In Search of the human
mind, 425469. Tokyo: Kaitakusha.
Mizutani, Osamu, and Nobuko Mizutani (1977). An introduction to modern Japanese. Tokyo:
Japan Times.
(1987). How to be polite in Japanese. Tokyo: Japan Times.
Nihon kokugo dai-jiten (The comprehensive dictionary of the Japanese language) (1978).
Tokyo: Shogakkan.
Ochs, Elinor (1993). Indexing gender. In B. D. Miller (ed.), Sex and gender hierarchies, 146
169. Cambridge, MA: Cambridge University Press.
Okamoto, Shigeko (1995). Tasteless Japanese: Less feminine speech among young Japa-
nese women. In K. Hall and M. Bucholtz (eds.), Gender articulated: Language and the
socially constructed self, 297325. New York: Routledge.
Okamoto, Shigeko, and Shie Sato (1992). Less feminine speech among young Japanese fe-
males. In K. Hall, M. Bucholtz, and B. Moonwomon (eds.), Locating power: Proceed-
ings of the Second Berkeley Women and Language Conference, 478488. Berkeley:
Berkeley Women and Language Group.
Reynolds, Katsue Akiba (1990). Female speakers of Japanese in transition. In S. Ide and N. H.
McGloin (eds.), Aspects of Japanese womens language, 129146. Tokyo: Kuroshio.
Rosenberger, Nancy R. (1996). Fragile resistance, signs of status: Women between state and
media in Japan. In A. E. Imamura (ed.), Re-imaging Japanese women, 1245. Berkeley:
University of California Press.
Shibamoto, Janet S. (1985). Japanese womens language. New York: Academic Press.
[Shibamoto] Smith, Janet S. (1992). Women in charge: Politeness and directives in the speech
of Japanese women. Language in Society 21(1): 5982.
Shibatani, Masayoshi (1990). The languages of Japan. Cambridge, MA: Cambridge Univer-
sity Press.
Tamanoi, Mariko Asano (1990). Womens voices: Their critique of the anthropology of Japan.
Annual Review of Anthropology 19: 1737.
Teramura, Hideo (1984). Nihongo no shintakusu to imi (Japanese syntax and meaning), vol. 2.
Tokyo: Kuroshio.
Uchida, Nobuko (1993). Kaiwa-koodoo ni mirareru seesa (Gender differences in conversa-
tion). Nihongogaku 12(6): 15668.
Uno, Kathleen S. (1993). The death of good wife, wise mother? In A. Gordon (ed.), Post-
war Japan as history, 293322. Berkeley: University of California Press.
Uyeno, Tazuko Y. (1971). A study of Japanese modality: A performative analysis of sen-
tence particles. Unpublished Ph. D. dissertation, University of Michigan.
(1972). Shuujoshi to sono shuuhen (Sentence-final particles and related issues).
Nihongo Kyooiku 17: 6277.
256 REAL LANGUAGE , REAL PEOPLE

14

AYUMI MIYAZAKI

Japanese Junior High School Girls

and Boys First-Person Pronoun Use
and Their Social World

Ndee, onna no ko ni kakomareteru tokii, nan to nakuu, ore tte icchau n da yo ne.
Tte yuu kaa, Aa, jibun ga otoko da ttara, kore zenbu jibun no onna na no ni naa
tte omottari suru toki ore nan da yo ne. Chotto kawatteru deshoo? . . . Atashi nee,
nan daroo nee, nnn, dotchi katte yuu to, boseehonnoo yori moo, otoko no ko ga
yowatchii onna no ko mamoritaku naru, ano kanji no hoo ga tsuyoi ka mo shinnai,
boseehonnoo yori.1
And when Im surrounded by girls, Ill say ore [a strongly masculine first-person
pronoun]. Or when I imagine, Ahh, if I were a boy, these girls would be all mine,
Ill say ore. Arent I strange? . . . I, I wonder, if I had to choose one, Id say that I
have more of a boys instinct to protect weak girls than a maternal instinct. (13-
year-old girl)
Ore ga niau hito tte iru n desu yo. Tatoeba, supootsukee ga dekiru hito toka . . .
Ore tte itte kimaru hito iru jan. Boku nanka zenzen kimannai jan.
Ore suits some people. For example, people who are good at sports . . . There are
people who sound cool with ore. I wouldnt sound cool at all if I used ore. (13-
year-old boy)

In this chapter, I analyze diverse linguistic practices of Japanese junior high

school students that go against the Japanese ideology of gender-divided language.
In particular, I explore girls and boys untraditional first-pronoun usage, as demon-
strated in the preceding comments of a girl who says that she uses the supposedly
vulgar masculine first-person pronoun, ore, and of a boy who reports that he eschews
this masculine pronoun. This chapter examines such linguistic practices by locating
them within students social world, namely, their peer-group relations, and within

256
 JAPANESE JUNIOR HIGH SCHOOL GIRLS AND BOYS 257

the power dynamics of their gakkyuu, the fundamental unit of Japanese schools.
From 1997 to 2001, I conducted a longitudinal ethnography of a largely middle-
class public junior high school in a city that is an hours train ride from Tokyo. The
data for this study is drawn from observations of various classes and activities at school
and interviews with students in individual and group settings when the students were
in the seventh grade, the first year of junior high school (April 1998 to April 1999).

14.1. Beyond a direct linkage between gender

and language

Many scholars (e.g., Shibamoto 1985, Ide & McGloin 1990, [Shibamoto] Smith 1992,
Reinoruzu [Reynolds]-Akiba 1993, Ide 1997) have pointed out that, in the Japanese
language, female and male speech diverges widely in self-reference and address
forms, sentence-final particles, vocabulary, pitch range, and intonation. For instance,
sentence-final particles, which signal speakers various sentiments in conversation
(e.g., McGloin 1990, 1997), directly index the gender of the speaker. Sentences that
end with the particle wa are considered to belong to womens speech, while those
that end with the particle ze are considered to belong to mens speech.
Another concrete example of this gendered framing is first-person pronouns
glossable as I in English. Table 14.1 shows the variety of first-person pronouns
within Standard Japanese.
As the table indicates, women and men are supposed to use different first-person
pronouns: atakushi and atashi are for women; boku and ore are for men. Moreover, in
womens speech there is no other-deprecatory first-person pronoun comparable to ore
in mens speech. Women are expected to avoid the deprecatory level of pronoun use
and to display a polite demeanor (Ide 1997:7374).2
Many studies, however, including those in part III of this volume, have con-
firmed the existence of significant variation in the actual speech of Japanese women.
Okamoto and Sato (1992) and Okamoto (1995, 1997), for instance, make it clear that
gendered speech does not simply indicate the gender of the speaker. Analyzing the
speech of Japanese women of various ages, these scholars observed that younger
women tended to use traditional masculine speech more frequently than older
women, although each age group exhibited wide variations. Sunaoshi (chapter 10,
this volume) found that farm women in Ibaraki Prefecture frequently used ore, a first-
person pronoun supposedly not available to women, during their product develop-

14.1 Gender-marked first-person pronouns

Mens Speech Womens Speech

Formal watakushi watakushi

watashi atakushi
Plain boku watashi
atashi
Deprecatory ore f
Ide (1997:73)
258 REAL LANGUAGE , REAL PEOPLE

ment meetings. In Abes study (chapter 11, this volume), young female employees
at a lesbian bar used jibun, a first-person pronoun that is normally associated with
men in sports or in militaristic groups.
Inoue (1994) calls into question the notion that a unified Japanese womens lan-
guage has existed throughout Japanese history. Through the analysis of voices of
female characters in early modern novels, Inoue demonstrates that Japanese womens
language as it is characterized today does not flow from actual usage since time
immemorial but was deliberately created in the process of Japans modernization in
the early twentieth century. The intensely researched gender division of the Japa-
nese language, therefore, is an ideology of how Japanese speakers should speak, not
rules that all Japanese speakers follow.
These findings coincide with gender and language studies in North American
and Western European settings, where a number of scholars have recently stressed
the need to examine seemingly direct and exclusive relationships between language
and social attributes, such as the gender and social status of speakers (e.g., Bing &
Bergvall 1996, Woolard 1998, Bucholtz 1999a), and have revealed, from ethno-
graphic perspectives, how women and girls negotiate language ideologies of gender
(e.g., Coates 1994, 1999; Mendoza-Denton 1994; Goodwin 1998, 1999; Bucholtz
1999b; Morgan 1999). Bucholtzs (1999b) ethnographic study, for instance, finds
that a group of nerdy California high school girls resist the hegemonic ideology of
femininity through creative linguistic practices such as punning. Morgan (1999),
through analyzing how African-American girls and women actively participate in
elaborate and skillful language games, opposes the idea that African-American ver-
nacular culture is represented only by African-American boys and men and that femi-
ninity is represented only by middle-class European American girls and women.
Goodwin (1998, 1999) observes that Latina girls playing hopscotch express strong
disagreement by using their body movements and linguistic resources. Goodwin
successfully challenges a series of earlier gender studies that dichotomized girls and
boys play in terms of cooperation versus competition.
Based on these studies, this chapter explores similarly dynamic relationships
between gender and language by examining Japanese girls and boys speech within
its specific social and pragmatic context. Ochs (1992) argues that to untangle the
seemingly direct connection between linguistic forms and gender one needs to ex-
amine the complex sets of pragmatic meanings that mediate them. The Japanese
sentence-final particle ze, for instance, does not always suggest that the speaker is
male, but the coarse intensity of this particle connects it to the image of masculinity.
The sentence-final particle wa, in similar fashion, does not determine the female
gender of its speaker; instead, its delicate intensity only indirectly constitutes its gender
meaning. Agha (1993) also notes that linguistic forms, in this case the honorifics of
Lhasa Tibetan, do not automatically encode social status. Rather, these honorifics
signal deference entitlements in myriad ways and cannot be understood outside of
their context. If someone says, Mother went to the house, with the word house in
an honorific form, the listener would not understand who is being deferred to: is it
the mother, the owner of the house, or the builder of the house? The listener needs to
understand the real-time occurrence of the events, discourses, and their contexts to
make sense of how the linguistic form signals social status. Only by examining such
 JAPANESE JUNIOR HIGH SCHOOL GIRLS AND BOYS 259

rich, complex constructions of local pragmatic meanings within a specific commu-

nity of practice (Eckert & McConnell-Ginet 1992) are we able to understand the
connections between language and gender and other social categories.
The specific context explored in this chapter is the Japanese junior high school
class, called gakkyuu (or kurasu, from the English word class). The gakkyuu orga-
nizes its students differently from U.S. schools. Gakkyuu are composed of a group
of up to 40 students and one teacher who is in charge of the gakkyuu. From elemen-
tary school to high school, Japanese students belong to a gakkyuu. Gakkyuu are gen-
erally reorganized every one to two years; between reorganizations students cannot
switch gakkyuu.3 Students in a gakkyuu spend most of their school life together in
the same classroom, follow the same curriculum, spend break time together, and
interact with one another in front of the lockers found in classrooms. Moreover, stu-
dents in a gakkyuu share various nonacademic activities such as morning and after-
noon meetings, serving and eating lunch in the classroom, performing cleaning duties
together, preparing as a group for inter-gakkyuu competitions of chorus or sports,
and taking on numerous tasks such as organizing cleaning and lunch duties, orga-
nizing the classroom work assignments, checking attendance and reporting it to the
health office of the school, and taking care of handouts and bulletin boards. A gakkyuu
is therefore more than just a class; it is a seekatsu kyoodootai school community
where tightly knit relationships form.
Students try to belong to a peer group within their gakkyuu not only because it
is important for their social lives in this closed community but also because both
academic and nonacademic activities are often organized around small groups of
students in a gakkyuu. In the school I studied, these small groups were usually com-
posed of peer groups.4 Students who do not belong to a peer group in their gakkyuu
have trouble finding their place on the daily or even hourly occasions in which a
gakkyuu divides itself into groups for performing activities. As a result, gakkyuu
become fields where students constantly negotiate their relationships, status, and
power.
Ethnographic methods allowed me to closely observe girls and boys who were
actively interacting and making sense of the meaning of gender, language, and iden-
tity in their gakkyuu. Based on such investigation, I analyze (1) what kinds of first-
person pronouns girls and boys actually use in their peer groups and how these
pronouns differ from the ideological map of feminine or masculine speech, (2) how
girls and boys form pragmatic meanings (Silverstein 1976) with respect to their first-
person pronoun uses, and (3) how these uses relate to students relationships within
their gakkyuu.
Although previous studies of preadolescents and adolescents actively experiment-
ing with their speech and identity greatly contribute to reconsidering the fixed rela-
tionships between language and social categories, only a few studies (e.g., Ogbay
2000) have investigated them in great depth in a non-Western context. The Japanese
ideology of gendered language provides a pivotal point from which to explore the
ways individuals negotiate, struggle with, and resist their gender constraints. Such
negotiations in the social world of the gakkyuu in turn mirror intricate relations be-
tween individuals, groups, and the dominant ideology. As Woolard says, ideologies
of language underpin not only linguistic form and use but also the very notion of
260 REAL LANGUAGE , REAL PEOPLE

the person and the social group (1998:3). Thus, I believe that the exploration of
Japanese junior high school students linguistic practices will contribute to a general
theory of the complex interplay between human interactions and ideology.

14.2. First-person pronoun uses of the girls

and boys in Gakkyuu A
14.2.1. Repertoires of first-person pronouns
Figure 14.1 shows a map of girls and boys peer-group relations and the first-per-
son pronouns in a gakkyuu I studied, which I call Gakkyuu A. Small ovals in the
diagram represent individual girls and small dotted ovals represent individual boys.5
The forms of the first-person pronouns within each small oval represent the first-
person pronoun most frequently used by the individual in question in informal set-
tings such as when talking to friends.6 The pronouns in the diagram are not drawn
from students own explanations but from my observations of students actual use
and from transcripts of group interviews. Larger ovals that surround these small ovals
represent peer groups, ovals with unbroken lines with the letter G represent girls
groups, and dotted lines with the letter B represent boys groups.
The masculine/feminine continuum at the top of the diagram compares the Japa-
nese ideology of gendered first-person pronouns reflected in table 14.1 with students
interpretations of the four main pronouns they use. Although students interpreta-
tions are by no means uniform, as I explain in the next section, the simple continuum
nonetheless serves as a point of reference both for the dominant ideology of gendered
first-person pronouns and for students actual use of first-person pronouns shown in
the diagram below the continuum.
According to table 14.1, girls should use watashi or atashi, an informal female
pronoun, among friends. No girl in this gakkyuu, however, regularly used watashi,
and many girls used boku, a plain male pronoun, or the other-deprecatory male pro-
noun ore. Many girls also used uchi, which does not even appear in table 14.1. Uchi,
a word that originally meant inside or home seems to be a newly created first-
person pronoun in the Tokyo area. Although I grew up in the city where the school
is located, I had never heard of anybody using uchi as a first-person pronoun before
I entered the research site. Of the seven mothers of students I interviewed in the year
2000, who were mostly in their forties, not one realized that such a first-person pro-
noun existed.
As table 14.1 shows, boys are supposed to use the plain form boku among friends.
Most boys in Gakkyuu A, however, used ore, and only one boy used boku exclu-
sively. One boy, who reported that he belonged to an otherwise all-girl group, often
used atashi. Two boys regularly used forms other than the four main forms the stu-
dents at the school usedatashi, uchi, boku, and ore. One boy used washi, which is
considered to be a pronoun for old men. Another boy used oresama Mr. I or the
honorable I.7
These descriptions of girls and boys first-person pronoun use demonstrate that
the students did not always follow the gendered language ideology reflected in table
14.1. In fact, it is difficult to mark these untraditional linguistic practices as mere
 JAPANESE JUNIOR HIGH SCHOOL GIRLS AND BOYS 261

Masculine Feminine

Gendered ideologies
reflected in Table 14.1: ore/boku watashi/atashi

Students interpretations: ore boku uchi atashi

G1 B1
ore ore
atashi ore ore ore
uchi

ore boku boku

boku ore

boku
atashi B2
G3
G4
ore ore ore
boku
atashi
uchi

atashi
atashi uchi B3
uchi
G4 oresama ore
ore

atashi
ore
G6 G2 B4
ore ore
atashi atashi atashi
uchi uchi

boku washi
atashi
uchi

B5
ore ore
G5 atashi
uchi

uchi

14.1 A map of girls and boys peer-group relations and first-person pronoun use
in Gakkyuu A
262 REAL LANGUAGE , REAL PEOPLE

deviations, since only 2 students out of 34 consistently followed the norm of gendered
contexuality indicated in table 14.11 girl out of 17 used atashi exclusively, and 1
boy out of 17 used boku exclusively. These two students were the minority within
the linguistic world of Gakkyuu A.

14.2.2. First-person pronouns within peer groups

According to my observations, and conversations and interviews with students, the
girls and boys in Gakkyuu A divided themselves into different peer groups, each of
which expressed different tendencies in the use of first-person pronouns and other
nonlinguistic features. Peer-group cultures are by no means monolithic or simple,
but in the following paragraphs I try to describe their salient features. The number of
each group corresponds with the numbers in figure 14.1.

Girls groups. Group 1 was the group of girls whose first-person pronouns were
most untraditional. The two core members of this group used ore, and three other
members used both ore and boku or boku alone. There was also one girl who used
atashi and uchi. This group had a distinctive subculture and language that both its
members and other students considered gehin vulgar. These girls invented dances
to express their disdain for the skirts they had to wear as part of their school uni-
forms, and they composed erotic songs to express their strong interest in sex and
sexuality. Their interest in their changing pubescent bodies also appeared in their
subculture, as Aki, one of the core members of this group, shows in example (1).8

(1) Kinoo no yoru ore no shikyuu wa itteki mo dasanakatta . . . ! Ore no shikyuu

shiboridashitai!
My [ore no] womb didnt produce a single drop of blood last night! I want to squeeze
my (ore no) womb . . . !

This statement certainly breaks with traditional Japanese constructions of femi-

nine behavior and language. The content of the statement, Akis disappointment at
the sudden end of her menstrual period, is a taboo topic. Moreover, in producing this
utterance, Aki used ore, a very masculine first-person pronoun.
This group of girls did not like the teacher in charge of their gakkyuu and re-
sisted her strongly. One member in this group did not resist, however. This girl, who
used atashi and uchi, belonged to the group because she and the core members par-
ticipated in the ping-pong club together, but her attitudes toward school and the
gakkyuu teacher differed from those of the other girls in this group. She later moved
to Group 2.
Group 2 comprised four studious, pro-school girls who used atashi. All but one
of these girls also used uchi. What differentiates these girls from the Group 1 girls is
their attitude toward school. The girls in Group 2 diligently performed various gakkyuu
tasks such as cleaning duties. They did not resist teachers and school rules, although
they often complained about both within their group.
Group 3 was created when one girl who constantly used boku left Group 1 and
formed a group with a girl who used mainly atashi and occasionally uchi. The social
 JAPANESE JUNIOR HIGH SCHOOL GIRLS AND BOYS 263

identities of Groups 3 and 4 fell somewhere between those of Groups 1 and 2. The
girls in Group 3 were less resistant to the teacher than the girls in Group 1 but were
not as accommodating to the rules or as active in official gakkyuu activities as the
girls in Group 2. One boy who used boku and atashi considered himself to belong to
this group.
Group 4 consisted of two girls who used atashi and uchi. These girls became
closer because they belonged to the same sports club. Like the Group 3 girls, they
did not oppose the teacher publicly but considered themselves to be less otonashii
reticent than Group 2 girls.
Group 5 had two members. One was a girl who constantly used uchi. She did
not belong to any of the groups described earlier, and boys bullied her. Another girl
who used atashi and uchi, and who was excluded from Group 2, interacted with other
girls but reluctantly ended up forming a group with this unpopular girl.
Finally, Group 6 had only one member: a girl who was a loner and did not
like belonging to a peer group. She used atashi and ore.9

Boys groups. Group 1 was a group of four boys who belonged to sports clubs, the
Japanese equivalent of jocks (Eckert 1989). These boys always used ore. They were
considered to be undooshinkei ga ii athletic, kakkoii cool, ninki ga aru popular,
ussai noisy, and tsuyoi strong. Three boys belonged to the soccer club and one to
the track club.
There were three groups (2, 3, and 4) whose social identities fell between Groups
1 and 5. Group 2 consisted of three boys. These boys also used ore and belonged to
sports clubs but did not get along with the boys in Group 1. One boy belonged to the
soccer club with Group 1 boys, but he said he did not like to be with them because
these boys, and one in particular, shikiru control too much and are jikochuushinteki
self-centered. The other two boys belonged to the baseball club, whose members
were considered weaker than those in the soccer club.
One boy in Group 3 used ore and the other boy used oresama, which as noted
earlier can be translated as Mr. I, or the honorable I, and ore. These boys belonged
to sports clubs but also drifted apart from the Group 1 boys.
The four boys who belonged to Group 4 used various first-person pronouns. Two
boys used ore, one used boku, and the other used washi. Three boys belonged to the
badminton club and one to the swimming club, which were considered less power-
ful and athletic than most of the boys sports clubs, though more powerful than the
boys ping-pong club. These boys were close to the Group 3 boys, with whom they
often played role-playing games during break time.
There was also a group of three boys (Group 5) who were quiet and marginal in
the gakkyuu. They shared an interest in playing the guitar. One boy rarely came to
school. Because the subjects of sentences are often omitted in Japanese, it was diffi-
cult to record which first-person pronouns these quiet boys used, but I heard two boys
in this group using ore when speaking to each other.
Thus, different groups have different nonlinguistic characteristics, such as their
attitudes toward school and the teacher (resistant, accommodating, etc.), their in-
terests (sports, games, the body and sex, etc.), their positions in the gakkyuu (cen-
tral, marginal, etc.), and their femininity or masculinity. The linguistic features of
264 REAL LANGUAGE , REAL PEOPLE

these groups were also a part of such subcultures (e.g., Eckert 1989, Mendoza-
Denton 1994). Members of each group tended to use the same first-person pro-
nouns or similar ones (ore and boku, or atashi and uchi). The first-person pronoun
use of the girls and boys was influenced by that of the other students in the same
peer groups. Girls in Group 1, for instance, used more masculine first-person pro-
nouns than those in Group 2. First-person pronouns are thus in part products of
group cultures.
It is important to note, however, that there were significant variations within
groups with respect to girls and boys first-person pronoun uses. The complex in-
terplay between individual and group identities is beyond the scope of this chapter
(see Miyazaki 2002), but these intragroup variations in the students pronoun uses
testify to the negotiated distance between an individual and her or his group. For
instance, the core members of girls Group 1, on the one hand, used ore. The other
three members, on the other hand, had been friends since elementary school and
continued to use boku, their habitual pronoun. One of the three members, however,
was influenced by the core members and often used ore with them, but the other boku
girls were not so influenced. Thus, individual girls and boys do not completely fol-
low the group norms of language, just as they do not always accommodate the
gendered language ideology.

14.2.3. Metapragmatics of first-person pronoun use

In addition to investigating girls and boys use of first-person pronouns, I also ex-
amined the metapragmatics of such usethat is, how students themselves perceived,
interpreted, and stereotyped their language use in the social world of peer-group
relations. Inoue (1995) notes that speakers interpretations and representations of
language use can be an important resource in understanding speech as a social pro-
cess. Similarly, Hanks (1993) says that although native metapragmatic discourses
often diverge from actual usage, they are nonetheless essential because they serve as
interpretive frames that guide everyday interactions.
Girls considered watashi, a plain feminine first-person pronoun according to
table 14.1, too formal to use in situations other than sakubun writing assignments
or presentations in class. Many girls regarded even atashi, a more informal pronoun
than watashi, as too formal and too feminine to use in informal settings. Students
interpretations of atashi, however, are by no means uniform. The girl who used atashi
exclusively explained that it is a normal girls pronoun and that it has nothing to do
with femininity. The use of atashi by a boy, however, was ridiculed and indexed as
unusual, or okama homosexual/transsexual, by both girls and boys.
Uchi is regarded as less feminine and formal than atashi and is preferred by many
girls for that reason. A number of girls also consider uchi a good pronoun to use among
friends because it is shorter and easier to say than atashi.
Ore is considered to be the most masculine pronoun. Although it is usually clas-
sified as other-deprecatory, boys who used it did not necessarily attach such a con-
notation to it; they said that ore is just a pronoun for boys. One boy who constantly
used boku, however, said that men fantasize about using ore and that only very cool
 JAPANESE JUNIOR HIGH SCHOOL GIRLS AND BOYS 265

men say ore. This boy did not think that he was cool enough to use ore, as he ex-
plains in the interview quoted at the beginning of this chapter. Another boy who used
boku and atashi and who belonged to a girls group also said that he did not use ore
because it was for cool boys such as those in Group 1.
Boys use of ore does not always arouse positive reactions in girls. Although
one girl said that the speech of Group 1 boys sounds imadoki trendy, some girls
said that ore sounds erasoo or ibatteru arrogant and that ore users employ this form
to try to cover up their inner weakness.
The use of ore by the resistant group of girls was received in various ways. When
I asked about the use of ore by these girls in an interview, a Group 1 boy told me that
these girls use ore because they are kichigai crazy. In general, however, these
girls masculine speech, along with their behavior, was usually accepted with laugh-
ter. Some academic-oriented girls told me that these girls were problematic, but that
they were so powerful that other students could not sakarau oppose them.
Boku is considered less masculine than ore and often has negative connotations
when used by boys: boku is for weak boys or for mamas boys. In all six seventh-
grade gakkyuu in the school, the groups of boys who used ore were regarded as stron-
ger and more powerful than those who used forms other than ore. It is interesting
that boku, a plain masculine pronoun, is denigrated in this way and is differentiated
from ore on the basis of masculinity, strength, or power in the social world of the
school. Girls use of boku, however, generally did not trigger negative images. A
girl who constantly used boku said that she used it simply because uchi is too femi-
nine for her, to say nothing of atashi.
Thus, first-person pronouns index complex sets of meanings in the context of
Gakkyuu A. Pronouns signal manifold pragmatic meanings not only regarding femi-
ninity and masculinity but also regarding many other factors such as power and soli-
darity and distance and intimacy (e.g., Brown & Gilman 1973, Friedrich 1979). Girls
and boys actively make sense of and negotiate these myriad, sometimes contested
meanings within and outside of their peer-group relations in the gakkyuu. These girls
and boys have ways of making meaning by using the traditional gender-differentiated
pronouns as well as others that fall outside the prescriptions of the dominant ideol-
ogy. Within the context of this gakkyuu, girls use of ore, which is supposed to be
illegitimate, at times has positive meanings and boys use of ore, which is supposed
to be a normal masculine pronoun, is sometimes severely criticized.
It bears noting, however, that gender nonetheless plays a significant role in stu-
dents language negotiations. For instance, feminine pronoun use by a boy was in-
dexed as unusual and discouraged among students. Ore and boku were differentiated
in terms of desired masculinity. Although many girls received girls use of ore posi-
tively, powerful boys perceived it as crazy. Moreover, although in this chapter I focus
on pragmatic meanings among students in the gakkyuu, ideologies promoted by teach-
ers and parents are an inseparable part of making meaning. These girls teachers and
parents often discouraged their masculine speech, which influenced girls linguistic
negotiations. This contrasts with boys use of ore, which was rarely singled out for
correction. In the next section, I turn to a weak boy, Taku, and his navigation of
the sea of pragmatic meanings of first-person pronouns. Although my focus in this
chapter is primarily on the dynamics of Gakkyuu A, I have chosen the following
266 REAL LANGUAGE , REAL PEOPLE

examples from another gakkyuu because they show well the field of gender and power
in which an individual student negotiates meanings.

14.3. Example: Takus negotiations

of first-person pronouns

Taku, a student in Gakkyuu B, was a short, thin boy with eyeglasses who was always
quiet in the gakkyuu. In this gakkyuu, a group of six boys, on the one hand, had the
most power: they dominated classroom conversation, occupied the most space, and
most effectively resisted the teacher in charge. Taku, on the other hand, along with
two other boys, occupied the opposite pole of the power spectrum. One boy in the
group told me in an individual interview that Taku was the boy least likely to join his
tsuyoi strong group.
All six boys in the powerful group used ore. One boy said that he had used boku
in elementary school, at a time when he was weak, but that in junior high school,
when he got close to Haruki, the leader of the group, he started using ore. Another
boy, who was the weakest of the six, at one point used oira, which in urban areas has
a funny, hillbilly connotation, and tried spreading it, but unsuccessfully.
Unlike the boys who used ore, Taku said in an interview with another powerless
boy in his gakkyuu, Hide, that he changes his first-person pronouns kanzen ni com-
pletely depending on the situation. Example 2 is Takus own explanation of how he
changes his first-person pronoun choices when he talks with Haruki.

(2) Taku = T, Hide =H, Interviewer = AM

1 T: Eettoo, mazu, (AM: Nn.).10 Konaka-kun (Harukis last name plus an honor-
ific suffix) mitai ni, kyooboo na hito no baai wa, (H: . . .) boku tutte . . .
(AM: Nn.)
(H: . . . ((whispering laughter))) ojigi suru n desu. (H: Ojigi . . . no?)
2 AM: Heee.
3 T: ((Turned to H)) A, ojigi shinee na, yappa Konaka no baai dakan na. (H: Aa,
soo da na . . .) Maa, boku tutte owari ni shite, de.
4 De, Uchida-kun (Hides last name plus an honorific suffix) mitai ni yasashii
hito wa, (H: . . .) (AM: Nn) nanka, nanka, nanka, ore tutte mo okonnasoo
dakaraa. (AM: Nn.)
5 Ne! Uchida!

1 T: Well, first, (AM: Nn.) to violent people like Konaka-kun (Harukis last name
plus an honorific suffix), (H: . . .) I use boku (AM: Nn.) (H: . . . ((whisper-
ing laughter))) and bow to him. (H: Bow . . . ?)
2 AM: Hmm.
3 T: ((Turned to H)) No, I dont bow to him; after all, its just Konaka. (H: Oh,
thats right . . .) Well, I just say boku and . . .
4 To somebody whos nice like Uchida-kun (Hides last name plus an honorific
suffix), (H: . . .) (AM: Nn) well, I guess, ore wouldnt upset him. (AM: Nn)
5 Right, Uchida?
 JAPANESE JUNIOR HIGH SCHOOL GIRLS AND BOYS 267

Taku explains that he uses boku to Haruki, the powerful boy in his gakkyuu, but
ore in front of Hide, whom he considers his equal. When I asked Taku which was
easier for him, he answered, Ore is easier to say, but when Im in front of Konaka-
kun I chicken out and end up using boku. Taku added that ore sounds erasoo arro-
gant and that if he uses it in front of the strong boys, he feels that they might try to
get revenge on him. Thus, Taku was aware that his status in the gakkyuu affected his
personal pronoun choices.
Right after Taku explained to me that he would use boku and bow to Haruki, he
turned to Hide, who had just murmured the question Ojigi . . . no? Bow?. In re-
sponse, Taku and denied what he had just said, that is, that he would bow to Haruki
(line 3). When Taku spoke to Hide, his speech suddenly shifted in a strongly mascu-
line direction: he omitted the honorific suffix on Harukis name and employed strongly
masculine forms such as shinee na and dakan na.11 This shift reinforces Takus as-
sertion that his use of gendered forms depends on audience.
The powerful boys in Takus gakkyuu lend further support to this assertion. When
I interviewed these six boys about their personal pronoun use, they said that they
unconsciously said ore. In example (3) I asked further.

(3) Haruki = H, Aiki = A, Shigeki = S, Yuji = Y, B = Bunta, Masa = M, Interviewer = AM

6 AM: Tatoeba saa, Konaka-kun (Harukis last name plus an honorific suffix) ga
boku to ka ittara imeeji awanai?
7 H: Ore iwanai.
[ ]
8 A?: . . . daro?
9 H: Ore iwanee yo.
10 A?: Awanee yo.
[ ]
11 Y?: Nanka Katchan (the nickname of a boy in another class) mitai jan.
12 H: Ore iwanee yo.
13 B: Kami no ke ga Katchan jan.
[ ]
14 H: Ore boku tte iwanee yo.
15 AM: E, boku tte doo yuu no?
[ ]
16 A: Ichiban au no wa nee Taku..
[ ]
17 S: (To H) Ima itta ja nee ka.
18 Y: Taku ka.
19 AM: Aa, aa soo, Taku-kun boku tte yuu yo ne.
[ ]
20 B: Taku wa, seekaku ga Taku mitai dattara, boku tte . . .
[ ]
21 Y: Taku wa nan ka saa,
joshi no mae dato ore tte to ka ittete saa, koko made kuru to saa =
[ ]
22 S: Taku kimochiwarui.
268 REAL LANGUAGE , REAL PEOPLE

23 Y: = boku (in a high, small voice) ttu n da yo ne.

24 H: Nakanishi (the last name of a girl who is in the same gakkyuu) kiratteru,
kiratte n desho.
[ ]
25 M Aa! Ano saa, ano shine tte yatsu desho?
[ ]
26 H: Soo soo.
27 B: E? Nani?
28 M: Furareta no.
6 AM: For example, if Konaka-kun (Harukis last name plus an honorific suffix)
said boku, would it be strange?
7 H: I (ore) wouldnt say boku.
[ ]
8 A?: . . . isnt it?
9 H: I (ore) wouldnt say it.
10 A?: It wouldnt fit.
[ ]
11 Y?: It sounds sort of like Katchan.
12 H: I (ore) wouldnt say it.
13 B: (Your?) hairstyle is like Katchans.
[ ]
14 H: I (ore) wouldnt say boku.
15 AM: What is boku like?
[ ]
16 A: The person boku fits best is Taku.
[ ]
17 S: (To H) You just said it (boku).
18 Y?: Ah, Taku.
19 AM: Ah, yeah, Taku-kun says boku. Thats right.
[ ]
20 B: Taku says boku because of his character.
[ ]
21 Y: Taku says ore in front of girls,
but in front of us, =
[ ]
22 S: Takus weird.
23 Y: = he says boku (in a high, small voice).
24 H: Nakanishi (the last name of a girl who is in the same gakkyuu) hates him,
right?
[ ]
25 M: Ah!
That one, (she told him,) I hope you die! Right?
[ ]
26 H: Yeah.
27 B: What?
28 M: Taku was dumped by her.
 JAPANESE JUNIOR HIGH SCHOOL GIRLS AND BOYS 269

This interview excerpt shows that Takus fear of the strong boys was not unfounded.
In this interview, they characterized him as kimochiwarui weird (line 22) and made
fun of him for being rejected by a girl he liked (lines 2428). From these boys per-
spective, boku fits Takus character (lines 16 and 20), but ore, which indexed mascu-
linity, arrogance and coolness in the pragmatic world of the school, does not. They
also made fun of Taku for saying boku in a high-pitched, frightened voice in front of
them (lines 21 and 23) and for using ore in front of girls (line 21). This accusation in-
dicates the power differences between these boys and Taku: they themselves uncon-
sciously and freely used ore to every classmate, whereas Taku did not enjoy such
freedom because of the strong boys pressure and intimidation. The way the strong
boys and Taku refer to each other in these interviews provides additional evidence of
power differences. Taku, on the one hand, called the powerful boys by their last names
plus the honorific suffix -kun; the powerful boys, on the other hand, called Taku by his
first name without any suffix, despite the fact that I used the honorific in my question
to them. Harukis group formed a powerful group culture represented by strongly
masculine, fast-paced speech, including the first-person pronoun ore, thereby setting
the terms of desired masculinity in Gakkyuu B and excluding Taku in the process.
Taku had to face such power dynamics in his gakkyuu every day. He was fully
aware of his situation and chose to shift his pronouns in front of the strong boys.
Takus personal history also seemed to affect his choice. Some girls informed me
that he had been severely bullied in elementary school; one girl suggested this might
have given him some bitter lessons in keeping a low profile. Hides pronoun choice,
however, provided an interesting contrast to Takus. Hide did not modulate his speech
in front of the strong boys, although he, too, was on the powerless side of social re-
lationships in the gakkyuu. Hide boasted to me, I could say ore even in front of the
principal. His persistence with ore may relate to the kind of identity he hoped to
construct. Hide explained that he himself was memeshii [a] sissy, by which he
meant that he was cowardly and easily frightened, had an amae no seishin the habit
of being indulged, and acted like a girl, although he wanted to change himself into
a otokorashii manly person. Hides steady expression of his desired identity, even
in front of Harukis group, however, came at a cost. The strong boys bullied him,
often physically. In the interview, Hide told me that when one of the strong boys
kicked the back of his leg he could not hold back tears. Such ijime bullying has
long been a serious problem in Japanese education.
Taku chose a path different from Hides, although he, too, wanted to use ore
and express the masculine side of his identity. The teacher in charge of their gakkyuu
told me that the strong boys generally left Taku alone because he was harmless to
them, but that Hides challenge to them resulted in bullying. Taku avoided being
bullied by altering his speech and attitude and by making himself harmless in front
of the strong boys. He repressed the part of his identity that could be expressed through
ore and reserved it for occasions such as conversations with Hide. In the interview
with Hide, Taku could constantly use ore, raise many topics of conversation he was
interested in, such as action movies and plastic models, or freely interrupt Hide. Takus
relationship with Hide allowed him to construct a different aspect of his identity, one
that he could not express in front of Haruki. Takus and Hides negotiation of their
speech and identity12 provides us with an example of the complex interplay between
270 REAL LANGUAGE , REAL PEOPLE

individuals language, identity, and social world. These boys, as well as the girls and
boys in Gakkyuu A, offer a valuable perspective on the linguistic dynamics of lan-
guage, power and gender.

14.4. Conclusions

I draw three conclusions from my analysis of first-person pronoun use within the
junior high school gakkyuu. First, gender as a fixed category cannot explain the com-
plex personal pronoun use of the students I studied. Girls and boys variously went
along with, contested, and continually negotiated the ideology of gendered language
in their daily gakkyuu interactions. Indeed, their negotiations often went beyond the
traditional grammar chart in table 14.1 that represented this language ideology. Some
girls used ore and boku, and one boy used atashi. Some girls created new pronouns,
such as uchi, which is not as feminine as atashi.
Second, students first-person pronoun use is deeply embedded in the specific
context of the gakkyuu. Students construct complex sets of pragmatic meanings about
their own and others use of first-person pronouns. Each student brings both individual
and group perspectives to the gakkyuu arena, creating a dynamic, changing assemblage
of meanings around gender, power, relationships, and identity. Nonlinguistic and lin-
guistic meanings fuse in this arena when, for example, girls convey a taboo topic with
a masculine pronoun and are positively received by other girls or when a plain mascu-
line pronoun, boku, is imbued with various negative meanings regarding power and
masculinity. Thus, rich, lively processes for making meaning sway students speech
away from the rigid traditional grammar chart.
Third, and finally, however, it is important to note that a student is not a completely
free agent, just as she or he is not an inactive object whose language is automatically
molded by language ideology. Although girls and boys first-person pronouns are often
at odds with the traditional language ideology, gender ideology nonetheless affected
the complex mass of pragmatic meanings of such pronouns. Girls masculine pronoun
use, for instance, was at times well received but at other times dismissed as crazy. A
boys feminine first-person pronoun use was ridiculed and sometimes severely pun-
ished. Girls and boys continually have to negotiate their speech and identity in a com-
plex field of gender and power, such as when Taku suppressed his desire to express the
masculine side of his identity in front of the strong boys.
The processes of creating actual speech and its meanings in a specific commu-
nity reach far beyond the imagination of the fixed, dichotomous picture of traditional
gender ideology. Such complex, contradictory, and unexpected processes can be
understood only by following the naturally occurring linguistic practices of a spe-
cific community. Ethnographic methods are best suited to capturing such richly
contextualized processes.

Notes
I appreciate the excellent guidance of the editors of this book, Shigeko Okamoto and Janet S.
Shibamoto Smith, and the series editor, Mary Bucholtz. I also would like to thank Profs. Sara
 JAPANESE JUNIOR HIGH SCHOOL GIRLS AND BOYS 271

Lawrence-Lightfoot, Marcyliena Morgan, and Miyako Inoue for their continued guidance
regarding this topic. The Japan Society for the Promotion of Science and the Tokyo Womens
Foundation provided financial support for the research described here. Special thanks go to
the girls, boys, and teachers who generously shared their lives with me.

1. Transcription conventions are as follows:

. end of intonation unit; falling intonation
? end of intonation unit; rising intonation
... omission of material by the author
[] beginning and end, respectively, of overlapping utterances
= places between utterances with no gaps, or to link different parts of a single speakers
utterance
. . . sounds that are inaudible or in doubt
(()) description of quality of talk and activity related to talk
2. The level of formality is said to differ between the genders. In Japanese, as in many
other languages, speakers are expected to use different language levels depending on the for-
mality of context and social relationship (e.g., Kondo 1990, Dunn 1996). Watashi indicates
a formal level for men but not necessarily for women; womens watashi sounds plainer than
mens watashi (Kondo 1990). Women, on the one hand, are said to use watashi or atashi
among friends or in other informal settings. Men, on the other hand, are usually said to use
watashi only in formal settings, such as when talking to superiors or older persons, or in an
official setting. This shows that women are expected to use more formal forms than men.
3. In elementary school, one teacher teaches all subjects to the gakkyuu, but in junior
high and high school various teachers enter the gakkyuu classroom to teach their subject of
specialization.
4. These groups are at times arranged according to lottery, other times according to the
class roster. One teacher told me that in another school he himself allocated students into groups
in ways that he thought would maximize educational outcomes.
5. Group formations underwent both dramatic and slight changes over the year. Figure
14.1 captures peer-group relations in the middle of the academic year.
6. The first-person pronouns listed here are the first-person pronouns the students used
regularly in their peer groups but do not completely exhaust their playful and creative prac-
tices. I observed many students occasionally using forms of first-person pronouns other than
those listed in the diagramfor instance, when girls who regularly used ore or boku occa-
sionally used atashi to make fun of femininity. In addition, students reported that they and
their gakkyuu mates sometimes used pronouns other than those I observed. These are not
included in the diagram.
7. There were many girls and boys at the school who creatively used pronouns other
than the four main ones. For instance, some girls used atai, oira (as noted earlier), their full
names (e.g., Miyazaki Ayumi is sleepy), their first names (e.g., Ayumi is going home),
and their first names plus the informal suffix -chan (Ayumi-chan doesnt want to study!).
There was even a girl with worldly pretensions who used the pronoun mii, which comes from
the English pronoun me. Atai, as well as oira, has a funny, hillbilly connotation in urban areas.
One girl explained to me that her friends use of atai conjures up images of a middle-aged
woman selling dango sweet rice cake balls in the Edo period. Self-referring with the suffix
chan has a childish connotation.
8. The names of the girls and boys that appear in this essay are all pseudonyms.
9. She reported that she had used ore in preschool, when she always played with boys,
but when she entered elementary school her girlfriends discouraged her from using ore. Scared
of being different from everybody else, she started using atashi and tried not to use ore throughout
272 REAL LANGUAGE , REAL PEOPLE

her elementary school years, although ore popped out whenever she talked to her preschool
friends. When she was a seventh grader (the first year of junior high school), she started using
ore more often, as she explains in the quotation at the beginning of this chapter. After this aca-
demic year, her use of ore increased over her junior high school years; by the time she was a
ninth grader (her final year of junior high school), she used ore more often than atashi.
10. In a Japanese conversation, listeners generally interject sounds such as hmm and
nn more often than English speakers would. In addition, Hide often made almost inaudible
backchanneling sounds. I put these small sounds in parentheses, instead of giving each sound
its own line.
11. According to McGloin (1990), the sentence-final particle na is characteristically used
by men, along with zo and ze, and signals strong insistence. Taku also used shinee instead of
its regular verb form, shinai. The phonological form ee instead of ai or oi is also considered
to be strongly masculine (Okamoto & Sato 1992, Okamoto 1995).
12. In the following year, when Taku became an eighth grader, however, he started to
challenge Haruki fiercely, and by the end of the ninth grade (the final year of junior high
school) Taku succeeded in using ore in front of him and other strong boys. Taku told me that
he was not afraid of Haruki anymore. Haruki admitted to me that Taku had become much
stronger and, in fact, almost like an equal.

References
Agha, Asif (1993). Grammatical and indexical convention in honorific discourse. Journal of
Linguistic Anthropology 3(2): 131163.
Bing, Janet M., and Victoria L. Bergvall (1996). The question of questions: Beyond binary
thinking. In V. L. Bergvall, J. M. Bing, and Alice F. Freed (eds.), Rethinking language
and gender research: Theory and practice, 130. London: Longman.
Brown, Roger, and Albert Gilman (1973). The pronouns of power and solidarity. In P. P.
Giglioli (ed.), Language and social context, 252282. Harmondsworth: Penguin.
Bucholtz, Mary (1999a). Bad examples: Transgression and progress in language and gender
studies. In M. Bucholtz, A. C. Liang, and L. A. Sutton (eds.), Reinventing identities:
The gendered self in discourse, 324. New York: Oxford University Press.
(1999b). Why be normal? Language and identity practices in a community of nerd
girls. Language in Society 28(2): 203223.
Coates, Jennifer (1994). Discourse, gender, and subjectivity: The talk of teenage girls. In
M. Bucholtz, A. C. Liang, L. A. Sutton, and C. Hines (eds.), Cultural performances:
Proceedings of the Third Berkeley Women and Language Conference, 116132. Berke-
ley: Berkeley Women and Language Group.
(1999). Changing femininities: The talk of teenage girls. In M. Bucholtz, A. C. Liang,
and L. A. Sutton (eds.), Reinventing identities: The gendered self in discourse, 123144.
New York: Oxford University Press.
Dunn, Cynthia D. (1996). Style and genre in Japanese womens discourse. Unpublished doc-
toral dissertation, University of Texas, Austin.
Eckert, Penelope (1989). Jocks and burnouts: Social categories and identity in the high school.
New York: Teachers College Press.
Eckert, Penelope, and Sally McConnell-Ginet (1992). Think practically and look locally:
Language and gender as community-based practice. Annual Review of Anthropology 21:
461490.
Friedrich, Paul. (1979). Structural implications of Russian pronominal usage. In A. S. Dil
(ed.), Language, context, and the imagination. Stanford, CA: Stanford University Press.
 JAPANESE JUNIOR HIGH SCHOOL GIRLS AND BOYS 273

Goodwin, Marjorie. H. (1998). Games of stance: Conflict and footing in hopscotch. In S. Hoyle
and C. T. Adger (eds.), Kids talk: Language practices of older children, 2346. New
York: Oxford University Press.
(1999). Constructing opposition within girls games. In M. Bucholtz, A. C. Liang,
and L. A. Sutton (eds.), Reinventing identities: The gendered self in discourse, 389409.
New York: Oxford University Press.
Hanks, William F. (1993). Metalanguage and pragmatics of deixis. In J. A. Lucy (ed.), Re-
flective language: Reported speech and metapragmatics, 127158. Cambridge, MA:
Cambridge University Press.
Ide, Sachiko (ed.) (1997). Joseego no sekai (The world of womens language). Tokyo: Meiji
Shoin.
Ide, Sachiko, and Naomi Hanaoka McGloin (eds.) (1990). Aspects of Japanese womens lan-
guage. Tokyo: Kuroshio.
Inoue, Miyako (1994). Gender and linguistic modernization: Historicizing Japanese womens
language. In M. Bucholtz, A. C. Liang, L. A. Sutton, and C. Hines (eds.), Cultural per-
formances: Proceedings of the Third Berkeley Women and Language Conference, 322
333. Berkeley: Berkeley Women and Language Group.
(1995). A woman manager and local metapragmatics in a Tokyo corporate office.
In R. Ide, R. Parker, and Y. Sunaoshi (eds.), SALSA: Proceedings of the Third Annual
Symposium about Language and SocietyAustin, vol. 36, Texas Linguistic Forum, 41
48. Austin: University of Texas, Department of Linguistics.
Kondo, Dorinne (1990). Crafting selves: Power, gender, and discourse of identity in a Japa-
nese workplace. Chicago: University of Chicago Press.
McGloin, Naomi Hanaoka (1990). Sex difference and sentence-final particles. In S. Ide and
N. H. McGloin (eds.), Aspects of Japanese womens language, 2341. Tokyo: Kuroshio.
(1997). Shuujoshi (Sentence-final particles). In S. Ide (ed.), Joseego no sekai (The
world of womens language), 3341. Tokyo: Meiji Shoin.
Mendoza-Denton, Norma (1994). Language attitudes and gang affiliation among California
Latina girls. In M. Bucholtz, A. C. Liang, L. A. Sutton, and C. Hines (eds.), Cultural
performances: Proceedings of the Third Berkeley Women and Language Conference,
478486. Berkeley: Berkeley Women and Language Group.
Miyazaki, Ayumi (2002). Relational shift: Japanese girls nontraditional first-person pronouns.
In S. Benor, M. Rose, D. Sharma, J. Sweetland, and Q. Zhang (eds.), Gendered practice
in language, 355374. Stanford, CA: CSLI Publications.
Morgan, Marcyliena (1999). No women, no cry: Claiming African American womens place.
In M. Bucholtz, A. C. Liang, and L. A. Sutton (eds.), Reinventing identities: The gendered
self in discourse, 2745. New York: Oxford University Press.
Ochs, Elinor (1992). Indexing gender. In A. Duranti and C. Goodwin (eds.), Rethinking con-
text: Language as an interactive phenomenon, 335358. Cambridge, MA: Cambridge
University Press.
Ogbay, Sarah (2000). The social and linguistic construction of identities of girls and boys in
Eritrea urban. Paper presented at the First IGALA Conference, Stanford University,
Stanford, CA.
Okamoto, Shigeko (1995). Tasteless Japanese: Less feminine speech among young Japa-
nese women. In K. Hall and M. Bucholtz (eds.), Gender articulated: Language and the
socially constructed self, 297325. New York: Routledge.
(1997). Social context, linguistic ideology, and indexical expressions in Japanese.
Journal of Pragmatics 28(6): 795817.
Okamoto, Shigeko, and Shie Sato (1992). Less feminine speech among young Japanese females.
In K. Hall, M. Bucholtz, and B. Moonwomon (eds.), Locating power: Proceedings of the
274 REAL LANGUAGE , REAL PEOPLE

Second Berkeley Women and Language Conference, 478488. Berkeley: Berkeley Women
and Language Group.
Reinoruzu [Reynolds]-Akiba, Katsue (ed.) (1993). Onna to nihongo (Women and the Japa-
nese Language). Tokyo: Yushindo.
Shibamoto, Janet S. (1985). Japanese womens language. New York: Academic Press.
[Shibamoto] Smith, Janet S. (1992). Women in charge: Politeness and directives in the speech
of Japanese women. Language in Society 21(1): 5982.
Silverstein, Michael (1976). Shifters, linguistic categories, and cultural description. In K. Basso
and H. Selby (eds.), Meaning in anthropology, 1155. Albuquerque: University of New
Mexico Press.
Woolard, Kathryn A. (1998). Introduction: Language ideology as a field of inquiry. In P. V.
Kroskrity, B. B. Schieffelin, and K. A. Woolard (eds.), Language ideologies: Practice
and theory, 347. New York: Oxford University Press.
 15

CINDI STURTZ SREETHARAN

Japanese Mens Linguistic

Stereotypes and Realities
Conversations from the Kansai and Kanto Regions

T here is very little empirical research into how men of any society use language at
the everyday, local level (but see Coates 2003; Johnson & Meinhof 1997). This is
certainly true in the Japanese case. The lack of investigation into Japanese men in
general and their language use in particular may give the impression that any Japa-
nese man on the street can stand as representative of the generic Japanese male and
that the language style he uses is both known and normatively spoken by all Japa-
nese men.
This chapter emphasizes the diversity among Japanese mens linguistic prac-
tices. I propose an alternative account based on ethnographically collected speech
data of how individual men negotiate their own ideas of normative or stereotypical
masculinity in expressing their identities through language. I examine casual con-
versations of men from two regionsKanto and Kansai. Specific features consid-
ered for investigation include sentence-final particles and discourse strategies. My
analysis will show that men use stereotypically masculine sentence-final particles
infrequently, and that even when they do, they use them in both ideology-consistent
and other ways to further particular discourse goals.

15.1. Background
15.2.1 Men, masculinity, and language in Japan
The literature on gender and speech styles in Japanese is very large. This literature can
be divided into two categories: research that deals with formal linguistic differences

275
276 REAL LANGUAGE , REAL PEOPLE

(phonological, morphological, lexical, syntactic, etc.) and research that deals with
interactional or discourse differences (backchannels, interruptions, emotional ex-
pressivity, etc.). The former are more numerous than the latter.
Using self-report surveys and prescriptive usage rather than empirical investi-
gation, the bulk of the studies on womens and mens differential use of formal lin-
guistic features report that women and men are thought to use different sets of first-
and second-person pronouns and sentence-final particles (SFPs) (e.g., Ide 1993,
Kanamaru 1993). Other research claims that Japanese men use polite forms of verbs
less frequently than women do and use a higher frequency of verb endings that are
blunt, assertive, and direct (e.g., Ide 1982, Reynolds 1985, Shibamoto 1987). Men
are also associated with reduced phonological forms such and dekee (< dekai big)
and umee (< umai delicious) (Ide 1982).
Previous studies have provided fairly clear differences between SFPs that women
and men use. McGloin (1990), Reynolds (1985), and Uchida (1993), among many
others, report that Japanese men use a set of SFPs that index aggression, authority,
masculinity, or intimacy, depending on interlocutor and context of interaction. Spe-
cifically, McGloin (1990) asserts that the SFPs zo and ze are used exclusively by men
and are typically characterized as strongly masculine. Zo, however, is considered the
most imposing of SFPs, conveying insistence, authority, aggressiveness, and a sta-
tus higher than ones interlocutor. While McGloins examples appear to derive from
prescriptive usage or native speaker intuition, it is important that she relates the prag-
matic meanings of SFPs to gender.
Uchida (1993), based on actual speech data of SFPs used by university students
in the Tokyo area, finds that ze and yo na are used exclusively by men; however, she
concludes that the SFPs used by women and men are becoming neutralized with re-
gard to gender differentiation.
Investigations that address the interactional aspects of Japanese language and
gender have found men to exhibit less pitch variation (Ohara 1992, chapter 12, this
volume), to use a smaller and less emotionally charged lexicon (Shigemitsu 1993,
Uchida 1993), to use fewer backchannels and tag questions (Kurozaki 1987, Horiguchi
1991), and to interrupt their conversational partners more frequently than women
(Shigemitsu 1993, Uchida 1993).
Neither the structural nor the discursive-interactional studies of gender dif-
ferences in Japanese have been linked to larger issues of language ideology, mas-
culinity, identity, or region, although they have been linked to femininity (cf.
Shibamoto 1987, Okamoto & Sato 1992, Inoue 1996). However, it has been fre-
quently noted that Japanese speakers have notions of what it means to talk like
a woman or man. Inoue (1996) and Okamoto (1995, chapter 2, this volume) both
note that women and men recognize idealized forms of Japanese and provide nu-
merous examples of popular (often negative) commentary about womens failure
to use Japanese womens language. Specific linguistic features that trigger these
responses need to be further investigated (but see chapter 12, this volume); never-
theless, it is evident that Japanese women and men have linguistic ideological
positions concerning the particular language forms that they hear and use (cf.
Silverstein 1979).
 JAPANESE MEN S LINGUISTIC STEREOTYPES AND REALITIES 277

Ethnographic and historical documents on Japanese men indicate that prior to

the Meiji era (18681912), the bushidoo warrior with his topknot and two swords
was the image of manliness; after 1868, a new man emerged, an updated samurai,
now showing off his masculinity by being enterprising and entrepreneurial (see
Kinmonth 1981). The bushidoo warrior ideals had not been replaced but rather re-
dressed, literally, in more stereotypical Western fashion and ideologically transformed
into a company warrior or entrepreneurial samurai (Vogel 1979). Characteriza-
tions of Japanese men as strong, silent, dependable, and so on are also abundant in
popular culture materials (e.g., novels, films, songs, and commercials; cf. Davis 1996).
However, systematic research on masculinity in Japan, in which the portrait of the
masculine man is emerging, is very recent.
More recent ethnographic accounts of Japanese men note that to be masculine
in Japan is to be hardworking and useful (Fujieda 1995, Gill 1999); it is to be strong
and dependable on the outside (Ito 1996, Seko 2000) and yet soft on the inside
(Yorifuji 1998). A masculine man excels at sports and is tsuyoi strong at drinking
alcohol (Toyoda 1997, Seko 2000); he is rhetorically adept (Rosenberger 1994) and
yet paradoxically lacks eloquence (Iwao 1993, Yamada 1997).
Descriptions of Japanese men as silent, taciturn, or at least verbally restrained
are not uncommon. Donahue (1998) notes that a Japanese man who is verbally seri-
ous and unexpressive has been traditionally more highly valued than one who is ef-
fusive and verbally outgoing. Similarly, Loveday claims that in formal settings of
social equality between males Japanese men take a low profile linguistically, un-
derstating, being terse, presenting an unemotional, self-restrained exterior (1986:95).
Seward comments that Japanese men emphasize the masculinity of their speech by
adopting a deep-voiced, guttural mode of speaking which is often accompanied by
stern faces and stiff postures (1968:111).
These studies give insight into different social (and sometimes sociolinguistic)
behaviors that men are expected to exhibit but are limiting because they do not de-
lineate the specific behavioral or linguistic strategies available to individual men to
construct hardworking, strong and dependable yet soft-on-the-inside selves
as well as to examine how individual men deal with the ideologies of normative
behavior in specific social contexts.

15.1.2 Regionality, language, and men

Until quite recently, much English language scholarship on modern Japan has fo-
cused on the Tokyo area. Japan has typically been assumed to be homogeneous at
many levels, including ethnicity, class, and language, and scholarship on Japan fo-
cuses on the characteristics of the middle class in Tokyo or Kanto, part of the eastern
region of Japan. However, as is increasingly recognized, a rich diversity of Japanese
identities exists and proliferates.
Kansai, part of the western region of Japan, is as representative of urban Japa-
nese people as Kanto, albeit in a slightly different way. Historically, Kantos popu-
lation was largely samurai families, while merchants were limited to the shitamachi
downtown regions (Otani 1994). Although it is currently a large commercial center,
278 REAL LANGUAGE , REAL PEOPLE

it is better known for its administrative function and political power. This is in con-
trast to Osaka, which has been historically and continues today to be a major com-
mercial center of Japan, enjoying strong economic power. Kansai, in particular Osaka,
is one of the main regions that dictate popular culture trends to the rest of the nation;
it is the area responsible for karaoke and most of Japans TV comedy entertainment
(manzai) (Sugimoto 1997).
The dialect of the Kansai region is popular throughout Japan; it is considered
one of the two prestige varieties used in Japan (the other being Tokyo standard;
see, e.g., Miyake 1995, Kunihiro, Inoue, & Long 1999). Scholars report that Kansai
residents do not hesitate to speak openly and publicly in their own language[s]1
(Sugimoto 1997:59); moreover, young people strive to mimic the dialect of Kansai
in lieu of their own Tokyo language, or Standard Japanese (SJ) (Onoe, Kasai, &
Wakaichi 2000). The dialect is particularly loved at home, that is, in the Kansai
region itself. While the Tokyo dialect is described as monologue-esque (monoroogu
muki; Sato 2000:65), the dialect of the Kansai region is said to be shitashimi yasui
friendly/familiar/affectionate and to have the ability to bind speakers together
(Peng 2000).
Images of men specific to Kansai are difficult to obtain, at least in the research
literature. Kansai natives are described as isogashii or sewashinai busy and always
in a hurry (Otani 1994). This image of hurriedness is matched linguistically by Pengs
description of the Kansai dialect as being spoken quickly, at a fast tempo (2000:75).
Whether this hurried image is particular to men is not clear, but images of Kansai
residents certainly provide alternative gendered identities to draw upon rather than
just the strong, silent, and slow sarariiman company man of Tokyo found in popu-
lar literature. We are left to wonder how ordinary Kansai men talk and how they might
utilize linguistic expressions of masculinity.

15.2. This study

This chapter, then, examines theories of masculinity and of linguistic difference as

they are instantiated in different regions of Japan. I maintain that masculinity, like
language, is a dynamic force that interacts and manifests itself differently across
speakers, spaces, and contexts. I do not assume that all Kansai men talk like the men
in my data, nor that these men talk this way at all times. Rather, I hope to begin to
make empirical inroads into potential links or connections between speech styles,
masculinity, and regionality.
I analyze the conversations of men from the Kanto and Kansai regions to find out
howor whetherthey express masculinity through particular linguistic features or
discourse strategies, and whether there are regional or situational differences in the way
they use these features and strategies. Rather than the term Kansai dialect, I use the
more specific term Hanshinkan dialect (HKD) to refer to the variety spoken by Osaka
and Kobe speakers. The Osaka and Kobe area in the Kansai region is commonly re-
ferred to as Hanshin, a Sino-Japanese term (Han refers to Osaka and Shin to Kobe),
and its dialect is called Hanshinkan dialect, although exactly what should be included
in this dialect has been debated (e.g., Wada & Kamata 1992, Hirayama 1997).
 JAPANESE MEN S LINGUISTIC STEREOTYPES AND REALITIES 279

15.3. Methods
15.3.1. The conversations
The main data analyzed in this chapter comes from three conversations (approxi-
mately 70 minutes each): two from the HKD area and one from Kanto. The HKD
conversations used in this chapter are part of a larger corpus of data (comprising over
45 hours of conversation) collected during my field research in the Kansai area of
Japan, including Kobe and Osaka, from July 1998 through January 2000. The Kanto
conversation is taken from the Shibamoto Smith Japanese Conversation Corpus,
comprising data collected in the late 1970s by Janet S. Shibamoto (as reported in
Shibamoto 1985). In each case of data collection, the participants knew the researchers
were interested in language use but were not guided toward discussions of language.
All HKD data was recorded on a MiniDisc portable recording device; I was not
present for any of the recordings. After being introduced (via a third party) to one
man, I would then explain my research to him and, if he agreed to help, would ask
him to gather one or two friends together to talk. I provided my contact with a re-
corder and recording instructions. I encouraged the men to have their conversations
anywhere they felt comfortable talking informally. The Kanto conversation was re-
corded on a Sony TC-800A open reel recorder with an attached microphone. The
investigator was present during this recording; however, based on the informality of
the forms used, the conversations were judged to sound natural by native speakers
of Japanese (Shibamoto 1985:74).
All of the men were in the career stage of their lives2 and, at the time of each
recording, all were employed by Japanese companies. The Kanto conversation has
three participants and takes place in the mens company lunchroom. The three men
Shibata, Mihara, and Kawamura3are each 34 years of age and are coworkers. The
Kobe conversation has four participants and takes place in a local okonomiyaki-ya,
a shop that sells a pizza-pancake kind of food found throughout Japan. The four men,
like their Kanto counterparts, are co-workers; they are Sato (42 years old), Yamada
(38 years old), Honda (45 years old), and Nakayama (29 years old). Despite their
slightly disparate ages, they are a closely knit group and have attended one anothers
weddings. The Osaka conversation has two participantsTanaka (40 years old) and
Honda, who also appeared in the Kobe conversation. It takes place in Hondas com-
pany office. The men have been friends for a long time and often join each other on
both private and company sporting excursions. They are coworkers, although in dif-
ferent sections of the same company.
Each conversation was transcribed and coded for a variety of features; in this
chapter I examine the use of SFPs as discourse strategies. SFPs were chosen be-
cause the recent findings by Inoue (1996), Okamoto (1996, 1998), and Ogawa and
[Shibamoto] Smith (1997) raise questions about how speakers actively use SFPs
to subvert or conform to traditional or stereotypical notions of gender.
One additional HKD conversation is analyzed later, but not with respect to SFPs.
In this conversation, I focus on the use of highly marked stereotypical masculine lin-
guistic practices, including phonological, lexical, and morphological features. This
conversation is by two menIto and Kadofrom the Kawachi region of Osaka
280 REAL LANGUAGE , REAL PEOPLE

Prefecture. The Kawachi dialect is stereotypically associated with rough or rude

speech and, perhaps consequently, with male speech. Both Ito and Kado are 67 years
old and retired. Itos son is married to Kados daughter. Thus, Ito and Kado have
known each other for as long as their children have been married, about 15 years,
and frequently interact via their children and grandchildren.

15.3.2. Stereotypically gendered categories of SFPs

The SFPs identified in this chapter are based on previous studies of SFPs and on
grammar texts (cf. Fujiwara 1982, 1985, 1986; Komatsu 1988; Okamoto & Sato 1992;
Kawashima 1999). Table 15.1 shows stereotypical gender categorizations of SFPs;
each category has two subcategories, HKD and SJ.
The categorizations for SJ are based on the vast literature on this topic (e.g.,
Kawaguchi 1987, Okamoto & Sato 1992); the categorizations for HKD are based on a
less vast and perhaps a less strictly academic literature (e.g., Kamata 1979; Fujiwara
1982, 1985, 1986; Makimura 1984; Wada & Kamata 1992; Yamamoto 1995; Hirayama
1997). Thus, it is important to note that for both HKD and SJ the categorizations are
ideological or prescriptive in nature.
While it is not always possible to separate dialect usage cleanly from SJ, the men
from Hanshin display quite typical HKD features at the lexical and morphological
levels. However, they also use SJ; the usage is highly mixed at the morphological,
lexical, and syntactic levels. Following are examples of two HKD features drawn
from my data. Because this chapter focuses on masculinity, examples 1 and 2 are
only of those SFPs that are associated with stereotypically masculine gender. (For
explanation of those HKD SFPs that are considered to be neutral with regard to gen-
der see, e.g., Hirayama 1997.)

(1) [HKD] moratteru no kai?

(cf. [SJ] moratte iru no ka?)
Do you receive it?

15.1 Stereotypical gender categorization of sentence-final particles in Standard

Japanese and Hanshinkan dialect
Strongly Moderately Moderately Strongly
masculine masculine Neutral feminine feminine

HKD yan ke ga na; wai; ya n(ai)(ka) ya wa

jai kai de; non
nen/ten (na)
shi; de ne; na
SJ zo da yo; kai + yo ne mon/o kashira
ze ka na; mon na + ne + ne + no(yo)(ne)
na + yo na wa + no + wa(yo)(ne)
sa; + yo Q no + na no(yo)
ja n; ke + yo; wa

= verb; adjective; = nominal form; Q = question; arrows indicate falling or rising intonation.
 JAPANESE MEN S LINGUISTIC STEREOTYPES AND REALITIES 281

The form kai, equivalent to the interrogative final particle ka in SJ, is considered to
be masculine in both SJ and HKD; it is used to express strong opposition to some-
thing said by an interlocutor (Makimura 1984, Kawashima 1999).

(2) [HKD] mainichi, nikkee o yomu no? erai ga na

(cf. [SJ] mainichi, nikkee o yomu no? rippa ja nai ka)
You read the Nikkei [Japanese Economic Newspaper] every day? Thats amaz-
ing, isnt it!

The ending ga na is an emphatic equivalent of the SJ form ja nai ka isnt it?.

It is considered stereotypically masculine in HKD (Makimura 1984).

15.4. Results and discussion

15.4.1. Quantitative analysis

Not all of the SFPs listed in table 15.1 were used within the segments under investi-
gation. The subset of SFPs included in my analysis is shown in table 15.2.
Table 15.3 presents the distribution of SFPs according to their gender categories.
The neutral count and percentage do not include the absence of SFPs, which is nor-
mally considered neutral; the percentage is calculated against the total SFPs possible.4
There are some notable differences across the conversations. First, I examine
the conversation of the Kanto men. They use neutral forms most frequently (40%),
followed by masculine forms (22%), and feminine forms the least (9.1%). The form
da yo is considered to be moderately masculine, at least ideologically (see table 15.1).
However, it is used rarely by the Kanto speakers (4 times, or 2.3%). Likewise, the
standard strongly masculine SFP na is used only 9 times (8%) by the Kanto speak-
ers. Zo and ze, two of the strongly masculine forms, are never used. Rather, the neu-
tral form ne is overwhelmingly the SFP of choice. The Kanto men do make use of
the relatively masculine SFP adjective/verb + yo and other moderately masculine
forms (29 times, or 17%). Among feminine forms, the moderate forms +
no((4%) and + ne (3.4%) were most favored.
Turning to the Osaka conversation next, we see a slightly different pattern emerg-
ing. These speakers overwhelmingly use neutral SFPs (49%with 27% in SJ and
22% in HKD). They split the rest of their SFP usage equally into moderately mascu-

15.2 SFPs used in the analysis

Strongly Strongly
masculine Masculine Neutral Feminine feminine

HKD ga na ya n(ai)(ka)
kai wa ; de
na; nen/ten
SJ na kai; da yo + yo ne mon/o + no ne
+ yo ~ + ne + ne + no yo
sa wa ; Q no no + yo
282 REAL LANGUAGE , REAL PEOPLE

15.3 Percentage of stereotypical gendered SFPs used in

each conversation
Kaato Osaka Kobe

Strongly masculine 9 (5.0%) 0 0

Moderately masculine 29 (17.0%) 9 (4.4%) 36 (13.0%)
Neutral 70 (40.0%) 101 (49.0%) 119 (43.3%)
Moderately feminine 13 (7.4%) 9 (4.4%) 8 (3.0%)
Strongly feminine 3 (1.7%) 0 6 (2.2%)

Total
Masculine 38 (22.0%) 9 (4.4%) 36 (13.0%)
Feminine 16 (9.1%) 9 (4.4%) 14 (5.2%)
Neutral 70 (40.0%) 101 (49.0%) 119 (43.3%)

line SFPs and moderately feminine forms (4.4% in each case, all in SJ). They make
use of neither the strongly masculine nor the strongly feminine SFPs.
Finally, looking at the Kobe speakers we find that they also use neutral forms
most frequently (43.3%); they use masculine forms 13% of the time (all moderately
masculine, with 10.5% in SJ and 2.5% in HKD) and feminine forms 5.2% of the time
(3% moderately feminine and 2.2% strongly feminine, all in SJ).
The speakers patterns of SFP use do not conform to one another; more crucially,
they do not correspond to those described in the literature for male speakers. Japa-
nese men, at least those in my data, are not taking advantage of SFPs as a place to
index their gender. Overwhelmingly, all speakers from each region use neutral SFPs
with a much higher frequency than they use moderately or strongly masculine ones.
Further, they occasionally use feminine particles.
While no speakers use strongly and moderately masculine forms frequently, the
Kanto speakers have a much higher usage (22%) than either the Osaka (4.4%) or the
Kobe speakers (13%). Further, gendered forms used by Osaka and Kobe speakers
are primarily SJ forms. These findings may be partially explained by the fact that the
ideology of gendered language has been more prominent in SJ than in other dialects.

15.4.2. Qualitative analysis

While it is true that these men do not use masculine SFPs the majority of the time, it
is important to investigate the contexts in which they are using these forms. The fol-
lowing examples illustrate how masculine SFPs are qualitatively used by both SJ
speakers and HKD speakers.5

(3) da yo iya nai nai, dakara komaru n da yo [CT, Mihara, #13]

Uh, no, no [it doesnt], thats why its a problem!

(4) da yo marukkiri onaji da yo [CT, Mihara, #43]

They are exactly the same.
 JAPANESE MEN S LINGUISTIC STEREOTYPES AND REALITIES 283

In examples (3) and (4), Mihara uses the stereotypically masculine sentence final
form da yo. The participants in the Tokyo conversation are discussing accents
(namari) in Japan. Shibata states that he is originally from Iwate Prefecture (in the
northeast of Japan) but that he does not find Iwate to be known for an accent nor
does he have an accent himself.6 Upon his saying this, Mihara responds with example
(3), a complaint that Shibatas lack of accent is a problem, and laughter by all speak-
ers follows. Mihara here seems to be using the masculine form da yo to express his
complaint in a direct and forceful manner. The complaint here is not a serious one; it
is a kind of banter only allowable when the appropriate degree of intimacy obtains.
Its humorously forceful expression is indexing friendship or solidarity rather than a
serious, forceful complaint.
In example (4), Mihara again uses da yo. The men are still discussing dialect
variation within Japan and how it correlates with gender differences in speech. Mihara
mentions that his wife is from Fukushima, where the regional dialect (according to
Mihara) does not exhibit gender differences. Mihara says that minna onaji yoo ni
shabetteru kedo everyone talks the same, to which Kawamura says dakedo onaji
ja naibut theyre not the same with falling intonation. Mihara responds with
example (4), confirming that, indeed, the language that women and men in the
Fukushima region speak is exactly the same. In this case, I suggest that Mihara is
using the masculine form da yo to make a strong disagreement, which in turn may
index his expert knowledge and authority, because he thinks he is most knowledge-
able about the Fukushima dialect, the dialect of his wife, and because he originally
established the topic of regional dialects and has contributed more to this topic than
the other two speakers. Consequently, when Kawamura opposes Miharas statement,
he linguistically points to himself as the expert by using the stereotypical masculine
form of authority: da yo.
I turn now to how HKD speakers use stereotypically masculine SFPs within the
context of their conversations. Examples 5 and 6 involve the forms ga na and kai.

(5) kai ee n kai na [CK, Sato, #82]

Thats good, isnt it.

(6) ga na ee kotcha ga naa [CK, Honda, #192]

Thats nice, isnt it? (sarcastically)

Examples (5) and (6) occur during a discussion of the free gifts that one is offered in
exchange for signing up for newspaper home delivery. Yamada begins the topic by
saying that recently someone came by his home offering the newspaper Yomiuri
Shimbun free for one year if he signed up for a three-year delivery contract. Sato re-
sponds with example (5). Yamada continues his topic by naming the items he has re-
ceived in the past (cookware and salad oil), noting that he has gotten over 10,000 yen
in goods from these deals. Nakayama joins in the conversation, saying that he is
unable to run out of dishwashing liquid even though he tells the sales representa-
tives not to bring it. At this point, Honda sarcastically utters (6), Thats nice, isnt it?
In these examples, Sato and Honda listen, give minimal responses, and offer some
kind of evaluation that utilizes stereotypical masculine final forms to index their
284 REAL LANGUAGE , REAL PEOPLE

sympathy and support for the two complaining men. The support is not to be con-
fused with nurturing; it is manly strong support, and this manliness is indexed
through the final forms and the dialect itself. The dialect bolsters the camaraderie
and friendship ties that the men have with one another.
In all cases, the men creatively and deftly use stereotypically masculine SFPs. They
are not using these final forms necessarily to jockey for position, establish a position
of authority, or show status; they are joking and giving support as well as showing
authority. It seems that the men are aware of the stereotypical functional values of these
final forms given their use of them, albeit infrequently, in an ideological fashion.
Although the men in these conversations do not seem to be relying heavily upon
SFPs as a resource for marking masculine linguistic behavior, this does not neces-
sarily mean that they are not or do not know how to be masculine. Other linguistic
sites need to be investigated to identify other potential places where men do gender
work, if, in fact, they do. What follows is a brief excerpt taken from a conversation
between two men from the southern part of Osaka. These men, like those whose
conversations are analyzed earlier, do not use SFPs as a place to mark their mascu-
line gender (Sturtz 2000). However, in example (7) they indicate that they are ca-
pable of using highly stylized stereotypical masculine linguistic features such as trilled
/r/ and rough command forms to enact masculinity. This excerpt underscores the need
for further investigation of other potential sites of gender work.
In this conversation, Ito and Kado are talking about how todays young people
have no manners and do not learn to say thank you, excuse me, I am sorry,
and other mannered formulas properly. They focus on the failure to say excuse me
when people collide with one another on foot or bicycle. At this point, Kado invites
Ito to join him in the performance of a bicycle accident. In the excerpt given here,
the two men display their ability to negotiate and manage conflict in a situationally
proper, manly mannerthat is, through vulgar, crude, and rough speech.7

(7) K: soohoo ga hashittemashite, gashan-tto yarimashita. taoremashita. <aita!>

I: <mazu> watashi wa omae nani shitton nen to yuu koto o mazu yaru
deshoo na.
K: soo deshoo na.
I: omae nani shitton nen to
K: ware maa, warui kotoba desu to warre [trilled r] nan ya sono hashirikata wa
t-to
I: sono hashirikata wa nan ya, to yuu koto de omae aho chau ka yuu kotoba ga,
mazu, dete kuru deshoo na, ee.
K: yarre [trilled r] sonna supiido dashiyagatte, sonna ni hashitorru kara ataru n
ya nai ka to
I: un, ataru n ya nai ka to yuu
K: mae mite hashire! doko muitotten?
I: un, soo yuu kotoba deshoo na.

English Gloss:
K: We both are riding our bikes, Gasshan! [sound of bikes colliding], we collide and
fall down. <aita!> [sound of shock/anger]
 JAPANESE MEN S LINGUISTIC STEREOTYPES AND REALITIES 285

I: <at the least> I would say, at the least, What the hell are you doing?!
K: Yeah, thats right, isnt it.
I: What the hell are you doing?! [I would say]
K: The word ware [you] is bad, but, Hey asshole, you drive too fast; what kind
of fucked-up bike riding is that [youre doing] [I would say]
I: Id say, What kind of bike riding is that?! and at least something like Youre
a fool/asshole! would probably come out [of my mouth].
K: If you speed that much and ride your bike, thats why you hit me!
I: uh-huh, Youll hit someone, youd say
K: Look in front of you when you ride! Where the hell were you looking?!
I: Yeah, thats the kind of words we would use, right.

In this conversation, Kado overtly invites Ito to participate in a role play of two bikes
colliding. Kado trills his /r/ sounds, which is not typical of the phonology of either
SJ or HKD but rather is stereotypically associated with working-class male speakers
or with TV gangsters. Moreover, prescriptively nongeminated forms are geminated
(for example, shitton doing). There appears potentially to be a correlation between
rude or gruff speech and hyper-geminated forms; further research across more
speakers is necessary to confirm this hypothesis. The boldface segments exhibit ste-
reotypical masculine speech style in their formsthat is, the verbs are marked nei-
ther for politeness nor for honorifics; in other words, the style is rude, coarse, and
quite rough. For example, in Mae mite hashire! Look in front [of you when you]
ride!, the verb hashiru ride (lit., run) is in the command form of hashire; there is
no mitigation of the on-record rebuke. The pronouns are stereotypically strongly
masculine forms; for example, the crude masculine second-person pronoun warre is
used. The overall style is unelaborated and very curt. In short, this metalinguistic
discussion is filled with boorish, vulgar, and rude utterances.
The conversation that precedes and follows the bicycle accident role play does
not make use of these extreme stereotypical masculine styles, but this metalinguistic
demonstration gives a clear indication that these men are capable of producing ste-
reotypical masculine styles, and of judging under what circumstances such styles are
appropriate and, indeed, effective.

15.5. Conclusion

The SFPs used and the way that they are used by the men in Kanto and HK are differ-
ent not only from one another but also from what would be expected from traditional
gendered classifications of SFPs (as seen in table 15.1). My findings challenge the idea
that there is a single danseego Japanese mens language, or (male) Standard Japa-
nese. None of the men in any of the conversations ever called upon the most marked
SFPs available to them, such as zo and ze. The moderately masculine form da yo was
used by the Kanto speakers only a few times and never by the HKD speakers. This
suggests that the traditional gendering of SFPs is ideological and that in reality (at least
in ordinary conversations) male speakers do not resort to the exclusive or frequent use
of masculine forms to express stereotypical (or old-fashioned) masculinity.
286 REAL LANGUAGE , REAL PEOPLE

It is clear, however, that men can use language to create and inhabit specific stances
(e.g., camaraderie, support, authority, anger, vulgarity, gangster-associated identity).
Examples (3) through (6) show that the men are able to use the stereotypically mas-
culine final particles both beyond and within their ideological uses. By using the forms
variously to enact a joking or an authoritative context, the men show how these final
forms can be used as creative as well as presupposing indexes (Silverstein 1976).
Furthermore, in example (7), Kado and Ito use rude and rough language in an imagi-
nary conversation to show how to be angry in appropriate situations. Although this
is hardly surprising, heretofore men (Japanese or otherwise) have rarely been awarded
recognition of this linguistic prowess.
Large differences were not observed across the Kanto and Kansai regions of
Japan. While it is true that Kanto speakers use more traditionally masculine final
forms, all speakers produce neutral forms (including the absence of an SFP) much
more than any gendered form available to them. The HKD speakers do favor dialect
final forms somewhat over equivalent SJ forms. It is not that the SJ forms are not
available to the HKD speakers (as is easily seen by the high use of other SJ forms),
but it is possible that for HKD speakers, SJ forms index something more than just
stereotypical masculinity. Using the SJ forms to create or display camaraderie, for
instance, may not be effective for HKD speakers.
This chapter has been a preliminary study only. Regionality still needs much
more attention (but see chapter 10, this volume). Class distinctions and identifica-
tion also demand much more consideration, for both female and male speakers. In
order to get an encompassing picture of what is going on with Japanese mens lin-
guistic practices, we must look across regions, class, genders, and ages to identify
and begin to sustain an understanding of how men and their dynamic identities are
arrayed across Japanese real space.

Notes
I would like to thank Janet S. Shibamoto Smith and Shigeko Okamoto for their careful and
valuable comments. This research was supported by the Kobe College Corporation Graduate
Fellowship and the National Science Foundation.

1. I add the plural languag[es] here to recognize the variation within the Kansai dialect.
2. Plath (1989), Skov and Moeran (1995), and Traphagan (2000) suggest that life stage
stratification is quite salient in Japanese society.
3. All of the names that appear in this chapter are pseudonyms.
4. The neutral total for the Osaka and Kobe speakers represents the combined total of
SJ and HKD. The Osaka speakers used 56 (27%) SJ forms and 45 (22%) HKD forms, total-
ing 49% combined; the Kobe speakers used 62 (23%) SJ forms and 57 (20.3%) HKD forms,
totaling 43.3% combined. All gendered SFPs produced by the HKD speakers from Osaka
were SJ, although Kobe speakers occasionally used an HKD-gendered SFP.
5. The information in brackets includes the conversation (C = Conversation, O/K/T =
Osaka/Kobe/Tokyo), the speakers name and the clause number from the conversation.
6. Although from Iwate Prefecture, Shibata has lived in Tokyo for several years and
exhibits Tokyo dialect throughout the conversation.
 JAPANESE MEN S LINGUISTIC STEREOTYPES AND REALITIES 287

7. Square brackets indicate authors note, angled brackets (< >) indicate overlap, of the
words within, and boldface text indicates metalinguistic speech under discussion. The trans-
lations are mine, done with help from male native speakers of HKD.

References
Coates, Jennifer (2003). Men talk: Stories in the making of masculinities. Oxford: Blackwell.
Davis, Darrell (1996). Picturing Japaneseness: Monumental style, national identity, Japa-
nese film. New York: Columbia University Press.
Donahue, Ray (1998). Japanese culture and communication: Critical cultural analysis. Lanham,
Md.: University Press of America.
Fujieda, Mioko (1995). Ehon ni miru onna (no ko) zoo, otoko (no ko) zoo (Images of women
(girls) and men (boys) in picture books). In T. Inoue, C. Ueno, and Y. Ehara (eds.),
Hyoogen to media (Expression and media), 173192. Tokyo: Iwanami Shoten.
Fujiwara, Yoichi (1982). Hoogen bunmatsushi bunmatsu joshi no kenkyuu (Research
on dialectal sentence-ending particles sentence-final particles), vol. 1. Tokyo:
Shunyodo.
(1985). Hoogen bunmatsushi bunmatsu joshi no kenkyuu, vol 2. Tokyo: Shunyodo.
(1986). Hoogen bunmatsushi bunmatsu joshi no kenkyuu, vol. 3. Tokyo: Shunyodo.
Gill, Tom (1999). Yoseba no otokotachi (Men of the yoseba). In Y. Nishikawa and M. Ogino
(eds.), Kyoodoo kenkyuu: Danseeron (Collaborative research on men), 1743. Kyoto:
Jinbun Shoin.
Hirayama, Teruo (1997). Osaka-fu no kotoba (Osaka Prefectures language). Tokyo: Meiji
Shoin.
Horiguchi, Sumiko (1991). Aizuchi kenkyuu no gendankai to kadai (The issues and stages of
aizuchi research). Nihongogaku (10): 3141.
Ide, Sachiko (1982). Japanese sociolinguistics: Politeness and womens language. Lingua 57:
357385.
(1993). Sekai no joseego, Nihon no joseego: Joseego kenkyuu no shintenkai o motomete
(Womens language of the world, womens language of Japan:Searching for new devel-
opments in womens language research). Nihongogaku 12(6): 412.
Inoue, Miyako (1996). The political economy of gender and language in Japan. Unpublished
Ph.D. dissertation, Washington University, St. Louis, MO.
Ito, Kimio (1996). Danseegaku nyuumon (Introduction to mens studies). Tokyo: Sakuhinsha.
Iwao, Sumiko (1993). The Japanese woman: Traditional image and changing reality. New
York: Free Press.
Johnson, Sally, and Ulrike Hanna Meinhof (eds). (1997). Language and masculinity. Oxford:
Blackwell.
Kamata, Ryoji (1979). Hyogoken hoogen bunpoo no kenkyuu (Research on the grammar of
Hyogo-ken dialect). Tokyo: Sakura Hosha.
Kanamaru, Fumi (1993). Ninshoo-daimeeshi, koshoo (Personal pronouns and address terms).
In S. Ide (ed.), Joseego no sekai (The world of womens language), 1532. Tokyo: Meiji
Shoin.
Kawashima, Sue (1999). A dictionary of Japanese particles. New York: Kodansha.
Kawaguchi, Yoko (1987). Majiriai danjo no kotoba: Jittai choosa ni yoru genjoo (The inter-
secting speech of men and women: The current situation as assessed by survey). Gengo
Seekatsu 429: 3439.
Kinmonth, Earl (1981). The self-made man in Meiji Japanese thought: From samurai to salary
man. Berkeley: University of California Press.
288 REAL LANGUAGE , REAL PEOPLE

Komatsu, Hisao (1988). Tokyogo ni okeru danjosa no keeshiki: Shuujoshi o chuushin to shite
(The form of gender differences in Tokyo dialect: Centering on the sentence-final par-
ticles). Kokugo to Kokubungaku 65(11): 94106.
Kunihiro, Tetsuya, Fumio Inoue, and Daniel Long (1999). Takesi Sibata: Sociolinguistics in
Japanese contexts. Berlin: Mouton de Gruyter.
Kurozaki, Yohei (1987). Danwashinkoojo no aizuchi no unyoo to kinoo (The function of
aizuchi to the advancement of conversation). Kokugogaku 150: 1528.
Loveday, Leo (1986). Explorations in Japanese sociolingusitics. Amsterdam: John Benjamins.
Makimura, Shiyo (1984). Osaka kotoba jiten (Dictionary of the Osaka language). Tokyo:
Kodansha.
McGloin, Naomi Hanaoka (1990). Sex difference and sentence-final particles. In S. Ide and
N. H. McGloin (eds.), Aspects of Japanese womens language, 2341. Tokyo: Kuroshio.
Miyake, Yoshimi (1995). A dialect in the face of the standard: A Japanese case study. In
J. Ahlers, L. Bilmes, J. S. Guenter, B. A. Kaiser, and J. Namkung (eds.), Proceedings of
the 21st Annual Meeting of the Berkeley Linguistics Society, 217225. Berkeley: Berke-
ley Linguistics Society.
Ogawa, Naoko, and Janet S. [Shibamoto] Smith (1997). The gendering of the gay male sex
class in Japan: A preliminary case study based on Rasen no Sobyoo. In A. Livia and
K. Hall (eds.), Queerly phrased: Language, gender, and sexuality, 402415. New York:
Oxford University Press.
Ohara, Yumiko (1992). Gender-dependent pitch levels: A comparative study in Japanese and
English. In K. Hall, M. Bucholtz, and B. Moonwomon (eds.), Locating power: Proceed-
ings of the Second Berkeley Women and Language Conference, 469477. Berkeley:
Berkeley Women and Language Group.
Okamoto, Shigeko (1995). Tasteless Japanese: Less feminine speech among young Japa-
nese women. In K. Hall and M. Bucholtz (eds.), Gender articulated: Language and the
socially constructed self, 297325. New York: Routledge.
(1996). Indexical meaning, linguistic ideology, and Japanese womens speech. In The
Proceedings of the Twenty-second Annual Meeting of the Berkeley Linguistics Society,
Berkeley: Berkeley Linguistics Society, 29030.
(1998). The use and non-use of honorifics in sales talk in Kyoto and Osaka: Are they
rude or friendly? Japanese/Korean Linguistics 7: 141157.
Okamoto, Shigeko, and Shie Sato (1992). Less feminine speech among young Japanese fe-
males. In K. Hall, M. Bucholtz, and B. Moonwomon (eds.), Locating power: Proceed-
ings of the Second Berkeley Women and Language Conference, 478488. Berkeley:
Berkeley Women and Language Group
Onoe, Keisuke, Seiji Kasai, and Kooji Wakaichi (2000). Interview: Osaka no kotoba, Osaka
no bunka (Osaka language, Osaka culture). Gengo 29(1): 1439.
Otani, Kooichi (1994). Osaka Gaku (The study of Osaka). Tokyo: Shinkosha.
Peng, Fei (2000). Gaikokujin kara mita Osaka kotoba (Osaka language from the perspective
of a foreigner). Gengo 29(1): 7379.
Plath, David (1989). Arc, circle, and sphere: Schedules for selfhood. In Y. Sugimoto and
R. Mouer (eds.), Constructs for understanding Japan, 6793. London: Kegan Paul
International.
Reynolds, Katsue Akiba (1985). Female speakers of Japanese. Feminist Issues 5: 1346.
Rosenberger, Nancy (1994). Indexing hierarchy through Japanese gender relations. In J. Bachnik
and C. Quinn (eds.), Situated meaning: Inside and oustide in Japanese self, society, and
language, 88112. Princeton, NJ: Princeton University Press.
Sato, Makoto (2000). Osakaben no seeshin (The spirit of Osaka dialect). Gengo 29(1): 6065.
 JAPANESE MEN S LINGUISTIC STEREOTYPES AND REALITIES 289

Seko, Koji (2000). Koo yuu otoko ni naritai (I want to be a man like that). Tokyo: Chikuma
Shobo.
Seward, Jack (1968). Japanese in action. New York: Weatherhill.
Shibamoto, Janet S. (1985). Japanese womens language. New York FL: Academic Press.
(1987). The womanly woman: Manipulation of stereotypical and non-stereotypical
features of Japanese female speech. In S. Philips, S. Steele, and C. Tanz (eds.), Lan-
guage, gender, and sex in comparative perspective, 2649. New York: Cambridge Uni-
versity Press.
Shigemitsu, Yuka. (1993). Kaiwa no pataan (Conversational patterns). Nihongogaku 12(6):
135140.
Silverstein, Michael (1976). Shifters, linguistic categories, and cultural description. In K. Basso
and H. Selby (eds.), Meaning in anthropology, 1156. Albuquerque: University of New
Mexico Press.
(1979). Language structure and linguistic ideology. In P. R. Clyne, W. Hanks, and
C. L. Hofbauer (eds.), The elements: A parasession on linguistic units and levels, 193
247. Chicago: Chicago Linguistic Society.
Skov, Lise, and Brian Moeran (1995). Introduction: Hiding in the light: From Oshin to
Yoshimoto Banana. In L. Skov and B. Moeran (eds.), Women, media, and consumption
in Japan, 174. Honolulu: University of Hawaii Press.
Sturtz, Cindi (2000). Being and becoming (Japanese) men at 67. Paper presented at the 99th
Annual Meeting of the American Anthropological Association, San Francisco, CA.
Sugimoto, Yoshio (1997). An introduction to Japanese society. Cambridge, MA: Cambridge
University Press.
Toyoda, Masayoshi (1997). Otoko ga otokorashisa o suteru toki (When men throw away
masculinity). Tokyo: Asuka Shinsha.
Traphagan, John (2000). Reproducing elder male power through ritual performance in Japan.
Journal of Cross-Cultural Gerontology 15: 8197.
Uchida, Nobuko (1993). Kaiwa-koodoo ni mirareru seesa (Sex differences seen in conversa-
tional patterns). Nihongogaku 12(6): 156168.
Vogel, Ezra (1979). Japan as No. 1: Lessons for America. Cambridge, MA: Harvard Univer-
sity Press.
Wada, Minoru, and Ryoji Kamata (1992). Hyogo no hoogen, rigen (Hyogos dialect and
slang). Kobe: Kobe Shimbun Sogo Shuppan Senta.
Yamada, Haru (1997). Different games, different rules: Why American and Japanese misun-
derstand each other. New York: Cambridge University Press.
Yamamoto, Toshiharu (1995). Osakaben no ruutsu o saguru. In NHK Osakaben Project (ed.),
Osakaben no sekai (The world of Osaka dialect: Searching for the roots of Osaka dia-
lect), 5463. Tokyo: Keidoo Shoin.
Yorifuji, Kazuhiro (1998). Ii otoko mitsuketa! (I found a good man!) Tokyo: Kaneko Shobo.
This page intentionally left blank
 INDEX

Abe, Hideko, 15, 46, 49, 94, 205218, Briggs, Charles L., 51
258 Bucholtz, Mary, 6, 24, 187, 258
Achilles, Nancy, 206 Butler, Judith, 93, 205
address terms, gender and the use of, 11, burakumin, 33, 179
1415, 120124, 155, 157 burikko phony girls, 12, 148162
adolescents, personal pronouns use and, assessments of burikko performances,
256270. See also age 155162
age features of burikko talk, 151155
age difference in speech, 46, 48, 52 media representations of burikko, 155
junior high school youth and gendered 160
speech style, 256270
middle-aged women, 240253 Cameron, Deborah, 150
age, regionality, and mens speech, 279 Cherry, Kittredge, 148
286 commercials/advertisements, language and
agency gender in, 2526
speakers as social agents, 17, 48, 105 communities
womens agency and speech, 27, 132, construction of imagined communities,
135, 223, 242, 252253 127, 6869, 142146
constraints on, 222223, 233235, 270 construction of (gendered) magazine
(see also norms) communities, 12, 131132, 137146
Agha, Asif, 258 communities of practice, local pragmatic
meanings in, 259
baby talk and cute femininity, 152 correlational studies on linguistic variation,
Barrett, Rusty, 206 48
Bergvall, Victoria L., 38 cultural constraints. See norms

291
292 INDEX

cultural ideologies. See ideologies entextualization, 61, 70

cultural models for Japanese women and essentialism
men, 5, 1112, 14, 113, 222223, in language and gender research, 4, 7,
240244, 248, 250, 275277. See also 2324, 26, 28
ideologies ethnographic studies
cuteness of language and gender, 7, 1516, 29,
cult of cuteness in Japan, 149 187200, 205218, 256270
display and assessment of cute exclamatory expressions
femininity, 148162 and construction of magazine
maru moji round characters and, 156157 communities, 137
young married women and, 252 gender and the use of, 4, 12, 26

Dan, Michiko, 82, 8586 Fairclough, Norman, 12

danseego mens language. See mens farm women
language farming families in Ibaraki, 194195
dialects. See regional dialects linguistic practices of, 27, 187200
dictionaries, sexism in, 13, 29, 172182 female personae, variability and, 241253
discourse analytic approach to language feminists
and gender, 7, 2930, 100105, 115 gender and feminist scholarship, 2831,
128, 137146, 194200, 208218, 150
242253, 266269, 280286 Japanese feminists on language issues,
discriminatory terms. See also sexism 8, 25, 97, 179181
racist expressions, 172 uuman ribu womens lib movement
diversity and language issues, 8, 25, 28
in mens speech (see mens speech) femininity. See also onnarashisa
in womens speech (see womens womanliness
speech) and class-status, 5153, 84, 87
in interpretations of (linguistic) behaviors, cute femininity (see cuteness)
9, 4953, 158159, 189, 233236 femininity training, 240241, 250 (see
(see also pragmatic meanings) also education of women)
linguistic and social diversity in Japan, 4, heterosexist femininity, 104105
1314, 33, 48, 63, 187 (see also mens hyperfeminine behavior, 123, 150,
speech and womens speech) 160
Donahue, Ray, 277 Japanese normative femininity, 24, 26,
Dunn, Cynthia, 45 115, 214, 223
male performance of, 95
Eckert, Penelope, 23, 38, 126, 251 media representation of. 1112, 113
education of women 128, 142145 (see also womens
education of women in the Edo Period, language)
43, 76 politeness and femininity, 5053, 87
education of women in the Meiji Period, (see also honorifics)
44, 60, 76 use of feminine speech patterns (see
education of women in the World War II womens language)
era, 7690 formality. See also honorifics
manner books for women, 4344, 77, expressions of, 16, 211212
159 (in)formal speech and femininity, 51
Reehoo yookoo Essentials of etiquette, Foucault, Michel, 58, 60, 67, 69
7780, 8788 fugen ladies speech, 86. See also
Endo, Orie, 13, 24, 29, 31, 43, 76, 94, 166 womens language
181 Futabatei, Shimei, 66
 INDEX 293

Gal, Susan, 23, 31 Harvey, Keith, 92

Gaudio, Rudolf P., 208 hegemonic ideologies. See ideologies
Gauntlett, Tsune, 82 heterosexism. See sexuality.
gay men. See also sexuality heterosexuality. See sexuality
gei, 9497, 99 Hirano, Hiroaki, 9697
gei baa, 95, 206207 Hollien, H., 151
hentai obscene/perverse/queer, 94 homogeneity
homo, 9495, 99 in Japanese society, 34
language use of, 10, 92100, 105106, 206 among Japanese women or men, 5, 28,
nyuuhaafu, 95, 208210 4748, 136
okama, 9496, 99, 106, 209 homosexuality. See gay men and sexuality
onee kotoba (see onee kotoba) Hongo, Akemi, 169
stereotypes of effeminate gay men, 94 honorifics
genbunitchi speech and writing gender and the use of, 4, 9, 14, 2425,
unification, 5966 3853, 62, 66, 79, 83, 87, 153, 193,
gender 197199, 249250
binary distinction of, 5, 31, 105, 132 addressee honorifics, 39, 4546, 193,
construction of gender identity (see 197199
identities) and class status (see social class)
feminist conception of, 132 and status difference, 47, 52, 197
gender differences in speech, 13, 2829, bika-go beautification words, 3940,
38, 42, 48, 57 (see also mens 44, 47, 153
language and womens language) honorific-less dialects, 192193, 201
gender differences in regional dialects, how-to-books on, 4142, 45, 50
193, 196200 in Lhasa Tibetan, 258259
gender ideologies (see ideology) kenjoo-go humble words, 39
gender in media, 11 (see also referent honorifics, 39, 4546, 193
commercials, magazines, newspapers, sonkee-go respectful words, 3940
novels, and romance fiction) hortatives
gender meanings, 23, 69 and the construction of (gendered)
gender norms (see norms) magazine communities, 12, 138, 143
gender roles and linguistic norms, 79, Hoshina, Koichi, 44, 8082, 8587
93, 162, 224 housewives
gender stratification/relations and language use of, 15, 240253
language, 43, 53, 79, 135, 145, 150
gendered language (use), 10, 12, 14, 30, Ibaraki dialect
44, 94, 102, 114115, 124, 132, 212, and female speakers, 14, 195200, 257
215, 222224, 240242, 244, 257, attitude toward, 190191
275276, 281282 features of, 191193
gendered speech norms (see norms) Ichikawa, Fusae, 8182
intragender variation in speech. See Ide, Sachiko, 2829, 38, 41, 193
mens speech and womens speech identities
Giddens, Anthony, 150 construction of, 6, 1415, 17, 29, 3132,
good wife and wise mother. See ryoosai kenbo 100, 106, 131132, 135137
Goodwin, Marjorie, 258 construction of gender identities, 12, 29,
Grice, Paul H., 252 32, 131, 113, 132, 135, 207, 235
construction of lesbian identities, 206,
Hall, Kira, 92 208218
Hanks, William, 59, 264 heterosexuality as, 93
Hanshinkan dialect. See Kansai dialect modern Japanese identity, 69
294 INDEX

identities (continued) Japanese female speech. See Japanese

social identity and language use, 29, 106, womens language
150 Japanese language and gender studies. See
subversive identities, 24 language and gender research
ideologies Japanese man(hood), 5, 28. See also
cultural ideologies in Japanese language cultural models and ideologies
and gender studies, 58, 2334, 124, Japanese mens language. See mens
150 language
distinguishing ideology and practice, 5, Japanese sociolinguistics. See
38, 53, 133, 136 sociolinguistics
gender and language ideologies, 1112, Japanese woman(hood), 5, 28, 58, 63, 67
25, 29, 3233, 128, 132, 135136, 68, 70. See also cultural models and
145, 148, 258 ideologies
hegemonic language and gender Japanese women
ideologies, 67, 26, 28, 3031, 38, 44, education of Japanese women (see
51, 53 education of women)
language/linguistic ideologies, 58, 11, language and diversity among (see
17, 2326, 29, 38, 49, 59, 84, 114 womens speech)
of heterosexuality and language, 128 (see modern Japanese women, 60, 63, 66, 69
also sexuality) 70
of mens language, 277, 284285 nationalizing Japanese women, 5960
of womens language, 7, 60, 84, 133, Japanese womens language, See womens
135 (see also the other entries under language
ideologies) jogakusee schoolgirls, 6667
imperatives and construction of (gendered) jogakusee kotoba the speech of
magazine communities, 12, 138, 143, schoolgirls, 66, 70
216 joseego womens language. See womens
indexicality, 6, 9, 14, 51, 61, 67, 71, 150 language
dialectic relationship between language Jugaku Akiko, 8, 23, 2627, 29, 50, 176
and social structure, 132, 135136 junior high school students
indexing gender, 70, 161162, 240, 244, language use of, 16, 256272
258
indirect speech acts, gender and, 34, 14, Kaibara, Ekken, 86
38, 41, 44, 51 Kanamaru, Fumi, 212
Inoue, Fumio, 149, 191, 201 Kansai dialect
Inoue, Hide, 82, 87 and male speakers, 16, 277278, 280286
Inoue, Kiyoshi, 8586 attitudes toward, 201, 278,
Inoue, Miyako, 9, 31, 5772, 94, 189, 222, kateeshoosetsu the domestic novel. See
258, 264, 276 novels
interpersonal relations. See also solidarity Kawanari, Mika, 41
construction of, 6, 1415, 17, 32, 211 Kikuchi, Kyo, 82
212, 215, 218 Kikuchi, Yuho, 68
interpretation of (linguistic) behavior. See Kikuzawa, Sueo, 58, 8485
diversity Kindaichi, Haruhiko, 193
Irvine, Judith T., 6, 31 Kindaichi, Kyosuke, 24, 58
Ishiguro, Yoshimi, 78, 82, 8586 Kinsella, Sharon, 149, 154
Ishimoto, Shizue, 8182 kokugo national language. See national
language
Japan/Japanese as ideological Kokugo Choosa Iinkai National Language
constructs, 24, 3132, 60, 69, 62 Research Council, 77
 INDEX 295

kokugogaku national language studies. lesbian bars, 15, 205209

See national language studies/scholars onabe, 208211
Kokugo Gakkai Society for the Study of rezu lesbian, 209211
Japanese Language, 78 stereotypes of masculine lesbians, 94
Kokugo Kyookai National Language linguistic ideologies. See ideologies
Association, 77 linguistic others, 188
Kokugo-shingikai a government council linguistic resources. See language as a resource
on the national language, 41, 78 linguistic variation and gender. See
Komatsu, Hisao, 31, 6465 diversity, mens speech, and womens
Komori, Yoichi, 33 speech
Kono, Kensuke, 33 Livia, Anna, 92
Konoe, Fumimaro, 7980, 89 Loveday, Leo, 151152, 235
Koyama, Shizuko, 89 Lunsing, Wim, 10, 92107
Kramarae, Cheris, 207
Kristof, Nicholas, 148 magazines
Kroskrity, Paul V., 5 construction of magazine communities
Kulick, Don, 114 (see communities)
Kyooiku Shingikai Education Council, 79 construction of magazine readers voice,
kyootsuugo common language, 77. See 138140
also Standard Japanese gender and language in magazines, 68
69, 1112, 2627, 68, 142147
Labov, William, 191 manners books for women. See education
Lakoff, Robin, 26, 28 of women
language masculinity
notion of, 60, 72 and male body, 160
as a resource, 6, 14, 16, 207, 213, 222 media and representation of, 1112,
223, 236238 113128, 142145 (see also mens
language ideologies (see ideologies) language)
language and gender research (normative/stereotypical) masculinity,
dynamic model of language and gender 115117, 277
studies, 132137 and linguistic practice, 9, 11, 16, 115,
Japanese language and gender studies, 7, 281286
2333, 113 use of masculine speech patterns (see
language and sexuality studies/queer mens language)
linguistics, 206 Maree, Claire, 10, 33, 92107
third wave language and gender research, Mashimo, Saburo, 58
23, 33, 258 Matsumoto, Yoshiko, 15, 94, 223, 240
language planning, 7778 253
language policy McConnell-Ginet, Sally, 23, 38, 126, 152,
and gendered language (see gender) 251
for women, 4, 78, 82 McGloin, Naomi H., 276
in the Meiji period, 5772, 7677 McVeigh, Brian, 149, 159
in World War II Era, 9, 44, 7690 media representations
Leap, William, 92 language and gender in, 1113, 30, 43
Lebra, Takie Sugiyama, 240241 (see also media representations of burikko (see
femininity training) burikko)
Lee, Yeonsuk, 33 media representations of femininity (see
lesbians. See also sexuality femininity and womens language)
language use of, 10, 15, 9294, 100107, media representations of heterosexuality
205218 (see femininity and sexuality )
296 INDEX

media representations (continued) Nakamura, Momoko, 12, 3031, 131146,

media representations of masculinity 201
(see masculinity) Nakamura, Yoichiro, 159
use of sexist terms in media, 169172, naming and identity. See identities
178181 narratives in Japanese writing, 6263, 66
Mendoza-Denton, Norma, 258 metadiscursive narratives, 101105
mens language 119, 124. See also metalinguistic functions of, 63, 66, 100
masculinity National Language Association, 10, 80
Japanese mens language as a cultural 88
construct/ideology, 4, 7, 16, 94, 115 and mass media, 8485
use of (normative/stereotypical) mens womens wing of, 8083
language, 15, 27, 95, 124, 206207, national language
212218, 276 Japanese (or kokugo), 3132, 44, 58, 61
representations of (normative/ 62, 7677, 86, 88
stereotypical) mens language in national language studies/scholars, 24, 27,
media, 115, 119, 121127, 142145 31, 58, 71, 78, 84, 88
mens speech nationalism
diversity in, 45, 1617, 4647, 52, 96 and language policy in Japan, 24, 44,
99, 126, 162, 226238, 263270, 275, 5961, 69
281286 Kokumin Seeshin Soodooin Chuuoo
linguistic prowess in, 16, 52, 236, 282 Renmee Central Federation for the
286 General Mobilization of the National
use of honorifics in, 46, 52 Spirit, 7984, 87, 89
metalinguistic discourse, 23, 93, 100105 Natsume, Soseki, 63
metapragmatic discourse, 1516, 31 negotiation
methodological issues. See sociolinguistics linguistic negotiation, 105
Miller, Laura, 12, 148162 of context, 100101, 105106
Minami, Hiroshi, 81 newspapers, language and gender in, 30
Ministry of Education, Japanese, 71, 76 Nihonjiron theories on the Japanese, 3
80, 85, 89 nonstandard dialects, speakers of, 28. See
Miwa, Akihiro, 97, 106 also regional dialects
Miyazaki, Ayumi, 16, 94, 97, 256270 norms
models of language and gender cultural norms and gender, 150, 158
dominance model, 23, 28 161, 174, 222224 (see also
dual-culture/difference model, 23 ideologies, cultural)
modernity, Japans, 5758 gendered language/speech norms, 7, 100,
modernization 115, 132, 135136, 187
and print capitalism, 5961, 67, 69 gender and sexuality norms, 92
linguistic modernization in Japan, 5859, gender norms, 10, 1416, 25, 38, 44, 92
6162 94
modern Japanese citizen, 63 heterosexist language norms, 92, 100101
modernizing Japanese women, 5960 linguistic norms, 45, 1415, 39, 41, 49
Moonwomon-Baird, Birch, 207 normative feminine speech as cultural
Morgan, Marcyliena, 258 constraints, 16, 50
Mulhern, Chieko Irie, 117118 novels
multiplicity of social/pragmatic meanings. and modernity, 59, 61, 65
See diversity, pragmatic meanings, gender and language in, 11, 66, 258
and social meanings kateeshoosetsu the domestic novel and
Muraoka, Hanako, 82 womens language, 6768, 189
 INDEX 297

romance novels (see romance fiction) power

teyo-dawa novels, 68 as a relational and processual term, 207
nyooboo kotoba court womens language, expressions of, 16
10, 24, 31, 44, 8384, 86, 89 hegemonic power, 23
masculine power, 213, 218
OBarr, William, 207 modern power, 60, 62
objectivism in linguistic research, 31 power dynamics, 257, 282284
occupation and language use, 46, 200 practice
Ochs, Elinor, 51, 161, 244, 258 discursive practice and identity
Ogawa, Naoko, 33, 92, 96, 201 construction, 6, 12, 135, 205
Ohara, Yumiko, 15, 32, 151152, 222237 distinction between ideology and
Okamoto, Shigeko, 317, 32, 3854, 94, linguistic practice, 7, 38, 53
126, 150, 188, 217, 222, 236, 257, semiotic practice and construction of
276 gender, 12, 113
onomatopoeic words and cute femininity, subversive linguistic/discourse practice,
153 8, 24, 2627, 132, 136
onee kotoba older sister speech, 9697, pragmatic meanings of (gendered)
99, 101 linguistic forms, 1416, 32, 61, 215
onnarashii kotoba feminine speech. See 218, 223236, 240253
femininity pronouns. See personal pronouns
onnarashisa womanliness, 24, 27, 5053, prosodic features,
83. See also femininity high-pitched voice and femininity, 4, 15,
Otsuma, Kotaka, 82, 87 27, 32, 148, 151152, 156, 158159,
Ozaki, Koyo, 67 161, 223224, 227, 232233, 237
high-pitched voice and politeness, 151,
Peak, Lois, 119 236
peer group relations nasality and cute femininity, 153
power dynamics in, 257, 282284 pragmatic meanings of voice-pitch
within a gakkyuu class, 259264 levels, 223235
performance
burikko performance (see burikko) queer. See sexuality
gender performance, 148, 150, 154,
205 Radway, Janice A., 115116
performative strategy (see strategy) Read, Kenneth, 206
subversive linguistic performance (see Reehoo yookoo Essentials of etiquette.
practice) See education of women
personal pronouns reference terms. See also personal pronouns
gender and the use of first-person gender and the use of, 30, 92, 119120,
pronouns, 16, 27, 33, 79, 83, 93, 96 196, 208211
107, 120124, 193, 211215, 256 gender and the use of self-reference
270 terms, 4, 10, 1415, 33, 93107, 120
gender and the use of second-person 124, 211215
pronouns/references, 26, 79, 120121, regional dialects
123124, 215216 gender and speakers of, 33, 5253, 187
politeness and gender, 3, 9, 1516, 28, 38 188, 200
53, 62, 87, 98, 197, 236238, 249 Hanshinkan dialect (see Kansai dialect)
250. See also femininity and hoogen bokumetsu undoo dialect
honorifics extermination movement, 188
Povinelli, Elizabeth, 114 Ibaraki dialect (see Ibaraki dialect)
298 INDEX

regional dialects (continued) sexist terms and the patrilineal family

Kansai dialect (see Kansai dialect) system, 167, 180
Kyoto dialect and honorifics, 192 sexist term onna woman, 176178
Reinoruzu [Reynolds]-Akiba, Katsue, 28, sexist term onna datera (ni) despite being
276 a woman, 167169, 173176, 181
reported speech, 9, 66, 68 sexist term oorudo misu old maid, 167
representations. See media representations 168, 171172, 177178, 181
romance fiction sexist term rooba hag/crone, 167168,
characterizations of hero/ines in, 115119 170171, 176177, 181
cultural codification of love in, 114, 127 sexuality. See also gay men and lesbians
language and gender in, 113128 conflation of gender and, 93
subcategories of Japanese romance expression of adult sexuality, 148, 156
fiction, 114115 heterosexism in society, 9295, 97, 205, 209
Rosenberger, Nancy, 11, 127, 241 heterosexism and language, 99100,
ryoosai kenbo good wife and wise 104105
mother, 31, 44, 60, 71, 76, 82, 8687, homophobia and language choice, 94, 99
89, 241, 250 homosexuality in Japan, 9394, 105
Rubin, Gayle S., 93 language, gender, and, 4, 17, 32, 51, 10
11, 60, 87, 92107, 113
Saito, Masami, 78, 2334 representations of heterosexuality, 113128
Sanshiro, 6366 Shalom, Celia, 92
Sato, Haruo, 66 Shibamoto Smith, Janet S., 317, 29, 32
Sato, Shie, 32, 257 33, 92, 96, 113128, 150, 152, 201
schoolgirls. See also jogakusee Shibatani, Masayoshi, 38
language practice of, 27 Shikitei Sanba, 63
Schieffelin, Bambi B, 38 Shimoda, Jiro, 86, 89
Schultz, Muriel, 207 Shimoonu, Fukayuki, 9697
script reform. See modernization Shimose, Kentaro, 81
second person references. See address Shuushin Morals. See education of
terms and personal pronouns women
Sedgwick, Eve Kosovsky, 93 Silverstein, Michael, 5, 38, 61
seekatsu kyoodootai, school community, Sino-Japanese characters, women and, 30
259 social class
self-reference terms. See reference terms and Standard Japanese, 44, 62, 77, 89,
sentence-final particles. See also verb- 187188
ending forms and womens speech, 44, 67, 8788, 159
and construction of (gendered) magazine honorifics and, 4044, 51, 53, 87
communities, 137, 142143 speech of middle(-to-upper)-class
gender and the use of, 4, 12, 1416, 24, women (in Tokyo), 13, 15, 28, 53, 66,
3233, 6468, 96, 119, 124127, 193, 68, 84, 89, 187189
196199, 201, 217218, 222, 240249 social context, language use and, 6, 93,
sex versus gender, 205 105106, 226232, 236237
sexism social deixis, 101
government responses to sexist language social factors, language use and interaction
practice, 179180 of gender and other. See diversity,
heterosexism (see sexuality) mens speech, and womens speech
sexism in linguistic expressions, 150 social meanings
sexism in Japanese language (practice), of linguistic forms, 6
8, 13, 25, 27, 29, 136, 148, 166181 multiplicity of, 15, 38
 INDEX 299

social personae, construction of, 6. See also Takahashi, Yoshiko, 46

identity Takasaki, Midori, 5051
sociolinguistics Tanaka, Kazuko, 29
Japanese sociolinguistics, 6, 28 Tanaka, Teru, 85
Japanese sociolinguistic phenomena, 56 Tsubouchi, Shoyo, 66
methodologies in Japanese Tokugawa, Yoshichika, 7980
sociolinguistics, 67, 29, 32, 38 transgendered groups 9395. See also gay
solidarity, expressions of 16, 49, 195199, men, lesbians, and sexuality
211212, 215, 218 transgendered language, 105
speech style shift. See style shifting and transgressive language, 105
contexts Treat, John, 149, 156
Spender, Dale, 28
stance Uchida, Nobuko, 276
delicate versus forceful stance, 242253 Uchida, Rohan, 68
linguistic forms and, 242 Ueda, Kazutoshi, 188
Standard Japanese Ueno Chizuko, 97, 205
and modernization of Japan, 33, 44, 60 Ukiyoburo, 6366
63, 7677, 187 (see also Uraji, D. K., 9899, 105106
modernization) Usami, Mayumi, 76
and Tokyo speech, 44, 62, 77, 188189 uuman ribu womens lib movement. See
and (normative/stereotypical) womens feminists
language, 9, 1315, 5253, 5772, 80,
88, 120, 136, 188189, 192 Van Bezooijen, 152, 224
stereotypes. See also cultural models Vanbaelen, Ruth, 92
stereotyped images of women and men, verb-ending forms
24, 26, 29 gender and the use of, 6162, 67, 83. See
stereotypes of language use, 94 also sentence-final particles
strategies voice
discourse strategies and gender, 14, 195 construction of magazine readers voice
199 (see magazines)
linguistic strategies and gender, 10, 14, variations in linguistic style and, 252
26, 32, 53, 9394, 100101, 104106, 253
152, 236 womens voice and language, 63, 66
performative strategies and gender, 6, 105 voice-pitch levels. See prosodic features
sexist discourse strategies, 30 Voloshinov, V. N., 66
style shifting and contexts, 213218, 226
232, 236237 Washi, Rumi, 9, 24, 3132, 44, 7690,
Sturtz Sreetharan, Cindi, 16, 126, 275286 94
subject positions and gender, 12, 113, 135 Watanabe, Mieko, 207
137, 145146 Wetzel, Patricia J. 41
Sugimoto, Tsutomu, 38, 76 womanliness. See onnarashisa and
Sunaoshi, Yukako, 14, 51, 94, 187200, femininity
257 womens language. See also femininity
Suzuki, Mutsumi, 38, 50 as a cultural construct/ideology/norm, 4
symbolic domination and language, 132. 5, 7, 9, 94, 115, 150, 187188, 200,
See also ideology 222223, 236237, 240
as a metapragmatic category, 69, 71
Takagi, Masayuki, 179 and Japanese culture, 24, 5758, 84,
Takahashi, Osamu, 33 136
300 INDEX

womens language (continued) womens speech

and Standard Japanese (see Standard diversity in, 45, 910, 1416, 27, 32,
Japanese) 38, 4547, 4953, 94, 101106, 126,
historical development of, 7, 9, 33, 57 132, 135136, 162, 187188, 195
72, 7690, 258 200, 211218, 222238, 240253,
ideology of (see ideology) 256265
use of (normative/stereotypical) womens studies in Japan, 2728
womens language, 810, 1317, 24 Woolard, Kathryn A, 5, 11, 38, 259260
28, 3133, 38, 52, 104, 257
representations of (normative/ Yamazawa, H., 151
stereotypical) womens language in Yanagi, Yae, 8586
media, 1113, 114115, 119127, Yonekawa, Akihiro, 149, 154
142145 (see also femininity and Yoshida, Sumio, 44, 8486
sexuality) Yoshioka, Yayoi, 82, 87
resistance to womens language as a Yukawa, Sumiyuki, 78, 2334
cultural norm, 241 yuujogo courtesans speech, 8384, 86