How Gender Shapes the World

How Gender Shapes the World

How Gender Shapes
the World
ALEXANDRA Y. AIKHENVALD
Language and Culture Research Centre
James Cook University
1

3
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Alexandra Y. Aikhenvald 2016
The moral rights of the author have been asserted
First Edition published in 2016
Impression: 1
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, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2016931923
ISBN 978–0–19–872375–2
Printed in Great Britain by
Clays Ltd, St Ives plc
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.

In loving memory of remarkable and independent women in whose
shadow I grew up—my great-aunt Frania S. Rosman who showed
what a woman can achieve against all odds, my grandmother Maria
S. Bonné who educated me concerning the nature of the status of
women, and my great-grandmother Nina K. Aikhenvald whose
indomitable spirit and strength inspired me.

Contents
Acknowledgements xi
Abbreviations and conventions xiii
List of boxes, figures, schemes, and tables
xv
1. The multifaceted Gender 1
1.1 Disentangling ‘Gender’ 1
1.2 What is special about Linguistic Gender 4
1.3 How this book is organized 6
1.4 The empirical basis, and a note on conventions 9
2. Linguistic Gender and its expression 13
2.1 Linguistic Gender in a nutshell 13
2.1.1 Marking Linguistic Gender 14
2.1.2 Gender agreement and anaphoric gender 15
2.1.3 Linguistic Gender and other linguistic categories 16
2.1.4 How many genders? 17
2.2 How to choose a Linguistic Gender 18
2.2.1 Gender choice by meaning 18
2.2.2 Gender choice by form 22
2.2.3 Meaning meets form: mixed principles of gender assignment 23
2.3 Markedness and Linguistic Genders 25
2.4 Linguistic Genders and their labels: envoi 29
3. Round women and long men: physical properties in Linguistic Gender 33
3.1 Small round women and long slender men 33
3.2 When ‘women’ are larger than ‘men’ 41
3.3 Physical properties in Linguistic Gender choice: contrasting
the two scenarios 43
3.4 Beyond mere physique: attitude, value, and importance
in Linguistic Gender 44
3.4.1 Endearment and disdain through Linguistic Gender 45
3.4.2 Value and importance in Linguistic Gender 47
4. What are Linguistic Genders good for? 52
4.1 Variable choice of Linguistic Gender 53
4.2 Linguistic Gender in discourse 57
4.3 The utility of Linguistic Gender 59

viii
Contents
5. Gender meanings in grammar and lexicon 63
5.1 Sex, humanness, and animacy in classifier systems 63
5.1.1 Numeral classifiers 64
5.1.2 Noun classifiers 65
5.1.3 Verbal classifiers and other classifier types 67
5.1.4 Linguistic Genders and classifiers as noun categorization devices:
commonalities and differences 68
5.2 Sex, humanness, and animacy in noun categories 69
5.3 Gender in gender-less languages 71
6. The rise and fall of Linguistic Genders 76
6.1 Developing Linguistic Gender 76
6.1.1 From generic nouns to Linguistic Genders 77
6.1.2 From generic nouns to noun classifiers and then to
Linguistic Genders 79
6.1.3 From anaphoric gender to agreement gender 81
6.1.4 From derivational gender to agreement gender 81
6.1.5 Linguistic Gender from other nominal categories 82
6.2 Reshaping Linguistic Genders 83
6.3 Partial loss and reinterpretation of gender: the story of English 85
6.4 Linguistic Gender in language contact 87
6.4.1 To lose a Linguistic Gender 87
6.4.2 Evolving a Linguistic Gender 88
6.4.3 Readjusting Linguistic Gender 90
6.4.4 Linguistic Gender in language obsolescence 91
6.5 Linguistic Gender in language acquisition and language
dissolution 92
6.6 Linguistic Gender and language reforms 93
6.7 Linguistic Genders, their development, demise,
and transformations 94
7. Manly women and womanly men: the effects of gender reversal 99
7.1 Reversing Linguistic Genders with jocular effects 100
7.2 Offence and praise in Linguistic Gender reversals 102
7.3 Linguistic Gender reversal: endearment and solidarity 105
7.4 Men as women, women as men: a summary of Linguistic
Gender reversals 108
7.5 Attitudes to Social Genders through Linguistic Gender reversals 109
7.6 ‘Women’ as a subtype of ‘men’? The overtones of masculine generics 112
7.7 Markedness, status, and power in Linguistic Gender choice 114

Contents
ix
8. The images of gender 120
8.1 Myth-and-belief in the choice of Linguistic Gender 120
8.2 The metaphors of Linguistic Gender 121
8.3 Does Linguistic Gender affect cognition? 126
8.4 What men and women look like 127
8.5 ‘Gendering’ the world: images, metaphors, and cognition 132
9. When women and men speak differently 136
9.1 ‘Male’ and ‘female’ dialects 136
9.1.1 How male and female dialects differ 138
9.1.2 Speakers and addressees of male and female dialects 141
9.1.3 Male and female dialects, and language history 142
9.1.4 Male and female speech on the way out 144
9.1.5 Beyond Natural Gender 145
9.1.6 Politeness, identity, and change: male and female dialects
in Japanese and Thai 148
9.2 Gender-variable skills: men’s and women’s speech practices 152
9.2.1 Social status, and women’s speech 152
9.2.2 Tokens of men and women in gender-variable languages 154
9.3 The other genders 158
9.4 Male speech, female speech: a summary 160
10. The rituals of gender 166
10.1 Social Gender, speech genres, and speech practices 166
10.2 The languages of manhood 168
10.3 Secrets, avoidance, and taboos: what women are not supposed
to know 169
10.4 Men, women, and language change 175
10.5 Language keepers or language killers? Women and language
maintenance 177
10.5.1 Women as language keepers 177
10.5.2 Women leading language shift 177
10.5.3 Women and modernity 178
10.5.4 Women as language killers 179
10.6 Summary: Social Gender through rituals, genres,
and speech practices 181
11. Gender in grammar and society 185
11.1 What Linguistic Gender can tell us about Social Gender 185
11.2 Social inequalities through gender asymmetries 186
11.3 The value of ‘man’ through gender in lexicon 190
11.4 How Linguistic Genders reflect social change 191

x
Contents
11.5 Thwarting ‘sexist language’ 194
11.5.1 Masculine bias through Linguistic Gender: pronouns
and agreement 195
11.5.2 Fighting the ‘generic masculine’ throughout the language 201
11.5.3 Bias in address terms and naming patterns 204
11.6 Expression of Linguistic Gender and social change: a summary 206
12. The heart of the matter: envoi 212
References 218
Index of languages, linguistic families, and peoples 251
Index of authors 259
Index of subjects 267

Acknowledgements
I have been working and publishing on gender for over two decades now, and am
indebted to many people, of different continents, backgrounds, and walks of life. My
gratitude goes to native speakers of Amazonian languages who taught me their
remarkable languages, especially my Tariana family—José, Jovino, Olívia, Rafael,
Leo, Maria, Diká, Emílio, Juvenal, the late Graciliano, Ismael, and Cândido Brito,
together with the Muniz family, and Afonso Fontes, Ilda Cardoso, and the late
Marcília Rodrigues from whom I learnt Baniwa.
I am immensely grateful to my adopted family in the Manambu-speaking Avatip
village (East Sepik Province, Papua New Guinea)—especially Jacklyn Yuamali Ala,
Pauline Yuaneng Agnes Luma Laki, James Sesu Laki, Dameliway, Jenny Kudapakw,
Motuway, the late Gaialiwag, Yuawalup, and John Sepaywus Angi. Special thanks go
to my new family in the Yalaku village of the East Sepik Province, especially Joel
Ukaia and his wife Rethi, Ritha Saun, Elsa Kasandemi, yafa Mark, yafa Solomon, and
David Kwaibori. Angela Filer, a Kwoma speaker from the East Sepik Province, was
the first one to draw my attention to the ways in which Papuan women lose their
identity by having to take their husband’s name.
I am grateful to my friends who taught me Estonian—and from whom I learnt that
speaking a language without a Linguistic Gender is not as boring as it may seem—Aet
Lees, Reet Bergman, Reet Vallak, and Krista Gardener in Australia, and Sana
Valliulina, together with the late Elsa Endemann, Maimu Endemann, and Lembit
Oiari back in Estonia. I owe bursts of revelation concerning Linguistic Gender in
Modern Hebrew to my dear cousin Lia Shaked (née Rosman). A debt of love and
gratitude goes to Emma Aikhenvald (née Breger), Tata Baeva, Ella Vainermann, and
indefatigable Tamara Margolina, for patiently answering my linguistic and other
questions concerning Russian as she is spoken today.
I am grateful for the support and comments of many colleagues, students, and
friends who allowed me to learn from their work, provided references, additional
sources, and patiently answered my questions concerning their areas of expertise—
especially Willem Adelaar, Angeliki Alvanoudi, Azeb Amha, Juliane Böttger, Nancy
Dorian, Sebastian Drude, Luke Fleming, Diana Forker, Valérie Guérin, Jenia
Gutova, Emi Ireland, Olga Kazakevitch, Pablo Kirtchuk, John Koontz, Maarten
Kossmann, Iwona Kraska-Szlenk, Maxim Kronhaus, Aet Lees, Mike Lu, Silvia
Luraghi, Lev Michael, Elena Mihas, Marianne Mithun, Edith Moravcsik, Heiko
Narrog, Simon Overall, Chia-jung Pan, Nick Piper, Vera Podlesskaya, Natasha
Pokrovsky, Renate Raffelsiefen, Nick Reid, David Rood, Hannah Sarvasy, Dineke
Schokkin, Glenn Shepard, Cácio and Elisângela Silva, Anne Storch, Marilena

xii
Acknowledgements
Thanassoula, Yishai Tobin, Rosa Vallejos, Louise Vasvári, Mary Ruth Wise, Kasia
Wojtylak, Sihong Zhang, and the late Bob Rankin.
Invaluable comments on just about every page came from R. M. W. Dixon,
without whose incisive criticism and ideas, and constant encouragement and support
this book would not have appeared. Special thanks go to Angeliki Alvanoudi for her
comments on the manuscript, to Nerida Jarkey for looking over the discusssion of
Japanese, to Sebastian Drude for materials and comments on Awetí, and to Rosa
Vallejos for materials and comments on Kokama-Kokamilla.
While working on this book, I made extensive use of the Oxford English Dictionary
Online—a wonderful and comprehensive resource on English and its history. I owe a
debt of gratitude to JCU Library who provided us with this resource, and especially to
Bronwyn Forster and Caroline Tredrea. The efficient interlibrary loan system worked
like clockwork—particular thanks go to Lynn Clarke, Janine Meakins, and Bridie
Soo, also at JCU Library. I owe a considerable debt to Brigitta Flick and Jolene Overall
for carefully reading through drafts of this book and making corrections, and to
Amanda Parsonage for looking after all things administrative while I was doing the
writing. This book was supported by the Australian Laureate Fellowship (from the
Australian Research Council) ‘How gender shapes the world: a linguistic perspective’.
This volume would have never been brought to fruition without encouragement
from Julia Steer, the Linguistics Editor of Oxford University Press. Her constant
support makes the whole enterprise worthwhile.

Abbreviations and conventions
1 first person (I)
2 second person (you)
3 third person (she, he, it, they)
A
transitive subject
ABS absolutive
ACC accusative
ADJ adjective
ANIM animate
ART article
AUX auxiliary
CL
classifier
dim
diminutive
DOM differential object marking
ERG ergative
fem, FEM feminine
GEN genitive
IMPERS impersonal
IMPV imperative
INANIM inanimate
inanim inanimate
INDEF indefinite
LK
linker
LOC locative
masc, MASC masculine
NCL noun class
NEG negation
NOM nominalization
NUM.CL numeral classifier
O
object
pl
plural
POSS possessive

xiv
Abbreviations and conventions
PRES
S
sg
VERT
VIS
present
intransitive subject
singular
vertical
visual
Numbers of examples, tables, boxes, figures, and schemes consist of the chapter number and
then are numbered consecutively. So, Table 6.1 is the first table in Chapter 6. The majority of
examples from different languages are glossed and then translated into English. I keep the
original orthography and also the glossing of the quoted sources.

List of boxes, figures, schemes,
and tables
Box 2.1 General properties of linguistic genders 14
Box 6.1 How to gain a linguistic gender: pathways of development 76
Figure 3.1 Gender assignment in Manambu 36
Figure 10.1 Elvia, a Tucano woman married to a Tariana man,
was supposed to ‘kill’ his language. Here she is doing a traditional woman’s
task: serving manioc beer to Tariana men at an Assembly of the Tariana
in the Tariana school (Aikhenvald 2013b) 180
Scheme 6.1 Gender and number in Zande personal pronouns 77
Scheme 12.1 How the three faces of Gender relate to each other 213
Table 2.1 Genders in Romanian and their marking 17
Table 2.2 Gender–number pairs in Bantu 21
Table 2.3 Semantic basis of gender choice in German: an illustration 24
Table 3.1 Semantic features in gender choice for nouns in Cantabrian Spanish 43
Table 3.2 Physical properties in Linguistic Gender choice 44
Table 3.3 Endearment and disdain through Linguistic Genders 47
Table 4.1 Variable Gender assignment in Maung 55
Table 5.1 Animate classifiers in Burmese 64
Table 5.2 Noun classifiers for humans and deities in Jacaltec (a selection) 66
Table 5.3 Mescalero Apache classificatory verb categories 67
Table 6.1 Noun classifiers for humans in Mam and nouns they originated from 78
Table 7.1 Personal pronouns in Gala compared with Manambu 99
Table 7.2 Treating ‘men’ as ‘women’: masculine to feminine Linguistic
Gender reversal 108
Table 7.3 Treating ‘women’ as ‘men’: feminine to masculine Linguistic
Gender reversal 109
Table 7.4 Meanings of ‘she’ vs ‘he’ in American English (Mathiot 1979b) 110
Table 9.1 Male and female forms of a selection of enclitics in Lakhota
(Trechter 1995: 57) 137
Table 9.2 Female versus male forms in Kokama-Kokamilla: a selection
(Vallejos 2010: 42) 140
Table 9.3 Female versus male forms in Awetí: a selection 140
Table 9.4 Male and female speech determined by speaker and addressee in Biloxi 141
Table 9.5 Male and female speech in Kũr̩ux: present tense of the verb bar- ‘come’ 142

xvi
List of boxes, figures, schemes, and tables
Table 9.6 Male and female speech chosen by the sex of the addressee in Tunica 142
Table 9.7 Personal pronouns in Japanese: men’s and women’s speech
(adapted from Ide 1991: 73) 148
Table 9.8 Traits of men’s and women’s speech among the Malagasy 154
Table 10.1 ‘Hidden from women’: tabooed nouns employed in traditional Tariana 171

1
The multifaceted Gender
The multifaceted notion of ‘gender’ pervades every aspect of life and of living. Gender
differences form the basis for family life, patterns of socialization, distribution of tasks,
spheres of responsibility, and occupational predilections. Understanding the nature
of ‘gender’ is central to many disciplines—anthropology, sociology (and, of course,
women’s studies), criminology, linguistics, and biology, to name a few. The way
gender is articulated shapes the world of individuals, and of the societies they live in.
To different people, the word ‘gender’ means different things. For a grammarian
and a linguist concerned with the structure of languages, ‘gender’ is a linguistic way of
categorizing nouns reflected in their form, the form of an adjective or a verb which
would agree with the noun, or a personal pronoun.
For a sociolinguist, a psychologist, and an anthropologist, ‘gender’ is a set of
norms, attitudes, feelings, and behaviours that a given culture or society associates
with the person’s biological sex (male or female). A philosopher defines gender as
‘social construction of male/female identity’ distinct from ‘sex, the biologically-based
distinction between men and women’. 1 Gender is also defined as a set of ideas about
relations and behaviours, and principles of social organization, to be understood
within a social context. 2 For some, ‘gender’ reflects a social and cultural elaboration
of sex differences, ‘a process that restricts our social roles, opportunities, and expectations’,
and also determines some ways in which we speak. 3 And when we fill in a
customs declaration, we need to state which gender we belong to—male or female.
That is, in day-to-day usage, the term ‘gender’ has expanded at the expense of ‘sex’:
then gender is a physiological distinction between men and women.
How to reconcile all the different meanings packaged into one word?
1.1 Disentangling ‘Gender’
The multifaceted concept of Gender has three faces.
• LINGUISTIC GENDER. This is the original sense of ‘gender’ as a linguistic term. One
class of nouns may be marked in a particular way, another class in another way.
That class which includes most words referring to females is called ‘feminine’,
How Gender Shapes the World. First edition. Alexandra Y. Aikhenvald.
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.

2 1 The multifaceted Gender
similarly for males and ‘masculine’. Gender classes are defined by their male and
female members but may extend beyond those. The ways in which animals,
birds, insects, plants, and natural phenomena (such as thunder and wind) are
assigned to genders may reflect their roles in legends and metaphors, and reveal
folk taxonomies. Linguistic Gender is integrated in the grammar, and is one of
the ways languages use to categorize nouns.
• NATURAL GENDER. This is what was until recently simply called ‘sex’—male versus
female. A female is able to bear children, a male is not. Natural Gender entails
anatomical and hormonal differences, linked to concomitant physiological and
psychological traits. In the day-to-day use, ‘gender’ has just about displaced the
term ‘sex’—perhaps felt to be too blunt and rude.
• SOCIAL GENDER. This reflects the social implications, and norms, of being a man
or a woman (or perhaps something in between). In Simone de Beauvoir’s(1949:
267) adage: ‘One is not born, but rather becomes, a woman.’ In many traditional
societies of New Guinea, social manhood is achieved, and defined, through male
initiation. Similarly, in other traditional societies, social womanhood used to be
achieved through female initiation. Social Gender relates to contrasting social
roles of the sexes, and how these are embodied in cultural practices and public
ritualized behaviour. (These patterns include conventions for the behaviour of men
and women, known as ‘gender etiquette’, social stereotypes associated with males
and females, and a traditional complex of knowledge and beliefs about mythical
women and mythical men. 4 )
The three faces of Gender interact. Investigations of Natural Gender focus on
innate biological differences between men and women. They are also played out in
the ways men and women communicate, within their Social Genders. As Labov
(1972: 304) puts it, ‘the sexual differentiation of speakers is . . . not a product of
physical factors alone’, but ‘rather an expressive posture which is socially more
appropriate for one sex or another’. In a ground-breaking study of physical features
of ‘women’s’ speech among the Tohono O'odham (a Uto-Aztecan group from
Mexico), Hill and Zepeda (1999) show how women (not men) use a pulmonic
ingressive airstream in order to construct a special atmosphere of conversational
intimacy, taking advantage of size differences between male and female vocal tracts.
Such sound production is easier to achieve with the smaller female larynx and
pharynx. Physical attributes—including high pitched voice—typical of female Natural
Gender come to be associated with ‘female talk’, and redeployed as tokens of
Social Gender and associated attitudes.
The difference between Natural Gender, or sex, and Linguistic Gender was captured
by Jespersen (1949: 174):
Sex is a natural quality shown primarily in the productive organs . . . Gender is a grammatical
category. Many languages have class distinctions of different characters. Gender in primary

1.1 Disentangling ‘Gender’ 3
words (chiefly substantives [nouns]) is not always shown by the form of the word itself, but it
may influence the form of other words (secondary) and is thus chiefly a syntactic category.
Languages vary very much with regard to the number of classes distinguished, also with regard
to the correspondence forums between these grammatical classes and natural distinctions such
as those of sex, between big and small, between living and lifeless, etc. Gender thus cannot be
defined as the grammatical expression of sex, but may relate to other things.
Natural Gender and Social Gender work together creating stereotypes of behaviour
in each society and culture. Ortner and Whitehead (1981: 1) put this as follows:
Natural features of gender, and natural processes of sex and reproduction, furnish only a
suggestive and ambiguous backdrop to the cultural organization of gender and sexuality. What
gender, what men and women are, what sorts of relations do or should obtain between them—
all of these notions do not simply reflect or elaborate upon biological ‘givens’, but are largely
products of social and cultural processes. 5
The division of biological, or ‘Natural’ Genders, goes beyond a male and female
dichotomy. Numerous traditional and modern societies have ‘groups whose gender
identities and enactments fall outside of sociocultural norms for women and men’—
these are the ones described as ‘a third sex’, ora‘third (Natural) Gender’. 6 ‘Transgender’
is another umbrella term which encompasses those whose behaviour shows
patterns associated with the opposite sex—including transsexuals, transvestites, and
drag queens and kings. Transsexuals are those whose Social Gender identity does not
correspond to the male or female Natural Gender characteristics they were born with.
Some undergo sex-reassignment surgery to change their biological features, so that it
should match their gender identity. Transvestites are men and women in terms of
their Natural Gender who dress and behave as members of the opposite gender. The
way gays, lesbians, and transgender people speak reflects their identity as special
groups, and highlights linguistic features perceived as characteristic of being ‘male’ or
‘female’. One of these features is manipulating Linguistic Gender. Brigitte Martel, a
transsexual male who became female, aptly captured this in the title of her autobiography
by changing Linguistic Genders which accompanied the change of her Natural
and Social Gender from male to female: Né homme, comment je suis devenue femme—
‘Born (masculine) a man, how I became (feminine) woman’. Hijras—womanly men in
India and Nepal—talk about themselves using feminine or masculine Linguistic
Gender depending on circumstances and attitudes. 7
Traditionally, transgender practices play a role in initiation and other rituals across
the world. These include Naven, made famous by Gregory Bateson, and further
explored in the literature on Sepik cultures, and across New Guinea. 8 Cross-dressing
of men as women during initiation ceremonies in the Sepik area and the Highlands of
New Guinea is thought to be a way of getting initiates to acquire a proper male social
identity—or masculine Social Gender—and rationalize sexual roles (usually characterized
by male dominance). Male and female transgender people have been documented

4 1 The multifaceted Gender
for numerous groups of the North American Plains. Just a few correlations with
language have so far been recorded for these cultural practices. 9
From a linguist’s perspective, Linguistic Gender occupies a central position in
shaping the role and the meanings of Gender in its three faces. Linguistic Gender as
a way of categorizing entities through language—and other realizations of Social
and Natural Gender in language—shapes the world we live in and the ways we
perceive and construct it. Other aspects of language use also set women and men apart.
These include ways of speaking, speech genres and speech practices, and often politeness
forms.
The multifaceted concept of ‘gender’ spans a linguistic category, a complex of
social norms, and a set of biological features. This book is about the ways in which
gender is reflected in language—and more specifically, the role of Linguistic Gender
in the expression of Social Gender and Natural Gender, their manipulations and
development.
1.2 What is special about Linguistic Gender
Some linguistic categories show strong correlations with cultural values, social
hierarchies, and their conceptualization. Imperatives and commands reflect relationships
between people: for instance, if a speaker of Dolakha Newar is considerably
younger than the addressee, or is talking to someone they particularly respect, they
will use special honorific imperatives. Simple imperatives will be reserved for their
equals. 10 Meanings encoded within possessive structures often reflect relationships
within a society, and change if the society changes.
Together with other ways of categorizing nouns, Linguistic Gender tends to mirror
social and cultural stereotypes and patterns of human perception. Linguistic Gender
is a repository of beliefs about what men and women are like and how they behave, and
features which are ‘male’-like or ‘female’-like. 11 Language planning, political correctness,
and societal changes shape various aspects of Linguistic Gender—especially with
regard to how humans are categorized. The ways in which people use Linguistic
Gender may mirror the status of Social Genders. For instance, in Jarawara, a small
Arawá language from southern Amazonia, a particularly respected woman can be
referred to with masculine gender, as if she were being ‘promoted’ to the male gender
status. Genders reflect the history of ideas and attitudes. The recent trend against the
generic masculine pronoun in English reflects the ways in which established stereotypes
can be gradually remoulded.
More than half the world’s languages have Linguistic Gender in their grammar. 12
As Franz Boas (a founding father of modern linguistics) put it, languages differ not in
what one can say but in what kind of information must be stated: ‘grammar . . .
determines those aspects of each experience that must be expressed’ (Boas 1938:
132). Having to be always conscious of which Linguistic Gender to apply—especially

1.2 What is special about Linguistic Gender 5
to humans—makes one alert to other faces of Gender, as a natural division of people
into male, female, and perhaps ‘other’, and as a cultural and social construct.
Linguistic Gender comes in many guises and serves many masters. It helps follow
the thread of communication, and figure out who or what is being talked about. Having
an obligatory Linguistic Gender allows for rich and expansive imagery, and makes the
language we speak more colourful and versatile. Linguistic gender, its choice and
associations, is something speakers are ready to discuss and argue about. This metalinguistic
perception of gender makes it central for metaphors—especially in poetry.
No term in linguistics is fully straightforward; there is some approximation in each.
Of all linguistic terms, ‘gender’ is uniquely polysemous in its coverage—it subsumes
‘gender’ in a strict grammatical meaning, and extends to a biological division of
humans into males and females, and to conventionalized differences in their social
status and linguistic behaviour. The diffuse nature of the term is sometimes
irritating—as every mention of the term has to be unpacked and explained. Turkish
can be considered ‘gender-less’—in terms of the absence of grammatical Linguistic
Gender. Yet meanings of Linguistic Gender can be expressed through some derivational
affixes on nouns, e.g. hoc-anIm ‘female teacher’. And the language is not
‘gender-neutral’ in the sense that Social Genders find their expression through other
means. A nurse is likely to be a woman, and a taxi driver a man. 13
On the other hand, the advantage of having a general term encompassing every
aspect of gender classification—linguistic, biological, and social—helps bring the
three together, and highlight their commonalities and the ways in which they may
influence each other. The ambiguity of the term ‘Gender’ alerts us to the existence of
an overarching concept behind it, spanning linguistic expression, social aspects, and
biological features. Linguistic Gender in its various guises and the expression of
Gender in language reflect and shape Social Gender stereotypes, associations, and
attitudes, in their relationship to Natural—or biological—Gender distinctions. 14 This
is what this book is about.
Linguistic Gender is a way of categorizing nouns. It always involves universal
features of sex, humanness, and animacy, and is a window into social life and
cognitive patterns. But Linguistic Gender classification goes beyond male and female.
As we will see throughout Chapter 2 (particularly in §2.4), Linguistic Genders are
also used to categorize inanimate entities. This is where we might find that a gender
labelled ‘feminine’ would include more than just females, and the one labelled
‘masculine’ more than just males. The term ‘neuter’ tends to refer to a gender
which includes inanimate (or irrational) beings, or a residue gender whose semantic
basis is difficult to capture. The choice of Linguistic Gender can be based on clues
other than just the meaning of a noun. Feminine and masculine genders often
include inanimate nouns with no connection to female or male sex, e.g. French
maison ‘house’ (feminine), château ‘castle’ (masculine).

6 1 The multifaceted Gender
In Antoine Meillet’s (1964: 164) words, gender provides an example of ‘a grammatical
category that plays (in a good many of the modern Indo-European languages)
a considerable role in morphology without answering, most of the time, to a
definite meaning’—especially where inanimate nouns are concerned.
The partial semantic opacity of Linguistic Gender has earned it a bad reputation,
among some linguists and lay people who complain that Linguistic Gender is hard
to learn. Jespersen (1972: 108) praised Modern English for losing complex agreement
in Linguistic Gender—as found in Old English—and making things easier and more
straightforward:
In Old English, as in all the old cognate languages, each substantive [noun], no matter whether
it referred to animate beings or thing or abstract notions, belonged to one or other of the three
gender-classes. Thus masculine pronouns and endings were found with names of a great many
things which had nothing to do with male sex (e.g. horn, ende ‘end’, ebba ‘ebb’, dæg ‘day’) and
similarly feminine pronouns and endings with many words without any relation to female sex
(e.g. sorh ‘sorrow’, glof ‘glove’, plume ‘plum’, pipe). Anyone acquainted with the intricacies of
the same system (or want of system) in German will feel how much English has gained in
clearness and simplicity by giving up these distinctions and applying he only to male, and she
only to female, living beings. The distinction between animate and inanimate now is much
more accentuated than it used to be.
Throughout this book we will see how important Linguistic Gender is for many
aspects of human communication. And Linguistic Gender in English is not as simple
and clear as it may seem. The ways in which the use of pronouns, especially the
generic he, have changed in recent years reflect social developments and the changing
status of Social Genders across the English-speaking world.
What is so special about the Linguistic Gender? How does it interact with Social
Gender and Natural Gender, across languages and cultures? What makes it a useful
linguistic resource rather than an encumbrance for poor language learners? And how
does the integrated complex of resources of Natural Gender, Linguistic Gender, and
the evolving images of Social Gender play out in view of societal changes? How can
Linguistic Gender and the semantic composition of categories related to ‘male’ and
‘female’ undergo restructuring in language planning? These are the questions we
approach in the present study.
Throughout the book, I have chosen to capitalize Linguistic Gender, Natural
Gender, and Social Gender—to stress the fact that all of these are ultimately just
nicknames which only partly capture the concepts and categories they cover.
1.3 How this book is organized
We start, in Chapter 2, with Linguistic Gender and its expression. Many languages
of the world have a gender system in their grammar. The size of the system varies.

1.3 How this book is organized 7
There are two genders in French, three in German, four in Dyirbal (from North
Queensland in Australia), more elsewhere. We seldom find an exact correspondence
between masculine/feminine (Linguistic Gender) and male/female sex (Natural
Gender). In German most nouns referring to females are feminine but Mädchen
‘girl’ is in neuter gender (because it contains the diminutive suffix -chen which is
always neuter). Linguistic Gender may span grammar and lexicon; vide he-man, tomboy
in English. Gender may be distinguished in personal pronouns only, as in
English. This chapter covers the formal properties of Linguistic Genders, gender
agreement, and anaphoric gender, and the ways in which Linguistic Gender interacts
with other linguistic categories. We then focus on different principles of Linguistic
Gender choice—by meaning and also by form—and look at the problem of markedness
in Linguistic Genders.
Linguistic Genders always include semantic parameters of animacy, humanness,
and sex, or Natural Gender. In a number of languages, the choice of Linguistic
Genders—especially for inanimate entities—is based on their shape and size. The
meanings of Linguistic Genders may involve value and importance—reflecting
associations with, and stereotypes of, Social Genders. This is the topic of Chapter 3,
‘Round women and long men: physical properties in Linguistic Gender’.
Linguistic Genders have a plethora of functions—they help highlight different
meanings of the same noun, track referents in discourse, and are a source of elaborate
metaphors. In a number of languages a noun can be assigned to more than one
Linguistic Gender with a change in meaning: these underscore the versatility, and the
utility of Linguistic Genders as a means of classifying entities of the world, debunking
the myth of gender as an arbitrary and redundant category. We discuss these in
Chapter 4, ‘What are Linguistic Genders good for?’
Meanings associated with Linguistic Genders—animacy, humanness, and sex—
can be expressed through a variety of other means. These include noun categorization
devices, or classifiers, and many noun categories, including case and number.
So-called ‘gender-less’ languages have ways of expressing gender meanings, through
using different words for males and females, or different affixes to distinguish sexes.
Attitudes to Social Genders—often downplaying the status of women—come to light
through the use of terms and forms in ‘gender-less’ languages. Chapter 5, ‘Gender
meanings in grammar and lexicon’, addresses these issues.
In Chapter 6, ‘The rise and fall of Linguistic Genders’, we turn to where Linguistic
Genders come from, how they may get restructured over time, and how they can be
lost altogether. If languages are in contact, they often come to share their Linguistic
Genders. Contact is often to blame for the demise of Genders. Adaptability of
Linguistic Gender in situations of language contact and language obsolescence
further attests to its vital importance, and functionality. This chapter also touches
upon the acquisition of Linguistic Genders by children, and their loss in language
dissolution. Linguistic Genders can be reshaped as part of conscious language

8 1 The multifaceted Gender
engineering: this issue is mentioned in Chapter 6, and then discussed in some more
detail in Chapter 11.
What happens when men are assigned to the feminine Linguistic Gender and
women to the masculine—that is, if Linguistic Genders are reversed? This may be
done for a joke, or the effect may imply offence, praise, solidarity, and endearment—
based on subtle overtones of value underlying Social Genders. Reversing Linguistic
Gender of humans—speaking of a man as if he were a woman, and of a woman as if she
were a man—highlights the stereotypes associated with male and female Social Gender.
We also discuss the overtones of the word meaning ‘man’ used to refer to people in
general, and markedness, status, and power—intrinsically associated with Social Gender
categories—as reflected in Linguistic Gender choice. These are the topics discussed
in Chapter 7, ‘Manly women and womanly men: the effects of gender reversal’.
Linguistic Genders are a source of metaphors and poetic imagery. They reflect
myths, beliefs, and traditions of the speakers. Linguistic Genders may mirror Social
Gender patterns, as we saw in Chapter 7. In addition, Linguistic Genders and their
meanings may affect cognition and the ways in which people perceive the world
around them. Men and women have different physical characteristics which relate to
their Social Genders. These are the topics of Chapter 8, ‘The images of gender’.
Social Gender finds its linguistic expression through the ways in which men and
women speak. Differences between male and female ways of speaking exist in any
language. Natural Gender properties—such as higher pitch and narrow vocal tract—
account for some of such differences. Male and female speech distinctions can be
paralinguistic (that is, differences may lie in higher pitch for women’s speech, and
other phonetic features, plus facial and bodily gestures). Or they may be conventionalized
as an integral part of grammar or lexicon: this is the case in ‘genderexclusive
languages’ where men and women have different phonemic systems, or
obligatorily use different words, or different sound correspondences. In a number
of languages—from linguistic minorities in North and South America, Siberia, and
India to Japanese and Thai—such differences between women’s and men’s dialects
are obligatory and striking. Conventionalized registers known as ‘women’s speech’
and ‘men’s speech’ where gender indexicals systematically span phonological, morphological,
and lexical domains have been described for numerous languages in
North America, the Chukotko-Kamchatkan family in Siberia, and a few in Amazonia.
The choice of code may be determined by a combination of Natural Gender and
Social Gender: In Koasati, a Muskogean language from North America, using male
code is a mark of authority. According to the male Atsinas, from the Algonquian
family, ‘male’ status is acquired by birth and social and cultural maturity; and it is
thus natural that a racial and cultural outsider be addressed with a female form. 15
‘Male’ and ‘female’ are an achieved status, and not an innate property.
Patterns of male and female speech, or male and female dialects, can be deployed
in constructing one’s identity. Members of the third gender—including hijras in

1.4 The empirical basis, and a note on conventions 9
India, and gays, lesbians, and transvestites in Western societies—deploy male and
female speech differences, Linguistic Gender, and many other linguistic features, to
project an image of a male or a female in a Social Gender sense. Chapter 9, ‘When
women and men speak differently’, focuses on distinct speech patterns corresponding
to the divisions in Social Gender, and gender-determined variation in speech
practices in gender-variable languages.
Men and women may have different speech styles and master different genres. In
Chapter 10, ‘The rituals of gender’, we further explore the roles of Social Genders in
traditional societies where women and men used to be associated with distinct
domains, and different speech styles. This is where the asymmetry, and the lack of
equality, between women and men is particularly apparent. Special languages and
language registers can come to be used in male-only rituals. A whole set of terms may
be forbidden to women. Women can be viewed either as keepers and promoters of
prestigious linguistic norm, and traditional language, or as dangerous ‘others’ which
lead the society in the wrong direction. Conventionalized speech practices reflect the
relative standing of, and the asymmetry between, Social Genders.
How does Linguistic Gender reflect social changes—and the improvement of the
woman’s place (using the title of Robin Lakoff ’s 1975 classic book)? In Chapter 11,
‘Gender in grammar and society’, we discuss the impact of social changes in the position
of women on the use of Linguistic Genders and classifiers, including the avoidance of
generic use for male terms and trends for European languages to become more ‘genderequal’.
The feminist movement plays a particular role in shaping language change.
Throughout the history of feminism and the backlash against it, language has been used
as a rhetorical weapon—reflecting power relationships, investigating social and linguistic
discrimination, and the embodiment and sexualization of women, and their traditional
activities, as a way of putting women down. Feminist theories have played a
substantial role in working out problems with women’s status and obtaining equality—
and even more, moving away from polarization and classification based on ‘sex’. We
focus on how language reflects gender in all its guises, and how the linguistic treatment
of ‘women’ can be seen as a barometer of social change.
The last chapter—‘The heart of the matter: envoi’—summarizes the main points—
the meanings and expression of Linguistic Gender and their correlations with gender
as a social construct, together with changes in Linguistic Gender choice, and form,
depending on changes in world view, cosmology, and social realities in flux.
1.4 The empirical basis, and a note on conventions
This book is focused on the three faces of gender across languages and cultures of
the world. It has an empirical inductive focus—relying on facts rather than on ad hoc
theories and hypotheses. 16 This study is based on an investigation of about 700
languages and—where possible—their social environment. Special attention has been

10 1 The multifaceted Gender
paid to data from languages on which I have first-hand expertise, and to minority
languages and groups which have not yet been given enough prominence in the
existing literature on aspects of gender. Giving prominence to minority languages
and cultures helps avoid a bias towards Western stereotypes and westernized perception
of women and men. For instance, a treatment of Linguistic Genders—which
deviates from what one is accustomed to in familiar Indo-European languages—as
‘non-canonical’ reflects the weight, and the bias, of post-colonial traditionalism. This
is what I have attempted to eschew.
Throughout this book, I have only been able to mention a portion of the available
literature, and only a selection of examples. A fair number of works on gender in its
various guises have not been mentioned here—either because they provide additional
instances and anthropological rather than language-oriented discussion of points
exemplified here, or because they are not exactly relevant to the ways linguistic
phenomena are used, or because they make claims which are not fully authenticated
or convincing, or contain mistakes and misinterpretations which make them unreliable.
I could not cite all the examples of every particular phenomenon—otherwise the
book would have become immense. I usually provide a particularly illustrative
example, and mention other similar ones (in a note). If a certain phenomenon is
found in more than half of the languages under consideration I call it ‘relatively
frequent’; if it is found in a restricted number of languages (one to ten), I cite all of
them and indicate its rarity. Note, however, that what appears rare to us at the present
stage of knowledge may turn out to be more frequent when we start learning more
about hitherto little-known languages and areas. This is the reason why I chose at this
stage not to give any statistical counts. Only about one-tenth of all human languages—
and societies in which they are spoken—have been documented so far; it therefore
seems most judicious to follow a qualitative approach, postponing quantitative analysis
some time in the future, when more data is available and can be assessed.
This book contains many examples from—and many mentions of—languages from
various areas and genetic groupings. When the language is introduced for the first
time, its affiliation and where it is spoken is given in brackets—for instance, ‘Kwami
(a Chadic language from Nigeria)’. Later mentions of the same language do not
include this information—which is summarized in the Index of languages at the end
of the book. At the end of each chapter the reader will find notes and sources.
This books aims at unravelling how Linguistic Gender, Social Gender, and Natural
Gender interact, viewed through the eyes of a linguist. As the societies move towards
greater equality of Social Genders in their attitudes and practices, the languages they
speak evolve. I have tried to capture some of the dynamics of how the expression of
multifaceted Gender reflects the world of perception, cognition, and social change. As
Bolinger (1991: 319) puts it, ‘no matter how wide the net is cast, a fish or two always
escapes’. There will always be room for upcoming enthusiasts to cast their nets wider.

1.4 The empirical basis, and a note on conventions 11
NOTES AND SOURCES
1. Bullock, Stallybrass, and Trombley (1977: 348).
2. This is the definition on the website of the American Psychological association. Many
sources focus on issues of Social Gender and individual aspects of its linguistic expression.
A full bibliography and critique will be a task on its own. A comprehensive bibliography
on Social Gender is in Upton (2012); see also definitions and references in Holmes and
Meyerhof (2003), Kramarae and Treichler (1992: 173); Baron (1986), Coates (1993, 2012),
Mills (2003a, 2003b, 2008), Talbot (2010), Coates and Pichler (2011), McConnell-Ginet
(2014), and also Romaine (1999); Aikhenvald (2015b) is an up-to-date bibliography on
Linguistic Gender. Silverstein (1985: 220–3) identifies a further meaning of ‘gender’—that
of an ideology of the ways in terms of which gendered language and the patterns of
variation are understood by speakers themselves.
3. Cheshire (2002: 423).
4. TheOxford English Dictionary Online describes ‘gender’ as ‘the state of being male or female’,
also stating that ‘gender’ is a linguistic term and refers to the ‘grammatical classification of
nouns and related words, which roughly corresponds to the two sexes and sexlessness’.
Terms ‘natural gender’,or‘biological gender’ are sometimes used interchangeably.
5. See also Moore (1994) on‘sex’ and ‘social gender’.
6. Zimman and Hall (forthcoming). The term ‘queer’ is an alternative, which covers gay men
and lesbian women, and ‘transgender’ individuals. The term ‘queer linguistics’ is used in
academic writing to refer to linguistic practices by gays, lesbians, and transgender people.
Barrett (2006) offers an overview of the history of the term, and of ‘queer’ speech; see also
Leap (2012).
7. See Zimman and Hall (forthcoming); Bucholz and Hall (1995b), Hall (2002), Hall and
O’Donovan (1996) and also McConnell-Ginet (2014). We return to this in §9.3.
8. Bateson (1958); see also Silverman (2001), Herdt (1987), Creed (1984).
9. Trechter (1995); see §9.3.
10. See Aikhenvald (2010: 215–16).
11. See Aikhenvald (2000: 347–50), and references there; and Chapters 4 and 11.
12. Linguistic Gender has been subject to many misconceptions and ‘linguistic myths’. One
is, in Philips’s(1980: 530) words, that ‘grammatical gender is relatively rare’, in contrast to
natural gender which is ‘a cultural and linguistic universal’. The statement about ‘rarity’ of
grammatical gender is utterly wrong—grammatical gender is one of the most widespread
categories in the world, covering much of Africa, Europe, North and South America, and
New Guinea; see Corbett (1991) and map 1 in Aikhenvald (2000: 78). The term ‘linguistic
gender’ has been used in a number of contradictory ways. English has no gender agreement
within a noun phrase. This has led some German scholars to state that English has
‘no gender’, forgetting about the gendered pronouns ‘she’, ‘he’, and ‘it’. Pitfalls of the term
‘natural gender’ and its current usage have been addressed by McConnell-Ginet (2014).
13. See a comprehensive discussion in Braun (1997b; 2000a; 2001).
14. English is rather unusual in having just one term to cover the three faces of ‘Gender’. In
French, genre is used for Linguistic Gender (and also in a number of other meanings,
including kind and genre); the term sexe covers Social and Natural Genders. In German,
both Genus and Geschlecht are used for Linguistic Gender; the term Geschlecht is used to
cover Social and Natural Gender. In Modern Greek, the term for linguistic gender is yénos
(γένος); social gender is referred to as cinonikó fílo (κοινωνικό φύλο), and natural gender
violoyikó fílo (βιολογικό φύλο) (Angeliki Alvanoudi, p.c.). Contemporary Russian is rather

12 1 The multifaceted Gender
striking in that each of the three meanings have a distinct one-word term: rod for
Linguistic Gender, pol for Natural Gender, and the recent borrowing from English,
ghender, for Social Gender. English has some alternatives to gender as a cover term.
Large systems of noun categorization with more than four terms in Bantu, Australian,
and South American languages are sometimes referred to as ‘noun classes’. An alternative
to Natural Gender is ‘sex’,or‘demographic gender’ (Silverstein 2015). Social Gender is also
referred to as sociocultural gender.
15. See Saville-Troike (1988) on Koasati and Taylor (1982) on Atsina.
16. As Leonard Bloomfield (1933: 20) put it: ‘The only useful generalizations about language
are inductive generalizations. Features which we think ought to be universal may be absent
from the very next language that becomes accessible . . . The fact that some features are, at
any rate, widespread, is worthy of notice and calls for an explanation; when we have
adequate data about many languages, we shall have to return to the problem of general
grammar and to explain these similarities and divergences, but this study, when it comes,
will not be speculative but inductive.’

2
Linguistic Gender and its expression
Linguistic Genders are grammatical classes of nouns, based on core semantic properties.
These include sex (female and male), animacy, humanness, and also shape and
size (especially for non-humans). We start with general features of Linguistic Gender
expressed through agreement and anaphoric personal pronouns.
2.1 Linguistic Gender in a nutshell
Linguistic Gender is one of the earliest grammatical categories to be recognized. The
Greek philosopher Protagoras (fifth century BC) was the first to use the term gender.
In his description of Ancient Greek, he divided Greek nouns into three classes:
‘feminine’, ‘masculine’, and ‘inanimate’ (nowadays called neuter). This typical gender
system is found in many Indo-European languages, including Latin. During historical
change Latin neuter nouns were redistributed between the other two genders,
giving the modern system of masculine and feminine in French and Italian.
Traditional grammar has continued using the term ‘gender’ for Indo-European
and Semitic languages with a typical division of nouns into masculine and feminine,
or masculine, feminine, and neuter. When Europeans came to study African languages,
they discovered larger gender-like systems with eight or more possibilities in
languages like Swahili. These did not necessarily involve just a masculine–feminine
distinction. The term ‘noun class’ came to be used for larger systems of this kind. 1
Relatively simple two-term gender systems include masculine and feminine, as in
French, Portuguese, Hausa, Hebrew, and many other Afroasiatic languages. Algonquian
languages of North America have one gender for animates, and another one
for inanimates. Tamil and a few other Dravidian languages divide nouns into two
gender classes. One is labelled ‘rational’ and covers humans, gods, and demons; it
further divides into masculine and feminine. The other one is labelled ‘non-rational’:
it includes all animates and inanimates which do not belong to the ‘rational’ class. 2
Latin, German, and Russian have three genders—masculine, feminine, and neuter.
Four genders broadly labelled ‘masculine’, ‘feminine’, ‘edible vegetable’, and ‘all else’
(or ‘neuter’) are a feature of a number of Australian languages, including Dyirbal
(to which we return in §2.2.1).
How Gender Shapes the World. First edition. Alexandra Y. Aikhenvald.
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.

14 2 Linguistic Gender and its expression
Box 2.1 sums up the general properties of linguistic genders.
BOX 2.1 General properties of linguistic genders
I. There is a limited, countable number of gender classes.
II. Each noun in the language belongs to one (or sometimes more than one) class.
III. There is always some semantic basis to the grouping of nouns into gender classes.
Languages vary in how much semantic basis there is. This usually includes animacy,
humanness, and sex, and sometimes also shape, size, and extent.
IV. A constituent outside the noun itself must agree in gender with a noun. Agreement can
be with other words in the noun phrase (adjectives, numbers, demonstratives, articles,
etc.) and/or with the predicate of the clause, or an adverb. Gender can be marked in
personal pronouns, reflecting ‘anaphoric’ gender agreement.
2.1.1 Marking Linguistic Gender
Linguistic Gender in a language can be marked on agreeing modifiers and verbs, or
personal and other pronouns. Gender can be marked on the noun itself: it is then said
to be marked ‘overtly’. Examples 2.1 and 2.2, from Portuguese, illustrate masculine and
feminine genders. Gender is marked on the noun and on the accompanying definite
article and the adjective ‘beautiful’, indicating their agreement in gender with the noun:
2.1 o menin-o lind-o
ARTICLE:MASC.SG child-MASC.SG beautiful-MASC.SG
‘the beautiful boy’
2.2 a menin-a lind-a
ARTICLE:FEM.SG child-FEM.SG beautiful-FEM.SG
‘the beautiful girl’
Portuguese also has two anaphoric genders in personal pronouns: ele ‘he’, ela ‘she’,
eles ‘they:men’ and elas ‘they:women’. Derivational genders are marked by suffixes
on nouns, e.g. senhor ‘mister’, senhora ‘lady’, patrão ‘owner of (masculine)’, patroa
‘owner of (feminine)’.
Gender of a noun may be ‘covert’. Then the form of the noun tells us nothing
about the gender it belongs to. The word Mond ‘moon’ in German is masculine, while
Gabel ‘fork’ is feminine. The form of the article they require will indicate their gender:
we say der Mond ‘the (masculine singular) moon’ but die Gabel ‘the (feminine
singular) fork’. The gender of a noun is recognizable through the agreeing forms.
The form of some German nouns gives us a hint of what gender they belong to: we
will see in §2.2.3 that gender in German is chosen based on a mixture of principles,
semantic and formal.

2.1 Linguistic Gender in a nutshell 15
There is no agreement gender in English: 3 the same forms of the article and the
adjective are used with ‘boy’ and with ‘girl’ in the translations of 2.1 and 2.2. Anaphoric
gender in English features in personal pronouns he for male humans, she for female
humans, and it for non-humans. The actual usage is not fully straightforward: a
non-human—for instance, a beloved cat or dog—can be personified and referred to
as ‘he’ or ‘she’ depending on their sex. ‘Ships’ and ‘countries’ are referred to as ‘she’
notwithstanding their being non-humans. In some varieties of English, using ‘he’ or
‘she’ with regard to an inanimate expresses the speaker’s attitude—endearment, contempt,
or familiarity. He used to be a pronoun with generic reference. However, it is no
longer the preferred generic pronoun (they is commonly used instead, to avoid what is
perceived as ‘male bias’). This change reflects the changing status of male and female
Social Genders (see §§7.5–7, and Chapter 11).
2.1.2 Gender agreement and anaphoric gender
English is among numerous languages with just the anaphoric gender in personal
pronouns. (We will see, in §6.3, how the Old English agreement gender was lost.)
Kaingáng is the only Jê language (of southern Brazil) to have masculine and feminine
distinction in third person pronoun. Many Carib languages in South America
(including Trio, Apalaí, and Hixkaryana) have animate and inanimate forms of
third person singular pronoun. In Finnish, the third person pronoun hän ‘he, she’
can only be used for human referents (the proximate demonstrative se is used for
non-humans). 4
Anaphoric gender tends to be found in languages with agreement gender, including
German, French, Portuguese, and Russian. But this is not uniformly the case.
Maithili, an Indo-Aryan language from Nepal and northern India, has four honorific
forms in its personal pronouns, but no gender specification. Masculine and feminine
genders are expressed through agreement on verbs and on definite adjectives. In
Abui, a Papuan language from eastern Indonesia, gender is marked in object prefixes
on verbs but not in personal pronouns. 5
In many languages, animate nouns tend to agree with modifiers based on their
meaning rather than on their formal allegiance to a gender. In Swahili ki-faru (gender
7-rhinoceros) belongs to the ‘inanimate’ class 7. But the noun can occur with an adjective
marked for gender 1 (which covers animates): ki-faru m-kubwa ‘a big rhinoceros’.This
is how semantic agreement overrides the expected agreement based just on form (also
known as ‘syntactic’ (or ‘mechanical’) agreement). 6
Anaphoric agreement in a language with complex rules of gender choice may show
more straightforward correlations with meaning than agreement with an article or an
adjective in a noun phrase. The word for ‘girl’ in German, das Mädchen, belongs to
the neuter gender on morphological grounds: as we will see in §2.2.2, the diminutive
suffix -chen it contains is always associated with the neuter. We can see that in 2.3 ‘the

16 2 Linguistic Gender and its expression
(neuter) girl’ requires neuter agreement of the article and the adjective. However, in
numerous instances a girl will be referred to with the feminine pronoun sie ‘she’
rather than es ‘it’. In this case, anaphoric agreement follows the meaning of the noun
rather than its form. This is known as semantic agreement. The following example
comes from Heinrich Mann’s novel Der Untertan (published in 1914):
2.3 Das junge Mädchen (neuter) gab ihm die Hand. Sie (fem) wollte wohl nett sein,
aber was war mit ihr (fem) anzufangen.
‘The young girl gave him her hand. Apparently, she intended to be nice, but
what could be done with her’ (p. 14).
The girl is introduced with a definite article, in the neuter form (reflecting the
neuter agreement gender of the noun itself). She is then referred to with the feminine
anaphoric pronoun as ‘she’. In a recent study, Braun and Haig (2010) showed that
speakers tend to favour the feminine pronoun sie (and not the neuter pronoun es) for
older girls (18 years of age), and neuter pronouns for 2-to12-year-olds who have not
yet attained the status of ‘women’. 7
2.1.3 Linguistic Gender and other linguistic categories
Genders can be marked in a variety of ways—with suffixes or prefixes, apophony (or
vowel changes), tone patterns, and change of stress. Linguistic Genders are a feature
of languages with complex morphology, and a fusional or agglutinating (but not an
isolating) profile. Gender and number can be combined into a single affix. In Bantu
languages, gender agreement markers come in singular/plural pairs. They appear on
modifiers of different kinds and on the predicate. In 2.4, from Swahili, the noun class
marker, ki-, also appears on the noun itself. That is, noun classes are overtly marked
(Welmers 1973: 171). Ki- marks the singular term of noun class 7/8 (see Table 2.2)
which covers inanimates. The plural counterpart of ki- will be vi-.
2.4 ki-kapu ki-kubwa ki-moja ki-li-anguka
NCL7:INANIM-basket
‘One large basket fell’
NCL7:INANIM-large NCL7:INANIM-one NCL7:INANIM-PAST-fall
There may be fewer gender distinctions in plural than in singular. In German and
Russian, genders are not distinguished in plural number. 8 Most frequently, anaphoric
genders are distinguished in third person only: this is what we find in familiar Indo-
European languages. Berber, Semitic, Cushitic, and many Chadic languages have
masculine and feminine forms of second person and third person pronouns, both
singular and plural. Manambu and other languages from the Ndu family in the Sepik
area of New Guinea have different forms for masculine and feminine genders in
second and third person singular pronouns. Gala, from the same family, has two
genders in first, second, and third persons (singular)—as we will see in Table 7.1. 9

2.1 Linguistic Gender in a nutshell 17
Linguistic Gender interacts with definiteness in Maithili. Definite forms of adjectives
modifying a noun agree with it in gender, e.g. mot̍-ka mərəd ‘the fat man’, and
mot̍-ki jənana ‘the fat woman’. Indefinite forms do not, e.g. pəigh mərəd ‘a tall man’,
pəigh məugi ‘a tall woman’. 10
2.1.4 How many genders?
A language may distinguish different sets of gender forms in different contexts.
Romanian, a Romance language, distinguishes two gender forms in the singular,
and two gender forms in the plural. We find three combinations:
(I) Nouns which take -Ø in the singular and -i in the plural (marked on an
agreeing adjective and other modifiers).
The core of this group of nouns are men. This is why this form is often called
‘masculine gender’. Examples are un prieten bun (INDEFINITE.ARTICLE:masc.sg friend:
masc.sg good:masc.sg) ‘a good male friend’; and prieten-i bun-i (friend-masc.pl goodmasc.pl)
‘good male friends’.
(II) Nouns which take -ă in the singular and -e in the plural.
The core of this group of nouns are women. This is why this form is described as
‘feminine gender’. Examples are un-ă prieten-ă bun-ă (INDEFINITE.ARTICLE-fem.sg
friend-fem.sg good-fem.sg) ‘a good female friend’; prieten-e bun-e (friend-fem.pl
good-fem.pl) ‘good female friends’.
(III) Nouns which take -Ø in the singular and -e in the plural on adjectives.
This group of nouns includes many inanimates which is why some scholars call
this ‘neuter’. Examples are un stilou bun (INDEFINITE.ARTICLE.masc.sg pencil good:
masc.sg) ‘a good pencil’, stilouri bun-e (pencil:pl good-fem.pl) ‘good pencils’.
The meanings of the three gender combinations are not entirely transparent. Not
all inanimates belong to the ‘neuter’ gender (group III). Table 2.1 summarizes the
marking of singular and plural feminine, masculine, and neuter forms. 11
TABLE 2.1. Genders in Romanian and their marking
Markers of singular
number
Distribution in gender
groups of nouns
-a feminine
-e
neuter
-Ø
masculine -i
Markers of plural
number

18 2 Linguistic Gender and its expression
Alternatively, a language may have two coexisting subsystems of genders used in
different contexts: one for agreeing pronouns (including demonstratives or verbal
cross-referencing) and one for adjectives and other modifiers. Gaaguju, an Australian
language, has a system of four genders used on adjectives as modifiers—one for
human males and some animates, one for human females and other animates, one
for plants and their parts, and for weapons, and a further one for abstract entities,
body parts, and fire. Just two classes—animate and inanimate—are marked on verbs.
Palikur, an Arawak language from Brazil and French Guyana, distinguishes three
genders in demonstratives and personal pronouns. Agreement markers on verbs and
adjectives have two gender forms: one for feminine and one for masculine/neuter.
Languages with several gender systems employed in different environments are
known as ‘split-gender’ systems. 12
2.2 How to choose a Linguistic Gender
Linguistic Gender always has some semantic basis, although not everything can be
explained in semantic terms. No matter how complicated the overall system of Linguistic
Gender assignment may be, its semantic core will always include reference to the
universal parameters of sex, humanness, and animacy. Non-humans and inanimates
can be assigned to gender classes based on further features—including size and shape
(see Chapter 3). Morphological and phonological form can also be at play.
2.2.1 Gender choice by meaning
Linguistic Gender can be assigned to a noun based entirely on its meaning. The
choice can be fairly transparent. Malto, Kolami, Ollari, and Parji, also from the
Dravidian family, have one Linguistic Gender for human males and the other
one for the rest. We can recall from §2.1 that all nouns in Tamil divide into two
genders labelled ‘rational’ and ‘non-rational’. Rational nouns include humans,
gods, and demons, with the remainder being ‘non-rational’. Diyari, an Australian
language, and Kaingang, a Jê language from southern Brazil, divide all nouns into
female humans and the rest. Nouns in Siouan languages from North America, Carib
languages of South America, and a few Austronesian languages spoken in eastern
Indonesia divide into animate and inanimate gender classes.
Sex, humanness, animacy, and rationality may combine in one gender system.
Zande and Ma (Ubangi languages from Central Africa) have four genders—feminine,
masculine, non-human animate, and inanimate. Godoberi, a north-east Caucasian
language, has three genders—feminine, masculine, and ‘non-rational’. 13
Gender can be chosen based on somewhat different principles for different groups
of nouns. Tunica, an isolate formerly spoken in Louisiana, had two genders—
masculine and feminine. Gender was marked in definite articles and in pronouns.

2.2 How to choose a Linguistic Gender 19
The gender of humans, birds, and most animals was assigned based on the sex:
females were feminine, and males masculine. Fish, shellfish, smaller reptiles, and
insects were masculine. Trees, fruits, vegetables, and tools were also masculine.
Inanimate feminines included terms for substances viewed as masses (such as sand,
salt, metal, cotton). Buildings and their parts were feminine, and so were abstract
nouns and nouns referring to processes, such as dance, war, noise, and sickness. 14
Systems with semantic gender assignment vary in their transparency. Gender
choice in Dyirbal 15 is semantically based, but not straightforward. Four genders are
expressed through article-like noun markers (but not on nouns themselves). Three of
them are associated with one or more concepts:
gender 1 (noun marker bayi)—male humans and non-human animates;
gender 2 (noun marker balan)—female humans; fire; drinkable liquids; fighting;
gender 3 (noun marker balam)—non-flesh food;
gender 4 (noun marker bala)—a residue gender covering everything else, including
body and other parts, place names, and flesh food (meat and fish).
All animates are distributed between gender 1 and gender 2 (except bees which are
in gender 4). Three general principles determine gender membership of a noun:
I. If a noun has a characteristic X (on the basis of which its gender will be chosen)
but is associated with characteristic Y through BELIEF OR LEGEND it will be then assigned
to a different gender based on characteristic Y. This is a principle of MYTHOLOGICAL
ASSOCIATION, orMYTH-AND-BELIEF PRINCIPLE.
It is believed that birds are, as a class, the spirits of dead women. Birds are classed
as members of gender 2 (‘feminine’, balan) rather than bayi on the basis of their nonhuman
animate status. There are exceptions to this, also based on beliefs. Willy
wagtail belongs to gender I (masculine), bayi jigirrjigirr, since he is believed to be the
metamorphosis of a legendary man (and the way the bird wiggles its tail is reminiscent
of how men dance in traditional dances, corroborees). Non-edible snakes are
members of gender I (bayi). An exception is balan bima ‘death adder’ who is also a
legendary woman, and thus belongs to gender 2. The sun was believed to be a
woman, and so belongs to gender 2 (balan garri). The moon is her husband, and
so the noun belongs to gender 1 (bayi gagara). Thunder, lightning, and rain are
believed to be legendary males, and are included in gender 1.
II. If the referent of a noun with a characteristic X is perceived to have a PHYSICAL
ASSOCIATION with the characteristic Y, then this may be reflected in the gender choice
for this noun. The association can be based on physical association. Terms for fruit
and vegetable food belong to gender 3; they are produced by plants which belong to
gender 4. Honey is balam girñjal (gender 3), and is produced by bees. Because of this
analogy, bees belong to gender 4—they are the only animates not in genders 1 or 2.
Physical similarity may be another reason for gender choice. Most insects belong to

20 2 Linguistic Gender and its expression
gender 1 (bayi). But the firefly, balan yugiyam, belongs to gender 2 because the
flashes of light it emits are similar to sparks from a fire.
Or the referent with a characteristic X can be used to make or catch Y. For
instance, matches and a match box used to hold them belong to gender 2 (balan)
since they produce fire which belongs to the same gender.
III. If a set of nouns belongs to a certain gender, and members of its subset share a
particular IMPORTANT PROPERTY, then this subset may be assigned to a different
gender. Most trees without edible parts belong to gender 4, but stinging trees are
placed in gender 2, due to their harmful nature. Most fishes belong to gender 1, as
non-human animates. Three species of fish belong to gender 2, since they are harmful
to humans. The jellyfish, balan jawayi, has tentacles that can inflict a painful sting.
The freshwater stonefish, balan jaŋgan, has venomous spines which may give painful
wounds if stepped on. And eating balan juruŋun, the toad fish, can be fatal. Pythons,
the only snakes that are eaten, are balan.
These principles of gender choice highlight what there is in the legends and what is
important for survival. They are living proof of functionality of genders.
Mythological association, beliefs, and legends play a role in gender in quite a few
languages. In the Western Torres Strait language, all nouns denoting males are
masculine, and all the rest are feminine. The moon is masculine, due to mythological
association with masculinity. In Tunica, thunder and lightning were personified as
mythical men, and were masculine for this reason. 16 We return to this in §8.1.
The meaning of a small gender system may be transparent to a limited extent.
Nouns in Fox and other Algonquian languages divide into animate and inanimate
genders. 17 Gender is marked on the noun, and on demonstratives and verbs. Those
nouns which denote humans and biologically animate beings belong to the animate
gender, e.g. neniwa ‘man’, ihkwewa ‘woman’, mahkwa ‘bear’, and maneto . wa ‘spirit,
snake’. Nouns denoting entities that move—including spirits and heavenly bodies—
are also grammatically animate. The noun ni . c᷈a . pa ‘doll’ is animate, since a doll
represents a human.
A number of nouns which refer to inanimates belong to the animate gender. Some
names of plants are inanimate, e.g. ahte . himini ‘strawberry’, and others are animate,
e.g. wi . tawi . ha ‘raspberry’. The ‘Implacable Raspberry’ in the title of a paper by Straus
and Brightman (1982) sums up the idea of unpredictability of Algonquian gender
when it comes to entities other than human beings or animals.
A few animate–inanimate pairs show regular differences in their meanings. In Fox,
‘grandfather’ is animate; but a ceremonial pole referred to as ‘our grandfather’ is
inanimate. A noun with general or collective reference may be inanimate, and individual
items animate. In Cree, another Algonquian language, mistik ‘tree’ is animate,
and mistik ‘stick’ is inanimate. In Fox s᷈o˞niya . hi ‘silver, money’ is inanimate, and
s᷈o˞niya . ha ‘a coin, a bill’ is animate; owiya . he . hi ‘animals, small game (collective)’ is
inanimate; and owiye . he . ha ‘a (small) animal’ is animate.

2.2 How to choose a Linguistic Gender 21
The animate gender is associated with spiritual power and agency. The noun
c᷈i . paya ‘corpse, ghost’ is animate: it is conceived as a former human being, and in
its meaning ‘ghost’ it denotes a spiritually powerful being. The principle of ‘power’
in the assignment of nouns to the animate class can be illustrated with gender shifts
in traditional narratives. A stone (aseny-) is generally inanimate. But in a traditional
story a man addresses a prayer to a stone which takes an animate agreement marker
on the verb. 18 Further semantic functions of a difficult-to-capture gender choice are
revealed through variable gender in discourse—we return to this in §4.2.
Noun classes in Bantu languages are a well-known example of a semantically
opaque system. Table 2.2 summarizes a basic semantic grid common to Bantu noun
class systems based on the interaction of shape, size, and humanness. These parameters
provide only a partial semantic motivation for the noun classes in individual
Bantu languages. 19
Gender assignment in modern Bantu languages is to some extent semantically
opaque, though its semantic nucleus is still discernible. Thus, in Babungo, class 1/2 is
basically human; however, it is a much bigger class than it was in Proto-Bantu, and
also contains many animals, some birds and insects, body parts, plants, and household
and other objects, e.g. necklace, pot, book, rainbow. Shape and size also appear
as semantic parameters: in ChiBemba, class 7/8 is associated with large size and
carries pejorative overtones, while class 12/13 includes small objects and has overtones
of endearment. 20
In Jingulu, an Australian language, nouns divide into four genders, which are only
partially semantically transparent. The ‘vegetable’ class mostly includes objects which
are long, thin, or pointed; this class happens to include most vegetables, as well as
body parts such as the colon, penis, and neck, instruments such as spears, fire-drills,
and barbed wire, natural phenomena such as lightning and rainbows, and also roads
and trenches. The ‘feminine’ class includes female humans and higher animates, and
TABLE 2.2. Gender–number pairs in Bantu
GENDER CLASS
SEMANTICS
1/2 humans, a few other animates
3/4 plants, plant parts, foods, non-paired body parts, miscellaneous
5/6 fruits, paired body parts, miscellaneous inanimates
7/8 miscellaneous inanimates
9/10 animals, miscellaneous inanimates, a few humans
11/10 long objects, abstract entities, miscellaneous inanimates
12/13 small objects, birds
6 masses
14 abstract qualities, states, masses, collectives
15 infinitives

22 2 Linguistic Gender and its expression
also words for axes, the sun, and most smaller songbirds. The semantic content of the
remaining two classes, ‘masculine’ and ‘neuter’, is harder to define: masculine is
mostly used for the rest of animates and neuter for the rest of inanimates, except that
flat and/or rounded inanimates (most trees and eggs, and body parts such as the liver
and eyebrow) are masculine. 21
A language can be thought of as a long-term repository of speakers’ culture and
lore, and it is unlikely that every detail could be explained. There is no synchronic
reason why ‘raspberry’ should be animate throughout Algonquian languages—any
mythological association which may have existed in the past has since been lost.
The great majority of gender choices in Dyirbal can be explained in terms of general
principles and rules of gender transfer and association. But, as Dixon (2015: 42) puts
it, ‘not all. I know of no reason why balan (Gender II) is used for prawns and crabs,
for the two egg-laying mammals (platypus and echidna), and for the dingo or native
dog. (There are a number of other individual eccentricities.)’
Most languages have such ‘eccentricities’—an unpredictable residue of semantically
unexplainable gender choice. In Dyirbal, ‘there may have been further legends
and beliefs, which I [Dixon] did not become familiar with, that would help explain
more about gender assignment. And there could be legends which had been forgotten
but left in their wake some special gender specification. As in any judicious enquiry
into the semantic basis of a grammatical system of genders, a lot can be explained—
but not everything’ (Dixon 2015: 42–3).
2.2.2 Gender choice by form
The form of a noun may play a role in Linguistic Gender choice. Numerous
derivational affixes in German are each associated with one gender. Nouns containing
the suffix -ung ‘action noun’ are feminine, e.g. Bedeutung ‘meaning’, Bewegung
‘movement’. Nouns which contain a diminutive suffix -lein or -chen are always
neuter. The noun Mäd-chen ‘girl’ belongs to the neuter gender because it contains
the neuter suffix. Similarly, korítsi ‘girl’ in Modern Greek is neuter thanks to the
neuter suffix -(í)tsi it contains. The augmentative form korítsaros ‘pretty or buxom
girl’ belongs to masculine gender (despite its female referent) because nouns with the
augmentative suffix -aros are always masculine.
Phonological form of a noun may correlate with Linguistic Gender. In Qafar, an
East Cushitic language, nouns with inanimate referent whose citation form ends in a
vowel are feminine, all the rest are masculine. In Hausa, all non-sex differentiable
nouns which end in ‐aa are feminine. In Katcha (Kordofanian), any noun beginning
with m- belongs to the feminine gender, unless it is ‘notionally masculine’. In
Punjabi, an Indo-Aryan language, ‘the rule of thumb is that inflected nouns ending
in -aa are usually assigned masculine gender whereas the nouns ending in -ii are
feminine’. 22

2.2 How to choose a Linguistic Gender 23
Phonological principles of gender choice help determine gender assignment of
loans from a French-based creole, French, and Portuguese in Palikur. The gender of
the source language is usually overridden by these phonological principles. Masculine
human nouns usually end in ‐e or -i. Feminine nouns end in -u or -o. Neuter nouns
can end in any vowel or consonant. Loan nouns which end in -o or -u are usually
feminine in gender, e.g. marto ‘hammer’ (from French Creole marto), sitru ‘lemon’
(from French Creole citrõ), tattu ‘armadillo’ (variant of tat; from Portuguese masculine
tatu). 23 Nouns which end in a consonant are usually given neuter gender, e.g.
simis ‘shirt’ (from French chemise).
Morphological and phonological features may help determine the Linguistic
Gender, but are never the only principles for its choice. We now turn to the ways
in which form and meaning work together in Linguistic Gender assignment.
2.2.3 Meaning meets form: mixed principles of gender assignment
Not infrequently, Linguistic Gender is chosen based on a combination of meaningbased
and form-based principles. The gender of humans is based on semantics.
Other, formal principles, are at work for the rest. Most Albanian varieties have two
genders, masculine and feminine. Nouns denoting females are feminine, and those
denoting males are masculine. Inanimate nouns ending in a consonant (e.g. gjak
‘blood’ or dhëmb ‘tooth’) or a stressed vowel are assigned to the masculine Linguistic
Gender. Nouns ending in an unstressed vowel tend to be feminine. The semantic
principle overrides the formal one: rojë ‘guard’ is masculine since it denotes a male
(although it ends in an unstressed ë typical of feminines). 24
As we saw in §2.2.2, in Greek, ‘to a certain extent, grammatical gender can be
considered to be a property of stem-formatives involved in the formation of particular
nouns. Thus, diminutives in -(í)tsi are always neuter, augmentatives in -aros are
always masculine, abstract nouns in -ja are always feminine, and so on’ (Joseph and
Philippaki-Warburton 1987: 152–3).
Semantic and phonological principles underlie the assignment of Linguistic Genders
in a number of Papuan languages of New Guinea. Yimas, from the Lower Sepik
family in Papua New Guinea, has eleven Linguistic Genders. Four are chosen by the
meaning of the nouns; these cover (i) human males; (ii) human females; (iii) animals;
and (iv) culturally important items. The rest are phonologically motivated: the agreeing
constituent repeats the last consonant of the noun root. Arapesh languages of the
East Sepik Province in Papua New Guinea work on a similar principle: they have a
noun class for male humans, and one for female humans. Nouns of other semantic
groups trigger similar ‘alliterative’ or phonologically based, agreement. This is always
overridden by semantics; so, loanwords with human referents in Arapesh languages
are assigned to classes according to whether they refer to a man or to a woman. Most
nouns which belong to the masculine gender contain a final or an initial segment n,

24 2 Linguistic Gender and its expression
e.g. aleman ‘man’ or Nekitel ‘personal name’ (Nekitel 1986, and p.c.). The English
loanword nes ‘nurse’ is treated as a member of the feminine class (whose members
are marked with ʔ or kw-). Words such as ankelo ‘angel’ and Spiritu Santu ‘Holy
Spirit’ are conceived of as males, and so treated as masculine. That is, meaning-based
choice may override the choice based on formal grounds. 25
It has been often stated that there is no real semantic basis for gender assignment
in some well-known Indo-European languages. In a seminal study, Zubin and
Köpcke (1986) provided a semantic rationale for gender assignment of nouns of
different semantic groups in German. In agreement with the natural sex principle,
masculine and feminine genders mark the terms for male and female adults of each
species of domestic and game animals, and neuter is assigned to non-sex-specific
generic and juvenile terms. Superordinate terms are often neuter. Masculine gender
is used for types of cloth, types of precipitation and wind, and types of minerals.
Types of knowledge (e.g. Kenntnis ‘knowledge’) and disciplines (e.g. Linguistik
‘linguistics’) have feminine gender, and games and types of metal have neuter
gender. Almost all nouns with generic reference, such as das Ding ‘thing’, das
Gerät ‘implement, apparatus’, das Gut ‘goods’, have neuter gender (with one
exception: die Sache ‘thing’ is feminine). Superordinate terms usually belong to
neuter gender, and items of a more basic level are feminine or masculine, more
rarely neuter. Table 2.3 illustrates the neuter gender of superordinate nouns and the
non-neuter (masculine and feminine) of nouns referring to ‘basic level’ objects (see
Zubin and Köpcke 1986: 147ff.).
A combination of meaning-based and form-based principles accounts for gender
choice in the majority of cases. But there is always a small residue of exceptions which
cannot be explained. For example, two superordinate terms are feminine: Pflanze
‘plant, herb’ and Farbe ‘colour’, going against the principle in Table 2.3.
Morphological and semantic principles interact in German gender assignment for
some groups of nouns. Bird names have masculine gender unless they end in a
derivational suffix which is feminine. Sailing vessels are assigned to the feminine
gender unless they end in a masculine derivational suffix (Zubin and Köpcke 1986:
TABLE 2.3. Semantic basis of gender choice in German: an illustration
SUPERORDINATE
Instrument (n) ‘musical instrument’
Obst (n) ‘fruit’
Spielzeug (n) ‘toy’
Land (n) ‘land’
BASIC LEVEL
Guitarre (f) ‘guitar’, Trompete (f) ‘trumpet’
Apfel (m) ‘apple’, Pflaume (f) ‘plum’
Bauklotz (m) ‘block’, Puppe (f) ‘doll’
Wald (m) ‘woods’, Sumpf (m) ‘swamp’,
Wiese (f) ‘meadow’

2.3 Markedness and Linguistic Genders 25
175). The loanword der Drink ‘alcoholic drink’ is masculine, and its gender assignment
is based on a complex of semantic and morphological and phonological
features (Zubin and Köpcke 1984: 44). These are:
(i) Phonological: Consonant clusters in initial and in final position; initial cluster
dr-; final nasal (typical for masculine nouns)
(ii) Morphological: -s ‘plural’ (characteristic of masculine or of neuter gender)
(iii) Semantic: class of drinkable liquids.
Semantic and formal principles can compete with each other. Gender of compounds
is normally determined by the last component. Streich-holz (strike-wood)
‘match’ is neuter because the last part, Holz, is neuter. But in north Germany
Streichholz is masculine because of its association with ‘stick-like objects’ which are
typically masculine (cf. Stock (masculine) ‘stick’). As expected, most compounds
containing the masculine noun Mut ‘courage, valour, spirit’ are masculine. These
include Freimut ‘frankness’, Heldenmut ‘heroism’, Hochmut ‘arrogance’, Lebensmut
‘exhilaration’, Unmut ‘bad temper’, and Übermut ‘bravado’. But some are feminine;
for instance Anmut ‘gracefulness’, Demut ‘humility’, Langmut ‘patience’, and Schwermut
‘melancholy’. An experimental study of -mut compounds (by Zubin and Köpcke
1984) showed that the -mut compounds assigned to the feminine gender were
associated with introverted feelings, while masculine gender was associated with
extroversion. The same principle appears to hold for nouns ending in the suffix
-sal. These are normally neuter, such as Schicksal ‘fate, destiny’. Those that are
feminine describe introverted emotional states, such as Trübsal ‘misery’.
Mixed principles of Linguistic Gender choice account for synchronically opaque
gender assignment which may seem close to arbitrary. The complexity, and semantic
opacity, in gender choice creates an impediment for second-language learners of
many European languages. Language learners—linguists included—complain that
Linguistic Genders are arbitrary and redundant and are nothing other than unnecessary
obstacles for the uninitiated. As we will see in Chapter 4, this is far from being
the case: Linguistic Genders are a useful and versatile device which helps organize
discourse, express subtle meanings, and reflect the status of Social Genders.
2.3 Markedness and Linguistic Genders
Linguistic Genders within a language may not have the same status. There is a
fundamental distinction between two kinds of markedness in language—formal
and functional. A formally unmarked term will be the only one in its system to
have zero realization (or a zero variant). The Spanish masculine singular pronoun el
is formally unmarked for gender. Its feminine counterpart, ella, is formally marked.
In Portuguese, ele ‘he’ and ela ‘she’ are equally formally marked.

26 2 Linguistic Gender and its expression
Functional markedness is different in nature. It relates to how the forms are used
(and not what they look like). The marked term(s) may be used each in a restricted,
specifiable situation, with the unmarked term being used in all other circumstances.
The unmarked category appears in neutralized contexts, or when one does not wish
to be specific. In Portuguese, the choice of a masculine or a feminine pronoun is
made when the sex of a human is known and unambiguous. But when there is a mix
of people of different sexes the masculine form is used. One distinguishes o filho
‘son’ and a filha ‘daughter’ in the singular. The plural ‘children’ will be referred to
as os filhos (lit. the sons). The corresponding feminine form, as filhas, willunambiguously
cover just ‘daughters’. Similarly, a irmã is ‘sister’ and oirmão‘brother’. If
I ask someone how many siblings they have, I will refer to ‘brothers and sisters’ as
irmãos, literally, brothers. That is, the masculine—not the feminine—form can be
considered a functionally unmarked choice.
In many familiar Indo-European languages the masculine form is used to refer to a
group of mixed sexes, to generic referents, or to those whose sex is unknown. In
Spanish, ellos ‘they (masculine)’ refers to a group consisting only of males and also a
group of mixed sex. The form ellas ‘they (feminine)’ refers exclusively to a group
consisting of females. Similarly, the masculine plural form les américains in French
can refer to a group of males or to a group of mixed sexes; its feminine counterpart,
les américaines, refers only to a group of female Americans. In Russian, question
words and indefinite pronouns require masculine singular agreement. I will say Kto
pris᷈el (who come+PAST.masc.sg) ‘Who came?’ even when asking about a woman. In
this case, the masculine gender is functionally unmarked. Until recently, the masculine
he was favoured as a generic pronoun in English, and thus as a functionally
unmarked choice; we return to this in §11.5.1. 26
If nouns belonging to different genders are coordinated, the gender form chosen
for agreement will be the unmarked one. In Portuguese, Hebrew, and French the
masculine plural form will be used. One would say, in Portuguese, um menino e uma
menina pequenos (indef.masc.sg boy and indef.fem.sg girl small.masc.pl) ‘a small boy
and girl’, using the masculine plural form of the adjective ‘small’. Hebrew distinguishes
feminine and masculine forms in singular and in plural. The masculine plural
form is used to refer to a group of mixed linguistic genders, for animates, and for
inanimates. For instance, sus means ‘(male) horse, stallion; horse in general’ and
sus-a (horse-fem.sg) refers to a mare. Sus-im (horse-masc.pl) could refer to either
horses in general or only to stallions. Sus-ot (horse-fem.pl) can refer only to mares. If
two humans—a male (of masculine Linguistic Gender) and a female (of feminine
Linguistic Gender)—are coordinated, the masculine plural forms are used. Example
2.5 illustrates the masculine plural pronoun hem ‘they’ (in the function of a ‘present
tense copula’), the masculine plural noun yeladim ‘boys, children’, and a masculine
plural adjective tovim ‘good’, to refer to a girl and a boy. 27

2.3 Markedness and Linguistic Genders 27
2.5 yael ve-xaim hem yelad-im tov-im
Yael(fem.sg) and-Haim(masc.sg) they.masc.pl child-masc.spl good-masc.pl
‘Yael (a girl) and Haim (a boy) are good children’
This criterion—known as gender resolution—will not apply if no genders are
distinguished in the plural (as in German, Russian, or Manambu).
In all-female groups, either feminine or masculine plural forms can be used. Yishai
Tobin—who used to teach a university course for speech pathology students—
comments that in classes consisting of just women both masculine and feminine
plural forms were used. This resonates with a preference for using masculine forms
(including the second person singular masculine pronoun ata) as generics by female
speakers of Hebrew and of Palestinian Arabic (we return to this in §11.5.2).
The generic use of masculine gender forms—especially anaphoric ones—is linked
to the issue of an ‘unmarked’‘generic’ man, and the presumed supremacy of the male
Social Gender. Linguistic Gender is seen as reflecting social attitudes: we return to
this in §7.6.
Feminine—rather than masculine—is the functionally unmarked choice in a few
languages, scattered around the world—in South America, North America, New
Guinea, and Australia. Jarawara, from the small Arawá family in Brazil, has two
genders: feminine and masculine. The gender of all humans is determined by their
sex. For other nouns, the principles of gender choice are complex, and only partly
predictable. For instance, all liquids are feminine (and so are the recently introduced
liquids such as gasoline or beer). Bahi ‘sun, thunder’, abariko ‘moon’, and amowa
‘star’ are masculine by mythological association: they are regarded as mythical men.
About 53 per cent of plant names are feminine. Most garden plants with sweet taste
are feminine (sami ‘pineapple’, jifari ‘banana’), and starch foods are masculine.
Gender is marked in agreement on adjectives, demonstratives, and verbs, and also
in possessive constructions. There is strong evidence for feminine being the functionally
unmarked gender choice.
First, all pronouns are marked on the verb with the feminine affix. R. M. W. Dixon
reports that having to refer to himself by a pronoun with feminine agreement was
one of the difficulties he found in speaking Jarawara. Secondly, the interrogative
himata ‘what?’ always takes feminine gender agreement (if the speaker does not
know the gender of the item they are asking about), as in
2.6 himata ama-ri?
what
‘What is it?’
be-CONTENT.INTERROGATIVE.FEMININE
Thirdly, if a general statement is made, a noun with no specific male or female
reference will take feminine gender agreement. For instance, the noun wahati ‘person’

28 2 Linguistic Gender and its expression
can refer to a man or woman of the Jamamadi tribe. A specific Jamamadi person will
be referred to with feminine or with masculine agreement, depending on their sex.
A generic Jamamadi person will require feminine agreement.
That is, the masculine gender in Jarawara is the marked choice, only used to refer
to a human male or something classified as belonging to the masculine gender. The
feminine—the functionally unmarked choice—is used to refer to human females, or
things classified as belonging to the feminine gender, and also where no gender is
specified. The feminine gender is functionally unmarked—yet it would be premature
to conclude that feminine Linguistic Gender as a default choice automatically accords
women a special Social Gender status. We will see, in §7.6, that a particularly important
woman can be referred to with the marked masculine gender, as if she were promoted
to masculinity. 28
Markedness relations may operate differently depending on a noun’s meaning. In
Manambu, the feminine gender is functionally unmarked with regard to non-human
nouns. A question ‘what is it?’ involves feminine agreement:
2.7 agwa ja:p-al?
what thing-3fem.sg.NOMINAL.CROSS.REFERENCING
‘What is it?’
Feminine forms are used in general statements. A common expression, al-al (that.
fem.sg-3fem.sg.NOMINAL.CROSS.REFERENCING), roughly translatable as ‘that’s it, that’s
how things are’, involves only feminine forms. Nominalized verbs and generic
statements always require feminine forms. So does the loanword tenkyu ‘thank
you’. Feminine forms of adjectives can be used as adverbs, e.g. kwasa ‘small (feminine
singular), a little bit’.
With human referents the situation is not the same. Plural is a functionally
unmarked choice for humans in content questions about the identity of a person
of unknown sex. In 2.8, a speaker asks who is going to come. Just one guest was
expected, and the answer was expected to be in the singular. Plural is a neutral, or
functionally unmarked choice.
2.8 sə ya-kəna-di?
who
‘Who will come?’
come-FUTURE-3pl.SUBJECT
Third person singular masculine or feminine pronoun is never used to refer to an
unknown human. But the masculine noun du ‘man’ has generic overtones of ‘a
human’ in general. ‘Humanness’ of non-prototypical humans is defined as being du
‘man’. Manambu women tell stories about underground villages where dead people
live in a manner similar to Europeans. In one such description, the speaker stressed
that those who live in these villages are humans, and said: du-adi (man-3pl) ‘they are
human’, using the word du in a generic sense ‘human’. When I asked my classificatory

2.4 Linguistic Genders and their labels: envoi 29
sister about mischievous spirits who live in fig trees and what they look like, I received
an answer: du-adi, ñan-a-pək (man-3pl, we-LINKER-LIKE) ‘they are human (literally
man), like us’.
As we will see in Chapter 3, masculine gender in Manambu is associated with
cultural significance. And the term for ‘man’, du, has a broader range of meanings
than ta:kw ‘woman’. Feminine gender is functionally unmarked for all effects and
purposes—except in reference to humans. It is tempting to correlate the generic use
of ‘man’ in Manambu with the male dominance across the board—we return to this
in §7.6 and §11.5.
Not every diagnostic context for ‘markedness’ can be applied to every language.
Dyirbal has a semantically based system of four genders, as we saw in §2.2.1. Neither
of them can be considered functionally unmarked. Dixon (2015: 42) elaborates on
this:
I tried to investigate every possibility. For a group of people of mixed sex it seems that either
bayi [Class I which includes male humans] or balan [Class II which includes female humans]
may be used. There may be a tendency to use balan if most of the group are female, or if a
senior member is a woman, and similarly for bayi and men. But these are only tendencies.
What about a baby in the womb whose sex is not known? In English, the pronoun it is often
used here…Bala (Class IV, residue or neuter) could not be employed, since in Dyirbal
animates are confined to bayi and balan.…I enquired about this and was told that as soon
as a baby was conceived the parents would think of it as male or female, thus referring to it by
bayi or balan.
In many languages, masculine Linguistic Gender is the functionally unmarked
choice—in §7.6 and §11.5.1 we return to the generic uses of masculine pronouns in
English, and masculine forms in other languages. How can the markedness of a
Linguistic Gender correlate with the status of a Social Gender? Markedness, status,
and power in Linguistic Gender choice are the topic of §7.7.
2.4 Linguistic Genders and their labels: envoi
Linguistic Genders are one of the means of categorizing nouns through language.
The choice of a Linguistic Gender may be semantically transparent only to a partial
extent, and correlate only in part with the Natural Gender, or sex. The assignment of
Linguistic Gender to a noun may be based on its form (rather than its meaning). This
creates opacity in gender choice, especially for non-humans where Natural Gender is
irrelevant. In languages where gender choice is opaque it is often hard—if not
impossible—to offer an exhaustive explanation for gender choice for every noun.
The terms used to ‘label’ genders are thus misleading—a gender labelled ‘feminine’
would include more than just females, and the one labelled ‘masculine’ more than
just males. The term ‘neuter’ tends to refer to a gender which includes inanimate (or

30 2 Linguistic Gender and its expression
irrational) beings, or a residue gender whose semantic basis is difficult to capture. In
each case, a linguist needs to carefully investigate the language-specific principles
obscured by the nickname-like terms.
In a number of languages from New Guinea and Africa, where gender is assigned
to humans based on their sex, physical properties play a role in choosing the
Linguistic Gender of non-humans. This is what Chapter 3 is about.
NOTES AND SOURCES
1. Noun class and gender are often used interchangeably. To avoid confusion, I will use the
term Linguistic Gender throughout. See Kilarski (2014: 59–272), on the history of study of
Linguistic Gender from antiquity onwards. Typological features of genders and noun
classes and the history of terms can be found in Aikhenvald (2012b, 2004, 2006, and
especially 2000: 18–80); and also Corbett (1991, 2014a, 2014b).
The term gender has been used in a different way. For instance, in the Athabaskan
linguistic tradition the term gender is used to refer to verbal classifiers which mark
agreement with intransitive subject or transitive object, and characterize the referent
noun in terms of shape and form (Aikhenvald 2000: 15). The term noun class may refer
to different ways of marking plural on nouns (something which involves classifying nouns
on formal rather than semantic grounds: an example is in Dimmendaal 2000).
2. See Lehmann (1993: 20); Asher (1985: 136–7).
3. There appears to be a misleading trend, especially among German speaking scholars,
to claim that English has no gender (e.g. Motschenbacher 2010). This shows disregard
of anaphoric gender in English and obscures the difference between English and those
languages where gender is restricted to derivation and lexicon, such as Hungarian, Estonian,
and Indonesian.
4. See Aikhenvald (2012a: chapter 10) for South American languages; Aikhenvald (2000: 369,
439) for Finnish and other languages.
5. See Schapper (2010a: 182–4, 2010b: 420–1).
6. See Heine (1982a: 194); further examples in Aikhenvald (2000: 38–9).
7. That anaphoric gender tends to be chosen on semantic rather than formal grounds, and that
the choice of agreement gender may follow formal principles (based on non-semantic
choice of gender) is captured by the Agreement Hierarchy (see Corbett 1991: 225–41).
This Hierarchy offers principles behind the choice for semantic rather than form-based
gender agreement. It identifies four types of agreement position:
attributive < predicate < relative pronoun < personal pronoun,
stating that ‘as we move rightwards along the hierarchy, the likelihood of semantic
agreement will increase monotonically’. This statement assumes that anaphoric gender is
a type of agreement gender (an issue which is debatable).
8. In a few languages, more genders are distinguished in non-singular numbers: see Steinhauer
(1986) on Biak; a discussion in Aikhenvald (2000: 243–5) and Aikhenvald and Dixon
(2011b: 181); Plank and Schellinger (1997). In Mangarrayi and Jawoyn, Australian languages’
gender markers are fused with case (Dixon 2002: 508–9).
9. This agrees with Greenberg’s universal 44, that ‘if a language has gender distinctions in the
first person, it always has gender distinctions in the second or third person’ (Greenberg

2.4 Linguistic Genders and their labels: envoi 31
1963: 95). Tocharian A, an extinct Indo-European language, is an exception: it is the only
language in the family to have feminine and masculine forms of first person full pronouns:
näş ‘I (male speaker)’ and ñuk ‘I (female speaker)’. There are no genders in second person
pronouns; two genders are distinguished in demonstratives used in lieu of third person
pronouns: see Matasović (2004: 63) and Jasanoff (1989).
10. See Yadav (1996: 63–5, 124–7).
11. See Matasović (2004: 51–2) on Romanian; Mallinson (1986: 244–6). Surface realizations
(or forms) of genders have been referred to as ‘target’ genders; and agreement genders as
‘controller genders’: see Aikhenvald (2000: 45–7), Corbett (1991: 151). A similar example
of two gender forms with three agreement possibilities comes from Telugu, a Dravidian
language (Krishnamurti and Gwynn: 1985: 56–8). In a number of Cushitic languages,
plural is recognized as ‘the third gender’: see Fraizyngier (2012: 523), Mous (2012: 364–9),
and Tsegaye, Mous, and Schiller (2014).
12. See Heine (1982a: 195) on different gender systems for pronouns and for other agreeing
modifiers in African languages, including Swahili; Harvey (1992, 2002) on Gaagudju,
Aikhenvald (2007) on Baniwa, and Aikhenvald and Green (2011: 405–7) on Palikur.
Further instances are discussed in Aikhenvald (2000: 68–75).
13. Aikhenvald (2000: 22–4, 2012a: 281–2) and references there; Kibrik (1996).
14. See Haas (1941: 65).
15. See Dixon (1972: 308–12; 2015: 27–43). Dyirbal has been the subject of numerous reanalyses
in secondary sources. That by Lakoff (1987) involves a severe misrepresentation.
16. Bani (1987), Haas (1941: 65); Mithun (1999: 101).
17. Algonquian languages are one of the large language families in the USA stretching from
Alberta and Montana in the west of the United States to the Atlantic, and from the northeast
(Labrador) to South Carolina; see Mithun (1999: 98–9), Kilarski (2007), Craik (1982),
Dahlstrom (1995), Darnell and Vanek (1976), Goddard (2002), Quinn (2001), Straus and
Brightman (1982), and Wolfart (1973).
18. See Dahlstrom (1995), on Fox; see also Wolfart (1973: 21–3) on gender in Plains Cree.
19. See Spitulnik (1989: 207).
20. See Schaub (1985: 175), Denny (1976), Aikhenvald (2000: 281–3).
21. See Pensalfini (2003: 159–68).
22. See Aikhenvald (2000: 25); Heine (1982a: 200); Bhatia (1993: 216–17).
23. There may be a historical explanation for this phonological principle of gender assignment.
Proto-Arawak had a feminine (non-masculine) affix *-u/o which is still preserved in
the form of Palikur gender-sensitive suffixes, and in some lexical items, e.g. tino ‘woman’,
cf. Proto-Arawak *c᷈ina-ru ‘woman’.
24. See Matasović (2004: 70); see also Heine (1982a: 200) on semantic principles of Linguistic
Gender assignment in Irakw, a South Cushitic language, and formal ones for the rest.
25. See Foley (1991: 119–64) on Yimas, Nekitel (1986) on Abu' Arapesh, and an overview
of Arapesh languages in Dobrin (2012: 84–109). Alliterative agreement has also been
described as one of the agreement techniques for a number of varieties of Baïnouk, a West
Atlantic language (Cobbinah 2010).
26. See the discussion of markedness in Schane (1970), Corbett (1991: 291), Aikhenvald (2000:
54–6), and Aikhenvald and Dixon (2011b: 177); Bulygina and Shmelev (1996: 103),
Rothstein (1973), Jakobson (1984) on Russian; see Wales (1996), Curzan (2003), Newman
(1997).
27. See Tobin (2001: 183–5) on the functionally unmarked masculine gender in Hebrew; Sa'ar
(2007) on masculine forms used by women speakers of Hebrew and Palestinian Arabic.

32 2 Linguistic Gender and its expression
See Sadiqi (2003) on masculine plural forms with reference to mixed-sex groups in
Moroccan Arabic.
28. See Dixon (2004: 285–7) on Jarawara (Arawá family), a few other South American
languages (Aikhenvald 2012a, 2012b), a few Australian languages (e.g. Wangkumara, the
western dialect of Wagaya, Kala Lagaw Ya, and Murrinhpatha: Alpher 1987: 174, Breen
1976a: 336, 1976b: 340, 590, Walsh 1976: 150–6), and a number of Northern Iroquoian
languages (Chafe 2004).

3
Round women and long men:
physical properties in Linguistic
Gender
Meanings of Linguistic Genders typically involve humanness, animacy, and sex—
female versus male. In a number of languages across the world, animates and inanimates
are assigned genders depending on their shape and size. We have identified two
scenarios.
SCENARIO 1. Masculine Linguistic Gender is associated with larger size and long slender
shape and feminine gender is associated with smaller size and roundish shape: this is
the topic of §3.1.
SCENARIO 2. Feminine Linguistic Gender is associated with larger size and masculine
gender is associated with smaller size; this is the topic of §3.2.
In §3.3 we reconcile these seemingly opposing scenarios. Linguistic Gender choice
can also correlate with value and importance of an object, partly reflecting Social
Gender stereotypes: this is the topic of §3.4.
3.1 Small round women and long slender men
Round shape and small size are associated with the feminine Linguistic Gender in a
dozen languages along the Sepik River in New Guinea. The ‘opposite’—long and
large—objects are considered ‘masculine’. Speakers of Manambu, a Ndu language
from this area, are ‘gender-proud’. When asked what is special about their language,
speakers reply: ‘in our language, everything in the world is either woman-like or
man-like’. 1 The language has two genders, masculine and feminine. Gender is
covert: it cannot be inferred from the form of the noun. One knows what gender a
noun belongs to by the form of an agreeing demonstrative, adjective, possessive
pronoun, or verb. Choice of gender for humans is always based on their Natural
Gender, or sex: asa:y ‘father’ and du ‘man’ are masculine and amaey ‘mother’ and
ta:kw ‘woman’ are feminine (Linguistic Genders for humans can be reversed under
How Gender Shapes the World. First edition. Alexandra Y. Aikhenvald.
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.

34 3 Round women and long men
exceptional circumstances—see §§7.1–3). If there is one word for both sexes,
gender on accompanying words helps tell them apart: ma:m is the word for
‘older sibling’. Amasculineforma-də ma:m (that-masc.sg elder.sibling) means
‘that older brother’, and a feminine form a ma:m (that+fem.sg elder sibling)
means ‘that older sister’.
Outside the realm of humans, gender choice is based on the object’s physique.
A big dog or a big pig will have masculine gender, no matter whether they are male or
female: a-də bal (that-masc.sg pig) means ‘that (big) pig’, and a bal (that+fem.sg pig)
‘that (small) pig’. One can always specify the sex by adding a noun, amaey (‘mother’)
for females, and asa:y ‘father’ for males: that is, amaey bal (mother pig) is a female
pig, and asa:y bal (father pig) is a male pig.
Some animals, birds, and heavenly bodies have special roles in myths. Məd
‘cassowary (a non-flying bird)’ is always feminine gender, because cassowaries are
conceived of as mythical women (who turned into cassowaries later). This association
is widespread in New Guinea. Similarly, mæ:n ‘bird of paradise’ is a woman in
disguise, and cannot be made masculine. In §8.1, we turn to mythological associations
in the choice of Linguistic Genders.
In choosing gender for inanimates, shape comes into play. Long or large objects
are treated as masculine, and small or round ones as feminine. Val ‘canoe’ is
masculine if big, and feminine if small. Væy ‘spear’ is masculine due to its inherent
long shape and size; it is feminine when the word is used to refer to a small spear or a
shotgun. A house of usual size is referred to as feminine; an unusually big house is
masculine. The Manambu people who have visited England all agree that Buckingham
Palace is a ‘masculine type’ house because it is really big. The House of
Parliament in Canberra was also judged big enough to be considered ‘masculine’.
Skyscrapers are ‘masculine’, and so are traditional big men’s houses which used to be
large and impressive (kara:b). A long road (ya:b) will be referred to with masculine
gender, and a short one with feminine gender.
Some inanimates have a ‘typical’ shape and size. For instance, ar ‘lake’, kabak ‘stone’,
ya:l ‘belly’, and ab ‘head’ are typically round, and so assigned to the feminine gender.
Changing gender of any of these implies their unusual, and even grotesque, size.
A woman has an affair with a snake, and becomes pregnant. Her belly gets bigger
and bigger and then becomes grotesquely huge. It is then referred to as ‘masculine’. The
change of gender is shown in example 3.1. Gender markers are in bold face.
3.1 lə-kə ya:l ata numa-ø məy ta:l
she-POSS+fem.sg belly then big-fem.sg very become+3fem.sgSUBJECT.PAST
a numa-də ya:l ta:d lə-kə-də ya:l
then big-masc.sg belly become+3masc.sgSUBJECT.PAST she-POSS-masc.sg belly
‘(Her) belly then became very big (feminine), here is an unusually big (masculine)
belly, her (masculine) belly’

3.1 Small round women and long slender men 35
In an English commentary this was described by a speaker as a ‘big man-type belly’.
Changing genders can describe unusual situations. ‘Head’ is feminine because of
its round shape. When a person has a headache, the head then feels heavy and
unusually big, and can be referred to with masculine gender. In 3.2, a speaker
complained of a headache, talking about her head with masculine Linguistic
Gender:
3.2 wuna-də ab kagəl yi-na-d
I-masc.sg head pain have-PRESENT-3masc.sg.SUBJECT
‘My head (masculine because of its unusual weight) is aching’
Linguistic Gender of nouns denoting natural phenomena and physical states is
based on their extent. A darkish night (ga:n) will be feminine, and a night which
is completely dark is masculine. Hunger (ka:m) is feminine; a very strong hunger is
masculine. Quantity determines the Linguistic Gender choice of mass nouns. A little
money or a little blood will be referred to with feminine gender; if there is a lot of
money, or blood, the Linguistic Gender will be masculine. The term gwalugw
‘patrilineal clan’ is masculine. A numerically small clan can be assigned feminine
gender. Figure 3.1 summarizes the principles of gender choice in Manambu.
Some entities have a typical shape—and so a typical Linguistic Gender. All trees
are masculine due to their height and vertical stance. Their fruit is feminine independent
of the shape, e.g. mi: ‘tree (generic)’ and ma:s ‘betel nut tree’ are masculine;
təkəmi ‘fruit (generic)’ and ma:s ‘betel nut fruit’ are feminine. An unusually big fruit
can be masculine.
If a noun refers to a concept central to the Manambu mythology, Linguistic
Gender is chosen based on ‘mythological association’ and important property
principles—something we saw for Dyirbal in §2.2.1. Moon (bap) is feminine, because
it is conceived of as a mythological woman; an alternative, endearing, name for it is
bap-a-ta:kw (moon-LK-woman) ‘lady moon’. Similarly, stars (kugar) are conceptualized
as women belonging to the clan group which is associated with sun, moon, and
light. No matter how big the moon or a star, they will never be referred to with
masculine gender.
Similar examples come from a number of neighbouring languages. Alamblak,
from the Sepik Hill family, is spoken to the south-east of Manambu. 2 Unlike
Manambu and its relatives, gender is marked on the noun itself. All nouns which
denote females and short, squat, or wide objects are feminine and have a form
marked with a feminine suffix -t. Tall, long, and slender objects are masculine gender,
and take the masculine suffix -r. Terms for inanimate objects which host the feminine
suffix include terms for house, fighting shield, and trees ‘which are typically relatively
shorter and more squat than other trees’. Those which occur with the masculine
suffix include ‘terms for arrows, signal trumpet, typically tall slender-growing trees,
large string bag varieties, etc.’

36 3 Round women and long men
human
adults
children and babies
Natural Gender
Natural Gender
age, size
males, and males by association:
masculine
females: feminine
male: masculine
female: feminine
relatively big and/or old: masculine
relatively small and/or young: feminine
non-human
animals
inanimates
lower animates
mass nouns
size
size and shape
quantity
large: masculine
small: feminine
long and large: masculine
round and small: feminine
big: masculine
small: feminine
natural phenomena
extent/intensity
FIGURE 3.1 Gender assignment in Manambu
complete/intense: masculine
non-complete/non-intense: feminine
A noun can occur with an ‘atypical’ gender suffix. This indicates that the object is
unusual as to its size. The word for ‘house’ usually occurs with feminine suffix. It can
occur with the masculine suffix(kuñ-r house-masc): then the house is perceived as an
unusually long one. If the term for a slit-gong drum is marked with the masculine
suffix(nërwi-r), this indicates that the drum is ‘unusually slender, which implies that
it was made incorrectly and does not sound good’. For humans and animals, gender
switch indicates change in Natural Gender, e.g. yima ‘person’, yima-r (person-masc)
‘man’, yima-t (person-fem) ‘woman’.
Sare (or Kapriman), another language from the Sepik Hill family, also has two
Linguistic Genders, masculine and feminine. 3 Linguistic Gender assignment to
nouns referring to lower animates (including rats, birds, and insects) depends on
shape and size: small, short, or rounded referents, such as mice, will be assigned to
the feminine Linguistic Gender, and those animates which are big, tall, or slender—

3.1 Small round women and long slender men 37
for instance, dogs—will belong to the masculine Linguistic Gender (unless the sex of
the animate needs to be differentiated).
The choice of the Linguistic Gender of an inanimate is determined by shape and
size: objects which are big, tall, long, or slender are treated as ‘masculine’, and those
which are small, short, or squat are feminine. Some examples are in 3.3.
3.3 MASCULINE: BIG, TALL, LONG, SLENDER FEMININE: SMALL, SHORT, SQUAT
wuni-r ‘big house’
xudari-r ‘big axe’
seboxu-r ‘high table’
wuni-s ‘small house’
xudari-s ‘small axe’
seboxu-s ‘squat table’
As Sumbuk (1999: 116) putsit,‘howbig,tallorhighanobjectmustbetobe
accorded one gender or the other is determined by the traditional expectation of
the speaker with regard to a particular object. With regard to the size of things
like houses, canoes and containers, it is the capacity of the referents of these nouns
to hold persons or things that determines what gender the noun takes. If a house,
for instance, has space enough only for a single family (parents, children and
grandparents), then it would normally be described as a small house, thus taking
the feminine gender. If on the other hand, the house has a capacity for several
families, then it would normally be described as a big house and be accorded the
masculine gender.’
Somewhat different principles of Linguistic Gender choice apply for different
semantic groups of nouns—similarly to Manambu, in Figure 3.1. Gender choice for
plants depends on whether the plant is young or mature. A term for a young plant
will belong to the masculine gender (despite its being typically small and short).
A term for a mature plant (normally big and tall) will belong to the feminine gender
(e.g. popo-r (pawpaw-masc) ‘young pawpaw tree’, popo-s (pawpaw-fem) ‘mature
pawpaw tree’): Sumbuk (1999: 117) suggests that mature plants ‘are regarded as
the female of the plants since they bear fruit’; young plants, ‘despite being short and
small, which are features that we would ordinarily associate with feminine gender,
are regarded as masculine by the Sare because of their inability to bear fruit’.
A similar principle applies to fruit and seeds. Mature fruit are feminine, and small
immature ones are masculine, e.g. gonxa-r (banana-masc) ‘immature banana’,
gonxa-s (banana-fem) ‘mature banana’, popoyioka-r (pawpaw-masc) ‘immature pawpaw’,
popoyioka-s (pawpaw-fem) ‘mature pawpaw’. The ‘important property’ of
maturity overrides the shape- and size-based choice of Linguistic Gender.
The choice of Linguistic Gender for inanimates depends on shape and size in a few
more languages across the island of New Guinea and surrounds. In Wära, an isolate
from the Fly River region in New Guinea, long objects are assigned to the masculine
gender and round objects, and objects consisting of multiple parts, are feminine. In
Olo, from the Torricelli grouping in New Guinea, feminine gender is associated with
the ‘small nature’ of the entity. The choice of Linguistic Gender for inanimates in

38 3 Round women and long men
Savosavo, a Papuan language from the Solomons, depends on the size of the object:
small objects are feminine, and large ones are masculine. 4
Correlations between the size and the Linguistic Gender go beyond New Guinea.
The majority of Afroasiatic languages of North Africa have two Linguistic Genders—
masculine for males and feminine for females. Gender is marked on the noun itself, in
agreeing modifiers, on verbs, and in personal pronouns. The marker of the feminine
Linguistic Gender is homophonous with diminutive marking across the family. And
the feminine Linguistic Gender tends to correlate with small size of an object. Ayt
Seghrushen (Moroccan Berber) men belong to the masculine Linguistic Gender and
women to the feminine gender (e.g. arba ‘male child’, t-arba-tt ‘female child’). For
inanimates and lower animates (whose sex is irrelevant and difficult to determine
anyway), a masculine form denotes ‘something bigger than the feminine’:
3.4 t-fus-tt (feminine) ‘little baby hand’ fus (masculine) ‘hand’
t-ams̍at̍t̍(feminine) ‘thigh’ amsad̍(masculine) ‘very big thigh’
t-axxuy-t (feminine) ‘louse’ axxuy (masculine) ‘very big louse’
Changing Linguistic Gender for humans and sex-differentiable animals implies
difference in size (and not in Natural Gender, or sex):
3.5 t-amət̍t̍ut̍t̍(feminine) ‘woman’ amət̍t̍ud̍(masculine) ‘very big woman’
t-yis-tt (feminine) ‘little stallion’ yis (masculine) ‘stallion, horse’
t-aymar-t (feminine) ‘mare’ aymar (masculine) ‘very big mare’
t-ɣat̍t̍(feminine) ‘goat (female)’ ɣad̍(masculine) ‘big female goat, difficult to
handle’
Masculine Linguistic Gender can have overtones of something big, or an ‘augmentative’.
Some nouns have three forms: augmentative (masculine), a neutral
(masculine or feminine), and diminutive (feminine). For instance, an Arabic loanword
lkursi ‘chair’ is masculine (and neutral in size). A feminine form t-akursi-tt
means ‘little chair’, and a masculine form a-kursi means ‘very big chair’ (the Arabic
article l- is replaced with Berber a-, an erstwhile article-like morpheme). 5
The association between the feminine Linguistic Gender and small size on the
one hand, and masculine Linguistic Gender and larger size, on the other hand, is
a pervasive feature of another subgroup within Afroasiatic—the Omotic languages
of Ethiopia (including Wolaitta, Maale, Benchnon, Sheko, Dime, and Dizin). In
3.6a, from Wolaitta, the masculine Linguistic Gender of a noun with a human
referent (‘child’) indicates male sex. The feminine Linguistic Gender in 3.6b indicates
female sex:
3.6a naʔá-y y-iísi
child-MASC:NOMINATIVE come-3masc.sg:PERFECTIVE
‘The boy came’

3.1 Small round women and long slender men 39
3.6b naʔ-íya y-aásu
child-FEM:NOMINATIVE come-3fem.sg:PERFECTIVE
‘The girl came’
An inanimate ‘tree’ in 3.7 is large in size—this is signalled through its masculine
Linguistic Gender:
3.7 mítta-y kúnd-iísi
tree-MASC:NOMINATIVE fall-3masc.sg:PERFECTIVE
‘The (big) tree fell’
The tree in 3.8 is small—and this is reflected in its feminine Linguistic Gender:
3.8 mítt-íya kúnd-aásu
tree-FEM:NOMINATIVE fall-3fem.sg:PERFECTIVE
‘The little (feminine) tree fell’
Feminine Linguistic Gender with nouns with inanimate referents in Benchnon, also
Omotic, indicates that the object is small, as in 3.9:
3.9 két-àɕ-ī versus két-àn-ā
house-THIS.MASC-NOMINATIVE.MASC
‘this (big) house’
house-THIS.FEM-NOMINATIVE.FEM
‘this (small) house’
The feminine Linguistic Gender may be used to vent a speaker’s ‘derogatory attitude’,
as in 3.10.
3.10 tá tʃˈāmá-ī versus tá tʃˈāmá-ā
I shoe-NOMINATIVE.MASC I shoe-NOMINATIVE.FEM
‘my shoe’
‘my shoe I don’t care about’
A similar principle applies to other subgroups of Afroasiatic. 6 In the Harar dialect
of Oromo, a Cushitic language, if the speaker wishes to ‘evaluate the referent as
major’, the noun acquires masculine gender marking. If a noun is assigned to
masculine gender but refers to something evaluated as minor, it may be treated as
feminine, via agreement. In eastern and southern varieties of Oromo, ‘if a speaker
wishes to indicate that the referent is marked as augmented with respect to a
particular aspect in the domain of discourse…the noun may be treated as masculine’
(Clamons 1993: 275–6).
The noun ablee ‘knife’ triggers feminine agreement on the modifier (tun ‘this:
feminine’) and the predicate (doom-tuu ‘dull-feminine’):
3.11 ablee-n tun doom-tuu
knife-SUBJECT.TOPIC this:fem dull-fem
‘This knife is dull’

40 3 Round women and long men
In 3.12, the same noun, ‘knife’, is masculine. This is reflected in the agreement
forms of the modifier and the predicate. The Linguistic Gender shift indicates that
the knife is considered big by the speaker:
3.12 ablee-n xun doom-aa
knife-SUBJECT.TOPIC this:masc dull-masc
‘This (big) knife is dull’ 7
Similar principles have been described for other languages in East Africa. In
Turkana, an East Nilotic language from Kenya, changing the gender of a noun
correlates with the size of a referent. 8 In 3.13, the same root can occur with the
masculine, the feminine, and the neuter prefix. The referent of the masculine noun is
typically big, the referent of the feminine noun is smaller in size, and the referent of
the neuter noun is very small:
3.13 e-mor-u` ‘rocky mountain, big stone’ —masculine
a-mor-u` ‘hill, stone’ —feminine
i-mor-u` ‘pebble’ —neuter
In Camus, from the Maa subgroup of East Nilotic, small objects are classed as
feminine, and big items are assigned to the masculine gender. In Khwe, a Central
Khoisan language spoken in Namibia (and also in Angola, Botswana, and Zambia),
inanimate referents are assigned to the masculine Linguistic Gender if they are
long, high, narrow, pointed, big, or strong, and to feminine Linguistic Gender if
they are short, small, round, thick, broad, or weak. In Katcha, a Kadugli-Krongo
language spoken in the Nuba mountains in the Sudan, masculine Linguistic Gender
includes long, thick, and solid objects; feminine Linguistic Gender subsumes
hollow, deep, flat, and thin objects. 9 In Piapoco, a North Arawak language from
Colombia, inanimate objects which are roundish, such as càyuwa ‘hat’, belong to
the feminine gender. Objects which are slender and long, such as énu ‘shotgun’, are
masculine. 10
Shape and size can be a gendering dimension in a language without Linguistic
Gender. Nungon is a Papuan language spoken in Morobe province. Natural Gender
is expressed in lexical items referring to man, woman, boy, girl, and a plethora of kin
terms. Hannah Sarvasy (2016) reports that ‘the only domain in which inanimate
objects are classed as male or female based on size is that of yams’. Referring to yams
as ‘man’ or as ‘woman’ is ‘largely based on the size and shape of the tubers—smaller
and rounder are female, while longer and larger are male’.
So far we have seen how the choice of feminine Linguistic Gender may correlate
with small size, and also round (and sometimes thin and narrow) shape. This is
especially so for inanimates and lower animates (whose Natural Gender is not known
or not important). Masculine Linguistic Gender is assigned to objects which are large,
long, and strong. We now turn to a different scenario.

3.2 When ‘women’ are larger than ‘men’ 41
3.2 When ‘women’ are larger than ‘men’
In a number of languages, large size (and also squat extended shape) is a correlate of
feminine gender. Small size and slender shape are a correlate of masculine gender.
The choice of Linguistic Gender in Tiwi, an Australian language, is based on the
Natural Gender of humans and animals: all males are assigned to the masculine
gender and all female human beings and animals are assigned to the feminine gender.
Masculine and feminine genders are marked on the noun, and also in agreement and
anaphora.
For inanimate entities, the choice of Linguistic Gender is determined by physical
size and shape. The semantic features correlating with the masculine gender are
‘small’, ‘straight’, and ‘thin’. The features correlating with the feminine gender
are ‘large’, ‘round’, and ‘ample’. These features are also used to assign genders to
those animals who play no role in the traditional mythology, and whose sex is hard to
determine. Similar kinds of objects or animals are distinguished through Linguistic
Gender: a smaller or thinner one is classed as masculine, and a larger one as feminine.
Some examples are in 3.14:
3.14 MASCULINE: SMALL, THIN FEMININE: LARGE, ROUND, AMPLE
waliwalini ‘small ant’
miaɹti ‘small pandanus’
muŋkwani ‘small stone axe’
man̪t̪ani ‘small stick’
t̪uŋkwaliti ‘narrow-headed spear’
waliwaliŋa ‘large ant’
miaɹiŋa ‘large pandanus’
muŋkwaŋa ‘large stone axe’
man̪t̪aŋa ‘large stick’
aɹawuniŋkiɹi ‘broad-headed spear’
In Tiwi, ‘all trees are feminine…because they are large, round and ample compared
with mere sticks which are all masculine’. 11
In Mali, a Baining language from East New Britain (Papua New Guinea), lower
animates and inanimates are assigned to the feminine Linguistic Gender if they are
large in size. If they are of average size, they are assigned to the masculine Linguistic
Gender. This is shown in 3.15.
3.15 MASCULINE: AVERAGE SIZE FEMININE: UNUSUALLY LARGE SIZE
thinem-ka ‘a fish’
bang-ka ‘a house’
lulen-ga ‘a dugout canoe’
thinem-ki ‘a big fish’
bang-ki ‘large building’
lulen-gi ‘a long dugout’
A three-way contrast between large (feminine), regular (masculine), and small
(diminutive) nouns can be seen in some borrowings from Tok Pisin, an Englishbased
Creole. A large serving spoon is a feminine sipun-ki, a dessert spoon (normal
size) is a masculine sipun-ka, and a tiny teaspoon is a diminutive sipun-ini. Feminine
plastik-ki refers to a largish plastic bag, or a large cocoa, flour, or rice bale. Masculine
plastik-ka is a plastic container of smaller capacity, or a plastic shopping bag. 12

42 3 Round women and long men
Hadza is a language isolate from central Tanzania. Masculine and feminine
genders are marked on nouns, and also in agreement and anaphora. Humans are
assigned to genders according to their sex. The choice of genders for other nouns
follows a number of principles; among them shape and size. In many instances, a
masculine noun refers to a smaller object, and a feminine noun (marked with the
feminine suffix -ko) to a larger one, as shown in 3.16.
3.16 MASCULINE NOUNS: SMALLER SIZE FEMININE NOUNS: LARGER SIZE
ʔato ‘small axe’
waɬi ‘testicle’
muʔa ‘twig, thin stick’
ʔato-ko ‘large axe’
waɬi-ko ‘swollen testicle’
muʔa-ko ‘stick’
Size is not the only physical property relevant for choosing a Linguistic Gender in
Hadza. Masculine nouns can refer to a longer and thinner object; their feminine
counterparts are short and stout, as shown in 3.17.
3.17 MASCULINE NOUNS: THIN, LONG, DEEP FEMININE NOUNS: SHORT, STOUT, SHALLOW
ts'ahu ‘long, thin tail’
ts'aho-ko ‘short, stout tail’
ʔutume ‘long spear’ ʔutume-ko ‘short spear’ 13
Just like other Omotic languages, Hamar has a feminine and a masculine gender
which are chosen on the base of sex (or Natural Gender) for humans and sexdifferentiable
animals, e.g. k'úl-ta ‘he goat’ and k'úllo ‘she-goat’. Hamar appears to
be unique in its family in one way. The feminine Linguistic Gender is associated with
large size and importance. Smaller items belong to the masculine Linguistic
Gender—see 3.18.
3.18 MASCULINE NOUNS: SMALL FEMININE NOUNS: LARGE
amā ‘small field’
nuta ‘individual fire’
ammo ‘large field’
nuno ‘large fire’
Inanimates assigned to the masculine Linguistic Gender are less significant than
feminine ones: for instance, goitā ‘path: masculine’ is a little used path, and goinno
‘path:feminine’ is the main path. 14 We return to Lydall’s explanation for this feature
of Hamar in §3.4.
In Cantabrian Spanish 15 Linguistic Gender assignment to higher animates, and
humans, generally follows their sex, or Natural Gender. We find pairs such as bellu
‘steer’, bella ‘heifer’; lobu ‘male wolf ’, loba ‘female wolf ’; oveju ‘ram’,andoveja ‘ewe’.
Inanimates and lower animates which are either tall or of smaller size, or of narrow
shape, or of vertical orientation are assigned to masculine Linguistic Gender. Objects
which are unusually large, wide, horizontal, or small and squat are feminine. Some
examples are in Table 3.1.
The choice of Linguistic Gender for inanimates correlates with size in various
forms of Standard Spanish. In many cases, the feminine noun refers to a larger entity,

3.3 Physical properties in Linguistic Gender choice 43
TABLE 3.1. Semantic features in gender choice for nouns in Cantabrian Spanish
MASCULINE
MALE, e.g. hiju ‘son’
SMALLER SIZE, e.g.
anguilu ‘very small eel-like fish’
tortu ‘small cake prepared in frying pan’
montón ‘stack (of hay)’
NARROW, e.g.
picu ‘metal spike’
VERTICAL, e.g.
coteru ‘rising mountain meadow’
castru ‘large protruding rock’
TALL, e.g. tree names:
manzanu ‘apple tree’
naranju ‘orange tree’
FEMININE
FEMALE, e.g. hija ‘daughter’
UNUSUALLY LARGE OR AVERAGE, e.g.
anguila ‘eel’
torta ‘larger cake prepared in oven’
montona ‘very large stack of hay’
WIDE, e.g.
pica ‘hammerlike instrument with a wide head’
HORIZONTAL, e.g.
cotera ‘mountain meadow (flat)’
castra ‘large flat rock (horizontal)’
SQUAT, SMALL, e.g. fruit names:
manzana ‘apple’
naranja ‘orange fruit’
e.g. garbanzo ‘chick-pea’ (masculine), garbanza ‘large species of chick-pea’ (feminine),
panero ‘small basket’ (masculine), panera ‘large oval basket’ (feminine), cesta
‘large hand basket’, cesto ‘smaller basket’. In some cases, a masculine rather than a
feminine noun would refer to a larger entity, e.g. barreno ‘large-size auger (a boring
tool)’, barrena ‘auger’. 16
3.3 Physical properties in Linguistic Gender choice: contrasting
the two scenarios
The choice of Linguistic Gender based on shape, size, and other physical properties of
an entity involves meanings summarized in Table 3.2.
A complex of physical features is at play in the choice of Linguistic Gender. They
include shape, dimensionality, and size. Size as a gendering parameter offers contradictory
results. In some languages the feminine Linguistic Gender is applied to larger
objects, in others, to smaller ones. Shape and dimensionality are more consistent.
Feminine Linguistic Gender subsumes objects that are round and squat, and also
horizontal. Masculine Linguistic gender includes narrow and elongated, and
also vertical things.
Shape and dimensionality are salient semantic parameters in all noun categorization
devices (or classifiers). 17 Linguistic Gender is one of these linguistic categorization
of entities. Orientation (vertical versus horizontal) is a preferred semantic
parameter for verbal classifiers, and especially classificatory verbs. A classificatory

44 3 Round women and long men
TABLE 3.2. Physical properties in Linguistic Gender choice
FEMININE GENDER MASCULINE GENDER EXAMPLES DISCUSSED HERE
short, squat, wide
objects
small, flat objects
large, wide, round
objects
big, tall, long, slender
objects
large, tridimensional
objects
small, straight, thin
objects
Languages of New Guinea
(Manambu, Iatmul, Gala, Yalaku,
Kwoma, Wosera, Boiken, Alamblak,
Sare, Olo, Wära, Savosavo);
Afroasiatic languages (including
Cushitic, Omotic, Amharic (Semitic));
East Nilotic;
Khwe (Central Khoisan)
Abau (isolate, New Guinea area)
Tiwi (Australian region)
larger objects smaller objects Mali (Baining); Hadza (isolate);
Hamar (Omotic)
larger, wide, horizontal,
squat objects
smaller, narrow,
vertical, tall objects
Cantabrian Spanish
positional verb ‘sit’ can be used with referents judged to be squat and horizontal,
including women. A classificatory verb ‘stand’ is used of tall, large, and vertical
referents, including men. This type of system has been described for numerous
Papuan languages. 18 In §5.1, we turn to the ways in which gender meanings are
expressed through noun categorization devices other than Linguistic Gender.
The shape and dimensionality associated with the choice of masculine Linguistic
Gender can be conceived as involving ‘phallic’ imagery. The objects associated with
feminine Linguistic Gender will then be in some sense ‘opposite’. In this way,
Linguistic Gender choice involves projecting physical characteristics typical of female
and male Natural Genders onto the inanimate world in general. Such Linguistic
Gender assignment is based on ‘important property’ (in line with examples in §2.2.1).
The roots of this ‘important property’ lie in conventionalized images associated with
the Natural Gender—to which we turn in Chapter 8.
3.4 Beyond mere physique: attitude, value, and importance
in Linguistic Gender
Changing Linguistic Gender of an inanimate may correlate with speakers’ attitudes.
Linguistic Gender may be a token of endearment, or disdain—see §3.4.1. Or it may
reflect value and cultural importance—see §3.4.2.

3.4 Beyond mere physique 45
3.4.1 Endearment and disdain through Linguistic Gender
Small roundish objects can be regarded as particularly cute, or dear to the speaker.
Overtones of endearment are common to many diminutive forms in the world’s
languages: in Portuguese, cabecinha ‘little head’ is a diminutive of cabeça ‘head’; this
word is often used to mean ‘dear little head’,or‘head of a dear little person’, such as a
child or a baby. 19
We can recall, from §3.1, that the feminine Linguistic Gender marker doubles as a
diminutive across the Afroasiatic language family. The association between the
feminine Linguistic Gender and endearment is pervasive. In Amharic, ‘almost any
noun can be treated as feminine if occasion calls for it…Thus, yih mas'haf “this book
(masculine)” is normal, while yicc mas'haf “this book (feminine)” might refer to a
favourite little booklet’ (Cowley et al. 1976: 84). In Oromo (see §3.1, and Clamons
1993: 276–8) something considered cute, or regarded affectionately, will be assigned
to the feminine Linguistic Gender. An example is at 3.19.
3.19 waan-ti tun jiidh-tuu
thing-FEM.SUBJECT.TOPIC this.fem wet-fem
‘This (cute little) thing is wet’
The masculine Linguistic Gender is the opposite of feminine: it will have pejorative
overtones. A ‘nasty thing’ belongs to the masculine Linguistic Gender:
3.20 waan-i xun jiidh-aa
thing-MASC.SUBJECT.TOPIC this.masc wet-masc
‘This (nasty) thing is wet’
The feminine Linguistic Gender in Ayt Seghrushen, a Berber language, has
overtones of affection: the masculine afus is ‘hand’, and the feminine t-fus-tt is ‘little
baby hand’ (as we can recall from example 3.4). The masculine Linguistic Gender
may indicate that something is grotesquely big, ugly, or hard to handle. The form
t-ɣat̍t̍(feminine) refers to a female goat. The masculine counterpart is ɣad̍(masculine)
which means ‘big female goat, difficult to handle’. The feminine noun tit̍t̍‘eye’
has a masculine counterpart at̍t̍aw ‘very big eye’ with pejorative overtones. The
feminine tamziyda ‘mosque’ can be made masculine; the resulting form amziyda
means ‘a ridiculously big mosque’.
We mentioned in §2.2.2, that gender choice in Palikur is determined by the form
of the noun, and also its meaning. Feminine Linguistic Gender for lower animates
can be associated with small size. Masculine Linguistic Gender implies larger size,
and may have overtones of unpleasantness. Rat is a small animal, but is assigned
masculine gender because it is looked upon as dirty and bad. But a cute little baby rat
would be referred to as feminine. Turtles are usually feminine; but a turtle which is a
nuisance and has to be got rid of would be referred to as masculine; all insects are

46 3 Round women and long men
masculine in spite of their small size, according to an explanation by a native
consultant, ‘because none of them are any good for food and all they do is bother
people, eat crops and cause sickness’. 20
In Cantabrian Spanish, the masculine Linguistic Gender has overtones of thinness,
meagre proportions, and generally smaller size than the feminine Linguistic Gender
(Table 3.1). Changing Linguistic Gender correlates with the speaker’s attitude.
A coarse and bumpy road would be referred to as masculine, and a smooth one as
feminine. The term oveju ‘male sheep’ can be used in a derogatory sense to refer to a
particularly meagre exemplar of the species: a cattle raiser was reminiscing about a
particularly bad meal he’d once had, consisting of a sheep’s head, a few green
vegetables, and a few potatoes, and remarked that the sight of the head of that
oveju (male sheep) ‘peering at him from the pot was something he would never
forget’. In Spanish, the feminine form oveja is normally used in a generic sense; and
here a ‘shock or humour’ has resulted from the selection of the male form based not
on the sex of the unfortunate animal, but as a ‘deprecative expression of the
revulsion’ felt by the speaker (Holmquist 1991: 60). Along similar lines, hiju míu,
literally, ‘my son’, was used to refer to a young girl about 12 years of age, a female not
yet developed, in a deprecatory reference. A little girl was treated on a par with a nonhuman.
The masculine Linguistic Gender of non-humans in Lokono, an Arawak language
from Guyana, has overly positive overtones. The feminine Linguistic Gender is the
opposite. Animals and birds which are thought of as having a ‘positive personality’
are masculine—they include turtles and hummingbirds. Domestic animals to which
speakers have a special attachment, for instance, a dog, are masculine; however, one’s
neighbour’s dog (whom one does not particularly like) is more likely to be feminine.
Nice and cute animals are masculine, while bigger animals are feminine (van Baarle
and Sabajo 1997). This is—so far—the only example of a language I have been able to
find where masculine Linguistic Gender is associated with endearment and special
affection. Linguistic Gender reversals for humans have similar overtones—we return
to these in §7.2.
Things can be seen differently. A small object may be treated as substandard and
not good enough. Feminine Linguistic Gender for some nouns in Ayt Seghrushen
has pejorative overtones. An Arabic loan (originally from French camion) lkamyun
(masculine) means ‘truck’. The Ayt Seghrushen form akaymun (masculine) means
‘gigantic truck’; the feminine form t-kamyun-t means ‘little worthless truck’. The
masculine form of ‘female goat’ has overtones of ‘a big female goat difficult to
handle’. 21 We saw in 3.10 how in Benchnon, an Omotic language, a shoe assigned
to the feminine Linguistic Gender is the one that the speaker doesn’t care about.
Table 3.3 summarizes the opposite attitudes of speakers, encoded in the feminine
and masculine Linguistic Genders of animals, inanimates, and—in Spanish—young
children.

3.4 Beyond mere physique 47
TABLE 3.3. Endearment and disdain through Linguistic Genders
FEMININE GENDER MASCULINE GENDER EXAMPLES
endearment, affection
negative overtones (being
grotesquely big, nasty, or
meagre and substandard)
Oromo, Palikur,
Ayt Seghrushen,
Cantabrian Spanish
substandard, derogatory neutral Ayt Seghrushen, Benchnon
negative personality, dislike positive personality, affection Lokono
In summary: small size and the associated Linguistic Gender can express cuteness
and endearment. Or it may involve ‘belittling’ something. The opposite Linguistic
Gender may then refer to a substandard and annoying entity. The overtones of
endearment or a negative attitude for shape- and size-based Linguistic Gender are
associated with an evaluative aspect of size: something small and thin can be nice and
dear, or no good at all. Something too big tends to be ridiculously grotesque. This is
what we have so far seen for non-humans; the effects of Linguistic Gender reversals
for men and women are the topic of §§7.2–3. We now turn to the associations
between Linguistic Gender, importance, and cultural value, and their possible links to
Social Gender concepts, and societal values.
3.4.2 Value and importance in Linguistic Gender
Typical features of Social Gender can be projected onto entities which have no Natural
Gender, in the form of Linguistic Gender assignment. The choice of Linguistic Gender
may reflect the place of the referent in the conceptual system of beliefs and rituals. We
start with an example of how the masculine Linguistic Gender—associated with large
size of an object—may reflect ritual life, and cultural importance.
Manambu society—just like other groups in the Sepik—is male-oriented. Traditional
ownership of land and of valuables is inherited through one’s father’s line. So
is membership of a clan. When a woman gets married, she is expected to move in
with her husband’s line. (A husband who moves in with his wife after marriage can
be ridiculed as a ‘female husband’ of sorts: see §7.2.) The centrepiece of Manambu
culture is, traditionally, male cults (including initiation rituals, now obsolete).
A man’s reputation and status used to depend primarily on esoteric knowledge of
totemic names, names belonging to the subclan, spells, and myths. Male proficiency
in these matters was crucial for men’s prowess in ritual debates—including debates
on ownership of names and of the land—and also magic and sorcery. In the past,
men’s central role used to depend on achievements in wars against neighbouring

48 3 Round women and long men
groups and in head-hunting. 22 According to the tradition, men have access to
esoteric information concerning secret names and spells, and have the upper hand
in political and ritual affairs. Women are not supposed to have access to highly
valued traditional knowledge (though some of them in fact do), and are denied active
participation in traditional activities such as the yam ritual and name debates.
This traditional importance of ‘male-hood’, or masculine Social Gender, is iconically
reflected in the assignment of masculine gender based on ‘important property’.
Names of rituals are masculine. It is not appropriate to refer to Saki ‘name debate’
with feminine gender (notwithstanding the fact that name debates may involve
female and male names, and can be relatively short in time). Similarly, Kəkətəp,
the ritual which involves mortuary payments, is also masculine (no matter whether
the deceased was a man or a woman, or how many valuables change hands).
Along similar lines, kara:b ‘large ceremonial house’ and sa:y ‘ceremonial house for
uninitiated men’ are both masculine. Ceremonial houses in present-day Avatip (the
main village of the Manambu) are smallish (much smaller than normal dwellings).
But referring to a ceremonial house with feminine gender is considered unacceptable
and even offensive.
Gender assignment based on importance is relevant for terms to do with the art of
speech. The term ma:j ‘story’ requires masculine agreement, if it is a traditional story,
or if it refers to head-hunting raids performed by men, no matter how long it is. If it
refers to a casual story or a biography of someone it is likely to be feminine. Similarly,
ba:gw ‘performance, dance’ is masculine only when it refers to a traditional activity.
A casual performance—such as a dance party organized by Manambu expatriates in
Canberra, or someone casually playing the guitar—is feminine. ‘Gender-proud’
speakers of Manambu refer to unimportant stories as ‘woman-type’ stories, and to
important ones as ‘men-type’.
Association with ‘male’ and ‘female’ Social Gender attributes determines the
Linguistic Gender of culturally important notions, and objects. Certain cultural
concepts have inherently ‘female’ reference, and are always assigned to the feminine
gender. Others are associated with masculinity, and trigger masculine agreement.
The notion of ja:m ‘a set of hereditary magical and ritual powers’ is personified as a
named female spirit, whose role is to punish incest and violation of the marriage
principles. Each subclan has a væy, its ancestor, with a literal meaning ‘spear’. The
connotations of væy are masculine and phallic (which goes together with its elongated,
‘masculine’, shape), while ja:m is represented as a womb. Together væy and ja:m
‘signify the “male” and “female” aspects of a group’s social identity’ (Harrison 1990:
33). This agrees with the assignment of masculine and feminine genders to them.
It is traditionally believed that human bones (ap) are formed from father’s semen
and transmitted through one’s father. In contrast, blood (ñiki) derives from mother’s
womb blood and is transmitted by matrifiliation (Harrison 1990: 33). Consequently,
ap ‘bone’ is assigned masculine gender, while ñiki ‘blood’ is usually feminine (unless

3.4 Beyond mere physique 49
it comes in large quantities). The noun ap also carries the connotations of centrality
and importance, another corollary of masculine gender. Thus, speakers derive the
name of Avatip, considered the most important of the four Manambu villages, from
ap-a təp (bone-LINKER village) ‘the strong, large, central village’ (lit. the bone-village).
Cultural roles—and the status of the masculine Social Gender—can be considered
the basis of Linguistic Gender choice. This is also how speakers explain them: an
important story relating a clan’s origin or wartime exploits will be referred to as
‘man-type’, and a biography or a fairy-tale as a ‘woman-type’. Similarly, in Yangoru
Boiken, a language related to Manambu, ‘size is a gendering dimension for referents’:
large objects are masculine and small ones are feminine. Masculine gender is
associated with strength, importance, and value, and gender is assigned to culturally
significant rituals, stories, and objects in numerous other languages spoken in
societies focused on male activities and cults. 23
The male ‘supremacy’ in cultural activities, and the status of masculine Social
Gender is reflected in Linguistic Gender even if there are no obvious correlations
between the assignment of Linguistic Gender and an object’s size. In Oyda, an
Omotic language, Linguistic Gender is assigned to ceremonial objects—drums—
based on their importance. A small roundish drum—described as ‘the major one,
the leader of all drums’ is assigned masculine Linguistic Gender; and a very large
oval-shaped drum which has less importance belongs to the feminine Linguistic
Gender (Amha 2013).
Hamar, from the Omotic subfamily of Afroasiatic, is unusual for its family in that
the feminine, rather than the masculine, Linguistic Gender correlates with the
referent’s importance—as we saw in §3.2. In Lydall’s words (1988: 78),
feminine gender is used to indicate large and major things rather than small and unimportant
ones, while masculine gender is used to indicate small and minor things rather than big and
important ones…Upon telling this to fellow linguists and anthropologists, I have been met
with expressions of disbelief or astonishment. How can the Hamar have arrived at such a
formula? How can a society in which men generally enjoy a higher status than women speak a
language in which masculine gender is equated with smallness?
To explain this apparent puzzle, Lydall (1988: 89) hypothesizes that in agricultural
and other tasks in the life of the Hamar,
any event or enterprise requires, on the one hand, male initiative and determination, and, on
the other hand, female performance and substantiation. Male initiative and determination is
conceived of as small in scale, and occasional, while female performance and substantiation
is large-scale and frequent. These ideas explain why the masculine form of nouns is used to
indicate small, infrequently used or seldom found items, while the feminine form is used
to indicate large, frequently used or commonly found items.
An association of a Linguistic Gender with cultural importance and value appears
to mirror stereotypes of Social Gender. But a word of warning is in order: as Mithun

50 3 Round women and long men
(2014: 159) put it, ‘though grammatical gender may indeed reflect culture, the
relationship is not necessarily immediate and direct’. In Manambu, nouns væy
‘spear’ and ap ‘bone’ are assigned masculine gender—and this also can be thought
of as based on their longish size and shape (and not necessarily associations with
‘masculinity’). Similarly, ya:l ‘womb, belly’—which is typically feminine—is associated
with women. Since it is also typically ‘round’ in shape, its gender assignment can
be attributed to a pervasive correlation between shape and feminine Linguistic
Gender. Moon and stars are mythological women, but they are also roundish in
their shape.
Which one comes first—gender choice by shape, or by mythology? Which one is
contingent upon the other? We can recall that ab ‘head’ is feminine, due to its round
shape, and despite its perceived importance for mental processes (which correlates
with ‘masculinity’ as exponent of ‘importance’). This may imply that shape is primary
in the gender assignment of ‘spear’, ‘bone’, ‘womb; belly’, ‘moon’, and‘star’, and that
the cultural extension to masculinity or femininity is, at least historically, a corollary of
the erstwhile shape-based association. But synchronically, we are faced with Linguistic
Gender choice which reflects a male orientation of values.
Correlations between lack of cultural importance and small size in Manambu do
not imply a subservient and secondary position of women in the society. The spheres
of men’s and women’s knowledge and experience are complementary and separate.
We return to this in §10.1 and §10.5.1.
NOTES AND SOURCES
1. Further details on gender assignment in Manambu are in Aikhenvald (2008; 2012a).
2. Bruce (1984: 97).
3. Sumbuk (1999: 109–23).
4. Riisto Sarsa, p.c. on Wära, McGregor and McGregor (1982: 55) on Olo, Wegener (2012:
63–5) on Savosavo. Similar systems have been described for Abau, a linguistic isolate
spoken to the west of the Sepik River in the adjacent Sandaun Province (Lock 2011:
47–52), Kwoma, from the small Kwoma-Nukuma family (Bowden 1997, Kooyers 1974),
and Iatmul, Gala, Wosera, and Boiken, from the Ndu family (see references in Aikhenvald
2012a). In Yalaku, closely related to Manambu, a small object, a bird, or an animal will be
referred to with feminine gender, and a larger one with masculine gender. A very small male
bat was described as feminine, and referred to as tu-il (man-3fem.sg), literally ‘she is a man’.
5. See Kossmann (2014), on Ayt Seghrushen, and further intricacies of the system; see
Diakonoff (1990), for an overview of Afroasiatic languages.
6. See Seyoum (2008: 43), Hellenthal (2010: 153) on Sheko, Amha (2001: 71–2) on Maale,
Amha (2012: 444) on Wolaitta; Rapold (2006: 182) on Benchnon. See Sasse (1974: 419) on
Galab (or Dasenech: Cushitic); Castellino (1975) on links between small size and feminine
gender across Cushitic languages, and Leger (1998: 207) on Kwami and other Chadic
languages.

3.4 Beyond mere physique 51
7. Hoben (1976: 287). Similarly, in Amharic, an Ethio-Semitic language from Ethiopia, a
noun can be ‘masculine or feminine according to whether its referent is considered to
be large and distant (masculine) or small and intimate (feminine)’: see Cowley et al. (1976:
84), and Pankhurst (1992: 168).
8. See Dimmendaal (1983: 219–21) for Turkana, and similar examples from the related
Maasai.
9. See Heine (1982a: 210–11), Kilian-Hatz (2008: 41–2), Köhler (1962) on Khwe, Heine
(1982a: 205) on Camus and Katcha.
10. See Klumpp (1990: 67).
11. See Osborne (1974: 51).
12. See Stebbins (2005: 102).
13. See Edenmyr (2004: 14–17).
14. See Lydall (1988: 78). Note that in this source she refers to the masculine as ‘minor’ and to
the feminine as ‘major’ gender.
15. Also known as Montañés. See Holmquist (1991); Pountain (2005); Bergen (1980: 52–4);
also see the historical explanation there and in Priestly (1983); see Butt and Benjamin
(2004: 1–15) for Linguistic Gender in Standard Spanish. Similar examples from Italian are
in Ferrari (2005: 39–40), Bonfante (1946: 847) (disputed by Ervin 1962).
16. That nouns belonging to the feminine Linguistic Gender (and ending in -a) denote objects
of larger size may be due to their origin: many such nouns may have originally been neuter
plural or collectives (which also ended in -a in Classical Latin) and later undergone
reinterpretation (Pountain 2005: 334–5); see also Kahane and Kahane (1948–9) on the
productivity of the ‘feminine augmentative’ in Romance languages.
17. An outline of preferred semantic parameters of various noun categorization devices is in
Aikhenvald (2000: 306).
18. See Lang (1975) on Enga, Brown (1981) on Waris; further references in Aikhenvald (2000:
158–60, 166–8).
19. See Mendes (2014a) on how diminutives in Brazilian Portuguese may convey the idea of
something nice and cosy; see also Grandi (2002) on meaning overtones of diminutives and
augmentatives across European languages.
20. See Aikhenvald and Green (2011).
21. See Kossmann (2014), on Ayt Seghrushen.
22. Aikhenvald (2009).
23. Nouns belonging to the feminine Linguistic Gender in Khwe, a Central Khosian language
(Kilian-Hatz 2008: 42), Galab (or Dasenech), a Cushitic language (Sasse 1974), and Angave,
an Angan language from Papua New Guinea (Speece n.d.) are characterized as ‘small, less
important’. Nouns of masculine Linguistic Gender are big and important.

4
What are Linguistic Genders
good for?
Linguistic Genders, as a way of grammatical categorization of nouns, impose a way of
speaking on men and women. The Natural Gender of a speaker of Modern Hebrew,
Russian, Modern Greek, and many other languages with pervasive gender agreement
will be obvious from the gendered forms they will use. A statement ‘I love you’ in
Modern Hebrew has to be specified for gender of ‘I’ (on the verb) and ‘you’ (in the
pronoun). A woman will say to a man, Ani ohevet otxa ‘I love(feminine) you(masculine)’,
and a man will say to a woman, Ani ohev otax ‘I love(masculine) you(feminine)’.
A failure to use the appropriate Linguistic Gender in a language can lead to
spectacular results. In a short story ‘The entertaining episode of the article in
question’ (1925), detective Lord Peter Wimsey solves a burglary mystery by use of
gender. At a London railway station, Lord Peter overhears what appears to be a
young French girl saying to her partner Me prends-tu pour un imbécile? (‘Do you take
me for an imbecile?). Lord Peter notices the masculine form of the indefinite article
‘un’ (rather than the feminine form ‘une’), and this serves to reveal that the supposed
girl is in reality a renowned male jewel thief. As Lord Peter explains, ‘in France, every
male child is brought up to use masculine adjectives about himself ’. For the thief, it
wasn’t possible to kick this habit. Gender gave the game away. 1
Linguistic Gender serves to differentiate the meanings of a noun (and help
distinguish males from females), as we will see in §4.1. It helps understand who
does what to whom—that is, track referents in discourse—this is what we will see in
§4.2. Gender individuates some participants and backgrounds others, and highlights
their properties. Linguistic Gender helps enrich the lexicon (for one thing, creating
male and female pairs). All these functions are shared with other noun categorization
devices—especially numeral classifiers, noun classifiers, and classifiers on verbs. 2
Linguistic Genders are a means of categorizing entities. In the first place, they
differentiate between men and women. In Portuguese, the forms menino ‘boy’ and
menina ‘girl’ differ in their linguistic gender marking and in agreement. Meanings
of genders in German reflect what Zubin and Köpcke (1986) called ‘hierarchical
categorization’ of nouns. The neuter gender tends to express general concepts, so
How Gender Shapes the World. First edition. Alexandra Y. Aikhenvald.
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.

4.1 Variable choice of Linguistic Gender 53
that superordinate nouns tend to be neuter. More specific and less general terms will
be assigned to non-neuter gender (see Table 2.3 in Chapter 2). So, the superordinate
das Tier ‘animal (fauna)’ is neuter. More basic level terms der Fisch ‘fish’ and der Vogel
‘bird’ are masculine, and the names of species, including die Eule ‘owl’ and der Karpfen
‘carp’, are feminine and masculine respectively. 3
In many familiar languages, each noun belongs to just one gender. Gender membership
is fixed—like a non-removable birthmark. In German, die Sonne ‘the sun’ is
feminine. It is nonsensical to try and refer to it as *der Sonne, with the masculine
gender. Yet in many languages gender of a noun can vary—and this highlights the
versatility of what the genders express. 4
4.1 Variable choice of Linguistic Gender
Some nouns have one form for females and for males. These are sometimes called
epicenes. 5 Different forms of Linguistic Gender—in agreement or in anaphora—will
disambiguate their sex. A baby can be referred to as ‘she’ or as ‘he’ depending on
whether it is a girl or a boy. In many languages, epicene nouns occur with forms of
different genders, as a means of their disambiguation. In Dyirbal, jaja ‘baby’ and
ñalŋga ‘young child’ can be specified as either masculine or feminine, depending on
their sex. Similarly, inamatewe ‘child’ in Jarawara, an Amazonian language with two
genders, can be cross-referenced as either masculine or feminine as appropriate.
Archi, a north-east Caucasian language, has four Linguistic Genders. The noun lo
‘youngster’ can be assigned to three of these. When assigned to the masculine class 1
it means ‘boy’. When assigned to the feminine class 2, it means ‘girl’, and when
assigned to class 4 (which includes animals), it means ‘young animal’.
In Dyirbal bimu is both ‘father’s elder sister’ (taking feminine gender marker,
balan) and ‘father’s elder brother’ (used with the masculine marker bayi). The terms
for ‘child’ in Dyirbal distinguish the sex of the parent: daman is a child (a son or a
daughter) of mother, or mother’s sibling, and galbin is child of father or father’s
sibling. Gender markers help distinguish between sons and daughters:
4.1 balan daman ‘mother’s daughter’ balan galbin ‘father’s daughter’
bayi daman ‘mother’s son’ bayi galbin ‘father’s son’
Many kinship terms in Manambu—including ma:m ‘elder sibling’, ñamus
‘younger sibling’, babay ‘maternal grandparent’, yanan ‘child of one’s daughter’
and gwa:l ‘child of one’s son’—can be masculine or feminine. Agreement gender
will disambiguate them: one will say wuna ma:m (my+feminine.sg elder sibling) ‘my
elder sister’ and wuna-də ma:m (my-masculine.sg elder sibling) ‘my elder brother’.
(A sister will be referred to as lə ‘she’, and a brother as də ‘he’.)
Similarly, nouns denoting professions may show variable agreement depending on
whether they refer to men or to women. Doctor in English can be referred to as he or

54 4 What are Linguistic Genders good for?
she depending on the sex of the person. In Russian, vrac᷈‘doctor’ or sudja ‘judge’ will
require feminine or masculine agreement depending on the person’s sex.
The degree of variability in Linguistic Gender depends on the language. In Dyirbal,
each name of an animal has a fixed class membership; however, exceptionally, noun
class assignment can be changed to stress the sex of a particular animal, e.g. ‘to point out
that a certain dog is male bayi guda can be used’ (Dixon 1982: 182). Usually, guda ‘dog’
belongs to Class II (Dixon 1982: 180), and so the ‘unmarked’ usage would be balan guda.
Very occasionally, changing noun class can create a pragmatic effect. In Dyirbal, yara
‘man’ belongs to Class I; and so would be referred to as bayi yara.However,Dixon(1982:
166) reports that a hermaphrodite was once jokingly referred to as balan yara, witha
feminine Class II marker, pointing out his female characteristics. In this case, the
manipulation of Linguistic Gender has a pragmatic, as well as semantic effect.
Variable Linguistic Gender may be important for ascertaining the meaning of a
noun. Kwami is a West Chadic language with two genders, feminine and masculine.
Gender is covert—that is, not marked on the noun itself. It is realized through
agreement on demonstratives (-mè ‘masculine’, -jè ‘feminine’). Gender choice for
human nouns is based on sex. The gender choice of inanimates is more complex.
Nouns which refer to the domestic sphere are typically feminine, and nouns referring
to objects outside the household are usually masculine. Some fruits and plants belong
to masculine gender, but only as long as they grow. As soon as they are harvested and
brought home for manufacturing, their gender changes to feminine, as in the
following examples: 6
4.2 GROWING ON A FARM HARVESTED OR AT HOME (FOR PREPARATION) MEANING
lòolów-mè (masculine) lòolów-jè (feminine) cotton
gùyà-mè (masculine) gùyà-jè (feminine) pumpkin
sháagúm-mè (masculine) sháagúm-jè (feminine) millet
Variable gender marking in Maung (Capell and Hinch 1970: 47–52) serves to
differentiate related meanings of polysemous nouns. Maung has five genders whose
choice is largely semantic. Gender I consists of names of male beings; Gender II of
names of female beings; Gender III includes objects associated with the ground
(except plants); Gender IV subsumes trees and their parts; and Gender V includes
vegetable foods, plants, and items associated with housing. Noun classes are overtly
marked on nouns, and realized through subject and object agreement on verbs,
adjectives, and possessive expressions. This is how variable gender works.
The noun -nimi expresses the idea of a ‘long bone’. Used with a ‘human male’
Gender I prefix, (j)i-nimi means ‘his backbone’; with a Gender II, female human
class, ninj-imi means ‘her backbone’. With a Gender IV prefix, ma-nimi means ‘trunk
of a tree’, and with gender V prefix ad-imi means ‘main radicle’ of a potato plant or
yam vine (Capell and Hinch 1970: 47). Table 4.1 illustrates how noun classes are used
with another noun root, mawur ‘arm, long part of ’.

4.1 Variable choice of Linguistic Gender 55
TABLE 4.1. Variable Gender assignment in Maung
Gender I i-mawur ‘man’s arm’
Gender II ninj-mawur ‘woman’s arm’
Gender III u-mawur ‘tributary of a river’
Gender IV ma-mawur ‘branch of a tree’
Gender V a-bawur ‘tendril of a vine’
In a follow-up study, Singer (2010: 397) gave further illustrations to the semantic
effect of changing genders. The noun jampakang ‘corrugated iron’ is usually
assigned to the masculine gender (Gender I) (like most other European-origin
artefacts). When used as building material, it occurs with Gender V, by analogy with
other items associated with housing which also belong to this gender. Linguistic
Genders allow the creation of new lexical items, and manipulate the meanings of the
existing ones.
Changing Linguistic Genders may have a special discourse effect. Algonquian
languages have two Linguistic Genders—animate and inanimate—whose meaning
is only partly predictable (see §2.2.1). Quite a few inanimate nouns belong to the
animate gender. In Goddard’s(2002: 225) words, ‘In discussing the phenomenon, the
usual and convenient explanation given by speakers is something like, “you talk about
it like it’s something living”, or“it sounds like it’s alive”’. Changing a noun’s gender
from inanimate to animate highlights its agency in a story. The comb (normally
inanimate) shifts to animate as it starts talking to its owner, in a Menominee story.
An inanimate skull becomes animate when it takes on the role of a Rolling Skull, a
mythical ogre. Some names for trees are usually inanimate, but may be referred to with
animate gender if used ‘in the winter story’ or ‘addressed religiously’. 7
In Ojibwe, a grammatically inanimate noun is often treated as ‘animate’ in a story,
‘due to its assuming animate properties, such as ability to speak to exert spiritual
power’, as Valentine (2001: 118) puts it. In one traditional story, a young man
receives a blessing from a mirror, which later on provides him with protection in
battle. When the mirror first visits the man, it is referred to with inanimate gender.
Later on, when it confers its blessing upon the man, it becomes animate. As a speaker
of Ojibwe put it, ‘in Odawa [a dialect of Ojibwe] the concept of animateness is
limitless. It can be altered by the mood of the moment, the mood of the speaker, the
context, the use, the circumstances, the very cosmos or our totality.’ 8
Bunaq, an Austronesian language spoken in Timor, has two genders—animate
and inanimate marked in object prefixes on verb and in demonstratives. The term for
‘clan’ and terms for vehicles are inanimate. But they can be reclassified as animate, if
the fact that they are carrying humans is emphasized. Nouns referring to natural
elements are inanimate. If human agency is involved in their creation or behaviour,

56 4 What are Linguistic Genders good for?
they can be reclassified as animate. So, il ‘water’ is inanimate if it refers to water
drawn from a river. If the same noun refers to water artificially flooded into a rice
paddy from a river, it is treated as animate. 9
Variable gender with humans may have overtones of value and of attitude.
Women in Oneida, a Northern Iroquoian language, may be referred to with feminine
gender if they are small and graceful, and with neuter, if they are big and aggressive. 10
In Swahili, a Bantu language, most nouns occur with prefixes of just one class, with a
few notable exceptions. M-zee means ‘old person’ (and contains the human class
prefix m-). This prefix can be replaced with ki- (inanimate class); the resulting form
ki-zee means ‘scruffy old person’. Switching genders can thus be a mark of intimacy
or derision. Gender choice may correlate with politeness. Addressing an older
woman with feminine gender in Lak, a north-east Caucasian language from Daghestan,
is not as polite as addressing her with a more general gender which subsumes a
larger class of nouns, including inanimates (see Khaidakov 1963: 49–50). This is
similar to what we find in Mohawk, an Iroquoian language (see §7.3, and Mithun
2014: 138).
We return to the effects of gender reversals for women and men in §§7.2–3.
Variable Linguistic Genders can help differentiate objects in terms of size and
shape. We saw in Chapter 3 how Linguistic Gender choice in Cantabrian Spanish
accords with the Natural Gender of humans and animates. Inanimates and lower
animates which are either tall or of smaller size, or of narrow shape, or of vertical
orientation are assigned to masculine gender. Objects which are unusually large,
wide, horizontal, or small and squat are feminine. Variable gender may reflect the
shape and the state of the object. In Ket, a language with masculine and feminine
gender, a growing tree is masculine, and a cut-down tree, or a log, is feminine; an
upright tree is masculine, and a tree with a curved trunk is feminine. In Khwe, a
Central Khoisan language, an inanimate noun can be allocated to masculine or
feminine gender depending on its shape: masculine is associated with big, long,
rectangular, and feminine with small, round, broad. 11
Savosavo, a Papuan language spoken in the Solomon Islands, has two genders
marked on verbs, pronouns, and demonstratives. Feminine gender includes humans,
birds, and animals. Masculine gender covers males and inanimates. An inanimate
noun can be assigned feminine gender, if it is very small. Molo ‘knife’ is usually
masculine. A small knife can be referred to as feminine. Gender variation may have a
further, pragmatic function: if a noun is particularly salient in a story, it then acquires
feminine agreement. 12 The meanings of shape and size in gender choice correlate
with stereotypical features associated with Natural Gender. They may also reflect the
Social Gender stereotypes—that is, the relative position of men and women within a
society, and their importance: this is what we saw in §3.4.2.
Variable gender can be of a different nature. A recent loan may not as yet have
settled into a gender. The word kivi ‘kiwi fruit’ in Russian is sometimes treated as

4.2 Linguistic Gender in discourse 57
neuter (based on its form and morphological features: it is an indeclinable noun
ending in a vowel), and sometimes as masculine (based on the analogy with frukt
‘fruit’ which is masculine).
Nouns with similar meanings may differ in gender only. In German, the masculine
der See means ‘the lake’, and the feminine die See means ‘the sea’. The masculine der
Band means ‘the volume (of a book)’, the neuter das Band means ‘the ribbon, tape,
band, wavelength (of a radio)’, and the feminine die Band means ‘the (rock)band’;
the masculine der Leiter means ‘the director, head, leader, manager’ and the feminine
die Leiter means ‘the ladder, stepladder’. 13 Nouns which differ only in their gender
can be semantically linked. The word gabal in Dyirbal is feminine (balan gabal) in
its meaning ‘egret, heron’. The same form gabal is neuter (bala gabal) with the
meaning ‘sand, sandbank’. Herons live and walk on sand—this could be the connection
between the two words which differ in gender only. The blossom on a tree is
maŋga—this is treated as belonging to the edible gender if it will develop into an
edible fruit (balam maŋga); any other blossom is neuter—bala maŋga.
Dyirbal had a special Mother-in-law, or Jalnguy, style used to communicate with
in-laws. This register was much less specific than the everyday register. In many
cases, the Mother-in-law register has just one term where the everyday language had
several. This is where gender helps disambiguate the homonyms in Jalnguy.
The everyday language style has three different words for ‘rainbow’, ‘wompoo
pigeon’, and ‘flame kurrajong’ tree. One term covers all the three in the Mother-inlaw
style. The wompoo pigeon was given the same name as the rainbow because of its
colour being like the green of the rainbow. The flame kurrajong tree was held to be
sacred to the rainbow snake—this explains it having the same name. Three gender
markers distinguish the three terms: the rainbow is masculine, the pigeon is feminine,
and the tree is neuter:
4.3 Everyday style Mother-in-law or Jalnguy style
rainbow bayi yamani bayi gagilbarra
wompoo pigeon balan bagamu balan gagilbarra
flame kurrajong tree bala dila bala gagilbarra
Linguistic Gender may correlate with how discourse is organized, and it may be
instrumental in identifying who is doing what.
4.2 Linguistic Gender in discourse
The use of Linguistic Gender may depend on the definiteness and topicality of the
noun. This is a feature Linguistic Gender shares with other systems of noun categorization.
14 In Motuna (a Papuan language from Bougainville in Papua New Guinea)
the verb has to agree in Linguistic Gender with its object only if the object is the topic

58 4 What are Linguistic Genders good for?
of discourse. The object nii ‘I (masculine)’ is topical: ‘I’ is what the text was about.
The verb bears the masculine agreement marker.
4.4 nii Aanih-ki tangumu-u-ng
I(MASC):ABS Aanih(FEM)-ERG slap+1sgO+3sgA-NEAR.PAST-MASC
‘Aanih (a female name) slapped me (topic).’
In 4.5, the subject, Aanih, is topical and its feminine gender appears on the verb:
4.5 Aanih nii tangumu-i-na
Aanih(FEM) I(MASC):ABS slap+1sgO+3sgA-NEAR.PAST-FEM
‘Aanih (a female name) (topic) slapped me.’ 15
Overt gender marking on a noun can signal its definiteness or specificity. In Gola,
a West Atlantic language of Liberia, gender-marking prefixes and suffixes act similarly
to definite articles. A definite noun is marked for gender, an indefinite noun is not,
e.g. kul ‘atree’, ke-kul ‘the tree’; gbalia ‘a dwarf antelope’, o-gbalia-a ‘the dwarf antelope’
(note that here noun class is marked by a combination of a prefix and a suffix).
The presence of an overt gender marker can correlate with a specific individuated
referent, and its absence with a more generic, or indefinite, referent. In Turkana, an
East Nilotic language from Kenya, genders are marked on nouns themselves (we can
recall from §3.1 how changing gender of a noun may correlate with the size and the
state of an entity). The overt gender prefix on nouns ‘is frequently omitted with
names of animals in folk tales when the names are used in their generic sense’. The
word for ‘lion’ is ɛ-ŋatùɲ, and for ostrich ɛ-kalèes, if used in isolation (Dimmendaal
1983: 221). At the start of a story about two friends, a lion and an ostrich, the two
words were used without their prefixes, as ŋatùɲ kà kalèes ‘lion and ostrich’ (generic
or indefinite). 16
Overt gender marking can depend on the function of a noun in discourse. In the
Australian languages Nunggubuyu and Warray the presence of a gender prefix ona
noun is correlated with definiteness or givenness, and its absence indicates focus and
foregrounding. In Wardaman noun class prefixes are used when introducing a new
participant and for maintaining reference to the noun throughout the story. 17
Linguistic Gender helps disambiguate what is being talked about. In German, Krug
‘jug’ is masculine, and Schale ‘vase’ is feminine. The jug is referred to as ‘he’ and the
vase as ‘she’. As a consequence, there is no ambiguity in German concerning what got
broken. 18
4.6 Der Krug (masculine) fiel in die Schale (feminine),
ART.MASC.SG jug fell into ART.FEM.SG bowl
aber er (masc) zerbrach nicht
but he broke NEG
‘The jug fell into the bowl but it (the jug, lit. he) didn’t break.’

4.3 The utility of Linguistic Gender 59
In English, ‘it’ would refer to both ‘jug’ and ‘bowl’, and the disambiguation will be
less straightforward. Different forms for different genders allow us to communicate in a
succinct fashion. Failing to understand gender cues may produce an embarrassment.
Dixon (2016: 83–4) tells an amusing anecdote about the discourse potential of
genders. He went to a shop in Porto Velho, a regional capital in Brazil, to buy one
mirror (um espelho, masculine) which one of his Indian friends had asked for. The
shopkeeper produced a whole box of mirrors (uma caixa, feminine), took one mirror
from it, and quoted the price, saying Dez reais por uma (ten real for one: feminine)
‘ten reais for one’. Dixon protested that ten reais (about five dollars) was too much for
one small mirror. The shopkeeper continued bargaining the price down, and then, in
Dixon’s words,
Then it hit me. My gender-starved brain has failed to pick up that he said por uma. The
feminine form of the indefinite article must be referring to the box, uma caixa. The price being
quoted was for a box of ten mirrors. If he had been giving the price for a mirror, um espelho,he
would have said por um. I purchased one mirror, for one Real, and left the shopkeeper shaking
his head at foreigners who can’t understand simple language. If the transaction had been
conducted in English, a price would have had to be quoted for a box or for a mirror. It can be
seen that marking gender on articles, in Portuguese, makes for a more succinct discourse.
Linguistic Genders help animate the world. A noun whose Linguistic Gender is
feminine can be depicted, and conceptualized, as a ‘woman’. In Portuguese and
Russian the word for ‘death’ is feminine. And so, in folk tales and pictures, Death
is represented as a woman. In Portuguese, it is politely referred as Dona Morte, Ms
Death. In German, the word for death is masculine—and it is depicted as an old man.
In §§8.1–2 we turn to the imagery associated with Linguistic Gender.
Not only does Linguistic Gender help in categorizing entities. Linguistic Gender is
a repository of beliefs and social practices. We can recall, from §2.2.1, that in Dyirbal
birds are classed as members of gender 2 (‘feminine’, balan) rather than bayi on the
basis of their non-human animate status. This is so because birds are believed to be
the spirits of dead women. In numerous languages of the Sepik area of New Guinea,
cassowary and bird of paradise are mythological women; and they are invariably
assigned to the feminine gender. More on this in §8.1.
4.3 The utility of Linguistic Gender
As Jakobson (1959: 116) put it, ‘languages differ essentially in what they must convey,
and not in what they may convey’. Like any grammatical category, Linguistic Gender
imposes partly semantic, and partly formal restrictions, on what one is to say in a
particular language. Some linguists, feminists, and language learners bemoan the fact
that Linguistic Genders are imposing and restrictive—especially so in languages where

60 4 What are Linguistic Genders good for?
their basis is not fully semantically transparent. We can recall from Chapter 2 that
Mädchen ‘girl’ in German belongs to neuter gender (based on its morphological make
up: a diminutive suffix -chen which automatically requires neuter gender choice). The
word for ‘turnip’, Rübe, belongs to the feminine gender (partly based on its phonological
form, as most nouns ending in -e are feminine). In his scurrilous attack on ‘The
Awful German Language’, Mark Twain complained that ‘In German, a young lady has
no sex, while a turnip has’. Too much is being made of a purely linguistic fact. And we
can recall, from §2.1.2, that a girl in German—especially if she is grown up—will be
referred to with the pronoun ‘she’ reflecting her Natural Gender and a mature female
Social Gender position (see the details in Braun and Haig 2010). That Linguistic
Genders have no function (other than hindering second-language learners) is simply
wrong. Their versatility, and the ways in which they may mirror myths and beliefs—
and can be a source of poetic metaphors—makes languages with Linguistic Gender far
more fun to work with than the ones which lack them. If languages are in contact, they
will be likely to come to share important features—Linguistic Genders among them, as
we will show in Chapter 6. We saw in Chapter 3 how positive features associated with
masculine gender can be projected onto culturally important entities. The meanings
associated with typical properties of males and females and their attributes get transferred
onto essentially gender-less, inanimate objects and phenomena. Expressive
power in poetry and metaphor is gained through deploying Linguistic Genders—as
we will see in Chapter 8. Those who accuse languages with Linguistic Genders of being
unnecessarily ‘irrational’ or ‘ornate’ simply haven’t looked far enough—or perhaps are
trying to be different for the sake of it. 19
Some feminists—speakers of languages with anaphoric gender—complain that the
obligatory choice between ‘he’ and ‘she’ in singular pronouns makes it difficult to
promote gender equality: speakers are forced to make a gender choice even if they
would rather avoid it. But the way the choice is made has another side to it. This
choice may give us an insight into the ways Social Gender stereotypes work, and how
speakers perceive them. The ways in which the choices change reflect changes in the
position of women within the society. These visible changes through language can be
monitored and traced more easily than if they are covert. Linguistic Gender is among
the lucky few grammatical categories which reflect social and cultural meanings, and
is amenable to linguistic engineering and conscious change, as a social environment
transforms. As we will see in subsequent chapters, Linguistic Gender offers a window
into the dynamics of social developments, and perhaps the human mind. Speakers
are aware of Linguistic Gender meanings and implications—this is something they
eagerly discuss, and may wish to reform—we return to this in §11.5. It is not the case
that languages with Linguistic Gender are more ‘sexist’ than those without it; we will
see in §5.3 and then in Chapter 11 how linguistic sexism can find its expression in
languages with scarcely any Linguistic Gender.

4.3 The utility of Linguistic Gender 61
Linguistic Gender is one of the means of categorizing nouns, and it shares its
many functions with other noun categorization devices of other kinds, especially
numeral, noun, and verbal classifiers. When applied to humans, Linguistic Gender
correlates with Natural and Social Gender. And these can be, and usually are,
expressed through other means. ‘Gender-less’ languages—for some, the lucky
escapees from the ‘prison-house’ of Linguistic Gender and its limitations—hardly
ever achieve the ‘ideal’ of gender neutrality. Igbo has no Linguistic Gender and, in
Amadiume’s (2015: 89) words,‘has not built up rigid associations’ between various
attributes, objects, and genders. ‘The genderless word mmadu, humankind, applies
to both sexes. There is no usage, as there is in English, of the word “man” to represent
both sexes, neither is there the cumbersome option of saying “he or she”, “his or
her”, “him or her”.’ However, this ‘linguistic gender’ neutrality did not stop men in
the traditional Igbo society from having more powers and more privileges than
women. With the advent of colonial powers, the inequality of Social Genders became
more drastic. 20
We now turn to how meanings of sex, animacy, humanness, and also Social Gender
find their expression beyond the strict division of nouns into Linguistic Gender classes.
NOTES AND SOURCES
1. Sayers (1995: 24, 33).
2. Functions of genders as part of a continuum of noun categorization devices (including
numeral classifiers, noun classifiers, verbal classifiers, possessive and relational classifiers,
and additional types) are discussed in Aikhenvald (2000: chapter 12), and summarized in
Aikhenvald (2006). The same functions are discussed (sometimes with different examples)
in Contini-Morava and Kilarski (2013). Additional discussion is in Weber (2000).
3. See Holmquist (1991) on Cantabrian Spanish, Ferrari (2005) on Italian, Pountain (2005)
on Spanish; Audring (2013) offers examples of the individuating functions of gender
choice in Dutch.
4. Typologists who choose to focus on the ‘canons’ of a few European languages relegate
systems with variable gender assignment to a questionable status of ‘non-canonical
systems’ (e.g. Corbett 2014b: 122). But if one ventures beyond one’s preconceptions,
languages with variable assignment of Linguistic Gender (which are not just the ‘exotic
few’) offer a fertile field for investigating the meanings of genders.
5. Aikhenvald (2000: 41); Matthews (2010: 124). And see §6.6 on one of the meanings of
‘common gender’.
6. See Leger (1998: 206) and Dinslage, Leger, and Storch (2000: 125).
7. Goddard (2002: 221).
8. Similar examples, from Fox, are Goddard (2002) and Dahlstrom (1995).
9. Schapper (2010a: 182–4).
10. See Abbott (1984), Mithun (1999: 100–1; 2014), Bonvillain (1973: 86–7), and §7.3.
11. Dul'son (1968: 62ff.) on Ket; Kilian-Hatz (2008: 41–2) for Khwe; also see Hagman (1977:
23) for similar phenomena in Nama.
12. Wegener (2012: 63–5).

62 4 What are Linguistic Genders good for?
13. See Zgusta (1971: 82–3) for more examples, and how to deal with them in dictionarymaking.
14. See a comprehensive discussion in chapter 12 of Aikhenvald (2000).
15. Masa Onishi (1994), p.c.
16. Westermann (1947: 17); Heine (1982a: 193) on Gola, Dimmendaal (1983: 219–21) on
Turkana.
17. See Heath (1984: 169–70) on Nunggubuyu, Harvey (1987: 53) on Warray, and Merlan
(1994) on Wardaman and Merlan, Roberts, and Rumsey (1997: 85) on other northern
Australian languages with prefixes marking genders; see Bruce (1984: 97) on Alamblak.
18. Example from Zubin and Köpcke (1986: 174).
19. See Deutscher’s(2010: 203–4) witty criticism of the misleading idea that Linguistic Gender
is ‘irrational’.
20. Amadiume (2015) offers a comprehensive analysis of the traditional Igbo society with its
division of labour and cultural and social roles for men and women, and the ways in which
women lost their place and their powers as a consequence of colonialization; we return to
this in Chapter 11.

5
Gender meanings in grammar
and lexicon
Linguistic Gender is one of many means of categorizing a noun, in terms of its
animacy, humanness, or sex, across the world’s languages. Classifiers of various
types—including numeral classifiers, noun classifiers, and classifiers on verbs—
encode the same set of meanings on a par with inherent properties of objects: their
shape, size, consistency, and dimensions. We briefly delve into these in §5.1. Other
categories of a noun—number, case, and declension type—can reflect meanings
associated with Linguistic Genders—see §5.2. Then, in §5.3, we turn to how gender
meanings can be expressed in ‘gender-less’ languages.
5.1 Sex, humanness, and animacy in classifier systems
Grammatical means for the linguistic categorization of nouns can be found in every
language. Noun categorization devices other than Linguistic Gender include a variety
of types. NUMERAL CLASSIFIERS appear next to a numeral or a quantifier, categorizing a
noun in terms of its meaning. These are typical of languages of East and South-East
Asia, North and South America, and New Guinea. NOUN CLASSIFIERS categorize the
noun on its own, independently of any other element in a phrase or a clause. They are
a feature of many Australian, Mayan, and Austronesian languages. VERBAL CLASSIFIERS
occur on a verb and categorize its object or intransitive subject. There can also be
classifiers in possessive constructions, and classifiers with locative and deictic expressions.
We will see, in Chapter 6, how Linguistic Gender can arise through reinterpretation
of noun categorization devices of other kinds (thus confirming their
ultimate unity). 1
One language can combine several types of noun categorization devices. In
‘multiple classifier’ languages, the same morphemes occur in several grammatical
contexts—with numerals, adjectives, demonstratives, and also verbs. Classifiers can
encode meanings typically associated with gender systems—sex, animacy, and
humanness. We start with numeral classifiers, as the most frequent type across the
world’s languages.
How Gender Shapes the World. First edition. Alexandra Y. Aikhenvald.
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.

64 5 Gender meanings in grammar and lexicon
5.1.1 Numeral classifiers
Most languages with a numeral classifier system have special terms for humans and
for animates. Jacaltec, a Mayan language, has three numeral classifiers: -wanʺ for
humans, -c'onʺ for animals, and -(e)b' for inanimates. The three most frequent
classifiers in Indonesian are orang for humans, as in se-orang guru (one-CL:HUMAN
teacher) ‘one teacher’, ekor for non-human animates, as in dua ekor ular (two CL:
NONHUMAN.ANIMATE snake) ‘two snakes’, and buah for inanimates, as in tiga buah
meja (three CL:INANIMATE table) ‘three tables’. There are a few more classifiers for
inanimates whose choice is based on shape and function, e.g. batang for ‘cylindrical
objects’ (Sneddon 1996: 135–6). Halkomelem, a Salish language of the Pacific Northwest
of North America, has a special classifier for counting grown up people, and
another one for counting children, e.g. lx w =elə ‘three people’, lx w =eyl ‘three children’.
Categorization of humans through numeral classifiers may involve more than just
sex. Korean has at least 154 obligatory numeral classifiers. Nouns with non-human
reference are classified based on their nature, shape, or arrangement. Humans are
categorized depending on their status and sex. The numeral classifier pwun refers to
respected persons (and also the Christian God, Buddha, and other gods); the classifier
myeng is a neutral way of referring to a person. The classifier nom refers to a
despicable or insignificant male, and nyen to a despicable or insignificant female,
e.g. kkangphay twu nom (gangster two NUM.CL:DESPICABLE.MALE) ‘two male gangsters’,
kkangphay twu nyen (gangster two NUM.CL:DESPICABLE.FEMALE) ‘two female gangsters’. 2
Burmese is a Tibeto-Burman language with several score numeral classifiers.
Table 5.1 features classifiers used to count humans and animates. 3
Large systems of numeral classifiers in languages of East and South-East Asia tend
to reflect societal organization in their classifiers for humans. Thai, a Tai-Kadai
language with several dozen numeral classifiers, is spoken in a monarchy. There is
a special classifier phráʔoŋ for ‘emperor, king’, ʔoŋ for ‘God, Buddha, Buddhist
monk, emperor, king’, rûup for Buddhist monk, thân for ‘prime minister and high
rank officer’, naay ‘policeman’, in addition to a general human classifier khon. 4
There is no monarchy in the Republic of Lao; and the system of human classifiers
lacks a term for ‘king’. A numeral classifier khon2 ‘person’ covers people other than
TABLE 5.1. Animate classifiers in Burmese
hsu
Buddha and his attributes: relics, images, the Law
pa
deities, saints, monks, royalty
u: people of status, teachers, scholars
jau' ordinary humans
kaun animals, ghosts, dead bodies, depraved people, children

5.1 Sex, humanness, and animacy in classifier systems 65
monks; classifiers gong3 and huup4 are used for monks. Other northern Tai languages,
including Zhuang and Maonan, have no classifiers whose choice would be
determined by the social rank, or religious position. Lu (2012: 112–13), himself a native
speaker of Maonan and of Zhuang, notes that the fact that Zhuang lacks classifiers for
various social and clerical ranks (such as ‘monk’,or‘high official’) is a direct reflection
of the social organization of the Zhuang people: there is no social stratum of clergy, and
the society is much more socially equal than in Thailand or in Laos.
What about special classifiers for women? Numeral classifiers in Zhuang categorize
humans in terms of sex and age. The classifier pou 4 simply states that the noun refers to
a ‘human being’, asinsa:m1 pou 4 lɯk 8 (three CL:HUMAN child) ‘three children’. The
classifier ta 4 is used for young women, as in sam 1 ta 4 lɯk 8 (three CL:YOUNG.FEMALE child)
‘three daughters’. The classifier tak 8 covers young males, as in sam 1 tak 8 lɯk 8 (three CL:
YOUNG.MALE child) ‘three sons’. Two further classifiers cover adults of each sex: me 6
‘adult female’, and kɔŋ 1 ‘adult male’, e.g. sa:m 1 me 6 a:ŋ 6 jɯ 1 (three CL:ADULT.FEMALE
doctor) ‘three female doctors’, and sa:m 1 kɔŋ 10 a:ŋ 6 jɯ 1 (three CL:ADULT.MALE doctor)
‘three male doctors’. The human classifier, rather than female-specific forms, in
Zhuang is used to refer to professional women (instead of female-specific forms
which have somewhat demeaning overtones).
The use of a human classifier mirrors woman’s place in society, and speakers’
attitudes, in Maonan. A human is typically referred to as ʔai 1 (CL:HUMAN) and a nonhuman
animal as tɔ 2 (CL:ANIMAL). Infants, children, and women are classified with tɔ 2 ,
the non-human animal category. As Tian-Qiao Lu, a native speaker of Maonan, put
it, ‘judging from the application of this nonhuman classifier to a child or a woman,
we can say that a child or a woman is deemed as undersocialized in the Maonan
society’. As more women become integrated into professional society, they start
being referred to with the human classifier ʔai 1 (Lu 2012: 83–4, 115). The use of
classifiers for women reflects their status in the society, and their exclusion—just as
Linguistic Gender often does. Classifiers reflect social changes—just as Linguistic
Genders may do. We return to this in §11.4.
Gender systems always have animacy, humanness, or sex as their core semantic
feature. Not so for numeral classifiers. An example of a language with numeral
classifiers with no animacy or humanness comes from Kana, a Cross-River language
from Nigeria. The language has nineteen numeral classifiers whose choice is based on
the shape of person or an object. So, a dog or a human will be referred to with a general
classifier. Kana does not have genders. Human nouns can be differentiated from the
rest by other means: for instance, emphatic pronouns usually refer just to humans. 5
5.1.2 Noun classifiers
Noun classifiers are independent words with generic meaning which categorize the
noun with which they co-occur. Three of more than twenty noun classifiers in Yidiñ,

66 5 Gender meanings in grammar and lexicon
an Australian language, are used with human nouns—these are waguja ‘man, human
male, masculine’, buña ‘woman, human female, feminine’, and bama ‘person’. One
would not generally say just ‘the girl dug up the yam’: it is more felicitous to say ‘the
person girl dug up the vegetable yam’: 6
5.1 [mayi jimirr] [bama-al yaburu-ŋgu] jula-al
vegetable+ABS yam+ABS CL:PERSON-ERG girl-ERG dig-PAST
‘The person girl dug up the vegetable yam’
Noun classifiers for humans in Australian languages can reflect socially significant
categories. Yankunytjatjara has a special noun classifier for ‘initiated men’. Murrinhpatha
has a classifier for Aboriginal people (which also covers human spirits) and
another one for non-Aboriginal people, which also covers all other animates and
their products. 7
Noun classifiers in Mayan languages typically refer to sex, age, and social status of
humans. In Jacaltec, noun classifiers contain information about kinship relations,
respect, and age—see Table 5.2. There are special classifiers for male and female
deities, and one for infants. Mam has classifiers for men and women; for young and
old men and women; for old men and women to whom respect is due, and for
someone of the same status as the speaker. 8
Noun classifiers typically have one or more forms for humans and animates. In
contrast, classifiers on verbs and possessive classifiers do not have to have animacy,
humanness, or sex as one of their meanings. We will see, in §6.1.2, how noun
classifiers can give rise to a system of Linguistic Genders.
TABLE 5.2. Noun classifiers for humans and deities in Jacaltec (a selection)
cumam
cumi7
ya7
naj
ix
naj ni7an
ix ni7an
ho7
xo7
xo7 ni7an
unin
male deity
female deity
respected human
male non-kin
female non-kin
young male non-kin
young female non-kin
male kin
young male kin
young female kin
infant

5.1 Sex, humanness, and animacy in classifier systems 67
5.1.3 Verbal classifiers and other classifier types
Verbal classifiers (also called verb-incorporated classifiers) occur on the verb, categorizing
a noun—typically in S (intransitive subject) or O (direct object) function—
in terms of its shape, size, structure, or position, and sometimes also animacy.
A subtype of verbal classifiers are special classificatory verbs. In Athabaskan languages,
these are regularly chosen depending on the inherent properties of the object,
subject, and even the oblique. In some languages, there is a special form for an
‘animate’ object. Mescalero Apache, an Athabaskan language, has ten classificatory
verbs. The choice involves animacy, shape, and form. A selection is in Table 5.3. The
stem given here means ‘be located’. 9
Other Athabaskan languages do not include animacy in their system of classificatory
verbs. This is the case in Carrier (where animacy is expressed in numerals and in
demonstratives, while classificatory verbs are chosen based on the shape of the noun).
Innu, an Algonquian language, has eight verbal classifiers, referring to shape, nature,
and consistency of S and O, but not animacy or sex. Palikur, an Arawak language
from Brazil and French Guyana, with three genders and several score numeral
classifiers, has a special set of verb classifiers with no term for animates or humans
(a classifier -pit ‘irregular shape’ is used on verbs to refer to a human, or an animate,
or something of irregular shape). 10
The choice of an existential, locative, and positional verb may depend on what
their subject is like—whether it is animate, human, round, and so on. Japanese
distinguishes animacy in its existential verbs iru ‘be:animate’ and aru ‘be:inanimate’.
Hua, a Papuan language from the Highlands of New Guinea, has a copula bai with
animate subjects and mo with inanimates. Small systems of existential verbs ‘be’ in
Tibeto-Burman languages tend to include a special verb ‘be’ for animates. Ersu has
five verbs ‘be’, one for animates, one for living plants, two for inanimates, and one for
abstract concepts. 11
In a number of Papuan languages of New Guinea, classificatory verbs categorize
the intransitive subject argument in terms of its orientation or stance in space, and
also its inherent properties. In Enga, a verb meaning ‘stand’ is used with referents
TABLE 5.3. Mescalero Apache classificatory verb categories
stem form ‘be located’
semantics of the intransitive subject
1. -'ą ‘single, solid, round inanimate object’
2. -tiį ‘single animate object’
3. -la ‘dual objects of any kind; a rope-like object’
4. -tą ‘elongated, rigid object; a stick-like object’
5. -ɬ-tsuu ‘flexible object; a cloth-like object’

68 5 Gender meanings in grammar and lexicon
judged to be tall, large, strong, powerful, standing, or supporting, e.g. men, houses,
trees; ‘sit’ is used with referents judged to be small, squat, horizontal, and weak, e.g.
women, possums, and ponds. This is reminiscent of shape and dimensionality
associated with masculine and feminine Linguistic Genders discussed in §3.3. Most
systems of classificatory verbs in Papuan languages have no special form for animates,
or for men and women: all of these are classified by what is conventionalized
as their typical position. 12
A number of languages have possessive classifiers—special morphemes which
characterize a possessed noun within possessive constructions. Palikur has five possessive
classifiers—one for domesticated animals, one for food, one for plants, one for
animals caught to eat, and one for child. So, a classifier construction nu-kamkayh tino
(1sg-CL:CHILD woman) is the only way of saying ‘my daughter’. A number of Uto-
Aztecan languages (including Cora and Papago) have animate and inanimate classifiers
in possessive constructions. Some Oceanic languages in New Guinea have special
possessive classifiers for domestic animals. Hmong, a Hmong-Mien language of Laos,
has a special classifier used for ‘possession’ of living beings illustrated in 5.2.
5.2 nws tus txiv ntxawn tus ntxhais
he CL:LIVING.BEING uncle CL:LIVING.BEING daughter
‘the daughter of his uncle’
Classifiers in Hmong are also used with numerals, demonstratives, and on nouns
themselves. It is a ‘multiple classifier’ language where classifiers are used in many
environments. 13 Kilivila, an Oceanic language with many classifiers used on
numerals, demonstratives, and adjectives, has classifiers te and to for humans and
males, and na for females (see §6.1.1 on their origin). 14 Classifiers of different types in
one language can classify humans and animates along different lines. In Jacaltec noun
classifiers for humans offer a detailed categorization based on sex, age, kinship
relations, and respect (Table 5.2).
Numeral classifiers may categorize nouns into humans, animals, and inanimates
only. Carrier, an Athabaskan language from British Columbia in Canada, has twelve
distinct kinds of noun categorization through classificatory verb stems, demonstratives,
numerals, and prefixes on verbs. Special forms for humans and non-humans
are distinguished for numerals, demonstratives, relativizers, and the quantifier ‘how
many’. 15
5.1.4 Linguistic Genders and classifiers as noun categorization devices:
commonalities and differences
Linguistic Gender and classifiers as noun categorization devices share their functions,
and can be related historically. They can combine in one language, each with
somewhat different meanings. Malto, a South Dravidian language, has two genders:

5.2 Sex, humanness, and animacy in noun categories 69
male human versus the rest in the singular number, and human versus non-human
in plural. Thirty numeral classifiers include a special form for humans (jen) and for
non-human animals (maq) which include ghosts (Mahapatra 1979: 120).
Dyirbal, an Australian language, has a system of four genders whose choice is
meaning-based. Yidiñ, a neighbouring language, has about twenty noun classifiers.
The two encode similar meaning, but in a different way. Dixon (2015: 58–9) offers a
comparison of the semantic fields covered by the two techniques in the two languages.
The semantic field of ‘people’ is covered by Yidiñ classifiers waguja ‘man’ and
buña ‘woman’, and by Dyirbal noun gender markers bayi ‘masculine’ and balan
‘feminine’. Yidiñ forms are restricted to living beings, while Dyirbal forms are also used
for male and female spirits and legendary people. We have also seen that balan has
other concepts associated with it: fire, drinkable liquids, and fighting. Two of these are
associated with different noun classifiers in Yidiñ: buri for fire, bana for drinkable
liquid. There is no special classifier in Yidiñ for fighting or fighting implement.
Every noun has to be assigned to a Linguistic Gender if a language has one.
Classifiers are less restrictive—in many languages, not every noun has to be accompanied
by a classifier. The choice of Linguistic Gender choice may be semantically
opaque. The choice of classifier tends to be more straightforward. For instance,
Palikur has three genders whose choice is mostly semantic, but with a degree of
opacity, especially for non-humans. There are a dozen numeral classifiers, which
include one classifier for humans with a masculine and a feminine form, chosen
exclusively by a person’s sex. Classifiers can never be chosen by a noun’s morphological
features or their form, unlike Linguistic Genders. We will see, in Chapter 6,
that classifiers can give rise to Linguistic Genders.
Just like Linguistic Genders, classifiers may reflect social values and stereotypes,
especially where women are concerned. This is what we saw in Maonan in §5.1.1—a
topic to which we return in §11.1.
5.2 Sex, humanness, and animacy in noun categories
The majority of languages have some grammatical distinctions between humans and
non-humans, or animates and inanimates. Uralic languages have no linguistic
genders or classifiers. All the languages of the family differentiate between question
words ‘who’ (Proto-Uralic *ki) and ‘what’ (Proto-Uralic *mi). Having different
forms for ‘who’ and ‘what’ is not universal: the same form covers both in Australian
languages Marra, Ngandi, and Kayardild and in Ewe, a Kwa language from Ghana. 16
Universal semantic parameters of sex, animacy, and humanness can find their
reflection in other noun categories—number, case, and declension classes.
There are typically more number distinctions made for human nouns and for
nouns referring to animates than for inanimates. In Koasati, a Muskogean language

70 5 Gender meanings in grammar and lexicon
from the USA, only human nouns can be optionally marked for plural; other nouns
do not take a plural marker at all. 17
A human or an animate noun is most likely to be in the function of a transitive
subject. If it appears in the object position it may acquire special marking. This is
known as Differential Object Marking (DOM). 18 A typical example comes from
Spanish. A direct object (O) occurs with a marker (the preposition a) if it is specific
and animate, as in 5.3.
5.3 Juan vio [a mi papá]
Juan saw DIRECT.OBJECT my Dad
‘Juan saw my Dad’
If it is inanimate it will be unmarked:
5.4 Juan vio [el libro]
Juan saw the book
‘Juan saw the book’
A division of nouns into declension classes may correlate with animacy and
humanness. Lezgian, a north-east Caucasian language, lost the genders reconstructed
for the proto-language. The choice of its oblique stem suffixes correlates with the
meaning of nouns: the suffix -Adi is used with nouns referring to a non-discrete mass,
and -rA is used with most nouns which denote animals and people. Armenian, an
Indo-European language without Linguistic Gender in anaphora or in agreement,
has the distinction between person and non-person in its inflectional paradigm. 19
Most Slavonic languages distinguish three genders—masculine, feminine, and
neuter, chosen on the basis of semantics and partly form (especially for inanimates).
In addition, many have developed a special ‘animate’ subgender. Animate nouns take
different agreement form from inanimates in the accusative case, e.g. Serbo-Croat
òvog sȉna (this.ACC.ANIM son.ACC.ANIM) ‘this son’ (object) and òvāj prózor (this.ACC.
INAN window.ACC.INAN) ‘this window’ (object). The inanimate object form is the same
as the nominative, and the animate object form is the same as genitive. The animacy
distinction is semantically transparent.
In addition to this, some West Slavic languages have developed special agreement
for nouns denoting males (known as ‘virile’ subgender). Three agreement genders
(masculine, feminine, and neuter) are distinguished in singular and not in plural in
Polish. Two different plural past tense agreement forms distinguish males from
everyone else, in byli ‘were’ (males) versus byly ‘were’ (other subjects, non-males).
Macedonian, a South Slavic language, has three genders (masculine, feminine, and
neuter) in agreement and anaphora. A special ‘virile’ form of cardinal numbers
covers men and mixed groups, e.g. trojca ‘three (male humans and mixed sex
groups)’ and tri ‘three (any other genders; non-virile)’. 20

5.3 Gender in gender-less languages 71
Animacy, humanness, and sex expressed through categories other than Linguistic
Gender are typically binary, and semantically transparent. Once developed, they can
evolve into Linguistic Genders: we turn to this in §6.1.
We now turn to ways of expressing gender-related meanings in languages which
have no Linguistic Gender classification in agreement, or in anaphora.
5.3 Gender in gender-less languages
In every language one can distinguish a bull from a cow, and a man from a woman,
using different words. Words for ‘man’ and ‘woman’ can be added to nouns, to
differentiate between sexes. Turkish and Indonesian have no anaphoric Linguistic
Gender and no gender agreement. The lexeme hanɪm ‘woman, lady’ in Turkish is
used in combination with terms referring to professions, e.g. hoca hanɪm ‘Lady
Teacher’. In Indonesian, the masculine form lelaki or laki-laki can be added to
a noun to refer to a male, e.g. orang laki-laki ‘man’, and perempuan to refer to a
woman, e.g. organg perempuan ‘woman’. The feminine betina can be added to a term
for an animal to denote the female, as in singa betina ‘lioness’, and djantan to denote
a male, e.g. singa djantan ‘lion’. 21
Languages without Linguistic Gender include the majority of Austronesian languages,
Turkic, Tungusic, Uralic, and most Tupí and Carib languages in South
America. Creole and Pidgin languages almost universally lack Linguistic Gender. 22
Standard Hausa, a major language of Northern Nigeria, has two genders, masculine
and feminine. These are expressed in personal pronouns and a variety of agreement
contexts, including possessive constructions. Pidginized Hausa (used as a means of
inter-ethnic communication by non-native speakers) uses just one, erstwhile ‘masculine’,
form in all contexts, without distinguishing masculine and feminine: 23
5.5 Standard Hausa
masc
fem
yā
tā
zō
zō
‘he has come’
‘she has come’
Pidgin Hausa
yā zō ‘she/he has come’
We can recall, from §2.1.2, that Swahili, just like many other Bantu languages, has
adjectival agreement markers which reflect gender and other properties of referents.
In the Kenya Pidgin Swahili, the zero gender marker has replaced all the agreement
markers. Fanagalo, a Bantu-based creole spoken in Botswana, has lost genders typical
for a Bantu language. Chinook Jargon was a pidgin based on a number of indigenous
languages of the American Northwest (especially Lower Chinook, Nootka, and
Chehalis), with elements of English, French, Hawaiian, and Chinese. Chinookan
languages have three genders in the singular (masculine, feminine, and neuter); this
feature is absent from the Chinook Jargon.

72 5 Gender meanings in grammar and lexicon
Most European-based Creole languages use one form for ‘she’ and ‘he’ (e.g. Tok
Pisin em, Kristang el, Papiamentu e, Nubi uwo ‘s/he’). A rare exception appears to be
the Negerhollands Creole Dutch, spoken until recently in the United States Virgin
Islands, which distinguishes between animate (h)am ‘he, she’ (cf. Dutch hem ‘him’)
and inanimate di ‘it’ (cf. Dutch dit ‘this: neuter gender’). Adjectives in Europeanbased
Creoles are typically based on a feminine form, but the same form is used for
men and for women. A feminine form of an adjective ‘fat’ has been preserved in
Guyanais Creole French with the meaning of ‘pregnant’ (grós, cf. French grosse); the
erstwhile masculine form, gro (French gros) has the meaning of ‘fat’.
Some languages without Linguistic Gender have derivational affixes producing
forms with male and female reference. Indonesian has a few masculine–feminine
pairs, like dew-a ‘god’, dew-i ‘goddess’, borrowed from Sanskrit. Estonian, a language
without Linguistic Gender, has many lexical pairs, including naine ‘woman’ and mees
‘man’, tütar ‘daughter’ and poeg ‘son’, and so on. Similar to other Uralic languages,
Estonian has no agreement gender or anaphoric gender. The same third person
pronoun tema is used for any sex or animacy. A few suffixes on nouns help distinguish
a man from a woman, e.g. kangelane ‘hero’ and kangelanna ‘heroine’, kuningas ‘king’
and kuninganna ‘queen’. We will see in §6.4.2 that the feminine derivational suffix -tar
in Estonian (and its Finnish counterpart -ttare) was borrowed from a Germanic source.
This suffix is now used with numerous terms (native and borrowed) referring to
women, e.g. runo-tar ‘poetess’, from runo ‘poem’ (native word), kuninga-tar ‘queen’.
The word sõber ‘friend’ can refer to a man or a woman, and can occur with two
feminine suffixes, with a slight change in meaning. The feminine sõbranna ‘female
friend’ is nowadays used among women talking about female friends, and the less
frequent sõbratar tends to refer to a man’s girlfriend. Ross (2012: 167) remarks that
sõbranna has overtones of someone ‘indisputably more lightweight and unreliable
than simply sõber’ and tends to occur with negative attributes such as ‘jealous’ or
‘grumpy’. This is reminiscent of negative overtones associated with many female
derivations in English—to which we return in §7.5.
The Eastern Maroon Creole of Suriname (also known as Ndyuka) is Englishbased.
Personal nouns can be marked as masculine (with man ‘man’) or as feminine
(uman ‘woman’). Nouns marked with uman may simply denote the functions
performed by women, and by men in a particular setting. Olo-man (grave-man) is
a grave digger; olo-uman is the one who prepares, brings, and distributes food to the
olo-man (Migge 2001: 99). Or a male counterpart may have a generic meaning, as in
waka-man ‘traveller’, and the female counterpart will have a more negative and oversexualized
connotation: waka-uman is not just a ‘travelling woman’: this has strong
connotations of a woman of easy virtue. Once again, this is reminiscent of negative
overtones of terms referring to women—see §7.5 and §11.2. 24
It is hard to translate expressions like ‘his and hers’ into a language where ‘he’ and
‘she’ share the same form. In order to remedy the gap, a new partly jocular form

5.3 Gender in gender-less languages 73
tema-ke has been recently introduced into Modern Estonian. Tema-ke consists of the
pronoun tema ‘he, she’ and the diminutive suffix. The form is far from being in active
use: it is found in jocular columns in newspapers and in the internet, and in
translations from English: the title of the film Her, by Spike Jonze (made in 2013),
is Temake. Web-pages advertise getaways for him and her, tema and temake. That
the diminutive suffix has been used to create a feminine form of a pronoun goes
together with small size associated with ‘femaleness’—something we mentioned in
the preceding chapter. 25
Languages without Linguistic Gender may have additional distinctions which
most languages do not make. The absence of Linguistic Gender is a feature of
Hungarian. A special suffix -nő marks a female referent, in pairs such as tanitó
‘teacher’ and tanitó-nő‘female teacher’. A further suffix -né refers to a ‘matrimonial
feminine’, someone married to who the suffix is attached to. Tanító is ‘teacher’, and
tanító-né is ‘teacher’s wife’. One can thus distinguish members of the royal family:
rather than using a general term ‘queen’, one will refer as király-né to king’s wife, and
as király-nőto a queen as the female head of state in her own right. 26
As Amadiume (2015: 89) points out for her native Igbo, no distinction in gender is
made in pronouns, ‘thus there is no reminder in speech to distinguish between the
sexes’. In Turkish, Estonian, and Hungarian nouns denoting persons may refer to
men and women. In Estonian, politseinik may refer to a policeman or a policewoman,
arst to a female or to a male doctor, and õpetaja to a male or a female teacher.
Speakers of languages without genders require extralinguistic—usage- and culturebased—clues
in distinguishing men from women, in terms of sex and social standing.
But the ways in which reference to men and to women is made remain asymmetrical
and essentially male-biased.
The word for ‘man’ is used as a generic noun meaning ‘person’ in many languages
without Linguistic Gender. In Indonesian, mahasiswa ‘student’ and karyawan ‘clerk’
often occur as generic terms, and their derivational female counterparts are used only
occasionally. Kuntjara (2001) describes this as a way of ‘excluding’ women. Braun’s
(2001) investigation of Turkish showed that forms which are gender-neutral can be
biased: an experiment with 130 university students (78 females and 52 males) showed
that nouns such as köylü ‘villager’, kis̹i ‘voyager’,andyoncu ‘passenger’, and the pronoun
birisi ‘someone’, were spontaneously conceived of as males in the first place.
This inherent gender bias reflects male dominance across Turkish society: ‘men
are simultaneously the privileged group and the leading figures’ in the Turkish
economy, labour market, politics, law, and religion. If one needs to specify that the
person is a woman, a word ‘woman’ or ‘lady’ will be added. Generally speaking,
Turkish terms for person reference contain information about the division of labour
in Turkish society. Futbolcu ‘football player’ will be immediately understood as
relating to a man, and c̹ocuk bakɪcɪsɪ ‘nursery school teacher’ as relating to a
woman. The use of gendered and generic nouns in Turkish discourse and in proverbs

74 5 Gender meanings in grammar and lexicon
provides clues about which terms are more ‘appropriate’ for a woman and which
ones for a man. Generic terms such as esi-mees (first-man) for ‘chairman’ used to
refer to a man and to a woman reflect the erstwhile dominance of ‘men’ in Estonian
(though the term esi-naine ‘chairwoman’ is now also used). Male—rather than
female—reference can still be considered functionally unmarked.
In Braun’s (2001: 305) words, in Turkish and in other languages, ‘men are central
to the “human” category’, and ‘a peripheral or secondary position is what remains for
females. This “extra-ordinary” status of women is linguistically communicated by a
pronounced tendency towards gender marking in female reference’: a child (çocuğu)
will be likely to be associated primarily with a male child, while female reference has
to be explicitly expressed, as in kɪz çocuğu ‘girl child’.
The stereotypes associated with Social Gender, and the male bias, will be reflected
in any language—with or without Linguistic Gender—as long as they are relevant for
the community which speaks it. Unravelling them will be a task as demanding as that
involved in understanding the social and cultural underpinnings in noun categorization
through dedicated means—including Linguistic Gender.
NOTES AND SOURCES
1. Advocated for in Aikhenvald (2000, 2006, and forthcoming).
2. See Lee (2014: 42–3) on Korean. See Adams (1989: 56–7) on examples of special classifiers
for men and women in Austroasiatic languages; see Gerdts and Hinkson (2004: 254) on
Halkomelem, Craig (1986a: 265) on Jacaltec, Sneddon (1996) on Indonesian, Thompson
(1987: 194) and Daley (1998: 96–7) on Vietnamese; Enfield (2007: 125, 142) on Lao;
Lu (2012: 83) on Maonan and Zhuang. A full list of human classifiers in Zhuang is in Lu
(2012: 114).
3. From Becker (1975: 116). See also Aikhenvald (2000: 315) on metaphorical extensions of
human classifiers.
4. See Lu (2012: 110), Diller (1985).
5. See Ikoro (1996: 90–1) on Kana; other systems of numeral classifiers are discussed in
Aikhenvald (2000: 287–8).
6. Dixon (1982: 185; 2015: 44).
7. See Goddard (1985: 94) on Yankunytjatjara, and Walsh (1997: 256) on Murinhpatha.
8. See Craig (1986a: 266–7), on Jacaltec; England (1983) on Mam. Number 7 indicates a
glottal stop.
9. See Rushforth (1991: 253). See Aikhenvald (2000: chapter 5), (2006) on how classificatory
verbs differ from lexical pairs like English eat (solid food) and drink (liquid food), and
further features of verbal classifiers.
10. See Poser (2005) on Carrier, Drapeau and Lambert-Brétière (2011: 302–4) on Innu;
Aikhenvald and Green (2011) on Palikur.
11. Haiman (1980) on Hua, Martin (1975) on Japanese, Zhang (2014: 445) on Ersu.
12. Similar systems have been described for Ku Waru and Kewa, also from the Papuan region.
See Lang (1975) on Enga, Franklin (1981) on Enga and Kewa, Merlan, Roberts, and
Rumsey (1997) on Ku Waru.

5.3 Gender in gender-less languages 75
13. Jarkey (2015), Bisang (1993: 29–30).
14. See Senft (1986: 76) for Kilivila.
15. See Poser (2005: 162–3).
16. See Collinder (1965: 138) for Uralic, Dixon (2002: 329) and references there for Australian
languages, Ameka (2012: 53–4) for Ewe.
17. This tendency was summarized as the principle of Nominal Hierarchy by Smith-Stark
(1974).
18. See Bossong (1985, 1991).
19. Haspelmath (1993: 76–7), Alexeyev (1985: 27ff.), Matasović (2004).
20. On Macedonian, see Friedman (1993: 267); for the development and characteristics of
animate subgenders in Slavonic languages see Comrie and Corbett (1993: 16); Schenker
(1993: 108).
21. See Kwee (1965: 55) on Indonesian, Braun (2000a: 53) for Turkish, Collinder (1965: 57)
for Finnish, Hasselblatt (2015) and Wiedemann (1875: 196–7) for Estonian.
22. Pidgins and Creoles arise as a makeshift means for filling communicative needs, rather
than evolving in the way natural, non-contact languages would. See Holm (2000: 216–17,
222) on general features of Creoles and their lack of genders; Silverstein (1972a: 397–8)on
traditional Chinook, and Silverstein (1972b: 618) on the Chinook Jargon, Owens (2001:
349, 355; 2014: 283) on Arabic-based Creoles, Anderson and Janson (1997: 183–5) on
Fanagalo. As pointed out in Aikhenvald (2000: 389), Creole and Pidgin languages uniformly
lack classifiers or other noun categorization devices.
23. See Heine and Reh (1984: 42–3); Hodge (1958).
24. Braun (2001) refers to the expression of a combination of social gender and natural gender
in languages without linguistic gender, and the social dichotomy of male and female roles
in language expressed through lexicon, as ‘covert’ gender. This is misleading, as such usage
can lead to confusion with covert linguistic gender (which is not marked on nouns
themselves).
25. See Ross (1996, 2012), Hasselblatt (2015) on Estonian.
26. See Vasvári (2015: 210–11), on the use and origin of these forms.

6
The rise and fall of Linguistic Genders
Linguistic Genders have their histories. They can develop out of lexical items or noun
categories with meanings of animacy or humanness. The meanings of Linguistic
Genders can change over time. Linguistic Genders can be gained, or lost, if languages
are in contact with each other. Social changes may affect the use of Linguistic
Genders applied to human beings. But losing—or gaining—a Linguistic Gender is
independent of increase or decrease in equality of Social Genders. We now turn to
the mechanisms of developing Linguistic Gender.
6.1 Developing Linguistic Gender
Nouns referring to humans—man, woman, person—can develop into a closed
system of anaphoric gender. Or they can give rise to derivational gender markers.
This process is known as grammaticalization. 1 Once a language has acquired a
marker of anaphoric or derivational gender, this can be reinterpreted and reanalysed
as agreement gender.
Pathways in the development of Linguistic Gender are summarized in Box 6.1.
BOX 6.1 How to gain a linguistic gender: pathways of development
PATHWAY I. Generic nouns > ANAPHORIC GENDER > AGREEMENT GENDER—§6.1.1 and
PATHWAY IV
PATHWAY II. Generic nouns > DERIVATIONAL GENDER—§6.1.1; then see also PATHWAY V
PATHWAY III. Generic nouns > closed system of NOUN CLASSIFIERS > DERIVATIONAL GENDER >
AGREEMENT GENDER—§6.1.2 and PATHWAY IV
PATHWAY IV. ANAPHORIC GENDER > AGREEMENT GENDER—§6.1.3
PATHWAY V. DERIVATIONAL GENDER > AGREEMENT GENDER—§6.1.4
Pathways I–III involve grammaticalization of nouns into exponents of Linguistic
Gender (and their subsequent reanalysis and reinterpretation). Pathways IV–V
How Gender Shapes the World. First edition. Alexandra Y. Aikhenvald.
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.

6.1 Developing Linguistic Gender 77
involve reanalysis and reinterpretation of markers of Linguistic Gender, and a change
from one kind of gender to another.
As we saw in §2.1.3, Linguistic Gender interrelates with other noun categories. Core
meanings of Linguistic Genders—humanness, sex, and animacy—can be expressed
through number, case, and classifiers of various types (§§5.1–2). Reinterpretation of
these categories may account for the emergence of Linguistic Gender classification.
6.1.1 From generic nouns to Linguistic Genders
Nouns referring to humans (‘man’, ‘person’, and ‘woman’) can give rise to a closed
grammatical system of anaphoric gender in personal pronouns. Kokota, an Oceanic
language spoken in the Solomon Islands, distinguishes masculine and feminine
genders in its third person pronouns. The masculine pronoun manei comes from
mane ‘man’. 2 In Bari and Toposa, East Nilotic languages, masculine personal pronoun
lo ‘he’ and its feminine counterpart na come from nouns, *lV ‘member/person
of ’ and *nyaa- ‘girl, daughter’. 3
Newly grammaticalized personal pronouns—that is, exponents of anaphoric
gender—can be reanalysed as agreement gender markers. Zande, an Ubangian
language spoken in the Democratic Republic of Congo, distinguishes four genders
in the singular, and three in plural personal pronouns. Feminine and masculine
genders in the singular collapse into one ‘human’ gender in the plural. In agreement
with a well-established typological tendency, there are fewer genders in plural than
in the singular (see §2.1.3). Gender and number distinctions in Zande personal
pronouns are shown in Scheme 6.1.
SINGULAR PLURAL
Masculine
Human
Feminine
Animal Animal
Inanimate Inanimate
SCHEME 6.1 Gender and number in Zande personal pronouns
A number of terms in the system come from independent nouns—as shown in 6.1:
6.1 Lexical origins of Zande pronouns
NOUN
GENDERED PRONOUN
man, male > masculine
person > human
animal, meat (pl) > animal (pl)
thing > inanimate

78 6 The rise and fall of Linguistic Genders
Some of the original nouns which gave rise to pronouns are still used as nouns in
related languages, e.g. Zande kɔ́‘he (masculine)’, Banda kɔ́, Ngala kwã ‘man’.
Personal pronouns (lexical in origin) have given rise to agreement markers used on
adjectives (as copula complements). In 6.2, the personal pronoun kɔ́appears as a
suffix and as a prefix marking agreement on the verb and then on the adjective ‘big’:
6.2 kɔ́-ni bakέrέ-kɔ́
he-be big-he
‘He is big’
Sometermswithinlargergendersystemsmayhavedevelopedoutofnouns.The
diminutive gender marker -pi in Bantu languages of Cameroon comes from Niger-Congo
*bi ‘child’, and the locative gender marker -ku comes from the noun *ku ‘outside’. 4
Nouns meaning ‘man’ or ‘person’ can give rise to numeral classifiers referring to
humans. For instance, the Korean numeral classifier salam ‘CLASSIFIER:PERSON/ADULT’
is related to the noun sālam. 5 Feminine and masculine terms within a system of
classifiers in multiple environments can come from grammaticalized forms for ‘man’
and ‘woman’. Kilivila, an Oceanic language spoken on the Trobriand Islands in
Papua New Guinea, has over a hundred classifiers which characterize a noun in
terms of its sex, animacy, shape, and other intrinsic features. Classifiers are prefixes
to numerals and adjectives and infixed to demonstratives. 6 One of the sex-based
classifiers, to/te- ‘man, human being in general’, comes from the noun tau ‘man’. The
other sex-based classifier na- ‘women, females’ comes from vivina ‘woman’.
Noun classifiers typically stem from grammaticalized nouns with generic meanings.
In Akatek, a Mayan language, naj, a noun classifier meaning ‘man’ developed
from the noun winaj ‘man’. The noun classifier ʔis᷈̍ ‘woman’ developed from ʔis᷈̍
‘woman’. In Mam, also Mayan, sex- and social gender-sensitive classifiers developed
from nouns referring to males and females—see Table 6.1.
TABLE 6.1. Noun classifiers for humans in Mam and nouns they originated from
CLASSIFIER
q'a ‘classifier: young man’
txin ‘classifier: young woman’
ma ‘classifier: man’
xu7j ‘classifier: woman’
swe7j ‘classifier: old man’
xhyaa7 ‘classifier: old woman’
xnuq ‘classifier: old man, respectfully’
xuj ‘classifier: old woman, respectfully’
COMMON NOUN
q'aa ‘young man’
txiin ‘young woman’
matiij ‘big’
xu7j ‘woman’
swe7j ‘old man’
yaab'aj ‘grandmother’
xiinaq ‘old man, respectfully’
xu7j ‘old woman, respectfully’

6.1 Developing Linguistic Gender 79
Noun classifiers in Coatzoquitengo Mixtec, a Mixtecan language from Mexico,
identify a referent as male, female, deity, animal, longish object, or liquid. Of these,
classifiers for ‘male’ and ‘female’ come from nouns meaning ‘man’ and ‘woman’:
Proto-Mixtec *tee > tà ‘noun classifier: male’, Proto-Mixtec *ñaha > ñà ‘noun
classifier: female’. Classifiers are used similarly to definite articles, e.g. tà loho (CL:MALE
young.person) ‘the boy’, ñà loho (CL:FEMALE young.person) ‘the girl’. 7
Historical development from ‘man’ to ‘person, human in general’ can be compared
to the use of the masculine form, or a masculine pronoun, to refer to humans as a
class. No language so far has a similar development, for a noun meaning ‘woman’ to
‘person, human in general’.
‘Inherent nature’ noun classifiers in Australian languages include terms for ‘man’,
‘woman’, and ‘person’. They transparently come from independent nouns with
generic reference. Yidiñ forms ‘person’, ‘woman’, ‘man’ can be used as full nouns
and as noun classifiers. In some languages these noun classifiers gave rise to derivational
gender markers and to agreement genders, as we will see in §6.1.2.
A noun meaning ‘person’, ‘man’, ‘woman’,or‘daughter’ can grammaticalize into a
derivational gender affix referring to a male or to a female. In Mupun, a West Chadic
language, masculine and gender can be marked by prefixes on proper names and
some common nouns, and on a pronominal element meaning ‘such and such’. The
feminine prefix nà- has developed via grammaticalization of naa ‘woman’. The
masculine prefix dà- comes from daa ‘man’, e.g. nà-kómtàk ‘such and such a female’,
dà-kómtàk ‘such and such a male’.
The suffix -(a)nɪm in Turkish appears in a few words with female reference, e.g.
hocanɪm ‘female teacher’. It clearly comes from grammaticalization of the lexeme
hanɪm ‘woman, lady’ in combination with terms referring to professions, especially in
address, e.g. hoca hanɪm ‘Lady Teacher’. 8
No examples of generic nouns developing into gender-sensitive derivational affixes
and then into agreement genders have been reported so far. We will see in §6.1.4 how
derivational gender markers can develop into agreement genders.
6.1.2 From generic nouns to noun classifiers and then to Linguistic Genders
Once generic nouns have developed into noun classifiers, they can give rise to
Linguistic Genders. This pathway starts with grammaticalization of nouns into a
closed system of noun classifiers. The next step involves reinterpreting noun classifiers
as affixes to a noun they occur with, and as agreement markers on accompanying
modifiers. The development of agreement gender does not have to pass through
the stage of derivational gender.
Ngan.gityemerri, a language from northern Australia, has fifteen noun classifiers. 9
Six of these are free words (which have grammaticalized out of generic nouns), five
are prefixes (or proclitics), and four are both free words and prefixes (or proclitics).

80 6 The rise and fall of Linguistic Genders
Ngan.gityemerri also has five gender classes whose origin lies in noun classifiers. The
development from noun classifiers to agreement genders involves several steps.
STEP 1: involves a generic-specific ‘pairing’ of nouns: a generic noun classifier precedes
the specific noun it refers to.
6.3 [gagu wamanggal] NOUN.CL-NOUN kerre ngeben-da
NCL:ANIMAL wallaby big 1sgS+AUX-shoot
‘I shot a big wallaby’
STEP 2: since noun classifiers are independent words, often favoured over specific
nouns especially to maintain reference within a story, a noun phrase can consist of a
generic noun plus a modifier, as in 6.4.
6.4 gagu kerre ngeben-da
NCL:ANIMAL big 1sgS+AUX-shoot
‘I shot a big animal [wallaby]’
STEP 3: if specific nouns are included, both the specific noun and modifiers tend to
‘attract’ generics. This ‘repetition’ of noun classifiers is the predecessor of agreement:
6.5 [gagu wamanggal] NOUN.CL-NOUN [gagu kerre] NOUN.CL-ADJ ngeben-da
NCL:ANIMAL wallaby animal big 1sgS+AUX-shoot
‘I shot a big wallaby’ (lit. I shot an animal-wallaby animal-big)
STEP 4: Repeated noun classifiers cliticize to the specific nouns, and are reduced to
proclitics which effectively mark agreement, as in 6.6.
6.6 wa=ngurmumba wa=ngayi darany-fipal-nyine
male=youth male=big 3sgS+AUX-return-FOCUS
‘My initiand son has just returned’
STEP 5: Noun class marking proclitics become obligatory prefixes on the nouns and
on agreeing modifiers. At this stage, they become fully grammaticalized:
6.7 wú-pidìrri wu=mákarri
NCL:CANINE-dingo
‘a bad dog’
NCL:CANINE=bad
STEP 6: Thisstageinvolves‘prefix absorption’: that is, a gender class prefix onanoun
becomes lexicalized as part of it. Some prefixed roots may be interpreted as stems which
can take further gender marking. This may lead to double class marking, e.g. wa-mumu
(male-police) ‘policeman’; wur-wa-mumu (female-male-police) ‘policewoman’.
Similar scenarios applied in other Australian languages. In Dyirbal, an Australian
language with no prefixes, the ‘edible vegetable’ gender developed out of a generic
noun *mayi ‘non-flesh food’. The generic noun reinterpreted as a prefix to

6.1 Developing Linguistic Gender 81
deictic determiners, and underwent morphological reduction to -m. 10 But unlike
Ngan.gityemerri, Dyirbal developed gender agreement markers without ever going
through the stage of a derivational gender.
6.1.3 From anaphoric gender to agreement gender
Agreement genders can arise via reinterpretation of anaphoric gender expressed
in demonstratives or third person pronouns. This scenario has been described for a
number of Australian languages. Warndarran has five Linguistic Genders. The
marker for the class of human males rna- transparently relates to the recurrent free
form pronoun masculine na. The marker for the class of human females (which also
includes some animals) ŋa- relates to the free pronoun ŋa(l). Along similar lines,
Greenberg (1978: 61ff.) proposed a chain of historical development, from demonstrative
to definite article to specific article to marker of a noun, or an overt gender
marker, in Voltaic languages. 11
Different terms in one gender system in one language may come from a variety of
sources. In Warndarran the prefix for the class of plants with edible parts ma- is
linked to the grammaticalization of noun classifier mayi ‘edible plant’.
6.1.4 From derivational gender to agreement gender
An already existing derivational gender marker can evolve into an exponent of gender
agreement. Gender systems are a salient feature of most Indo-European languages. The
Anatolian languages are thought to be the most archaic. Hittite, the first Anatolian
language to be deciphered and still the best understood, had just two genders—neuter
(or ‘inanimate’) and‘common’ (or ‘animate’). It is now the consensus in the Indo-
European scholarship that the feminine gender in the branches of Indo-European
(other than Anatolian) evolved after the Anatolian languages split off.
A later origin for feminine gender is supported by a number of arguments. For
instance, the Proto-Indo-European question word *k w is ‘who?’ referred to both
feminine and masculine, and a separate form *k w id ‘what?’ was used for the neuter.
Some compound adjectives in Greek have one form for masculine and feminine, e.g.
rhodo-dáktulos ‘rosy-fingered (he or she)’, and a separate one for neuter, e.g. rhododáktulon
‘rosy-fingered (it)’. 12 Exactly how Indo-European languages acquired their
feminine gender is a matter of some debate. Most scholars believe that agreement
gender came about through reinterpretation of derivational affix *-h 2 used to form
abstract and action nouns. This may also have involved analogy with the term for
‘woman’, *g w enh 2 (the noun which gave rise to words for ‘woman’ in many languages,
including Greek gine ‘woman’, and incidentally queen in English). Due to its
ending in the laryngeal *h 2 this noun may have become associated with the abstract
and collective nouns with the same ending, which was reanalysed as a derivational
marker for the new feminine gender. Once reanalysed as a gender marker, it
extended to mark agreement. 13

82 6 The rise and fall of Linguistic Genders
6.1.5 Linguistic Gender from other nominal categories
Linguistic Gender distinctions can originate in reinterpretation of some other nominal
category—typically number and case.
Languages tend to have more number distinctions for human and animate nouns
than for inanimates (see §2.1.3). As a consequence, categorization based on animacy
may be encoded through number marking. A combination of number and animacy
forms the basis of gender-like agreement classes in Kiowa (a Kiowa-Tanoan language
of Oklahoma). Nouns of Class 1 are animate (and some inanimates which have
animacy in legends, e.g. ‘sun’). Nouns of Class 2 are mostly inanimates, while those of
Class 3 are inherently paired (and thus dual) (Class 4 can be considered a residue)
(Watkins 1984: 81–91).
The marking of the agreement classes is on nouns themselves and in pronominal
prefixes on verbs. Balochi, a Western Iranian language, has no gender distinctions—a
feature shared with its relatives Persian, Kurdish, and Zazaki. The category ‘human’
is marked just on one plural form: a special marker of group plurality kadag
(originally meaning ‘settlement, village’) occurs only with humans. 14
A development of a special ‘masculine’ (or ‘virile’) form out of reinterpreted case–
number combination has been reported for Polish. The dual genitive-accusative
came to be used with nouns of masculine gender exclusively, giving rise to ‘virile’
forms initially used just with number word ‘two’ and the modifier ‘both’. Over the
course of the fourteenth–seventeenth centuries, dual forms were gradually replaced
with plural forms and extended to constructions with other numbers. The virile
genitive construction in Polish appears in number words in 6.8, for ‘men’, and 6.9, for
‘women’:
6.8 pie̹ciu/dzie̹sieciu/wielu panów
five:VIRILE/ten:VIRILE/many:VIRILE
‘five/ten/many men’
6.9 pie̹ć̩/dzie̹siḛć/wiele kobiet
five/ten/many women:GEN.PL
‘five/ten/many women’ 15
men:GEN.PL
New gender distinctions can be developed through reanalysis of existing case
forms. 16 We can recall, from §5.2, how Spanish has come to distinguish animate
and inanimate nouns through marking objects (examples 5.3–5.4).
During the Common Slavic period, the nominative and the accusative singular
forms of masculine nouns fell together, due to phonological shortening and erosion
of final segments. The masculine gender thus became similar to neuter (where
nominative and accusative forms were the same). The homonymy of two different
cases (or case ‘syncretism’) was solved by substituting the genitive for the accusative
for masculine animate nouns. Originally, this may have started just from nouns

6.2 Reshaping Linguistic Genders 83
referring to singular males. In thirteenth- and fourteenth-century Russian, only free
adult males were covered by the ‘animate’ declension—thus adding a sociolinguistic
dimension to the classification of nouns. The pattern of different declension for
animate and inanimate masculine nouns in Slavic languages can be accounted for by
the principles behind Differential Object Marking (mentioned in §5.2): animate, and
especially human nouns tend to acquire special marking when used as objects. 17
Proto-Indo-European gender has been lost in a few languages of the family. In
some, this loss was accompanied by emergence of a new, gender-like, distinction.
Armenian lost gender distinctions before its first attestation. There is now a distinction
between animate versus inanimate, or person–non-person, in case paradigms. 18
Bengali, an Indo-Aryan language, does not have any gender distinctions in agreement.
A new, human versus non-human, distinction appears in personal pronouns.
Third person pronouns typically refer to humans. To talk about a non-human, one
uses a demonstrative. 19
6.2 Reshaping Linguistic Genders
Linguistic Genders change, expand, and contract in a variety of ways. Derivational
gender markers on nouns can start being used with personal pronouns, and then
develop into an anaphoric gender. South Dravidian languages (which originally had a
masculine/non-masculine gender distinction) innovated a special feminine gender
distinction in personal pronouns by extending the feminine derivation suffix *-al̍to
the pronominal root *aw. The feminine personal pronoun awal̍was created following
the analogy of pairs such as *mak-antu ‘son’, *mak-al̍
‘daughter’. 20 This new
anaphoric gender has its roots in the erstwhile derivational marker.
The principles of Linguistic Gender choice may change over time. In Hausa,
feminine gender is now chosen on the basis of the noun’s phonological form:
nouns in -aa are feminine. This ending goes back to a fossilized derivational suffix
used to form feminine nouns. What used to be a morphological principle of assignment
is now phonological. 21
Linguistic Genders may develop new meanings. Konkani, an Indo-Aryan language
spoken on the west coast of India, has three Linguistic Genders—masculine, feminine,
and neuter. There is anaphoric gender in pronouns, and gender agreement on
some adjectives and verbs. In some dialects, a young woman can be referred to with
the neuter pronoun tε̃. The erstwhile feminine pronoun ti refers to an older female. If
a romantic novel or a movie with a young heroine has the title She its appropriate
translation would involve the erstwhile neuter pronoun tε̃‘(younger) she, formerly
it’, and not ti ‘she’. 22
Markers of Linguistic Genders may be lost altogether. Then nouns regroup, based
on their meanings and on their form. Examples come from the history of Romance
languages.

84 6 The rise and fall of Linguistic Genders
Latin had three genders—masculine, feminine, and neuter. These were marked on
the noun itself, and also by agreement on adjectives and a variety of pronouns. Most
Romance languages lost the neuter gender, and now have just masculine and
feminine. 23 The erstwhile Latin neuter nouns were redistributed between masculine
and feminine. So, neuter nouns of the second declension (which contained numerous
masculine nouns) became masculine, e.g. Latin ferrum ‘iron’, Italian and Portuguese
ferro, Spanish hierro, French fer, Catalan ferre, Sardinian ferru. Some neuter plurals
ending in -a were drawn into the feminine gender, as -a was reinterpreted as a feminine
gender marker, e.g. arma ‘arms’, Italian, Sardinian, Catalan, Spanish, Portuguese arma,
French arme.
Some neuter nouns had variable gender in Latin; they became masculine in some
languages and feminine in others. The descendant of the Latin neuter noun mare ‘sea’
is feminine in French (la mer), but masculine in Portuguese (o mar) and in Spanish
(el mar; but old Spanish la mar). Analogy played a role. According to Pope (1934:
304), mer ‘sea’ in French was co-opted into the feminine gender by analogy with its
counterpart, terre ‘earth, ground’ (Latin terra, also feminine). Similarly, Old French
vallis ‘valley’ changed its gender from masculine to feminine under the influence of
mons ‘mountain’. Latin aestas ‘summer’ (French été) was originally feminine, but
became masculine under analogical pressure of other terms for seasons. The Old
French dent ‘tooth’ changed its gender from masculine to feminine, under the
influence of bouche ‘mouth’. The history of ‘redistribution’ of neuter nouns boils
down to a history of individual words.
The loss of neuter occurred in other Indo-European languages. In Celtic, the
neuter disappeared fairly early. Old Irish had three genders. The neuter was lost in
the Middle Irish period, c. tenth–eleventh centuries AD. Hardly any traces of neuter
gender have been recorded for Welsh and Breton. Latvian, a Baltic language, has only
masculine and feminine genders. Neuter has been lost in a few Slavic varieties, among
them the Sele Fara dialect of Slovene spoken in Austria.
In a number of languages, the neuter gender was marked by an unstressed vowel at
the end of a noun and of agreeing forms. The reduction of the final unstressed o to -ə
in neuter nouns made them undistinguishable from masculine nouns. The ‘death’ of
the neuter gender (as Priestly 1983 puts it) in Sele Fara took a few decades. During
his fieldwork in 1935, Isačenko (1939) documented a tendency towards a ‘masculinization’
of neuter nouns. The neuter gender was completely gone by the time Priestly
did his own fieldwork in 1979 and 1982. Most neuter nouns became masculine (both
in form and in the agreement they take), and just a few shifted to the feminine
gender. 24
Phonological erosion of nouns’ endings—where genders are marked—may put the
whole category in peril. East Indo-Aryan languages Bengali, Assamese, and Oriya lost
all the traces of masculine, feminine, and neuter genders so prominent in Sanskrit,
their oldest Indo-Aryan relative. In Masica’s (1991: 221) words, ‘far from being

6.3 Partial loss and reinterpretation of gender 85
aberrant, this Eastern Old Indo-Aryan loss of gender can be seen as the most natural
outcome of the phonological erosion’. The Old Iranian languages Avestan and Old
Persian had three genders, just as did Sanskrit. The final syllables of nouns and
agreeing modifiers were weakened and lost in Middle West Iranian languages—
Parthian and Pahlavi—and in numerous modern Iranian languages, including Modern
Persian, Baluchi, Tat, and Talysh; and the genders were lost. 25
We now turn to the partial loss of Linguistic Gender in English, and its transformations
throughout the history of the language.
6.3 Partial loss and reinterpretation of gender: the story of English
Gender in English—if compared to its direct ancestor, Old English (and also to
Proto-Germanic)—is distinctly impoverished. Old English had three genders—
masculine, feminine, and neuter. Genders were marked on nouns themselves, and
on agreeing modifiers. Different agreement forms of the demonstrative ‘this’ feature
in 6.10. They have all been replaced with one form for all in the modern language, as
can be seen from the translation.
6.10 se cyning
this.MASC king.MASC
‘this king’
sēō cwēn
this.FEM queen.FEM
‘this queen’
Þæt sćip
this.NEUTER
‘this ship’
ship.NEUTER
Most nouns denoting men in Old English were grammatically masculine (e.g.
man, secg, wer ‘man’) and nouns denoting women were feminine (e.g. mægð ‘maid’,
cwen ‘queen’). There were a number of notable exceptions for nouns referring to
females—the masculine noun wifmann ‘woman’; and neuter nouns wif ‘woman’,
mægden ‘maiden’. Anaphoric gender—marked in personal pronouns—was mostly
based on the grammatical gender of a noun. So, wif ‘woman’, grammatically neuter,
would be referred to with a neuter pronoun. Studies of Old English show that the
seeds of semantic agreement (similar to what we saw in §2.1.2 and 2.3 for German)
were already visible as early as the eleventh century. As Curzan (2003: 45) puts it, in
Old English documents:
Within the noun phrase, grammatical gender agreement in the inflectional endings of the noun
and its modifiers is obligatory (with occasional exceptions for unnaturally gendered animate

86 6 The rise and fall of Linguistic Genders
nouns such as masculine wifmann ‘woman’). Outside the noun phrase, however, grammatical
gender agreement between anaphoric pronouns and their antecedents is often variable. In the
few cases in Old English where the grammatical gender of a noun referring to a human being
conflicts with its gender (e.g. wif ‘woman’ [neuter]), the anaphoric pronoun almost always
agrees with the referent’s sex. In other words, nouns such as wif and wifmann are referred back
to with forms of she. 26
In later documents, inanimate objects started being referred to with ‘it’ independently
of their grammatical genders. By the end of the fourteenth century agreement
gender and gendered forms of nominal declensions were all but lost, as part of
general decay of case and agreement endings.
The three anaphoric genders survived in personal pronouns (and also in reflexives,
such as herself and himself ). Their choice shifted to being almost exclusively based
on sex and animacy: nouns denoting male humans were referred to by he, nouns
denoting female humans referred to by she, and nouns denoting non-humans
referred to by it (irrespective of their sex). This is of course a rough approximation:
additional uses of gendered pronouns (such as ‘she’ for ships, cities, and countries)
make the system less straightforward.
The history of gender loss and restructuring in English involved two processes:
A. Loss of agreement gender and gendered declensions was due to the decay of
case and agreement ending (the result of phonological attrition of word-final
syllables). This process was complete by mid- to late fourteenth century.
B. The anaphoric gender has never been lost. Gender agreement became almost
straightforwardly semantic—from that based on the grammatical gender of a
noun (independently of the referent’s sex or animacy) to that based on animacy
and sex. The process started in Old English (before declensional distinctions
and agreement were lost), and must have become complete by the late Middle
English period. 27
In all likelihood, the two processes took place simultaneously. The exact reason for
the demise of agreement gender and gendered declensions in English (point A)
remains a matter of some debate. As Curzan (2003: 48) puts it, ‘the facts of the
gender shift in English potentially argue for at least some external causes’, as stress
shift ‘alone cannot account for why English lost its complex inflectional system and
grammatical gender system’. 28
The changing status of Social Genders has affected the use of gendered pronouns
in English, and especially of the ‘generic’ he; this is the topic of §11.5.1.
A number of other languages followed the same path as English—losing their
agreement gender, but maintaining Linguistic Gender in anaphoric pronouns. This is
what happened in Cross-River languages, from Nigeria. In Yazgulami and Farizandi,
two Iranian languages, masculine and feminine genders remain in personal pronouns
only (Edelman 1966). A number of languages went the other way. In the Dardic

6.4 Linguistic Gender in language contact 87
languages Garvi and Gavar, and in Punjabi, an Indo-Aryan language, the masculine
and feminine survive in nouns, some adjectives and verbal forms, but not in personal
pronouns. 29
So far we have focused on language-internal reasons for changes in Linguistic
Genders. We now turn to the role of language contact in how genders are shaped.
6.4 Linguistic Gender in language contact
As speakers of different languages come in contact with each other, their languages
become similar in various ways. Linguistic Gender is among the categories which
tend to spread from one language to the next in a contact situation. If a language with
Linguistic Genders comes in contact with a gender-less language, chances are that
Linguistic Gender is likely to fall into disuse. 30
6.4.1 To lose a Linguistic Gender
Intensive language contact has resulted in the loss of Linguistic Genders in a few
well-documented instances. Literary Latvian (from the Baltic subgroup of Indo-
European) has two genders, masculine and feminine. The Tamian dialect of Latvian
used to be in close contact with Livonian, a Balto-Finnic language with no genders.
The loss of grammatical gender in Tamian Latvian is the outcome of Livonian
influence.
Brahui is unusual for a Dravidian language in that it has no gender distinctions. In
the last centuries, the Brahui people appear to have been in close contact with the
Balochi, speakers of a West Iranian language with no gender. Earlier on, the Brahui
may have been in contact with Persians (whose language also has no gender). Once
again, the complete loss of gender in Brahui results from language contact.
Along similar lines, the Albanian dialect of Mandres in northern Greece completely
lost the system of two genders (feminine and masculine) found in all other
Albanian dialects. The entire village moved to its present location from the adjacent
areas of Turkey. In all likelihood the loss of gender is due to Turkish influence.
Similarly, Tangale, a Chadic language, lost gender distinctions as a result of contact
with unrelated Adamawa, with no Linguistic Gender. The loss of gender in Arawak
languages Mawayana, Amuesha, and Chamicuro was due to the impact from the
surrounding and unrelated languages with no genders. Two genders—masculine and
feminine—are a typical feature of Omotic languages of Ethiopia; Northern Mao is the
only exception. Its gender was lost due to the influence from their immediate
neighbours—speakers of unrelated Surmic languages. 31
When languages are in contact, with people speaking several languages, their
linguistic features adjust to each other. This is a way of reducing processing and
cognitive overload. Gender plays a role in cognition and conceptualization of the

88 6 The rise and fall of Linguistic Genders
outside world. It is also a source of metaphors, and the basis of cultural stereotypes
which determine human behaviour and relationship. What makes Linguistic Gender
so susceptible to contact is its role in cognition and communication.
The loss of Linguistic Genders may be conditioned by a number of factors—
linguistic contact being just one of them. Greek varieties, spoken in Turkey, were very
much unlike Modern Greek in their grammar. As a result of contact with Turkish,
they lost many of the typical Greek features, and developed Turkish-like ways of
saying things. Agreement gender and anaphoric gender (also absent from Turkish)
were lost. But was Turkish influence the only culprit? In all likelihood, Turkish
influence was an accelerating factor in the process of loss of a three-gender system,
already started before contact with Turkish. 32
6.4.2 Evolving a Linguistic Gender
A reverse situation may occur. Suppose that the majority of languages in a family do
not have any grammatical gender distinction, and just one language does. If this
language is in contact with, or located close to, an unrelated language with grammatical
gender, it is fair to suggest that gender has probably been developed independently, as a
result of contact-induced change. Tualatin was the only member of the Kalapuyan
language family (formerly spoken in Western Oregon) to have two genders marked in
demonstratives. One gender subsumed human females, the other everything else.
Tualatin bordered Clackamas Chinook (a language with three genders), and in all
likelihood the emergence of gender in Tualatin was due to language contact. 33
In languages with no other linguistic means of expressing gender, derivational
gender may come about through borrowed forms. A borrowed noun can be used in
compounds, and then reinterpreted as derivational gender.
Finnish has very little in terms of linguistic gender. Its major marker of derivational
gender is of a foreign origin. One of the oldest Germanic loanwords into
Finnish is tytär ‘daughter’, from Germanic *tuttare ‘daughter’. The borrowed form
appeared in many compounds and was later grammaticalized as a suffix -ttare. This
suffix is now used with numerous terms (native and borrowed) referring to women,
e.g. runo-tar ‘poetess’, from runo ‘poem’ (native word), kuninga-tar ‘queen’, from
kuningas ‘king’ (a Germanic loan). This form was borrowed from Finnish into closely
related Estonian by Johannes Aavik, the major proponent of the Estonian language
reform in the early twentieth century (it is most used in occupational terms, e.g.
laulja ‘singer’, laulja-tar ‘female singer’). 34
Developing a gender marker does not have to involve grammaticalization. Suppose
a language without Linguistic Gender comes in contact with another language with
productive masculine and feminine forms. It may then borrow gendered words.
Gender markers may get reanalysed as separate morphemes and extended to native
words, echoing the reinterpretation in the history of Agreement Gender (see §6.1).

6.4 Linguistic Gender in language contact 89
This is what happened in Indonesian, a language with no Linguistic Gender. Some
loanwords with human reference from Sanskrit, an Indo-Aryan language with
genders, come in pairs: forms denoting males end in -a and forms denoting females
end in -i—for instance, putra ‘son’ and putri ‘daughter’, saudara ‘male sibling or
relative’ and saudari ‘female sibling or relative’, dewa ‘god’ and dewi ‘goddess’. This
principle has been extended to a few native words. Pemuda originally meant ‘young
person’, but nowadays this refers to ‘young man’, and a new word, permud-i ‘young
woman’, was created as its female counterpart.
Masculine and feminine in Turkish can be expressed through a handful of
derivational markers. Feminine gender markers -ic̹e and -c̹a are borrowings, from
the Serbo-Croat feminine derivational gender marker -ica. They occur in a few Slavic
borrowings into Turkish, all referring to women, e.g. kralic̹e ‘queen’ (masculine kral
‘king’), c̹aric̹e ‘tsarina’ (masculine c̹ar ‘tsar’). The neologism tanric̹a ‘goddess’ was
formed by analogy from the name of the old Turkish sky-god Tanrɪ. Just like in
Indonesian, borrowed nouns have been reanalysed, and the loan suffix expanded to
further nouns.
The derivational gender marker -nna in Estonian is currently used with loanwords
and with native words. It penetrated the language from German loans, such as
kuninganna ‘queen’ (from German Königin) with its masculine counterpart kuningas.
As a result of reanalysis, it was expanded to native words, such as sõbranna
‘female friend’ (sõber ‘male friend, friend in general’). 35
Foreign markers of derivational gender are common in languages which have
linguistic genders of other kinds. The feminine suffix -ess in English goes back to
Greek -issa from where it was borrowed into Latin as -issa and French as -isse. Pairs
such as host and host-ess and patron and patron-ess were borrowed into Middle
English. From about the late fourteenth century onwards, the suffix expanded to
native English nouns, creating forms like goddess, danceress, and dwelleress. The
suffix is quite productive—we often hear hostess, waitress, governess, and stewardess.
But some -ess nouns shifted their meanings: mistress is not just a female counterpart
of a master. Many a female writer of poetry would object to being called a poetess,asa
generic poet is devoid of ‘woman-like’ overtones and is thus more desirable in
status—similarly to the treatment of ‘gender-specific’ terms for professions with
the same suffix in other languages—§7.5.
In Old English, a derivational gender marker could be added to nouns referring to
a person who made (or was concerned with) the item referred to by the noun:
6.11 BASE NOUN MALE PERFORMER FEMALE PERFORMER
sang ‘song’ sang-ere ‘male singer’ sang-estre ‘female singer’
sēam ‘seam’ sēam-ere ‘male tailor’ sēam-estre ‘seamstress, female dressmaker’
The original masculine suffix -ere has developed into modern -er (now used with any
gender reference). Suffix -estre gave rise to the modern -ster. This suffix lost its

90 6 The rise and fall of Linguistic Genders
exclusively feminine reference beginning from late Old English times. Once this
happened, the suffix -ess—already established as a marker of derivational gender—
could follow it. Seamster originally referred to a woman, then to someone of either
sex, and then the suffix -ess was added, to refer to a female maker of garments,
creating seamstress. The form seamster dropped out of use, but the pair songster and
songstress remains.
Another foreign feminine suffix borrowed from the Old French is -ette, first as a
diminutive as in kitchenette, and then, from the early twentieth century, as a
feminine: usher-ette in a cinema, or undergradu-ette at a university. Many feminines
in -ette have a negative feel about them; some were created as pejoratives in the first
place. The term suffragette was coined in 1906 by a British journalist writing for The
Daily Mail who was sceptical of the activities of a suffragist—a man or a woman
advocating equal rights to vote for both sexes. 36 We return to the overtones of
feminine forms in §7.5, and §11.3.
Borrowed gendered forms are not limited to derivation. Romanian, a Romance
language, was influenced by Slavic languages. The vocative singular feminine ending
-o was borrowed from Bulgarian into Romanian. Examples include soră ‘sister’,
vocative soro!, and mamă ‘mum’, vocative mamo (Puşcariu 1943: 435).
6.4.3 Readjusting Linguistic Gender
If languages are in contact, their genders may become similar. A prime example
comes from the language contact situation described by Gumperz and Wilson (1971:
156) in the Indian village of Kupwar. The languages spoken in Kupwar are Kannada,
a Dravidian language, and Marathi and Hindi-Urdu, two Indo-Aryan languages. The
Kupwar Kannada gender system is identical to that of standard Kannada. There are
three genders whose assignment is semantically based: nouns referring to males
belong to the masculine gender, nouns referring to females belong to the feminine
gender, and all other animate and inanimate nouns are neuter. The standard Marathi
has three genders (masculine, feminine, and neuter). Inanimates can belong to any
gender. Not so in the Kupwar Marathi: here non-human nouns which correspond to
Marathi masculines or feminines are neuter, replicating the Kannada pattern. Only
human males are masculine, and human females are feminine. The gender choice has
been restructured and adjusted to the dominant language—the Kupwar Kannada.
The noun nadī ‘river’ belongs to the feminine gender in standard Marathi; in Kupwar
Marathi it is treated as neuter. 37
Adjacent languages in the Australian area typically have similar gender systems. In
the Mayali dialect chain, the western-most dialects are becoming more like their
neighbour Jawoyn, with only three, instead of four agreement classes, and the
eastern-most dialects are tending to become more like their neighbours Rembarrnga
and Dangbon, both of which lack noun class agreement (Dixon 2002: 500). The

6.4 Linguistic Gender in language contact 91
Australian language Yanyuwa developed a five-term gender system marked on nouns
and on verbs (human masculine, non-human masculine, feminine, vegetable, and
trees). The terms in the system come from different sources. The system itself is
remarkably similar to Yanyuwa’s westerly neighbours, which had such a system
(Dixon 2002: 500–1).
Gender can be particularly stable in the history of a language if the speech
community is aware of it as a particularly salient feature. Gender—viewed as dividing
all the referents into ‘masculine-like’ and ‘feminine-like’—is what speakers of two
languages from the Sepik area in New Guinea—Manambu (Ndu family) and
Kwoma—pinpoint as a striking feature of their language. Gender is an emblematic
feature in Nilotic languages Luwo and Labwor—which has helped maintain it in
language contact. 38
6.4.4 Linguistic Gender in language obsolescence
As a minority language becomes obsolescent and loses ground to a dominant
language, its Linguistic Gender is likely to be affected. Dyirbal, a vibrant language
in the 1960s, is now declining. Traditional Dyirbal had four agreement genders based
on a complex network of semantic features, as we can recall from §2.2.1. Young
People’s Dyirbal has gradually adjusted its gender system towards that of English and
made the gender choice more like that for English anaphoric gender. Dyirbal’s
‘typical’ Australian gender covering edible non-flesh food simply disappeared: the
speakers of Young People’s Dyirbal simply stopped using it. The scope of ‘feminine’
gender 2 became reduced and became reserved only for females (it used to include
drinkable liquids, fire, and things associated with fighting). Gender assignment by
mythic association was lost; exceptions became regularized; and the use of the residue
‘neuter’ gender expanded—subsuming all the inanimates, covered by ‘it’ in English. 39
Scottish Gaelic is another example of a language in decline. Gaelic has masculine
and feminine agreement genders and also anaphoric genders. Agreement genders are
marked in very specific ways—by initial mutation of the noun after the definite
article, by agreement of an adjective (again using initial mutation), by the form of
the definite article. There is also a gender-marked diminutive suffix. Younger and
not-so-fluent speakers tend to overuse the masculine pronoun a for all inanimate
nouns. Adjective agreement via consonantal lenition shows variation. Dorian (1981:
148) predicts that if Eastern Sutherland Gaelic had ‘future generations of fluent
speakers’, gender marking in pronouns and agreement on adjectives ‘would surely
disappear’, but head-marking gender on nouns would survive, ‘producing a kind of
lexicalisation of gender specification linked to specific nouns’. 40
If speakers are shifting to a language with no linguistic gender, they may well start
losing their own. The last speakers of Baré, a North Arawak language from Venezuela
and adjacent regions of Brazil, were fluent speakers of Língua Geral, a Tupí-Guaraní

92 6 The rise and fall of Linguistic Genders
lingua franca. Língua Geral has no gender distinctions. As a consequence, masculine
and feminine agreement on adjectives, and even anaphoric gender, in Baré became
optional. The last speaker of Baré would often use ‘he’ to refer to a woman.
The demise of Linguistic Gender in language attrition is not universal. In Arvanitika,
an endangered Albanian variety spoken by Albanians who emigrated to Greece
in the eleventh and fifteenth centuries, the three genders remain distinct. This may be
due to the fact that the three-way gender distinctions in Arvanitika and the dominant
Greek are structurally similar. 41
6.5 Linguistic Gender in language acquisition and language dissolution
When children learn their first language, Linguistic Gender is among the first features
they master. By the age of 3, most Hebrew-speaking children have gained full
command of agreement gender. Similar results have been obtained for speakers of
German, Polish, and Sesotho, a Bantu language. 42
Complex marking of Linguistic Gender may slightly delay its full acquisition.
Polish children acquire three genders earlier than Russian children because Linguistic
Gender in Russian may be confusing: in the spoken language, the typical feminine
ending -a sounds very similar to the typical neuter ending -o since both are reduced
if they are not stressed. This is not the case in Polish. Russian also has a number of
diminutive and endearing forms referring to boys and men, but ending in -a, e.g.
Kolja ‘diminutive for Nicholas’, orVitja ‘diminutive for Victor’. And quite a few
common nouns refer to men, but end in -a (and decline as feminines), including
djadja ‘uncle’. Young children would say djadja side-l-a (uncle sit-PAST-feminine.
singular) instead of djadja side-l (uncle sit-PAST+masculine.singular) ‘uncle was
sitting’. These problems disappear by the age of about 3.
In addition to this, diminutives of female names can take masculine endings
‐ok/-ik, asinSveta ‘diminutive of Svetlana’, Svetik ‘diminutive of Sveta’. And there
are nouns which can refer to men and to women, with a masculine form, such as vrac᷈
‘doctor’. A recent study by Rodina (2014) showed that even when caregivers preferred
feminine forms, 2- and 3-year-olds tend to use masculine forms for these
cases—being led by the form of the noun rather than by the meaning. This overgeneralization
appears to have continued until about the age of 7. That is, the
acquisition of complex agreement rules can be delayed by their complexity and a
less than straightforward input—but only slightly. 43
Gender in German is marked in many different contexts—personal pronouns,
adjectives, articles, and so on. This extensive system of expressing grammatical
gender gives the German child ample opportunities to practise it, seek regularities
in the system, and produce the forms accurately at an early age. German children
have an advantage over English children who are exposed just to the anaphoric
gender in personal pronouns. And indeed, a study conducted by Mills (1986: 103–6)

6.6 Linguistic Gender and language reforms 93
showed that German children ‘are clearly in advance in their production of the
masculine and feminine pronouns’, compared with the English children within the
same age bracket. The German children make hardly any mistakes by the age of 3,
while English children do until the age of 4.
Learners of a foreign language project the Linguistic Gender of the language they
know best onto the system they are trying to master. Speakers of Russian learning
Hebrew tend to transfer the gender of their native language to Hebrew. Speakers of
English (which only has anaphoric gender) reply on the meaning of nouns, and
largely disregard the form of nouns (Armon-Lotem and Amiram 2014).
The human brain can be damaged—as a result of a stroke, or an accident.
Language skills may then be impaired or lost. Those affected by it can sometimes
produce non-grammatical sentences, or have difficulties understanding their first
language. Studies which looked at the behaviour of gender agreement forms in
German, Italian, and Polish found that sufferers of ‘agrammatical’ aphasia often
use the wrong gender form—typically, masculine for feminine. But no matter how
serious the condition, Linguistic Genders persist in some form even in those patients
who are most severely affected. 44
6.6 Linguistic Gender and language reforms
Speakers of languages with Linguistic Gender are typically aware of it as a salient
feature, and are often prepared to discuss it (see §4.3 and §7.7). Speakers of Manambu,
a Papuan language of New Guinea, are ‘gender-proud’: the two genders
(masculine and feminine) are pinpointed as a salient feature of the language. We
recall from §6.4.2 that Johannes Aavik borrowed a derivational marker for feminine
gender into Estonian, to fill what he felt to be a gap.
Linguistic Gender has been the target of language reform in the history of a number
of European languages. Norwegian is a case in point. Until the early twentieth
century, literary Norwegian had a system of two Linguistic Genders: a ‘common’
gender and a neuter gender, as in Danish and in Swedish, rather than the original Old
Norse three genders (masculine, feminine, and neuter). The feminine gender was
restored with the development of Nynorsk, a language largely created by Ivar
Aasen, in the context of the rise of Norwegian nationalism. The reform drew on
south-west dialects of Norwegian, where the feminine gender remained in use. The
1938 language reform made the feminine form obligatory for about 1,000 nouns. As
Gregersen (1979: 11) remarks, ‘nationalism rather than feminism was the motivation
for the change: the feminine gender was felt to be a particularly Norwegian trait’. At
present Nynorsk is one of the two official languages of Norway. 45
The forms and the uses of Linguistic Genders for humans may change as the status
of Social Genders evolves. The growing equality of women in many Western societies
has brought about an increase in derivational forms with female reference. Language

94 6 The rise and fall of Linguistic Genders
reforms have targeted ‘sexist’ language—ensuring the visibility of women, and
avoiding the generic use of masculine pronouns and terms for ‘man’. Social motivations
for change in Linguistic Genders highlight its functionality, and overall
importance. We turn to this in §§11.4–5.
6.7 Linguistic Genders, their development, demise, and transformations
Linguistic genders can come from a number of sources. A language can gain an
anaphoric gender by making a gender affix out of a noun meaning ‘person’ or ‘man’.
This can then be expanded into derivational gender, or into agreement gender.
A gender-sensitive pronoun can become a marker of gender agreement. Noun classifiers
can give rise to genders. Nominal categories of number and case are intrinsically
prone to reflecting meanings to do with animacy and humanness.
The existing genders can change their meanings. Genders can be lost, and nouns
regrouped based on their phonological shape or meaning.
Agreement gender and gender-based declensions were rife in Old English. By the
end of the Middle English period, they were as good as gone. The anaphoric gender
survived in personal pronouns. But its choice gradually came to be based on the
animacy and sex of the noun—rather than its agreement gender, now lost. The seeds
of this change were present in Old English: a neuter noun wif, the precursor of
modern-day wife, could be referred to as ‘it’ or as ‘she’. The process of making gender
in English semantically transparent appears now to be complete. But its simplicity is
illusory: Linguistic Gender—especially applied to human beings—is a target for language
reform, and a barometer of social status, and equality, for women (we return to
these issues in §§11.5.1–2).
If languages are in contact with each other, their genders are likely to be affected.
They can be gained, or lost. Or their meanings can readjust to each other. Languages
with no grammatical genders often borrow gendered forms. These are reanalysed,
and a derivational gender emerges. This is a means which many languages exploit,
expanding derivational genders they may already have.
If a language becomes obsolescent and gradually falls into disuse, it will adjust to
the language which is taking control. If the dominant language has no grammatical
gender, the obsolescent language will lose it. Or its genders can get restructured, to fit
the mould of the dominant system. Gender is acquired early and stays on in many
forms of language dissolution in aphasia.
Linguistic Gender can be a target of language reform, especially inasmuch as it
concerns the perceived asymmetries of Social Genders reflected through linguistic
expression of genders. Social motivations for changes in Linguistic Genders point
towards speakers’ awareness of them, and their functionality and importance in the
social life of many peoples.

6.7 Linguistic Genders, their development, demise 95
Gender—once it exists in a language—is an inextricable part of the linguistic
machinery, the way the language works. The mastery of genders offers speakers an
array of expressive devices, making the language more colourful and versatile. This is
what we turn to next.
NOTES AND SOURCES
1. See Heine and Kuteva (2002), Heine and Reh (1984), and Trask (2000) on grammaticalization.
Grammaticalization is often accompanied by a reanalysis of forms and by their
reinterpretation (when they acquire new and different meanings).
2. See Palmer (2009: 69), and van den Berg (2015), for further examples.
3. In a few other Eastern Nilotic languages, these markers of anaphoric gender develop into
agreement markers (step 2 in Pathway I); see Heine and Vossen (1983); cf. Heine and Reh
(1984: 219–20, 230–2), on Nilotic languages.
4. Heine (1982b: 214), Claudi (1985).
5. See Martin (1992: 180, 768), Lee (2014: 42) on Korean; see also Aikhenvald (2000: 442–6).
6. The demonstratives used as anaphoric gender pronouns are mtona ‘he’ and minana ‘she’,
and contain the classifier forms (see Senft 1986, 1996: 18–22, 353 and van den Berg 2015).
7. See Zavala (1992: 152–3, 2000: 134–5) on Akatek; England (1983: 158) on Mam; de León
(1987: 148, 157–8), and also Craig (1986a: 266–7) on classifiers in Jacaltec. Number 7
stands for a glottal stop.
8. The derivational markers are not used in agreement. See Frajzyngier (1993: 49) on Mupun;
similar examples have been described for !Xoo, the northern dialect of !Xun, a Khoisan
language, and Hunde, a Bantu language: Heine and Kuteva (2002: 133, 209, 213, 315). See
Braun (2000a: 53) on Turkish.
9. See Reid (1997: 215–18); a somewhat similar example is in Green (1997). A general
historical analysis is in Dixon (2002: 467ff.).
10. A similar scenario could apply to Bandjalang (Dixon 2002: 464–6). Details on these
scenarios can be found in Dixon (1982: 171; 2002: 464–6, 497–506), Reid (1997); a brief
summary is in Aikhenvald (2000: 394–400).
11. See Dixon (2002: 499), and Heine and Reh (1984: 232–4), on the mechanisms of development
of gender agreement based on anaphoric gender.
12. See, for instance, Clackson (2007: 91–113), Luraghi (2011). Note that three genders in
Latin quis ‘who (masculine)?’, quae ‘who (feminine)?’, and quid ‘who (neuter)?’ introduced
by analogy with other forms: Clackson (2007: 105).
13. A similar scenario is sketched out by Meillet (1931: 17–20), a major classic in Indo-
European scholarship. He suggested that the new gender-marking suffix-h 2 was extended
to the stem of the animate demonstratives, creating a subcategorization between masculine
and feminine nouns. Further discussion and suggestions are in Luraghi (2009, 2011).
A somewhat different scenario and explanation are in Brosman (1979) and Melchert
(1994); see also Fodor (1959: 213).
14. Axenov (2006: 62, 69). Sex differences of animates can be expressed with nouns nar ‘male’
and maādag ‘female’.
15. The scenario is outlined by Janda (1999: 80–2); see also Klenin (1983).
16. The development of genders in Iroquoian languages has been linked to prefixes marking
‘agent’ and ‘patient’. The non-specific agent marker in Proto-Northern-Iroquioan came to

96 6 The rise and fall of Linguistic Genders
be used just for humans, possibly, based on the frequency of its use. Its specific counterpart
developed into a non-human marker. A further non-human agent prefix has its origins in
reinterpretation of a number marker (*wa): see Chafe (1977: 515–16), and Mithun (2014).
17. Huntley (1980) contains a snap-shot of the history of the ‘animate’ and ‘inanimate’
declensions in Slavic languages; see also Luraghi (2014: 223–4), and Priestly (1983:
357–8). Some semantically inanimate nouns which follow the animate conjugation in
Polish, and a further division between masculine ‘person’ and masculine ‘animal’ are
discussed by Wertz (1977: 50, 53, 57–9). More examples of gender variation in Contemporary
Polish are in Kryk-Kastovsky (2000). Tocharian A, an extinct Indo-European
language, is the only language in the family with feminine and masculine forms of first
person full pronouns: näs̍‘I (male speaker)’ and ñuk ‘I (female speaker)’. Jasanoff (1989)
demonstrated that the masculine–feminine gender distinctions developed through reinterpretation
of the old accusative and nominative case forms.
18. Garibian (1976), Priestly (1983: 357); Clackson (2007).
19. See Onishi (1996); and Priestly (1983: 357); animate and inanimate pronouns are also
distinguished in Assamese in Oriya (further Indo-Aryan languages which lost Indo-
European genders).
20. See Krishnamurti (2003: 211).
21. Newman (2000: 208).
22. In a formal context ti can also refer to a young female. See Miranda (1975: 209–11). In all
likelihood, the change in the use of gendered forms was triggered by change in meaning of
a neuter noun c᷈ed̩ũ from ‘child’ to ‘girl’. See also Masica (1991: 220–1) on how the threegender
system in Sinhalese was restructured under Tamil influence, as basically animate
(with a subdivision into masculine and feminine) and inanimate.
23. See also Posner (1966: 132–6), Pope (1934: 304), Matasović (2004: 49–52) on gender in
Romance languages. See discussion of the neuter gender in Romanian in Corbett (1991:
150–2), and Loporcaro, Faraoni, and Gardani (2014) for the discussion of neuter in Italian
dialects.
24. Masculine and feminine genders have fallen together into a ‘common’ gender in a number
of Germanic languages—including Dutch, Frisian, and standard varieties of Swedish,
Norwegian, and Danish. The common gender now contrasts with neuter gender, for
reasons which are largely open to discussion. Harbert (2007: 93–103) offers a comprehensive
discussion of the history of gender in Germanic; a summary is in Matasović (2004:
59–61).
25. See Matasović (2004: 41–6), on the redistribution and loss of gender in individual Indo-
Aryan and Iranian languages. A phonological explanation for gender loss and restructuring
is echoed by the discussion by Ibrahim (1973: 86–7).
26. As Quirk and Wrenn (1957: 75) put it, ‘strict concord in grammatical gender is the rule in
O(ld) E(nglish). It is particularly regular with demonstratives and modifiers…But pronouns,
particularly when relating to human beings, are sometimes used with natural
gender in spite of the normal requirements of grammatical concord: Þæd maēden ˜ hēō
wearð “the maiden ˜ (n) she was”.’
27. Many scholars refer to A as ‘the loss of grammatical gender’, and to B as ‘the spread of
natural gender’ (e.g. Baron 1971). This terminology is confusing: it implies that anaphoric
gender is not grammatical, and does not account for ‘unnatural’, that is, non-sex-based,
uses of personal pronouns in English.
28. A comprehensive, source-based investigation of the use of anaphoric gendered pronouns
and the gradual shift from grammatical to semantic basis in anaphoric gender is in Curzan

6.7 Linguistic Genders, their development, demise 97
(2003: 58–132); see also Baron (1971) and Karatsareas (2009: 205–6, 2013), and a summary
in Corbett (1991: 101–2). The demise of grammatical agreement gender and the loss of
inflectional endings is discussed in Curzan (2003: 29–57); see also Kastovsky (2000) and
Harbert (2007: 93–101), Matasović (2004: 58–9). Some scholars suggested influence from
Old Norse, or Norman French as a likely trigger; Curzan (2003: 47–54) addresses the
possible external causes in the demise of gender agreement in English (a further theory of
‘creolization’ in the Middle English period is dismissed by her as ‘glib’).
29. See Demuth, Faraclas, and Marchese (1986: 459) and Heine (1982b: 196) on Cross-River
and Kru languages, Edelman (1966) on Yazgulami and Farizandi; further examples are
in Priestly (1983: 346). See Bhatia (1993: 216–17, 222–9) on Punjabi, Edelman (1999) on
Garvi, Grünberg (1999) on Gawar. Further generalizations, and examples, on loss of
genders and gender agreement are in Aikhenvald (2000: 398–9). Dardic and Punjabi
provide exceptions to Greenberg’s universal (1963: universal 43), stating that ‘if a language
has gender categories in the noun, it has gender categories in the pronoun’.
30. See Aikhenvald (2006) and Storch (2006) on spread of Linguistic Gender in situations of
language contact.
31. Matthews (1956: 316). See Emeneau (1980: 157–9) on Brahui; Hamp (1965: 146) on the
Albanian of Mantres; Jungraithmayr (1995: 200–1) on Tangale, Adelaar (2006) and
Aikhenvald (2012a) on Mawayana, Amuesha, and Chamicuro, and Ahland (2012: 41)
and Amha (forthcoming) on Northern Mao. Haugen (1976: §11.4.6) describes dialects of
Swedish which apparently lost genders under the influence of Finnish.
32. Karatsareas (2009: 207); see Janse (2002: 366), Winford (2005: 405), and Dawkins (1916).
33. See Mithun (1999: 432). The emergence of Linguistic Genders in Austronesian languages
of Alor-Pantar area in Indonesia and in a couple of Oceanic languages in West New
Britain are attributed to language contact (Schapper 2010a, 2010b, van den Berg 2015: 39).
34. See Hasselblatt (2015: 131–3) for more examples and the etymology.
35. See Abondolo (1998: 178–9) on Finnish, Tadmor (2007: 311–13), Sneddon (1996) on
Indonesian, Braun (1997a, 1997b), Lewis (2000: 23) on Turkish. Wiedemann (1875: 196)
offers further examples and an explanation of the historical path of the development of this
borrowed feminine affix in Estonian.
36. See Dixon (2015: 320–4), for a detailed analysis, and also Baron (1986: 120–1). The dates
for the words quoted here are based on the Oxford English Dictionary Online.
37. See Gumperz and Wilson (1971) for further details; Pandharipande (1997: 368–9) on
standard Marathi.
38. See Storch (2006) on Nilotic languages.
39. Details on Young People’s Dyirbal are in Schmidt (1985); see also Dixon (2015).
40. See Dorian (1981: 124–9, 147–8).
41. See Trudgill (1977: 35); Sasse (1985). As we recall from §3.2, gender choice in Traditional
Tiwi was partly determined by the size of objects. Young people and children—who speak
a simplified and partly Anglicized ‘Modern Tiwi’—tend to assign gender on the basis of
sex, and are inconsistent in gender assignment to nouns other than those denoting
humans and dogs (Lee 1987: 84).
42. See Berman (1981, 1985: 299–301), Levy (1983a, 1983b), Demuth (1988); similar results
were discussed by Popova (1973) for Russian, French, and German (Karmiloff-Smith
1979).
43. See Popova (1973), Rodina (2014) on the child language acquisition of Russian. Pérez-
Pereira (1991: 585–8) notes delayed acquisition of genders whose marking is ‘ambiguous,
barely transparent and scarcely predictable’, as is the case with three genders in German,

98 6 The rise and fall of Linguistic Genders
Czech, Serbo-Croatian, and Russian. This is in contrast to languages like Hebrew, Egyptian
Arabic, Spanish, or French, with clearly differentiated and regular marking of just two
genders; see also Aikhenvald (2000: 414–16).
44. See Menn, O’Connor, and Holland (1995: 87, 113); Perlak and Jarema (2003) on Polish;
Herbert (1991) on the maintenance of gender differentiation in Bantu-speaking aphasic
patients.
45. See §6.6 on the label ‘common gender’. See Gregersen (1979: 11) for Norwegian and
similar changes in Danish and Swedish, and the introduction of feminine pronouns in
written Chinese. See Hagège (2004: 109–10) on Norwegian and other Germanic languages.

7
Manly women and womanly men:
the effects of gender reversal
Using the opposite Linguistic Gender—feminine for a man, and masculine for a
woman—can be downright wrong. The main reason for my visit to the remote
Swakap in the Sepik area of New Guinea, the only village where the Gala language
is spoken, was to find out whether they really do have different forms for ‘me:
woman’ and ‘me:man’. I travelled with a group of Manambu speakers, and we
spoke Manambu. The two languages—Manambu and Gala—are related (as close as
English and German). The first thing I had to face was the pronouns. Manambu
uses the form wun for ‘me’, no matter what the sex of the speaker. ‘You’ distinguishes
gender of the addressee—mən ‘you man’ and ñən ‘you woman’. Personal
pronouns in the two languages are similar in form, but not in meaning—Table 7.1
summarizes this.
The form meaning ‘you feminine’ in Manambu is used as ‘me feminine’ in Gala.
My attempts at speaking Gala were met with some derision. As a language learner
with some Manambu language background, I found it difficult to use the Gala form
ñən to refer to myself (I, a woman, speaking): it sounded to me as if I was using
‘you feminine’ for myself. The Gala speakers laughed at me when I referred to
myself as wun—which means ‘me’ in Manambu, but just ‘me masculine’ in Gala—
TABLE 7.1. Personal pronouns in Gala 1 compared with Manambu
Gala Manambu
1 person singular masculine
wun
1 person singular feminine ñәn
wun
2 person singular masculine mәn mәn
2 person singular feminine yin ñәn
3 person singular masculine
kәl, kәr dә
3 person singular feminine ki
lә
How Gender Shapes the World. First edition. Alexandra Y. Aikhenvald.
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.

100 7 Manly women and womanly men
saying that I was trying to pretend I was a man. ‘Gender reversal’ was out of the
question.
Not so for female speakers of Palestinian Arabic in Tire (Israel): a woman
may refer to herself with a masculine form if she is talking about being tired, sad,
unhappy, or nervous, as in ʔana mažru:ħ min illi Sa:r (I hurt.masc.sg from that.
which happened) ‘I am hurt by what happened’. A normal feminine form—expected
to be used by a female speaker—is mažru:ħa (hurt.fem.sg). Switching Linguistic
Genders in self-reference is a mark of something unpleasant. No changes the other
way round have been observed: men always refer to themselves with masculine
Linguistic Gender. 2
Gender reversal can serve a different purpose—that of protection, and deception.
Pankhurst (1992: 169) describes how a family of Amharic speakers use pronouns
to address their little daughter—sometimes with the masculine pronoun, and
sometimes the feminine. In Pankhurst’s words,
when applied to children, this blurring of gender identity is used as a strategy to defeat the evil
eye, and other malignant forces conspiring against the life of a child. By referring to the boy as
‘she’ or the girl as ‘he’, people believe they can trick the ‘death-wishers’ and increase the likelihood
of survival. 3
We can recall, from §2.3, that in a system of Linguistic Gender one term may be
functionally unmarked. The masculine Linguistic Gender is the unmarked choice in
Hebrew (see example 2.5 and discussion there). Female speakers of Hebrew tend to
use masculine forms to refer to themselves and to other women extending this
unmarked usage. As 'Anat, a Hebrew-speaking hairdresser, put it, ‘talking in the
masculine frames the conversation as more general and less personal’. 4 Markedness
in Linguistic Gender systems is the topic of §7.7. We return to an association between
the ‘unmarked masculine’ and sexist language in §11.5.2.
Changes in Linguistic Gender assignment to humans may reflect role reversals in
traditional jocular relationships—see §7.1. Gender reversals may be offensive, or
carry overtones of praise—see §7.2. Or they can imply endearment and solidarity—
see §7.3.
7.1 Reversing Linguistic Genders with jocular effects
Among the traditional Manambu, classificatory sisters-in-law—female ego’s brothers’
wives (kajal)—would taunt and tease each other. For a woman, her brothers’ wives are
intruders and rivals of sorts. They would come from a different clan group from their
husband and his sisters, and are like a foreign element in a family. A conventional way
of jokingly greeting a brother’s wife is by saying:

7.1 Reversing Linguistic Genders with jocular effects 101
7.1 ñən wun-a ta:kw-añən
you.fem.sg I-LK+fem.sg wife-2fem.sg.NOMINAL.PREDICATE
wun ñən-a la:n-adəwun
I you.fem.sg-LK+fem.sg husband-1masc.sg.NOMINAL.PREDICATE
‘You (feminine) are my wife, I (masculine) am your husband (feminine)’ (said
by a woman to a woman)
The first clause in this sentence is fully grammatical—the feminine agreement with
the second person feminine pronoun is correct. The second clause is nonsensical,
from the point of view of Manambu grammar. The Linguistic Gender mismatch here
is as follows. The masculine gender is marked on the predicate of the second clause
(‘I am your husband’), but the agreement on the possessive ‘your’ is feminine. The
‘correct’ way to say ‘I (man) am your husband’ is ñən-a-də la:n-adəwun (you.fem.sg-
LK-masc.sg husband-1masc.sg.NOMINAL.PREDICATE). The feminine cross-referencing
here is part of the joke—a woman saying 7.1 presents herself as a kind of ‘female
husband’. She is virtually assuming the Social Gender identity of a patrilineal
representative of her brother’s—and her own—clan. By jokingly assuming the
masculine Linguistic Gender and thus flaunting her ‘quasi-male’ Social Gender
identity, she takes the upper hand, and demonstrates her mock-superiority. This is
achieved through an anomalous gender mismatch, unacceptable in any other circumstances.
Such pronouncements are typically accompanied by roars of laughter
from the speaker and the audience (and often followed by further nasty comments,
e.g. ‘I (masculine) will kick your fat stomach’, ‘I (masculine) will come at night and
strangle you’, and so on).
Switching genders in this context is not insulting. It is, rather, a way of manipulating
one’s Social Gender identity reflected in anomalous gender use. Traditional
women in the villages pride themselves on being able to produce elaborate ‘brother’s
sister to brother’s wife’ jokes: this is gradually becoming a mark of proficiency in
cultural knowledge (which is nowadays on the wane). 5 Culturally permitted ‘gender
switch’ is reminiscent of the Naven ritual described by Bateson (1958) for the closely
related Iatmul. The ritual involved cross-dressing and mock Social Gender switch
(we have no record of how this used to correlate with the use of Linguistic Gender;
the ritual is hardly performed at present). 6
The female version of Naven was described by Hauser-Schäublin (1995: 49, 1977:
83–95). Naven used to be performed whenever a male or a female child would
undertake a gender-specific task for the first time. During the all-women Naven
ceremony, women would dress as men, and imitate aggressive and proud male
behaviour. As Hauser-Schäublin (1995: 49) puts it, ‘Naven rites celebrate . . . sociocultural
achievements’,and‘their transvestite aspects involve women structurally identifying
themselves with their husbands’. We know nothing about the use of Linguistic
Gender in such rituals. Their existence confirms the cultural appropriateness of mock

102 7 Manly women and womanly men
Social Gender shifts in order to emphasize the person’s Social Gender identity, and
patrilineal allegiance. Manambu joking relationships point towards a similar principle.
In Machiguenga, a Peruvian Arawak language, the non-masculine Linguistic
Gender covers females and inanimates. It can be used by men in jocular reference
to a ‘third man’, asa‘part of typical male joking behaviour in which men feminise
one another with sexual, and especially homosexual, comments and jokes’ (Shepard
1997: 53). In 7.2 the verb is marked for feminine gender, and there is a feminine
pronoun. This can be part of a funny story told by men about another man who is
either not present, or who is within earshot but not participating actively in the
conversation. A personal pronoun iro can be added.
7.2 tata-ka iro o-ogo-ge-t-ak-a
what-INDEF she 3fem-know/thinking-DISTRIBUTIVE-EPENTHETIC-PERFECTIVE-REALIS
‘What in the world is she thinking?’ (referring to something funny the man did)
Glenn Shepard (p.c.) comments that ‘the function of the feminine seems to be to
actively exclude them from the conversation and make them the “object” of the joke
or anecdote. They emphasize the “O” on each feminine pronominal marker, in fact
emphasis on the first syllable of every word in the phrase with a sort of downward
tone with vocal cry, the way women talk to young children to scold them.’ Jocular
switches of Linguistic Gender make people laugh, because they draw attention to
something out of the ordinary and essentially grotesque.
7.2 Offence and praise in Linguistic Gender reversals
In the day-to-day Manambu, changing the Linguistic Gender of humans means
trouble. The Linguistic Gender of inanimates or non-humans depends on their size
and shape (as we saw in §3.1). If a woman is referred to with masculine Linguistic
Gender, and a man with feminine Linguistic Gender, they are downgraded to the
level of inanimate things. In casual conversations (but hardly ever at open meetings),
a smallish fat woman-like man can be referred to with the feminine Linguistic
Gender. But kə-ø numa-ø du (this-fem.sg big-fem.sg man) ‘this fat round man’
(smallish) can only be said behind the man’s back: this is very rude and offensive. 7
A man who displays wrong social behaviour can also be referred to as ‘woman’—but
again, behind his back, not to cause trouble. Someone who stayed in his wife’s village
rather than taking her away, and thus breached the normal patrilocal practices, was
described to me as:
7.3 kə-ø ta:kw la:n-ad
this-fem.sg woman husband-3masc.SG.NOMINAL.PREDICATE
‘This (feminine) woman is (masculine) a husband’
Here, the masculine Linguistic Gender is used to mark agreement with a feminine
referent (instead of feminine Linguistic Gender). This mismatch makes speakers

7.2 Offence and praise in Linguistic Gender reversals 103
laugh—it emphasizes the grotesqueness of the culturally inappropriate situation of
Social Gender switch: a man shows patterns of social behaviour associated with a
woman. The man is a kind of ‘female husband’—in fact, someone who has failed to
live up to his Social Gender status. We can recall, from §3.4.2, that a man’s house, no
matter how small in size, cannot be referred to with feminine Linguistic Gender—this
is potentially demeaning.
A woman who is too boisterous, large in size, and also imposing and ‘too big for
her boots’ can be called kə-də numa-də ta:kw (this-masc.sg big-masc.sg woman) ‘this
(unusually) big, boisterous, or bossy woman’. This is derogatory, and is not something
one would say to that woman’s face. A woman is classified by her size, as if she
were downgraded to the status of an inanimate referent, reflecting behaviour
inappropriate for her Social Gender.
On another occasion, kə-də numa-də ta:kw (this-masc.sg big-masc.sg woman)
‘this (unusually) big, boisterous or bossy woman’ was said of a woman who sported
an unusually extensive knowledge of totemic names which is traditionally the
domain of a male. In terms of her Social Gender status, she behaved inappropriately;
hence negative connotations of applying the masculine Linguistic Gender to her.
Once again, classifying a human being by their size is demeaning: this implies
bringing a human down to the level of a mere object.
In a number of Berber languages from the Middle Atlas area of Morocco, referring
to a woman with a masculine Linguistic Gender form is equally offensive. In Aït
Mguild and Aït Wirra dialects, the form marked for feminine derivational Linguistic
Gender, tamt̍t̍ut̍t̍, means ‘woman’. The masculine form, amet̍t̍u, denotes a woman
with man-like manners and has derogatory overtones of French femme hommasse
‘overly masculine woman’. 8
In many proverbs from across the world assembled by Schipper (2003) a woman
who is taller than her partner, or who displays male-type behaviour, is looked down
upon. In Schipper’s (2003: 20) words,
an Ashanti proverb presents the problem as follows, ‘When a tall woman carries palm nuts,
birds eat them off her head’. A tall woman who proudly carries nuts in a bowl on her head is
presented as one who is showing off, and she is warned that she will be punished for it. The
explanation given is that such ‘male’ behaviour is condemned in a woman. The proverb is
warningly quoted to girls who display what society considers to be male traits, and the palm
nuts are a metaphor for ‘the male world’: as long as men are alive and around, a woman is not
supposed to crack nuts. The proverb reminds women to refrain from getting involved in
designated male roles.
The title of Schipper’s book, Never marry a woman with big feet,reflects that same
attitude: women with big man-like feet are physically and socially inappropriate.
An example of how treating a woman as if she were a man may have a demeaning
effect comes from one incident of the male initiation applied to a woman, in the Iatmul

104 7 Manly women and womanly men
village of Palimbei, from the same Sepik region as Manambu (Hauser-Schäublin 1995:
50–2). A girl who had inadvertently seen men blowing long flutes in a fenced-off
enclosure (something a woman is not allowed to see) was gang-raped, and then
subjected to scarification and a shortened version of male initiation. The initiation
was meant as a severe punishment for a ‘crime’ and stigmatization: the girl felt
degraded and shamed, despite the fact that she had gained what was considered
important ritual knowledge. It was as if she’d lost her appropriate Social Gender status.
The girl never recovered from that experience and soon died.
Changing Linguistic Gender for humans can have overtones of praise, and solidarity.
The masculine Linguistic Gender in Lokono Arawak has connotations of
something good for animates, inanimates, and also humans (as we can recall from
§3.4.1). A man from the speaker’s tribe would normally be referred to with masculine
Linguistic Gender. But if he is despised, the gender will switch to feminine. Males of a
different tribe are referred to with the feminine Linguistic Gender. If one is friends
with them, they become ‘masculine’ in their Linguistic Gender. 9
An Amharic-speaking man can address another man with a feminine pronoun ‘as
a term of insult, to belittle’ him (Wołk 2009: 131). Such ‘derogatory’ connotations
come about when an elderly person can address a male youngster as ‘feminine’. One
can refer to a male enemy with a feminine pronoun, to show one’s superiority. Hoben
(1976: 287) mentions how the second person singular masculine pronoun antɛ can
be replaced with the corresponding feminine pronoun anci, ‘if the speaker wishes to
insult a male’. This form was used ‘to refer to the rebel groups as a put down’, and
also to express social distance—‘in the context of an older man using the feminine to
address a younger boy’ (Pankhurst 1992: 169). The form anci can also be used ‘in a
humorously belittling sense for the smallest in a group of friends or for the clown of a
group’ (Hoben 1976: 286).
Addressing women as if they were men has an opposite effect. To use a masculine
pronoun addressing a woman in Amharic means praising her—the implication being
that a woman is acting ‘like a man’ and is thus ‘promoted’ to a manly status
(Pankhurst 1992: 169). In Figuig, a North Berber language from Morocco, tamət̍t̍ut
‘woman’ can be made masculine (by removing feminine derivational gender marking).
The resulting masculine form, amət̍t̍u, means ‘a courageous or a corpulent
woman’. This implies approbation (in contrast to the examples from other Berber
languages where the overtones of a ‘masculine woman’ are distinctly pejorative). 10
In a few South American languages, reversing Linguistic Gender is a mark of
higher Social Gender status. Jarawara is an Arawá language from southern Amazonia
(with two genders, masculine and feminine). A woman can be referred to with
masculine gender if she is particularly important in the society, or close to the
speaker. A narrator referred to his wife as ‘he’, as‘a mark of affinity and of respect’
towards her (Dixon 2004: 287).

7.3 Linguistic Gender reversal: endearment and solidarity 105
Among the Tariana and their Tucanoan neighbours from the Vaupés River
Basin in Brazilian Amazonia, women were the erstwhile owners of the magic
Yurupary flutes but lost them and cannot even look at them any more. Nowadays,
women are denied access to the magic powers associated with the flute; any word
that sounds similar to the flute’s nameusedtobea‘secret’ from women (we return
to the anti-women taboo register, and the status of women among the Tariana, in
Chapter 10). But a particularly respected and important woman is promoted to an
honorary ‘manhood’. A woman in Tariana is normally referred to with feminine
gender, on the verb, or the pronoun duha ‘she’ or a classifier for females (‐ma). The
Woman-Creator or a particularly powerful woman is consistently talked about as diha
‘he’ and requires a human classifier (-ite), just as any man would. So was a particularly
knowledgeable woman (the mother of my major consultants). A common English
expression man up! meaning ‘get your act together in such a way as a real man would’—
especially if applied to a woman—would be an uncanny analogy here. ‘Man’ in the
Tariana lore epitomizes strength and prowess, while women are to blame for many
mishaps, past and present. We turn to this again in §11.3. This is very different from
Linguistic Gender switches in Manambu which are derogatory in essence: they reflect
behaviour inappropriate for one’sSocialGender.
7.3 Linguistic Gender reversal: endearment and solidarity
Overtones of Linguistic Gender reversal may differ depending on the context. Using
a feminine pronoun to a man in colloquial Amharic does not always imply an insult.
Second person feminine pronouns are widely employed by men to address other
men, as a term of rapprochement, or endearment (Wołk 2009: 131–2, Pankhurst
1992). According to Hoben (1976: 287), in Addis Ababa, but not in the countryside,
male friends may address each other as anci ‘you singular feminine’, to express
affection. 11 This reminds us of how feminine Linguistic Gender assignment to
lower animates and inanimates may have overtones of endearment and diminution
(as we saw for Oromo, in §3.4.1). The masculine pronoun can also be used between
women as a mark of ‘attachment and closeness’ (Pankhurst 1992: 170).
In Arabic, a feminine noun, pronoun, adjective, or verb form can be used to a boy,
and a masculine one to a girl. So, one can say, in Arabic, wēn ruħti yā,ʼ binti ‘Where
did you go, little girl?’ to a boy, and inta žuʔān ‘Are you (masculine) hungry
(masculine)?’ to a girl. 12
In Russian, a woman can be affectionately addressed with a form marked for
masculine gender: the author, a woman, has often been addressed as moj horoshij ‘my
good one-masculine-singular’ or moj malenjkij ‘my little one-masculine singular’ by
older native speakers of the language (all of them women). A younger woman can
address a group of women as Rebjata ‘Guys!’ (using a masculine Linguistic Gender

106 7 Manly women and womanly men
form). This is a mark of camaraderie, similar to girl-only groups addressing each
other as Hey, guys! in the USA. 13
In Modern Hebrew, a man can address close female friends, relatives, associates,
and partners with masculine pronouns, and verbal and adjectival forms, as a sign of
affection, intimacy, and solidarity. Close female friends and relatives can do the same
to each other, with similar effect. Tobin (2001: 187–91) describes Linguistic Gender
reversal in a family of native Hebrew speakers, with two non-identical twin daughters
8½ years of age: one twin is ‘bigger’ and the other one is ‘smaller’. The smaller twin is
the one who most frequently uses the masculine Linguistic Gender forms to refer to
herself; others also use the masculine Linguistic Gender to talk to her. In one
example, the smaller twin talks about herself as if she were a boy, asking her
grandmother to take care of her rather than of her bigger sister:
7.4 ki hi gdolah ve-ani katan
because she big.fem.sg and-I small.masc.sg
‘Because she is big (feminine) and I am small (masculine)’
The twins were once sent to their room as a punishment. They both reversed
genders while talking to each other as a sign of solidarity, as they negotiated their
uncomfortable situation:
7.5a rotseh lesaxek?
want.masc.sg to.play
‘Ya wanna play?’
7.5b lo, aval tixtov mixtav le-aba ve-tivakesh slixa
no but write.IMPV.masc.sg letter to-daddy and-ask.IMPV.masc.sg sorry
‘No, but write a letter to Daddy and apologize’
The father was telling the smaller twin what she was like as a baby using feminine
forms. He switched to the masculine form when he tells her that she had to stay in the
hospital alone to become stronger before they could take her home:
7.6 ki hayita tsarix lehithazek ktsat
because you.were.masc.sg necessary.masc.sg to.become.strong little
‘Because you had to become a little stronger’
Masculine Linguistic Gender marks solidarity, intimacy, and protection. Feminine
Linguistic Gender is not used this way.
The neuter Linguistic Gender in Bulgarian includes inanimates. But it may have
overtones of joking endearment if applied to humans. In colloquial Bulgarian a neuter
gender form can be used to refer to a man or to a woman, marking condescension
or endearment. A young teacher (daskal, masculine) can be addressed as daskalče
(neuter) by an older and a wealthier man. A woman behaving in a silly way can be
addressed as prosto (simple+neuter gender) ‘dear silly one’. 14

7.3 Linguistic Gender reversal: endearment and solidarity 107
A somewhat different kind of Linguistic Gender reversal has been described for the
Iroquoian languages Mohawk, Oneida, and Onondaga. A woman can be referred to
with either the special feminine, or the neuter form. The feminine form is also used in
the meaning of ‘people in general, and when sex is not specified. The neuter form
covers animals.’ 15 The differences between the two ways of talking about a woman
are subtle. The choice of the prefix in Oneida correlates with what the woman is like:
a small and graceful woman will be referred to as ‘feminine’, and a large and
aggressive one would be treated as ‘neuter’. According to Bonvillain (1973: 86–7),
the feminine pronouns and agreement prefixes (ye-/ya-) in Mohawk are used for
‘female human beings who are regarded in some prestigeful or respectful way by the
speaker’, or‘for whom the speaker has the feeling of affection or closeness’. The
alternative, the ‘neuter’ (ka-/wa-), is used otherwise. In Bonvillain’s words, ‘it may or
may not reflect a momentary or general negative feeling on the part of the speaker
toward the female who is spoken about. Some speakers feel that . . . a girl or a woman
who is awkward or aggressive may more readily fall into this category, whereas one
who is graceful and quiet may tend to be classed in the other way.’ 16
The two ways of referring to women vary across Mohawk communities. Some
people employ the feminine-indefinite form as a mark of endearment, when talking
about ‘gentler’ speakers, and the ‘neuter’ form for ‘someone who is really rough’. The
feminine form ie-ksa'tí:io ‘she is pretty’ would be used for someone ‘a bit more
ladylike’, and the neuter form ka-ksa'tí:io ‘she is pretty’ would be preferred for someone
attractive the man would ‘like to go out with’ (but perhaps not to marry). Some
comment that a neuter form ‘would be used for someone you don’t respect: an animal
or a stranger’. And indeed, these forms are common for non-Mohawk women.
A neuter form may imply that the speaker does not like the woman. In general,
feminine-indefinite forms may signal endearment toward family members, and neuter
forms may signal familiarity among close friends. The feminine-indefinite forms mark
respect, but the neuter forms are the ones used to talk about people one does not know
well. The history, and reconstruction, of Northern Iroquoian languages and Mohawk
shows that the erstwhile indefinite pronoun and prefix on verbs came to be used to
refer to women as a sign of respect—reflected in some of the current usage. The
feminine-neuter is the result of reinterpretation of the third person category once the
new masculine was introduced. 17
Once again, a word of warning. In Mithun’s words (2014: 140), the feminineindefinite
category ‘did not emerge from a view of women as quintessential or
prototypical human beings’, nor did the ‘feminine-neuter’ category ‘develop from a
view of women as akin to animals’. Mithun (2014: 137) reports that ‘speakers are
usually surprised when it is brought to their attention. Some are horrified to realize
that they are using the same forms for women and animals. They report that they
never thought about it, and that “No one notices”.’ We return to the status of
‘women’ in Iroquoian languages in §7.7. 18

108 7 Manly women and womanly men
Reversal in self-reference may serve as marker of gender identity, for transgender
and transsexual people. We can recall, from §1.1, how the title of the autobiography
of Brigitte Martel, a transsexual male who became female, captured the change of her
Natural and Social Gender from male to female: Né homme, comment je suis devenue
femme—‘Born (masculine) a man, how I became (feminine) woman’. We return to
men’s and women’s speech and transgender practices in §9.3.
7.4 Men as women, women as men: a summary
of Linguistic Gender reversals
Changing Linguistic Gender of inanimates may highlight their physical features,
mark endearment or value. Changing Linguistic Gender of humans reflects changing
Social Gender relationships. When Linguistic Genders are reversed in jocular contexts
in Machiguenga and Manambu, the grotesque effect of it makes people laugh.
Reversing genders may result in mortal offence, or in ‘promoting’ a woman to a
higher manly status. Addressing a woman as if she were a man may signal affection
or solidarity. Addressing a man as if he were a woman may be a sign of endearment.
Or it may imply dragging him down to a ‘woman’s’ level. Linguistic Gender reversal
can have an opposite effect in the same language: addressing men as women in
Amharic may sound insulting in one context, and endearing in another (depending
on the relative age of people and their relationships). We can recall, from earlier in
this chapter, that women speakers of Palestinian Arabic refer to themselves as ‘men’
if there is something wrong with them—changing Linguistic Genders implies an
unusual and an uncomfortable state of affairs.
Linguistic Gender reversals highlight positive and negative associations with men
and women as social constructs. Here, Linguistic Gender reflects the stereotypes of
Social Gender entombed in the language. Tables 7.2 and 7.3 summarize examples
discussed so far.
Idiomatic expressions reflect attitudes to Social Genders and their comparative
value. English drama queen or old wife when mockingly applied to a non-manly man
have a similar pejorative and insulting effect as do gender switches in Manambu,
Amharic, and Lokono in Table 7.2. Saying ‘man up!’ to a woman urges a woman to
act in a manly way and to be as good as a man—similar to Lokono, Amharic, Figuig,
TABLE 7.2. Treating ‘men’ as ‘women’: masculine to feminine Linguistic Gender
reversal
Masculine to feminine reversal
pejorative and insulting
endearment and solidarity
Language examples discussed
Manambu, Amharic, Lokono
Amharic, Arabic, and Marathi baby talk

7.5 Attitudes to Social Genders through Linguistic Gender reversals 109
TABLE 7.3. Treating ‘women’ as ‘men’: feminine to masculine Linguistic Gender
reversal
Feminine to masculine reversal
pejorative: downgrading a woman as if she
were too bossy and ‘too big for her boots’
adding value: ‘promoting’ a woman to male
status
endearment and solidarity
Language examples discussed
Manambu; Aït Mguild, Aït Wirra (North
Berber)
Lokono, Amharic, Figuig, Jarawara, Tariana
Arabic and Marathi baby talk, Russian,
Modern Hebrew
Jarawara, and Tariana in Table 7.3. We now turn to further ways of expressing
attitudes to Social Genders through the use of Linguistic Genders.
7.5 Attitudes to Social Genders through Linguistic Gender reversals
Social Gender stereotypes associated with Natural Gender can be seen through the
use of pronouns marked for Linguistic Gender in English. Traditional grammarians
insist that the choice of gender-sensitive third person singular pronouns is based
on Natural Gender: ‘she’ for women, ‘he’ for men, and ‘it’ for inanimates. However,
in many varieties of English the situation is not that straightforward. ‘She’ can have
overtones of ‘smallness’ (in agreement with the principles we saw in Chapter 3): a bee
was referred to as ‘she’ as early as John Lily’s Euphues (1578). In the following
modern conversation, an ant was referred to as he—emphasizing the speaker’s
empathy with the insect who is referred to as if it was a human. 19
7.7 Speaker A: Ah there’s an ant
Speaker B: Well catch it
Speaker A: Well put him outside/ let him go on to—/ Look he’s on the toaster /
Now put him outside nicely .../he may be someone else’s ant you
know
In some varieties of Australian English, ‘she’ can be used to refer to unruly objects for
which one has antipathy (Wales 1996: 139):
7.8 She’s an absolute bastard, this truck.
In the English poetic tradition, ‘“masculine”-marked words were grouped according
to supposedly “manly” attributes (mostly positive) and “feminine”-marked words
according to “womanly” attributes (some positive, many negative): such as strong,
active, aggressive, powerful, clever, big fierce, giving (“manly”); versus weak, timid,
passive, loving, soft, helpful, beautiful, small, moral, receptive (“womanly”)’ (Wales

110 7 Manly women and womanly men
1996: 148). This constellation of semantic features of Natural and Social Genders as
reflected in Linguistic Gender usage can be conceived as a kind of ‘sexist symbolism’. 20
Mathiot (1979b) conducted a comprehensive study of sex roles of men and women
(that is, their Social Gender) as revealed through Linguistic Gender reference, to see
how the images of males and females can be reflected in the use of personal pronouns
in English. The use of the pronouns he and she observed with inanimate referents in
American English was found to correlate with a number of stereotyped features—
part of the inherent image and role image American men and women have of
themselves, and of each other.
The semantic opposition BEAUTIFUL versus UGLY manifests men’s conception of
women’s versus men’s appearance; and the semantic opposition manifesting men’s
conception of women’s and men’s achievement potential is INCOMPETENT versus
COMPETENT. Thus a beautiful flower is referred to as ‘she’, and an ugly cactus as
‘he’. 21 A car referred to as a prized possession is ‘she’. An unruly storm is also ‘she’;
and so is the troublesome refrigerator.
In contrast, the inherent image and role image American women have of themselves
and of men can be formulated in one semantic opposition: MATURE versus
INFANTILE (Mathiot 1979b: 25). A chair which has lost its leg is a ‘he’, and so is a
typewriter that would not work. Table 7.4 summarizes features that consistently
came about in men talking about women and about themselves, and women talking
about men and themselves.
English spoken in Tasmania (described by Pawley 2004) has certain rules governing
the choice of pronouns he and she with non-human referents. When the sex of a
higher animal is not known, the animal is referred to as ‘he’. He is used with plants
or parts of living plants, and any item of goods or portable property (other than
vehicles) that is viewed as trade goods rather than a personal possession. As Pawley
(2004: 116; 127) puts it, ‘using “she” for nouns referring to portable goods (not
vehicles) is a marker of “attachment”, or something one manipulates’. For instance, a
soccer ball has been kicked into a tree. Player, shaking the branch, says: ‘She’s stuck.
Come down, you bitch!’.
TABLE 7.4. Meanings of ‘she’ vs ‘he’ in American English (Mathiot 1979b) 22
Men’s conception of men and women
men
women
ugly
beautiful
competent incompetent
brave
a challenge to, or a reward for, men
good-natured men’s prized possessions
Women’s conception of
men and women
men women
infantile mature

7.5 Attitudes to Social Genders through Linguistic Gender reversals 111
The negative connotations of feminine Linguistic Gender surface in many derivational
forms belonging to feminine Linguistic Gender with feminine Natural Gender
reference. As Baron (1986: 114–15) puts it,
feminine English nouns tend to acquire negative connotations at a much faster rate than
masculine or neuter ones, creating semantic imbalances in originally parallel masculine/
feminine pairs like fox—vixen and governor—governess. Efforts on the part of feminists and
usage critics to eliminate feminine nouns like authoress in favour of unmarked equivalents on
the grounds that the marked terms are demeaning have been only partially successful.
Many feminine derivations in ‐ette in English have a negative feel about them; some,
like suffragette, were created as somewhat derogatory in the first place (as we recall
from §6.4.2). The word spin-ster originally meant ‘a woman (or, rarely, a man) who
spins’ (Oxford English Dictionary). In the seventeenth century, it came to be used as
‘the proper legal title of one still unmarried’ (presumably by association of women with
the needle-trade). In the eighteenth century, spinster acquired the pejorative meaning
it has today, of ‘a woman still unmarried; especially one beyond the usual age for
marriage, an old maid’. In their feminist dictionary, Kramarae and Treichler (1992:
429–30) further add to the description of spinster: ‘Like most terms connected with
women, it became a euphemism for mistress, or prostitute’,a‘figure of fun and ridicule
to those men who see themselves as essential to a woman’s existence’. 23
Deprecatory overtones of ‘femininity’ appear in further English expressions such
as female logic, old wives’ tale, old woman. In his letter to his friend James Hogg on 24
March 1814, Lord Byron referred depreciatingly to his fellow poets Coleridge and
Southey as ‘mere old wives’. 24 One can say He is a drama queen about a particularly
fussy man. He is an old woman was once used to refer to a now deceased linguist who
was incapable of fulfilling his editorial responsibilities and was worrying about every
minute detail of a task he was not up to. This is reminiscent of the deprecatory use of
feminine Linguistic Gender in Manambu (in §7.2), when talking about a man who is
not up to the standards of the Social Gender expectations in the society. Unlike
Manambu, English does not have a corresponding masculine equivalent: saying *She
is a drama king or *She is an old man is not idiomatic.
English is not unique in having negative overtones ascribed to derivational forms of
feminine Linguistic Gender. The general term for ‘dean’ in Russian is dekan (masculine
Linguistic Gender). Feminine derivations—dekanša, or dekanixa—have distinctly
pejorative overtones. The general term for postgraduate student is masculine aspirant.
Zemskaya, Kitajgorodskaja, and Rozanova (1993: 126) cite Saratov as a city where the
corresponding feminine form aspirantka is perceived as slighting (this overtone is
absent from the standard language). 25 The female version of poet ‘poet’, poètessa, did
not have slighting overtones in the nineteenth century. However, in the 1960s Anna
Akhmatova, one of the greatest Russian poets, told Yuri Aikhenvald that she should be
referred to as a ‘poet’, and not downgraded to a mere ‘poètessa’. 26

112 7 Manly women and womanly men
Many professions in Brazilian Portuguese have a generic masculine term and a
feminine counterpart, with no negative feelings about them. A male teacher (from
kindergarten to university) is professor, and the female one is professor-a, a male
judge is juiz, and a female judge is juiz-a; a male lawyer is advogado, and a female is
advogad-a. A male president is presidente. On the surface of it, the female form
president-a refers either to a president’s wife, or to a woman who ‘presides’. However,
in common usage presidenta ‘female president’ used to have strong overtones of a
bossy and merciless lady—as big for her boots as a masculine woman among the
Manambu. Things changed drastically when Dilma Rousseff was elected the first
female president of Brazil. A special law of the Brazilian government passed on
3 April 2012 ‘determined an obligatory usage of gender inflexion for professions
and degrees’, leaving no option but to use the term presidenta for the female
president. 27 This is one way in which Linguistic Gender reflects social changes and
can be sensitive to language planning: we return to changes in anaphoric and
derivational gender as targets in the struggle against ‘sexist’ language in §11.5.
Words of the masculine Linguistic Gender and masculine pronouns can have
generic senses and be used as a functionally unmarked option. This is what we turn
to now.
7.6 ‘Women’ as a subtype of ‘men’? The overtones
of masculine generics
Traditional practice in English was to employ the masculine Linguistic Gender
pronoun he or man when no Linguistic Gender specification was intended, e.g.
Man is mortal. The use of the masculine pronoun he in English for generic referents
and those whose Natural Gender is not known has been a controversial issue for
some time. Until recently, he in English was generally used as a term for human
reference and also as one subordinate term, for male reference (see the discussion in
Alpher 1987). The meanings of quite a few nouns in English show a bias: high-status
occupational terms such as lawyer, physician, orscientist are traditionally referred to
by the masculine pronoun he even in the contexts where Natural Gender (or sex) is
irrelevant or not known. In contrast, a nurse, a primary schoolteacher, or a secretary
would be more likely to be a woman, and referred to as she. The Social Gender
stereotypes are encroaching upon the use of Linguistic Gender.
During recent years, it has become the custom that the generic unmarked pronoun
‘they’ should be generally used, to avoid what was perceived as linguistic sexism. (We
return to the story of ‘he’ and ‘man’ in the context of language planning and social
change in Chapter 11.) A similar concern applies to the use of ‘man’ as a generic term
in modern English: Baron (1986: 137) offers a history of attempts to get rid of ‘man’
as a generic term seen as a reflection of Social Gender-based male dominance:

7.6 ‘Women’ as a subtype of ‘men’? 113
Perhaps most troublesome to linguists, feminists and usage critics alike has been the use of the
word man itself, in phrases such as the man in the street and compounds like mankind and
chairman. Some authorities argue that man, atfirst a word in which both genders were
combined, now refers primarily to males, while others claim that the neutral sense of man is
not dead and that the word still retains the primary meaning ‘human being’. In any case, one
must admit that in actual use it is often unclear whether man refers to people in general or to
men only.
Masculine forms used as generics may be seen as inherently sexist—for some, they
automatically reflect the assumption that the male of the species represents the
species itself. The male interpretation of referents is hard to avoid if we use the
generic masculine ‘he’ (as McConnell-Ginet 1979: 77 puts it). The generic ‘he’ and
‘man’ reduce the ‘visibility’ of women. Susan Sontag (1973: 186) offers a stronger
statement: ‘grammar, the ultimate arena of sexist brainwashing, conceals the very
existence of women—except in special situations.’ Many sources on English indirectly
confirm the view. Samuel Johnson, in his 1755 classic dictionary of English,
defines ‘man’ as a ‘human being, not a woman, not a boy’; a woman is ‘the female of
the human race’. The 1828 Webster dictionary defines ‘man’ as ‘mankind, the human
race, the whole species of human beings’, while woman is ‘the female of the human
race, grown to adult years’. 28
It may appear that the issue of the generic masculine pronoun does not arise in
languages with pervasive agreement gender. In Portuguese, Greek, Russian, Spanish,
French, Manambu, and many other languages a pronoun used to refer to a person, or
an entity, will be strictly chosen by its Linguistic Gender. A turnip in German (die
Rübe) is feminine, and will have to be referred to as ‘she’. A man (der Mensch) is
masculine, and a woman (die Frau) is feminine: they will be referred to as ‘he’ and
‘she’ respectively. We saw in §2.1.2 that a girl (das Mädchen) is neuter, and is often
referred to as es ‘it’; however, grown up girls may be talked about as sie ‘she’. But
the generic term for humans is either the same as ‘man’ (German Mensch, French
homme) or is masculine, in its Linguistic Gender (as in Russian čelovek). Vinogradov
(1947: 62) stressed the idea of a ‘general human being’ expressed through masculine
forms in Russian.
Experimental studies point towards sexist overtones of masculine generic in
German. 29 Masculine terms der Leser ‘reader’, der Student ‘student’, der Bewerber
‘applicant’ tend to be interpreted by speakers as predominantly referring to ‘man’.
This was shown in an experiment conducted by Braun, Sczesny, and Stahlberg
(2005). To avoid a choice between plural forms like Studenten ‘masculine students’
and Studentinnen ‘female students’, ‘feminized’ generics were introduced—similar in
form to feminine plurals but with the capital letter I, e.g. StudentInnen, subsuming
male and female students. Feminized generics with capital letter I (visible only in
written language) were perceived by speakers as being more inclusive of women than
male generics. Other studies yielded different results. Nissen (2013) shows that

114 7 Manly women and womanly men
generic masculine forms in Spanish did show a sexist bias in 1995, but not in 2005.
During the span of ten years, the visibility of women in social life has increased
significantly—and as a consequence, generic terms have grown to be more inclusive
of women. The interpretation of generics can be linked to recent social changes (we
return to this in Chapter 11). The generic masculine seen as an inherently ‘sexist’
reflection of Social Gender stereotypes makes Linguistic Gender reference a phenomenon
ripe for linguistic reforms. But do languages with a generic feminine ‘she’
and the feminine Linguistic Gender as an unmarked choice reflect a better world?
7.7 Markedness, status, and power in Linguistic Gender choice
Speakers—and linguists—of many European languages are acutely aware of a connection
between the usage of form with male Natural Gender reference, and male
dominance—viewed as ‘sexist’ language. The linguistic phenomenon—functional
unmarkedness of masculine Linguistic Gender in its many forms—is associated
with male dominance in the sphere of Social Gender: we can recall, from §2.3, that
the masculine, or the feminine, gender can be an unmarked choice. This is the case
in many Indo-European and Semitic languages—so much so that all speakers of
Modern Hebrew have no qualms about using second singular masculine pronoun to
refer to people in general (men and women—as we will see in §11.5.2). At the same
time, language reforms aimed at restoring the balance between the use of masculine
and feminine forms reflect the intertwined societal perception of Natural Gender,
Linguistic Gender, and Social Gender (we return to this in §11.5). In other words,
lack of markedness—or having the masculine form as a default choice—is viewed as
the epitome of Social Gender status, and power.
In a few languages of the world, the feminine Linguistic Gender is functionally
unmarked. That is, if I don’t know whether a man or a woman is coming, I will use
feminine Linguistic Gender to refer to the ‘unknown’. Feminine Linguistic Gender is
the unmarked choice in Manambu anaphoric reference and agreement (as we saw in
§2.3). But in the lexicon things are different. The noun du ‘man’ is used as a generic
term for ‘human beings’. Some of the spirits of the jungle are human-like; and so are
the spirits of the deceased people. When telling stories about them, speakers stress that
they are humans, by saying du-adi (man-3plural), ‘they are people’. The expression
du kui (man meat) ‘human flesh’—which reflects earlier cannibalistic practices—is
another instance of a ‘generic’ man. The noun du ‘man’ can also mean ‘human body’.
The word for woman is not used this way. 30 We can recall that valuable and important
objects belong to the masculine Linguistic Gender—which is congruent with the
Manambu focus on male-oriented cults (§3.4.2).
In Jarawara, masculine is the functionally marked gender (see Dixon 2004: 186–7,
for discussion). It is used for reference to human males (and other referents assigned
to masculine Linguistic Gender). The feminine Linguistic Gender is the default

7.7 Markedness, status, and power 115
choice (as we can recall from §2.3). All personal pronouns require feminine agreement
forms—R. M. W. Dixon reports that he found it rather difficult to refer to
himself ‘as if ’ he were a woman. The functionally unmarked choice—feminine, or
‘non-masculine’—is used for reference to human females (and referents belonging to
the feminine Linguistic Gender), and also when there is no gender specification. But
this does not imply that women are higher in status than men: as we recall from §7.2,
an important woman can be referred to with masculine gender—as if she were
promoted to the status of an honorary ‘male’. There is no reason to believe that
having feminine Linguistic Gender as a default choice automatically accords them a
privileged status in the society.
In Mohawk and other North Iroquoian languages, the same Linguistic Gender
category and forms are used for women and generic reference to humans (as we saw
in §7.3). As Chafe (1967: 13–16) puts it, for Seneca:
the masculine [third person subject prefix] morpheme . . . denotes a male human being. The
feminine morpheme . . . either denotes a female human being or is an indefinite reference to
people in general, translatable as people, they, one. The neuter morpheme . . . denotes either an
animal or something inanimate.
How does the ‘distinctiveness’ of the masculine gender forms, as the functionally
and formally marked choice in Northern Iroquoian languages, correlate with the
cultural patterns and the Social Gender status of women? 31 Sources on traditional
practices in Northern Iroquoian societies (including the Huron, the Seneca, and also
the Onondaga) suggest that
sex roles were distributed in Iroquois society in such a way that men were conspicuous, often
even flamboyant, and invested with decision-making powers, whereas women stayed in the
background, a position from which they nevertheless exerted considerable influence on what
men did. Women were neither unimportant nor undervalued. On the contrary, they were
responsible for keeping life going, both from day to day and from generation to generation.
The importance of women in Iroquoian culture has been emphasized by the anthropologist
Cara Richards, who went so far as to exclaim, ‘If you must be born a woman, try to be an
Onondaga’.
According to Richards (1974: 401), the ‘relatively high status of Iroquois women’
was reflected in the matrilineal descent, and also the fact that land belonged to women,
and women were the ones who appointed the chiefs. The men ‘stood out as highly
visible figures against this essentially female background’. This matches the principles
of Linguistic Gender marking across Northern Iroquoian languages—with their special
marking for masculine gender and ‘its association of women with undifferentiated
people in general’ (Chafe 2004: 106). This is echoed by Mithun (2014: 137):
The use of the same forms for female persons, generics, and indefinites suggests a link between
grammar and culture. Iroquoian culture is characterized by longstanding matrilineal and

116 7 Manly women and womanly men
matrilocal traditions. Clan membership is inherited through the mother. Land was traditionally
passed down from mothers to daughters. Women have always been highly respected, and they
serve important community roles. Chiefs were traditionally men, but they were appointed (and
potentially recalled) by the clan mothers. It is perhaps no surprise that generic persons should be
represented by Feminine gender forms.
But in actual fact, ‘the associations between language and culture suggested by use
of one category of women and generics on the one hand, and a separate category for
some women and animals on the other, are not as direct as might be assumed’
(Mithun 2014: 140; see also note 16 in Chapter 6 for their history). That the same
gender category happens to be used ‘for women and generic human beings’ is the
result of historical development, from the use ‘of original indefinite category to
certain women as a sign of respect’ (p. 159).
A challenge to a straightforward correlation between Linguistic Gender markedness
and women’s societal status comes from further quarters. Social Gender roles in
many indigenous societies in North America can be described in terms similar to
Northern Iroquoian, with men having high ‘visibility’. However, far from all of them
have developed feminine as the unmarked Linguistic Gender (see Chafe 2004:
106–7). A culture pattern may at some point in time have motivated the linguistic
pattern; however, the exact causal relationship is hard to argue.
Correlations between functional markedness of Linguistic Gender and Social
Gender roles can be interpreted in two, mutually exclusive ways:
1. Having masculine gender forms as a functionally ‘unmarked’ category can be
understood as a token of ‘male dominance’ (and fought against). This is often
assumed to be the case in Modern English, and is reflected in the contemporary
debate against the use of terms with masculine referents as generic ones—see
§11.5.
2. Alternatively, having masculine gender as a ‘special’, marked one may be
understood as a token of special importance and particular ‘visibility’ of
males in cultural practices. This appears to be the case in Jarawara, and in the
Northern Iroquoian examples.
As Mithun (2014: 159) puts it, there is no doubt that ‘relations do exist between
grammatical gender categories and culture’. But they are rarely straightforward and
simplistic. Cultural norms and the Social Gender status cannot be directly correlated
with, or inferred from, the linguistic form and markedness relations in Linguistic
Gender. However, the choice of Linguistic Gender may reflect stereotypes and
expectations associated with Social Gender and with Natural Gender. This is especially
salient for humans with their defined social roles, and particularly so in
languages whose speakers are aware of the meanings of genders.

7.7 Markedness, status, and power 117
NOTES AND SOURCES
1. Based on my notes 2004. Laycock (1965: 133) wrote down the first person forms correctly;
he erroneously interpreted second feminine form as having third feminine reference.
2. Rosenhouse and Dbayyat (2006: 174–5); see there for further details and additional
explanation of the gender switch.
3. Hoben (1976: 287–8) tells a similar story: ‘Once in a while in the countryside, parents call a
son anci (you feminine) or a daughter antɛ (you masculine—A.Y.A.). The usual explanation
is that earlier children of the couple died in infancy. The pronoun switch is made in
an effort to conceal the true identity of this child from the evil forces that attacked the
other babies. This is not a case of expressive variation: it is rather one of deception.’ See
also Repp (1996: 49), for comparable practices in Arabic.
4. See Sa'ar (2007: 424–5); the same principle applies in Palestinian Arabic.
5. See Aikhenvald (2008: 618–23), on loss of cultural knowledge among the Manambu.
6. Cf. also Houseman and Severi (1998). The Manambu used to have a reduced version of
Naven (Harrison 1990), now all but forgotten.
7. We can recall, from §4.1, thatinDyirbalyara ‘man’ can be used with the feminine class
marker, instead of masculine, to point out the female characteristics of a hermaphrodite
(see Dixon 1972: 306–12; 1982: 178–83). I am not aware of such examples in Manambu:
there do not appear to be any hermaphrodites, nor homosexuals, in any of the
settlements.
8. See Taïfi (1991: 53–4), Oussikoum (2013: 820–1). I am grateful to Maarten Kossmann for
these references, and his comments on the meanings of derivational gender across Berber
languages.
9. See Pet (1987: 26–7). We do not know enough about the traditional Lokono society to
establish any association with the Social Gender patterns.
10. See Benamara (2013: 411). I am grateful to Maarten Kossmann for making me aware of
this, and other references on Berber.
11. Using a feminine form in Amharic may also have an opposite function: men may use the
feminine address form towards men whom they consider high in status (Wołk 2009:
131–2).
12. See Ferguson (1964: 106, 109) on affectionate Linguistic Gender reversals in Arabic and
Marathi. Similar instances of gender switch in Egyptian Arabic are described by Woidich
(2005).
13. See Yokoyama (1999: 423, 427). Based on Zemskaya, Kitajgorodskaja, and Rozanova
(1993), Yokoyama (1999: 423) reports that masculine forms used by both sexes to refer
to women carry affectionate connotations. Diminutives of female names may take masculine
endings in both Russian and Polish (Kasia Wojtylak, p.c.), with specially endearing
overtones. So, my Russian friend Masha can be lovingly addressed as Mashik (Masha
+diminutive.masculine.singular); the resulting form may trigger masculine rather than
feminine agreement. Both Milyj Mashik (dear+masc.sg Mashik) and Milaya Mashik (dear
+fem.sg Mashik) are possible (see also Doleschal and Schmid 2001: 265, and references
there). Rothstein (1973: 464) comments on ‘the expressive use of nouns of feminine
gender, especially hostile epithets, to apply to male human beings. Thus, it is more effective
to call a man dura “idiot”, the form with feminine gender, than durak (a corresponding
form of masculine gender)’ (a similar example is in Doleschal and Schmid 2001: 265). This

118 7 Manly women and womanly men
example is attributed to Roman Jakobson and is ‘often cited in the oral tradition of
Jakobson’s students’ (Yokoyama 1999: 423–7).
14. See Mladenova (2001: 37–9) on what she refers to as ‘value-laden neuters’ in Bulgarian and
also Greek, and their use in nineteenth-century literature and modern languages.
15. Iroquoianists call it ‘Feminine-zoic’: Mithun (1999: 100; 2014: 137).
16. Chafe (1977: 514) adds to that: ‘the situation in Onondaga is similar, though speakers of
the language have emphasized to me that ka-/wa- is not derogatory, but only detached and
impersonal’.
17. See Chafe (1977), Mithun (2014: 138, 141–3), and §4.1. A somewhat similar example of
Linguistic Gender reversal as a mark of politeness was described by Khaidakov (1963:
49–50) for Lak.
18. Changing genders may correlate with the participant’s agency: we can recall, from §4.1,
how an inanimate Comb is treated as ‘animate’ as soon as it starts speaking and acting on
its own, in stories in Fox and other Algonquian languages (see Goddard 2002). Polish has
three genders (masculine, feminine, and neuter). Masculine nouns further divide into
animate/inanimate (in the singular) and personal/non-personal (in plural). Non-personal
forms of personal nouns can be used for a pejorative effect, as a way of downgrading a
human to a non-human status; occasionally they may have positive overtones (see
Rothstein 1973: 697).
19. See Wales (1996: 143, 146–52); the example has been shortened. Similarly, speakers of
American English may use he to refer to animals they have empathy for (Morris 1999: 188).
20. See Baron (1986: 94); also see Yaguello (1979), for a similar approach to French.
Svartengren (1927 and 1954: 263–83) provides numerous examples, in American English,
of ‘she’ applied to inanimates in English (including tools, hollow objects, bridges, and
man-made tools and machinery, clothing, small objects, furniture, water, seasons, and a
few more).
21. See Mathiot (1979b: 18–19). See also Clamons (1993: 277) and §3.1 on Oromo.
22. See Mathiot (1979b: 12–29).
23. See Baron (1986: 117–20) on the history of the word; Dixon (2014: 323–4) on the history
of the suffix -ster and its negative overtones in other formations, such as gang-ster and
mob-ster.
24. See p. 163 of Byron’s letters and journals, ed. Richard Lansdown (Oxford: Oxford University
Press, 2015).
25. This is contrary to Connors’s(1971: 598) claim that the terms for female occupations have
no derogatory or facetious connotations, thanks to the egalitarian structure of Soviet
society and equal rights for men and women. See Yokoyama (1999: 420, 427), Zemskaya,
Kitajgorodskaja, and Rozanova (1993).
26. Yuri A. Aikhenvald, p.c.
27. Law N o 12.605, 3 April 2012: . This law reinforces a similar Federal law, of 2 April 1956 (Law 2.749
proposed by the Senator Mozart Lago and sanctioned by Juscelino Kubitschek, the then
president of Brazil), to the same effect. The term presidenta is morphologically unusual in
the language. The majority of agentive nouns in -ente, including agente ‘agent’, cliente
‘client’, and paciente ‘patent’, in Portuguese do not have a special feminine form (*clienta,
*agenta, *pacienta are ungrammatical). Just a few do: pairs such as sargente ‘sergeant’ and
sargenta ‘woman sargent; wife of a sergeant’ and parente ‘male relative, relative in general’
and parenta ‘female relative’ are in general use. The ‘legalization’ of the term presidenta—
whose referent is for now uniquely identifiable—may result in the expansion of productivity

7.7 Markedness, status, and power 119
of the feminine -enta (see, for instance, the discussion at ).
28. Sexism in English has been discussed at length: a useful summary is in Mills (2008: 145–7);
see also Hellinger and Pauwels (2007), McConnell-Ginet (2003), Eckert and McConnell-
Ginet (2003, 2006), and Coates (1993).
29. See Brauer and Landry (2008), for similar results in French.
30. See Aikhenvald (2015a) on subtle differences in grammatical behaviour between du ‘man,
human being’ and du ‘human body’.
31. See Chafe (2004: 99–104) for further discussion.

8
The images of gender
The meanings of Linguistic Genders mirror myths and beliefs. Linguistic Genders are
a rich source of poetic metaphors. What physical and other properties do Linguistic
Genders correlate with? What role do Linguistic Genders play in human cognition,
and in the creation of mental images and representations? Do Linguistic Genders
affect the ways in which speakers perceive and construe the world around them?
These are the questions we address in this chapter.
8.1 Myth-and-belief in the choice of Linguistic Gender
As Dixon (2015: 29) puts it, ‘the tribes’ repertoire of legends and beliefs, and their
general perception of the world also help explain the ways in which genders are
chosen’. Linguistic Gender of an object or an animal can correlate with its role in
myth. We can recall, from §3.1, that ‘moon’ in Manambu is a mythical woman, and is
always assigned to the feminine gender (no matter what its size or brightness). In
Alamblak, yam-t ‘moon’ is feminine (and bears the feminine suffix -t). In myths, the
moon is a woman, and the sun (mar-r) is her son. The ‘sun’ is masculine and bears
the masculine suffix -r (despite its round shape typical for feminines; Bruce 1984: 97).
Similarly, in Tunica, thunder and lightning were personified as mythical men, and
were masculine for this reason. 1 In Dyirbal, the sun was believed to be a woman, and
thus belongs to gender 2 which includes females (balan garri). The moon is her
husband, and so the noun belongs to gender 1 which includes males (bayi gagara).
Thunder, lightning, and rain are believed to be legendary males, and are also included
in gender 1.
In many Australian languages, the moon is conceived of as a man and the sun
as a woman. 2 In many languages in New Guinea, the moon is believed to be a
mythological man who engages in sexual intercourse with women. The moon is
responsible for women’s menstruation: it is said to inflict ‘moon sickness’ upon
women. Coincidentally, in other societies, e.g. among the Vaupés Indians of northwest
Amazonia, Moon is also held responsible for female menstruation. But, since
Moon is a mythical male (referred to with non-feminine Linguistic Gender), it is said
to have sexual intercourse with women when they menstruate (Aikhenvald 2012b,
How Gender Shapes the World. First edition. Alexandra Y. Aikhenvald.
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.

8.2 The metaphors of Linguistic Gender 121
and my own fieldwork). In Manambu and many other languages of the Sepik area,
‘moon’ and all stars belong to the feminine gender (we recall, from §3.1, that
feminine gender is associated with round shape, which is also the shape of the
moon). An alternative way of referring to the moon is bapa-ta:kw (moon-woman),
literally ‘The Lady moon’. 3
In Asheninca and Ashaninka Campa (Arawak languages from Peru), all inanimates
belong to the feminine gender. All animates are treated as masculine. But the
Moon, cashiri, is masculine: this is a mythical man (and so is his son, the Sun, who is
also a male). 4 Similarly, in Palikur, heavenly bodies (sun, moon, stars, planets),
thunder, and lightning belong to masculine gender, because according to traditional
legends they were once men.
Linguistic Gender assignment of birds can also be determined by their role in
myth. The traditional Dyirbal people believed that birds are, as a class, the spirits of
dead women. Birds are classed as members of gender 2 (balan) which includes
females. One snake, balan bima ‘death adder’, belongs to gender 2, as it is also a
legendary woman. (More examples were given in §2.2.1.)
In numerous languages of the Sepik region of New Guinea, the cassowary (a large
non-flying bird) is feminine. In many Sepik cultures, including Manambu, the
cassowary is a totemic woman. In an important Abau creation story, a cassowary
gives birth to a human child, and is thus an ‘honorary female’. Along similar lines,
‘cassowary’ in Dyirbal is feminine: it appears in the shape of a woman in myths
(R. M. W. Dixon, p.c.).
Conversely, the Linguistic Gender of a noun may be projected onto beliefs,
superstitions, and gender imagery.
8.2 The metaphors of Linguistic Gender
Linguistic Genders influence beliefs, superstitions, and poetic metaphors revolved
around an inanimate referent, which is then promoted to the status of an ‘honorary’
human, or ‘anthropomorphized’. Features of Natural Gender (male and female) are
then projected onto the items in accordance with their Linguistic Genders, endowing
them with features typical of Social Genders.
As Jakobson (1959: 117) pointed out, grammatical gender
plays a great role in the mythological attitudes of a speech community. In Russian the feminine
cannot designate a male person, nor the masculine specify a female. Ways of personifying
or metaphorically interpreting inanimate nouns are prompted by their gender. A test in the
Moscow Psychological Institute (1915) showed that Russians, prone to personify the weekdays,
consistently represented Monday, Tuesday, and Thursday as males and Wednesday, Friday,
and Saturday as females, without realizing that this distribution was due to the masculine
gender of the first three names as against the feminine gender of the others. ...

122 8 The images of gender
The overtones of Natural Gender and associated meanings of Social Gender are
often imputed to non-human referents depending on their Linguistic Gender, as a
kind of ‘gender metaphor’. As McConnell-Ginet (1979: 77) puts it, ‘poets and
painters convey messages by incarnating abstract concepts’. In both Russian and
Portuguese, the noun ‘death’ belongs to the feminine Linguistic Gender. Jakobson
continues:
The Russian painter Repin was baffled as to why Sin had been depicted as a woman by German
artists: he did not realize that ‘Sin’ is feminine in German (die Sünde), but masculine in
Russian (greh). Likewise a Russian child, while reading a translation of German tales, was
astounded to find that Death, obviously a woman (Russian smert', fem.), was pictured as an old
man (German der Tod, masc.).
In Portuguese, as in Russian, morte ‘death’ belongs to the feminine Linguistic
Gender and is often respectfully referred to as Dona Morte ‘Ms Death’. The identification
of śmierć ‘death’ with ‘woman’ in Polish culture and tradition goes back a
long way: a fifteenth-century Polish poem, Rozmowa Mistrza Polikarpa ze śmercią
‘Master Polycarpus’ Colloquy with death’, explicitly states that Master Polycarpus
‘saw a naked person of the female sex’ (‘uzrzał człowieka nagiego, przyrodzenia
niewieściego’) (Studzińska 2012). In his 1894 picture, the German painter Franz
von Stuck depicted the gruesome war as a man—in German der Krieg ‘war’ is
masculine. In Polish wojna ‘war’ belongs to the feminine Linguistic Gender. The
cycle of pictures ‘War’ (Wojna, feminine Linguistic Gender), by the Polish painter
Arthur Grottger, contains female images—so much so that Baudouin de Courtenay
described them as representing the war as a ‘woman’ (de Courtenay 1929: 246). None
of this makes much sense to an English speaker, and even less so to a speaker of
Estonian or Hungarian, with no Linguistic Gender.
Linguistic Gender of inanimate objects can be deployed in popular beliefs. As a child
in the 1960s I used to spend summer holidays with my grandmother in a rather remote
place, away from our native Moscow. Russian has three genders: masculine ‘he’,
feminine ‘she’, and neuter ‘it’. Every noun is assigned to a gender. If I drop a spoon
or a fork, I will be expecting a female visitor: lozhka ‘spoon’ and vilka ‘fork’ are
feminine. And if I drop a knife, this implies a male visitor: nozh ‘knife’ is masculine.
We had no telephone, and the post office was miles away. The belief seemed to work
quite well—and was of some help to us as to whom we might expect for lunch that day.
In Portuguese of the north-east of Brazil, if one drops a spoon colher (feminine), a
woman (mulher, a word which rhymes with colher and is thus reminiscent of it) will
come; and if one drops a fork garfo (‘masculine’), one expects a man (Eduardo
Ribeiro, p.c.). It goes without saying that such correlations and superstitions are hard
to explain to speakers of other languages. In German, spoon is masculine (der Löffel),
fork is feminine (die Gabel), and knife is neuter (das Messer)—so, when I tell my
German friends about the Russian superstition, they just shrug their shoulders at

8.2 The metaphors of Linguistic Gender 123
what looks to them like a ‘barbarian’ custom. For those who speak Estonian—a
language with no genders, and just one word for ‘he’, ‘she’, and ‘it’—this custom
makes even less sense.
Gender metaphors and personification make the task of a translator rather difficult.
Again, in Jakobson’s words, My sister Life, the title of a book of poems by Boris
Pasternak, is quite natural in Russian, where life is feminine (zhiznj), but was enough
to reduce to despair the Czech poet Josef Hora in his attempt to translate these
poems, since in Czech the noun zivot ‘life’ is masculine. 5
Linguistic Gender choice may be deployed as a poetic device. Then, the Linguistic
Gender, and the imputed Natural and Social Gender, of the referent form one
conceptual whole. We can recall, from Chapter 2, that the assignment of Linguistic
Gender to German nouns is subject to a set of fairly complex morphological and
semantic principles (Zubin and Köpcke 1986). Male and female adults of each species
of domestic and game animals belong to masculine and feminine Linguistic Genders.
Types of cloth, precipitation, and minerals belong to masculine Linguistic Gender. In
many instances, a generic term belongs to one gender, and specific terms to another.
The generic term for ‘tree’, Baum, is masculine (and so are all tree names containing
-baum as their part). Specific names for trees (which do not contain -baum) are
feminine, e.g. Palme ‘palm tree’, Linde ‘linden’, Weide ‘willow’, Pappel ‘poplar’. This
linguistic fact has been deployed in poetry, with trees being personified through
interpreting their Linguistic Gender as their Natural Gender, and then endowing
them with feelings and behaviour typical for Social males and Social females.
The famous poem by Heinrich Heine Fichtenbaum ‘Pine-tree’ reads as follows
(Linguistic Gender-sensitive forms are in bold type, and glossed in brackets after each
word):
8.1 Ein Fichtenbaum (one:masculine pine-tree) steht einsam
Im hohen Norden auf kahler Höh.
Ihn (to him) schläfert, mit weisser Decke
Umhüllen ihn (him) Eis und Schnee.
Er (he) träumt von einer Palme (one:feminine palm tree),
Die (which:feminine), fern im Morgenland,
Einsam und schweigend trauert
Auf brennender Felsenwand.
The noun ‘pine-tree’ belongs to the masculine Linguistic Gender, and the word for
‘palm-tree’ belongs to the feminine Linguistic Gender. There are obvious romantic
connotations to these Linguistic Genders—a lonely pine-tree-man is longing for a
lonely and silent palm-tree-woman. This play is completely lost in the English
translation, in 8.2: English has no means for the same Linguistic Gender play. Both
trees (in caps) are referred to as ‘it’ (in a well-known translation by James Thompson
(1834–88)):

124 8 The images of gender
8.2 A PINE-TREE standeth lonely
In the North on an upland bare.
It standeth whitely shrouded
With snow, and sleepeth there.
It dreameth of a PALM TREE
Which far in the East alone,
In mournful silence standeth
On its ridge of burning stone.
The gender-play effect is equally lost in translations of this same poem into
other languages. In Russian, sosna ‘pine-tree’ belongs to the feminine gender, and
so does paljma ‘palm-tree’. The romantic connotations of the original poem are
lost in its most famous Russian translation, by Mikhail Lermontov. The French
translation of the same poem, by Amélie Ernst (1873), involves le sapin ‘fir-tree’
and le palmier ‘palm-tree’. BothbelongtothemasculineLinguisticgender.The
feeling of loneliness and mutual yearning is there—but the subtle romantic innuendo
is lost.
Not that it cannot be captured. Emma Lazarus (1849–87), an American poet,
translated Heine’s poem into English, reflecting the German Linguistic Genders
through English anaphoric genders (in bold):
8.3 There stands a lonely pine-tree
In the north, on a barren height;
He sleeps while the ice and snow flakes
Swathe him in folds of white.
He dreameth of a palm-tree
Far in the sunrise land,
Lonely and silent longing
On her burning bank of sand.
As Guy Deutscher (2010: 196) puts it, ‘the price Lazarus pays for this faithfulness is
that her translation sounds somewhat arch, or at least artificially poetic, since in
English it is not natural to speak of trees in this way’. In his version of the poem,
Fyodor Tjutschev kept the romantic overtones by using kedr ‘cedar’, of masculine
Linguistic Gender in Russian, instead of the feminine sosna, ‘pine-tree’. 6
Die Lotusblume, the main character of Heinrich Heine’s poem ‘Lotus flower’,
belongs to the feminine Linguistic Gender in German. Its lover is the masculine
der Mond (the moon), in 8.4a. The forms relevant for Linguistic Genders are in bold.
8.4a Die Lotusblume ängstigt
Sich vor der Sonne Pracht,
Und mit gesenktem Haupte
Erwartet sie träumend die Nacht.

8.2 The metaphors of Linguistic Gender 125
Der Mond, der ist ihr Buhle,
Er weckt sie mit seinem Licht,
Und ihm entschleiert sie freundlich
Ihr frommes Blumengesicht.
Sie blüht und glüht und leuchtet,
Und starret stumm in die Höh;
Sie duftet und weinet und zittert
Vor Liebe und Liebesweh.
The English translation in 8.4b—by Emma Lazarus (1867)—reflects the loving
heterosexual relationship between the she-flower and the he-moon.
8.4b The Lotus-flower trembles
Before the sun’s gold light;
And, with her head low drooping,
Waits, dreamily, the night.
The Moon, he is her lover,
He wakes her with his light,
And unto him reveals she
Her flower-face so bright.
She blooms and glows and brightens,
And dumbly looks above;
She weeps and sighs and trembles
With love and the woes of love. 7
In his translation of this poem, the Russian poet Apollon Maikov had to replace lotus
with lily (since lotus belongs to the masculine Linguistic Gender, and lily to the
feminine Linguistic Gender). Instead of luna ‘moon’ (feminine Linguistic Gender in
Russian) he used mesjac ‘young moon, month’, of masculine Linguistic Gender.
A popular Russian song features a thin and vulnerable rowan tree (rjabina,
feminine Linguistic Gender), in her loneliness yearning for the company of a big
strong oak tree (dub, masculine Linguistic Gender) across the river. 8 As Fodor (1959:
206) puts it, the masculine gender of dub ‘determines the content as well as the mood
of the poem, and constitutes a stylistic device, of which genderless languages would
be incapable, the same thought and feeling having to be rendered by different means
or in roundabout ways’. 9
The Linguistic Genders offer access to a fixed range of metaphors: the masculine
Linguistic Gender evokes a man, and the feminine gender evokes a woman. In a
famous song by Bulat Okudzhava, the author requests forgiveness from his three
protectresses—sisters and wives at the same time—Faith, Hope, and Love. The terms
(Vera, Nadezhda, Ljubovj, which also happen to be used as female names) belong to
the feminine Linguistic Gender. The language imposes the female imagery on the
poet—and he makes use of this. 10

126 8 The images of gender
Morton Bloomfield (1963: 162) commented on a comparative freedom of gender
personification in English, with just anaphoric gender, as compared to other languages:
The subject is particularly interesting in English, because a writer of English after about 1300
[presumably, after the loss of agreement gender], unlike his German or French counterparts,
had more or less freedom in choosing masculine or feminine gender for his personified figures.
The Germans seem to have been particularly fascinated with this freedom of choice denied
them. The gender of their personifications, and even the very possibility of personification, was
determined by forces outside their control. One might say that languages with grammatical
gender, unlike English, have automatically built-in personification of some sort.
Do these ‘automatically built-in’ associations affect the ways in which speakers
perceive, and categorize, the world of inanimates?
8.3 Does Linguistic Gender affect cognition?
Connections between the Linguistic Gender of a noun and a female—or a male—
image behind it sparked the interest of psychologists. A few have attempted to test
how pervasive the associations are. Konishi (1993) conducted an experiment with
two dozen or so German and Spanish speakers who were asked to rate a number of
inanimate nouns for ‘potency’ (among other features)—a feature associated with
‘masculinity’, or male Natural Gender. Half the nouns were of masculine Linguistic
Gender, and half were of feminine Linguistic Gender. A number of nouns used in the
experiment have different genders in the two languages. For instance, die Brücke
‘bridge’ and die Luft ‘air’ are feminine in German, and their Spanish counterparts el
puente and el aire are masculine. Der Apfel ‘apple’ in German is masculine, and
Spanish la manzana is feminine. Nouns which are masculine in German (such as
chairs and keys) were rated as more potent by German speakers. Nouns which are
masculine in Spanish (such as bridges and clocks) were rated as more ‘potent’.
Sera, Berge, and del Castillo (1994) asked a group of Spanish and French children
to rate pictures of objects as ‘masculine’ or ‘feminine’. Speakers consistently classified
objects in agreement with their Linguistic Gender: masculine as men, feminine as
women. (The effect was even stronger when pictures were accompanied by Spanish
and French labels.) When asked to attribute a man’s or a woman’s voice to each
picture, Spanish-speaking children assigned voices in accordance with the Linguistic
Gender of each item. So when French speakers saw a picture of a fork, most of them
attributed a woman’s voice to it: la fourchette in French is feminine. In contrast,
Spanish speakers tended to attribute a male voice to the Spanish el tenedor ‘fork’, of
masculine Linguistic Gender.
To try and determine whether Linguistic Gender influences speakers’ mental
representation of an object, Boroditsky, Schmidt, and Phillips (2003) developed a
memory game for a group of German and Spanish speakers (the game was conducted

8.4 What men and women look like 127
in English, to avoid explicit mention of overt Linguistic Genders in these two languages).
Speakers were taught proper names for twenty-four objects; for instance, an
apple would have been called Patrick, or Patricia). Names for all the objects had
opposite Linguistic Gender in Spanish and in German. The participants were given a
fixed time to remember the names associated with objects, and then tested on the
results. The results were counted, and there appeared to be a strong bias towards a
correlation between the name (male or female) and the Linguistic Gender of the
noun: participants were better at remembering the male name associated with the
object of masculine Linguistic Gender, and the female name of an object of
feminine Linguistic Gender. That is, German speakers easily remembered the
name Patrick for an apple (der Apfel, masculine), and Spanish speakers remembered
the apple’s namePatricia (la manzana, feminine). Spanish speakers found it
easier to remember the name of a bridge (el puente, masculine) if it was Claudio, but
not if it was Claudia. It appeared that the Natural Gender assignment of inanimates
in the game context was influenced by Linguistic Gender of the language spoken
by the participants.
One should keep in mind that all such experiments are artificial—how many times
in real life do speakers of any language have to bother with female or male names
assigned to bridges and apples? They may, however, point towards an interesting
direction: that some people may attribute ‘male’ and ‘female’ labels to inanimate
objects in agreement with their Linguistic Gender assignment. This is echoed by
spontaneous remarks by people who speak languages with Linguistic Genders outside
Europe. The Manambu state that, in their language, everything in the world is either
male or female—thus drawing a parallel between the Linguistic Gender of an entity and
its Natural Gender associations. This anecdotal evidence points toward a certain
metalinguistic awareness of gender (something we mentioned in Chapters 1 and 4). 11
In many languages and cultures, special physical and psychological properties are
associated with males and females. And can ‘typical’ characteristics of men and
women be transferred to any entity in the world, following its Linguistic-cum-
Natural-Gender association? This is our next topic.
8.4 What men and women look like
The ways in which Natural Gender is attributed to inanimates through words for man
and woman reveal physical features associated with males and females. Ilocano, a
Western Austronesian language (Rubino 1997: 75–6), has no Linguistic Genders. The
Natural Gender of humans is distinguished lexically, through nouns lalaki ‘boy’, babai
‘girl’. These lexemes are also used to distinguish Natural Gender of hybrid nouns, e.g.
kabsat a lalaki ‘brother’, kabsat a babai ‘sister’. They can also co-occur with inanimates,
highlighting the overtones of ‘strength’ and ‘sweetness’ associated with ‘masculinity’

128 8 The images of gender
and ‘femininity’.So,basi a lalaki ‘strong sugar cane wine’ is associated with masculinity;
and basi a babai ‘sweet sugar cane wine’ with femininity. 12
Men’s and women’s activities and their descriptions can reflect the combined
images of Natural Gender and of Social Gender. In Nungon, a Papuan language
with no Linguistic Gender, the notion of nging ‘hot, hurtful, powerful, sharp’ relates
to male work and practices. These include hunting with bows and arrows, building
houses, and thatching roofs. The notion of hum ‘cold, pacific, weak, dull’ refers to
what women do (forage in the forest, dig up small animals’ burrows, and perform
domestic duties). Back in the old days, men used to have taboos and food abstinences
before hunting, so as not to become hum-o ‘dull, powerless’—or women-like. An
important dimension in masculine or feminine gender choice in Yangoru Boiken is
‘hardness’ (an esteemed male quality in metaphorical terms): ‘ironwood is masculine
because it is “hard”, hard-textured yam varieties are masculine, and softer tuber ones
are female’ (Roscoe 2001, p.c.). The hard natural fibres used to make traditional string
bags in Manambu are called du-a-maej ‘men fibres’. These are jokingly contrasted with
takwa-maej, lit. women fibres, to refer to imported wool which is much less strong. 13
Ket, from the Yeniseian family, distinguishes three genders: masculine, feminine,
and neuter. Gender choice is only partially predictable from the meaning of the noun
(see also §2.3). In Vajda’s (2004: 16) words, ‘in general, masculine-class items are
perceived as having “greater vital force” than feminine-class nouns’. 14 A growing tree
belongs to the masculine class, and a log or a dry tree is feminine. Some trees (e.g. the
birch) are feminine if crooked and masculine if straight. The meaning of gender
choice shows correlation with ‘cultural relevance, or economic importance, with
terms for animals and things, which are valued higher, or are viewed as having
more cultural or societal significance than others’, more often ending up in the
masculine than in the feminine class (Georg 2007: 88–90). Typical features of
feminine Social Gender are projected onto entities which have no Natural Gender,
in the form of Linguistic Gender assignment.
Gender assignment to inanimates in Amharic is based on size: in Pankhurst’s(1992:
168) words, the feminine being used to suggest a smaller, the masculine a larger object.
In addition to this, ‘a sexual distinction is conferred on an inanimate object which
reveals the gender construction in a society. Thus male soils, yewend merét, are rich and
fertile, whilst female soils, yesét merét, are light and relatively infertile.’ Masculine
Linguistic Gender in Oromo is associated with larger size, power, and importance.
Feminine Linguistic Gender has the opposite connotations. And nouns denoting
undesirable social characters—e.g. a thief, or vagabond—tend to be assigned to the
feminine gender. 15 This is reminiscent of our discussion in §7.5 and §11.3.
Physical properties of women in traditional stories may be associated with women’s
place in ritual, and specific female ‘pollution’ and impurity due to menstrual
blood. One of the myths of the Tariana of north-west Amazonia concerns libertine
ancestral women who had ‘misbehaved’ with a smelly mucura-rat (an Amazonian

8.4 What men and women look like 129
marsupial, of genus Didelphus, with a characteristic pungent smell). The women
used to smell nice. As a consequence of sexual intercourse with the animal, they
became smelly (i:-peni). As we will see in Chapter 10, many more things are wrong
with women in the Tariana lore.
Throughout the world, proverbs reflect the ‘traditional wisdom’ of maledominated
societies. Small size and relative insignificance are the physical properties
attributed to women. An Oromo proverb says, ‘A wife and a plough are best when
shorter than the man’. A Kalmyk proverb comments on women’s lack of intelligence—‘though
a girl’s hair be long, her brain is short’. Women are depicted as
passive objects, containers, animals, consumables, mentally and physically weak, and
generally inferior beings. ‘The hen is no bird, the woman is no person’, says an
Estonian proverb. An overwhelming majority of proverbs collected by Schipper
(2003) comment on women from a dominant male perspective, defending male
privileges and interests. In Schipper’s (2003: 374) words, ‘proverbial metaphors
reproduce a gender hierarchy as an imagined world in which it is self-evident who
are the speakers and who the silenced ones’. 16
Directions ‘left’ and ‘right’ may be associated with strength and weakness, and
extend to ‘female’, or‘feminine’, and ‘male’, or‘masculine’. For instance, in Khoe,
‘besides expressing distinctions of natural sex, masculine and feminine have a clear
cut secondary meaning’, with masculine being associated with ‘long, high, narrow,
big, strong items, and the right hand side’, and the feminine gender with ‘short, small,
round, thick, broad, weak items, and the left hand side’. An association of ‘right’, or
‘male’, side associated with being ‘strong’, and left, or female, with ‘being weak’, isa
feature of many languages. 17 The terms ‘left’ and ‘right’ in Manambu are used to
describe sides of people’s bodies. The term ‘left hand’ (aki-ta:b) is also used to refer to
one’s mother’s side in traditional payments, and the term ‘right hand’ (mapa-ta:b) is
used to refer to father’s side. (A synonym for the left hand is kaykwapa-ta:b, lit. lazy
hand, and for the right hand vyavya-ta:b, hitting hand).
Linguistic Gender choice for inanimates may be associated with the idea of
femininity as ‘motherhood’. We can recall, from §3.1, that in Sare, from the Sepik
region of Papua New Guinea, a mature plant or fruit is assigned to the feminine
gender while a young plant or fruit will be masculine. This consistent principle of
Linguistic Gender assignment in Sare can be seen as a projection of Natural Gender
properties (that is, femininity associated with procreation), as formulated by Sumbuk
(1999: 117), a native speaker of the language. Similarly, Beachy (2005: 63) suggests
that ‘sun’ and ‘blood’ in Dizin, an Omotic language, are assigned to the feminine
gender because they are ‘seen as life-giving’. The association of Linguistic Gender
with women and motherhood underlies the gender assignment.
A nature-versus-artefact dimension has also been aired, as a correlate of female
and male. In an experiment conducted by Mullen (1990), twenty-two 7-year-old
schoolchildren (all English speakers) were asked to classify names or pictures of

130 8 The images of gender
objects into ‘female’ and ‘male’. It turned out that objects pertaining to ‘nature’—such
as plant, seashell, leaf, and fish—were classified as ‘woman-like’ by the majority of
children. In contrast, artefacts—such as street sign, sailboat, television, and clock—
were considered ‘man-like’. The female–nature association was a bit stronger than
the male–artefact association. This reflects conventional practices of referring to
nature as ‘she’; as in the English expressions ‘mother nature’ and ‘mother earth’.
These associations originate in cultural stereotypes of the Social Gender (especially in
Western societies). As Mullen (1990: 585) puts it, ‘the realm of nature, as opposed to
the realm of artefacts, is thought to be less controllable, less predictable, less rational,
and more balanced, more whole, more nurturing. Each of these characteristics is
generally associated with females as well. The male–artefact domain is associated
with the opposite set of characteristics.’ 18 This whole way of looking at males and
females is a ‘construct of culture rather than a fact of nature’ (as Ortner 1972: 87) puts
it, and relies on an evaluation of women across many cultures as being somehow
‘second-class citizens’ (p. 68). What can be seen as Natural Gender features and
associations reflects the ingrained stereotypes of Social Gender.
Among the Kaluli (a Papuan community in the New Guinea Highlands), the
physique and capabilities of women and men are seen through the prism of Social
Gender stereotypes. The Kaluli language has no Linguistic Gender. As Schieffelin
(1977: 121–7) puts it,
‘men’ epitomise the ‘Kaluli’ behaviour, partly because of the high visibility of male activity. It is
the men who tend to be dramatic and flamboyant, who act publicly, who compel attention.
Most of the behavior that the Kaluli themselves consider culturally important is performed in
the male role...In relation to femaleness, maleness encompasses the dominant Kaluli values,
and the figure of a man is something of a general cultural ideal.
The Kaluli men characterize themselves as ‘wiry, alert, fast’, and see ‘women as heavy
and slow, unable to travel quickly or far’; they are viewed as ‘weaker and less dynamic’.
These images essentially reflect the Social gender, correlating them with features of
Natural Gender, so much so that even an outsider (in this case, the anthropologist
Edward Schieffelin) sees Kaluli women as ‘being generally bulkier in appearance than
men, and carrying themselves with a certain unwieldiness that is enhanced by their
swaying skirts and ample breasts’ (p. 122).
The ways in which men and women are talked about reflect what scholars call
‘hegemonic masculinity’—an ideal ‘man’ able to subordinate those who may not be
quite up to it (including gay, disabled, and working-class men, and all women).
Caldas-Coulthard and Moon (2010) examined adjectives occurring with ‘man’,
‘woman’, ‘boy’ and ‘girl’ in British tabloid and broadsheet corpora, and found that
men were more often categorized in terms of age, status, and personality, and for
women the focus was more on physical appearance and sexuality—that is, as sexual
‘objects’. Baker (2014) investigated how the use of adjectives with men and women

8.4 What men and women look like 131
changed throughout the Historical Corpus of American English which covers the
period 1810–2009. The texts there include fiction, magazines, newspapers, movies,
and play scripts. Throughout the nineteenth century, ‘man’ but not ‘woman’ would
be accompanied by adjectives like honest, brave, and good-hearted. Morality was the
‘province of men’ (p. 48), and a woman would be characterized in terms of looks, age,
and breeding. Towards the end of the twentieth century, the focus shifted to
adjectives relating to ‘bodies’: men appear as able-bodied, burly, and stocky, and
women as good-looking, sweet-faced, comely, and so on. A move from referring to
males and females in terms of moral qualities towards focus on their bodies may be
linked to men being regarded by women as sexual objects. This may indeed represent
an equalizing trend: men, as well as women, can be seen as sexual objects as women
become more independent and empowered.
The attributes of men and women, and their behaviour, are strongly linked to their
Social Gender roles, and the associated rituals and practices.
For many years, cognitive linguists have been attempting to conduct experiments and
uncover associations between physical properties and Natural Genders (male and
female). Results published by Osgood (1960) showed strong correlations between the
concept of ‘man’ and ‘woman’ and properties such as angular versus rounded, dark
versus light, colourless/colourful, sharp/blunt, thick/thin, large/small, near/far, and
also vertical/horizontal, heterogeneous/homogenous, and clear/hazy. The experiments
included speakers of English, Navajo, Mexican Spanish, and Japanese. Ervin
(1962) reported further support for ‘a tendency to ascribe different connotations to
masculine and feminine words in Italian’. She asked thirty bilinguals in Italian and
English to attribute Linguistic Genders to nonsense words, and then rate them with
regard to four parameters: bello/brutto (beautiful/ugly), buono/cattivo (good/bad),
delicato/forte (weak/strong), piccolo/grosso (small/big). The nonsense words assigned
to feminine Linguistic Gender were consistently rated as prettier, weaker, and smaller
than those assigned to the masculine Linguistic Gender.
Boroditsky, Schmidt, and Phillips (2003) put speakers of Spanish, German, and
English through yet another test, aimed at proving whether Linguistic Gender
influences the way people think about objects. Speakers of Spanish and of German
(all highly proficient in English) were given a list of twenty-four object names and
then asked to write down the first three adjectives that came to mind to describe
them. Then a group of speakers were asked to rate the adjectives as describing
masculine or feminine properties. As predicted by the authors of the experiment,
both Spanish and German speakers generated adjectives that were rated as more
‘masculine’ for items with masculine Linguistic Gender than for those with feminine.
All object names had opposite genders in Spanish and in German, and so Spanish
and German speakers produced different adjectives to describe the same object. The
word for ‘key’ (der Schlüssel) is masculine in German, and feminine in Spanish (la

132 8 The images of gender
llave). So, German speakers described keys as hard, heavy, jagged, metal, serrated, and
useful, while for Spanish speakers they were golden, intricate, little, lovely, shiny, and
tiny. The word for ‘bridge’ is masculine in Spanish—so speakers described bridges as
big, dangerous, long, strong, sturdy, and towering. In German, ‘bridge’ belongs to the
feminine Linguistic Gender—and so bridges were beautiful, elegant, fragile, peaceful,
pretty, and slender. What better proof that people’s thinking about objects is influenced
by the grammar of the language they speak? Grammatical genders might even
affect the design of artefacts—a closer look at German bridges may show that they
differ from Spanish bridges in the way they are built. And so on.
This astonishingly neat correlation between Linguistic Gender and its Natural
Gender associations was tested by Landor (2012). 19 The study involved 1,290 participants
from 24 countries (native speakers of German, Spanish, English, Hebrew, and
Hungarian). They were asked to describe twenty-four nouns (covering humans and
inanimates) with adjectives (‘hard’, ‘soft’, ‘big’, ‘small’, and so on), and then rate them
as ‘male’ or ‘female’. The outcome of the study was negative—no positive correlations
to speak of were found between Linguistic Gender of nouns and their Natural Gender
associations. Many speakers just shrugged the questions off with a simple ‘can’t decide’.
Further doubt was cast upon the whole enterprise of ‘experimental studies’ by Guy
Deutscher (2010: 211–12). In his words,
all the experiments described so far suffer from one underlying problem, namely that they forced
the participants to exercise their imaginations. A sceptic could argue with some justification that
the only thing the experiments proved was that grammatical genders affect associations when the
participants are coerced unnaturally to dream up properties for various inanimate objects. ... In
other words, if you force Spanish speakers to be on-the-spot poets, and extract properties of
bridges out of them, the gender system will indeed affect their choice of properties. But how can
we tell whether the masculine gender has any influence on speakers’ spontaneous conceptions of
bridges, even outside such exercises in poetry on demand?
And indeed, we cannot. All we can say with a degree of assurance is that Linguistic
Genders may open up an avenue for metaphors, and may reflect the roles and images
associated with Social Genders—but do not have to. From behind the smoke-screen
of ‘scientific experiment’ bridges in Spain may look slenderer and smaller than
bridges in Germany—a good topic fit only for popular magazines and small talk.
8.5 ‘Gendering’ the world: images, metaphors, and cognition
In Jakobson’s (1959: 116) words, ‘languages differ essentially in what they must
convey and not in what they may convey...In its cognitive function, language is
minimally dependent on the grammatical pattern...But in jest, in dreams, in magic,
briefly, in what one would call everyday verbal mythology and in poetry above all, the
grammatical categories carry a high semantic import.’ This is where special meanings

8.5 ‘Gendering’ the world 133
of Linguistic Genders—laden with images of Natural and Social Genders—step in.
And this is also where languages with Linguistic Gender allow us to see the world
through a different lens, creating difficulties for translators.
Linguistic Gender is a treasure chest of metaphors—as Guy Deutscher puts it,
‘language’s gift to poets’. Again, in Deutscher’s words, ‘the chains of associations
imposed by the genders of one’s language are all but impossible to cast off ’. Linguistic
Genders reflect myths and rich traditions of peoples’ heritage embalmed in the
language. Creole and Pidgin languages with shallow histories and little oral and
written tradition simply lack these. 20
Natural Gender-laden images permeate the world of myths and beliefs: the moon
may be a mythical man or a woman, and so can be the sun. Features of Natural and
Social Gender roles can be revealed through the ways men and women are talked
about. Physical properties typically associated with men and with women may draw
on their Social Gender roles, or cultural norms to do with men and women. For the
Kaluli (described by a male anthropologist), men are wiry and alert, and women slow
and bulky. Men tend to be conceived of as big, hard, and strong, and women as the
opposite. Attributes of women—and not so much of men—have negative overtones.
So do words which have an exclusively female reference. We saw in §3.4.2 that
masculine Linguistic Gender correlates with ‘value’ and ‘importance’ in many patrilineal
societies where the male Social Gender is high in status. This reflects traditional
power relationships, to which we return again in subsequent chapters.
There is no hard-core evidence that all inanimate objects are automatically
assigned to a Natural Gender by speakers of languages with Linguistic Gender. But
Linguistic Gender interacts with cognitive mechanisms. In Dan Slobin’s (1996: 91)
words, each language ‘is a subjective orientation to the world of human experience,
and this orientation affects the ways in which we think while we are speaking’. This is
part of ‘thinking for speaking’—the multitude of unconscious decisions a speaker has
to make, in order to express what the grammars require.
Having a Linguistic Gender in your language forces you to make a choice in gender
forms. And this choice, and the meanings of the Linguistic Gender, in turn, may
reflect the ways in which men and women are seen and behave—in their Natural and
Social Gender guises. But there is no reason to believe that the overtones of Natural
and Social Gender are always projected onto every Linguistic Gender choice—
German bridges are no more ‘woman-like’ than bridges in Spanish are ‘man-like’.
Taken to its extreme (and coupled with quasi-scientific experiments), any attempt to
find a perfect one-to-one correspondence between Linguistic Gender, Social Gender,
and Natural Gender verges on the grotesque.
To summarize: having Linguistic Gender in a language opens up an avenue for
metaphors—deployed in poetry, beliefs, and images. Properties of men and of
women—as seen in adjectives and other expressions associated with them—reflect
stereotypes of Social Gender. As a society changes, men and women may come to be

134 8 The images of gender
seen differently: Linguistic Genders come to reflect social changes and language
reforms. Different patterns of male and female speech further reflect differences in
Social Gender. This is what we turn to next.
NOTES AND SOURCES
1. See Haas (1941: 65), Mithun (1999: 101) and §2.2.1.
2. See Alpher (1987: 180).
3. The Manambu have no stories which might involve attributing the Natural Gender to the
sun. Based on its round shape, it is typically assigned feminine gender. But if it is
particularly strong and hot, it can be referred to as masculine (following the principles
in Figure 3.1).
4. See Payne (1989: 130), Romani Miranda (2004).
5. Nissen (2002) offers further examples and some hints on how to bridge the abyss of ‘nontranslatability’.
6. In Heine’s original, the male pine-tree and the female palm-tree imagery is thought to
go beyond unrequited romantic love: for a summary, see Deutscher (2010: 196). Nikolaj
Glazkov, a rather controversial and amusing Russian poet, offered his own version of
Heine’s poem: an ever-green larch (listvennica, feminine Linguistic Gender) transplanted
into the Far North (evoking Stalin’s concentration camps) is dreaming of a ‘fantastic
southerly’ oak-tree (dub, masculine) (1976, Voksal. Moscow: Sovetskij pisatelj).
7. Emma Lazarus, title 1: ‘Lotus blossom’, from Poems and translations, published 1867.
Another well-known example is a humorous poem by Joachim Ringelnatz about a male
stamp (masculine in German) falling in love with a (female) princess. The sexual Natural
Gender imagery is reflected in a translation by Ernest Seemann . The German text is: Ein männlicher Briefmark erlebte | was Schönes,
bevor er klebte | Er war von einer | Prinzessin beleckt | Da war die Liebe in ihm erweckt | Er
wollte sie wiederküssen | Doch hat er verreisen müssen | So liebte er sie vergebens/ | Das ist
die Tragic des Lebens. And the English translation is: ‘A postage stamp, male, was elated |
With joy, before he was dated | A princess licked him, by Jove! | Which did awaken his love
| He wanted to kiss her back | But had to go on a trek | His love was thus unavailing | So sad
is often life’s failing.’
8. The song ‘Why are you standing swaying, you thin Rowan’ (Shto stoish, kachajasj, tonkaja
rjabina...) is based on a poem by Ivan Surikov (1864). Further illustrative examples from
classical Russian authors and extensive commentary are in Vinogradov (1947: 64–70).
9. The ‘gender symbolism’, with special stylistic effect, can be achieved in languages without
Linguistic Gender, by using nouns with Natural Gender reference—man and woman. The
young moon is referred to as kuu-poeg ‘moon-son’ in a widely known Estonian lullaby, by
Heljo Mänd (Vikeraare all. Tallinn, Eesti Raamat, 1965, p.26).
10. The two largest cities in Russia translated into ‘a woman’ and ‘a man’, for Nikolaj Gogol:
‘Moscow (Russian Moskva) is of feminine gender, St Petersburg of masculine gender.
Moscow has brides, St Petersburg has bridegrooms.’ He goes on to describe Moscow (of
feminine Linguistic Gender) as a bulky woman making pancakes or sitting in her armchair,
never venturing anywhere—in contrast to St Petersburg, a dapper and curious
young man, eager to show himself off to foreigners (Peterburg notes 1836). There are
many more examples of hard-to-translate Linguistic Gender associations in poetry—some

8.5 ‘Gendering’ the world 135
examples are Charles Baudelaire’s ‘L’homme et la mer’ (the man:masculine and the sea:
feminine) and Pablo Neruda’s ‘Ode to the sea’ where the sea (el mar, masculine) strikes a
stone (una piedra, feminine) and ‘caresses her, kisses her, pounds his chest, repeating his
own name’ (see also Deutscher 2010: 215).
11. Having to always mark Linguistic Gender agreement imposes a necessity to always make a
Linguistic Gender choice. Having to choose a masculine or a feminine Linguistic Gender
form in Greek conversations helps foreground male or female identity. The grammatical
marking of masculine and feminine Linguistic Genders helps create the ‘sociocultural’—or
Social Gender—world which participants build step by step in their interaction. We turn to
this in Chapter 7 (see Alvanoudi 2014, 2015).
12. See Rubino (1997: 75–6).
13. This is reminiscent of Julius Caesar’s description of the brave Belgae, who were the bravest
in Gaul, since they had less access to objects which tend to ‘effeminate’ the mind, Caesar’s
Gallic Wars Book 1: ‘The Belgae are the bravest, because they are furthest from the
civilization and refinement of [our] Province, and merchants least frequently resort to
them, and import those things which tend to effeminate the mind’ (‘Horum omnium
fortissimi sunt Belgae, propterea quod a cultu atque humanitate provinciae longissime
absunt, minimeque ad eos mercatores saepe commeant atque ea quae ad effeminandos
animos pertinent important...’).
14. See also Dul'son (1968: 64), Werner (1994, 1997: 88–96).
15. Raga and Woldemariam (2011: 166–7).
16. From Schipper (2003: 19, 27, 386).
17. See, for instance, Raffelsiefen (2011).
18. To what extent such associations are indeed universal as claimed by Mullen (1990) isan
open question. In some non-Western societies, feelings and irrationality are a feature of
women, and rational behaviour is linked to men (see Kulick 1992a, 1992b, on the Taiap
people in the village of Gapun in the Sepik area of Papua New Guinea).
19. Landor (2012: 93–6) contains a description of the experiment. Later in the thesis, he
concludes that the Linguistic Gender of nouns may have a ‘clouding effect’ on how, and
whether, objects can be conceptualized as ‘male’ or ‘female’. Gendered conceptualization
of objects: those who speak languages with Linguistic Gender cannot decide on an entity’s
male or female attributes, as the labels of the genders (masculine and feminine) ‘suppress
the confidence with which participants ascribe gendered characteristics to the objects’
(2012: 193–5). These results go against the ‘feel-good’ and suggestive correlations between
Linguistic Gender choice and indisputably male or female characteristics of entities. Note
that in the majority of ‘experimental’ quantitative studies of cognitive effects of Linguistic
Genders, the size of a sample is too small to provide statistically significant results for a
whole population of speakers of German, Spanish, English, and other languages investigated
(see Dixon 2010: 257–61, on the principles of sampling). Landor (2012)—with a
careful statistic back-up—stands out as an exception.
20. No wonder that McWhorter (2014), primarily a Creole scholar, shows little appreciation of
gender as an ‘ornate’ category. Creoles lack many other linguistic features which reflect social
patterns and peoples’ perception—including evidentials (see Aikhenvald 2014c), possession
(Aikhenvald 2013c), and complex demonstrative systems (Dixon 2012: 189–261). This may
well be why all languages look the same for the likes of McWhorter—as they are likely to be, if
looked at through the lens of Creoles.

9
When women and men speak
differently
Differences between male and female speech vary, from language to language and
from society to society. Ways of speaking differentiate Social, and Natural Genders.
In a number of languages from North and South America, Asia, Australia, and the
Pacific male and female dialects show regular and salient differences. As Jean Kirton
(1988: 111) put it, the ‘separate dialects for the men and for the women’ in Yanyuwa,
an Australian language, have ‘differences of a kind normally associated with language
dialects in separate locations’. This is what we turn to now.
9.1 ‘Male’ and ‘female’ dialects
Sara Trechter, a major figure in the field of male and female dialects in the North
American context, describes her experience with the Lakhota of the North American
south-west:
Lakhota speakers who were unaware of my interest in gender have often told me that women and
men end their sentences in Lakhota differently. My Pine Ridge ‘permission to record’ agreement
contained a statement alluding to my interest in gendered speech styles. On reading this agreement,
potential consultants directly addressed what they perceived to be my research question.
Their explanation took the form of metapragmatic judgements or maxims about appropriate
speech. ...‘Men say yo and women say ye’ or ‘Men say lo and women say le’.(1999: 107)
Similarly to other Siouan languages, Lakhota has a set of enclitics which express
questions, commands, and emphatic statements. The clitics ye (male) and yo (female)
mark commands, and the clitics lo (male) and le (female) mark opinion or emphasis.
Table 9.1 features a selection.
A man will say 9.1 and a woman will say 9.2, telling someone to get up (Trechter
1995: 14):
9.1 kikta ye
get.up female.speech.imperative
‘Get up!’ (woman speaking)
How Gender Shapes the World. First edition. Alexandra Y. Aikhenvald.
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.

9.1 ‘Male’ and ‘female’ dialects 137
TABLE 9.1. Male and female forms of a selection of enclitics in
Lakhota (Trechter 1995: 57)
MEANING MALE FEMALE
formal question hu̹wo hu̹we (obsolete)
command yo ye
familiar command yet h o nit h o
opinion/emphasis/fact yelo yele (obsolete); ye
emphatic statement ks᷈t ks᷈to
entreaty ye, yee na
surprise/opinion yewa̹ yema̹
9.2 kikta yo
get.up male.speech.imperative
‘Get up!’ (man speaking)
The following fieldwork anecdote shows how real the distinction between male
and female registers used to be among the Gros Ventre, from the Algonquian
language family, at the time when Regina Flannery (1946: 134) did her fieldwork:
When an old woman asked me what name I had received in the naming ceremony, I repeated
as an old man, my namer, had pronounced it iθénædjæ (woman chief). Whereupon the old
woman said indignantly: ‘Your name is iθénækyæ. I knew that woman for whom you are
named. Can’t these men ever do anything right!’
Regina Flannery had repeated the name as the man had pronounced it, instead of
converting it to the form a woman would use.
We look into the differences between male and female dialects, their place in
language history, and correlations with the Natural and the Social Gender of both
speaker and addressee, in §§9.1.1–5.
In two well-known cases, male and female dialects correlate with politeness
registers, speakers’ status, and social identity. In Japanese, men and women employ
different personal pronouns, and markers of politeness and ‘elegance’ of speech. The
choice of forms depends on the relationship between the speaker and the addressee.
As the women’s place in Japanese society changes, the rules for male and female
speech use transform. Male and female speakers of Thai, the national language of
Thailand, use different final particles marking respect for the addressee. The choice of
the pronoun ‘I’ for men and for women depends on status of the speaker and the
addressee, and their relationships—see §9.1.6.

138 9 When women and men speak differently
Languages with distinct male and female dialects have been accorded the label of
‘gender-exclusive’. Languages in which differences between men’s speech and women’s
speech are more subtle and harder to capture are known as ‘gender-variable’—they
include most familiar European languages, among them English. As Southerland and
Katamba (1996: 550–1) put it, ‘gender-exclusive differentiation refers to the radically
different speech varieties used by men and women in particular societies. In these
societies, a woman or man may, except in special circumstances, not be allowed to
speak the variety of the other gender...Gender-variable differentiation is much more
common in the languages of the world than is gender exclusivity. This phenomenon is
reflected in the relative frequency with which men and women use the same lexical
items or other linguistic features. If, as is often asserted, female English speakers use
words such as lovely and nice more often than do male speakers we can claim that in
this respect English speakers exhibit gender-variable differentiation.’ 1
A gender-exclusive language imposes the choice between the male or the female
dialect on every speaker. As we will see throughout this section, either dialect can
come to be associated with a Social Gender, formality, and social status. A gendervariable
language operates the other way round: its linguistic forms are deployed in
different ways, reflecting social conditions, attitudes, and behaviours which then
come to be associated with an image of a man or a woman. These associations are
fluid and hardly obligatory. Unlike gender-exclusive languages, men and women
draw upon the same repertoire of forms—see §9.2. As gender-exclusive languages fall
into disuse, the boundaries between gender-exclusive and gender-variable languages
become blurred.
9.1.1 How male and female dialects differ
Differences between the male and the female varieties may be purely phonological. In
his grammar of the Inuktikut (Eskimo), Boas (1911: 79) remarked:
Among some Eskimo tribes the men pronounce the terminal p, t, k, and q distinctly, while the
women always transform these sounds into m, n, ñ and ñ̩.
Men’s speech in Karajá, a Macro-Jê language from Brazil, lacks the velar glottal
stop k (which is present in women’s speech). 2 A woman will say ruku, and a man will
say ru ‘night’; a woman will say dıkarə̃, and a man will say dıarə̃‘I’. The principle
affects recent borrowings from Portuguese, the main language of the country: a
woman will say kawaɾu, and a man will say awaɾu, from Portuguese cavalo ‘horse’.
A woman will say bicikreta, and a man will say bicileta, from Portuguese bicicleta
‘bicycle’. Female speech appears to be more conservative—we return to this in §9.1.3
and then in §10.4.
Male and female dialects of Gros Ventre show systematic differences in their
phonology. The sounds tc and dj in men’s speech correspond to k and ky in women’s
speech. Men would say itcénibitc ‘his gum’, and women would say ikébinik. 3 That

9.1 ‘Male’ and ‘female’ dialects 139
men’s andwomen’s speech sounds differently in Chukchi, a highly endangered
Chukotko-Kamchatkan language from the extreme north-east of the Russian Federation,
was first noted by Vladimir Bogoras. There still remain regular correspondences
between sounds in men’s and women’s speech:awomanwouldsay
mcen,andamanmren,for‘mosquito’; a woman would say cajkok and a man sajkok
for ‘teapot’. 4
Systematic differences between male and female speech can extend into morphology,
as we saw for Lakhota in Table 9.1. In her classic paper, Mary Haas (1944)
described differences between male and female speech in imperative and indicative
verbs from Koasati, a Muskogean language from Louisiana. A woman’s form would
end in a nasalized vowel, and a man’s form would replace a nasalized vowel with an
oral vowel followed by s. A woman would say /lakawtakkõ/ and a man would say
/lakawtakkós/ ‘I am not lifting it’.
Another prime example of male and female dialects comes from Yana, a now
extinct isolate formerly spoken in northern California and described by Sapir
(1929a). Many Yana words had a ‘male’ and a ‘female’ form. A male form used to
be longer, and a female form reduced. The male forms ‘au-na ‘fire’ and ‘au-‘nidja ‘my
fire’ corresponded to female forms ‘au‘ ‘fire’, ‘au-‘nitc‘ ‘my fire’. There were further
morphological differences. The demonstratives ending in ‐e would add -‘e in the male
forms. A man will say aidje’e ‘that one’, and a woman will say aidje. Men and women
use different forms to mark questions: to form an emphatic interrogative, a man
would use an enclitic nā and a woman an enclitic gā . 5
Male and female speech is a striking feature of Kokama-Kokamilla, a Tupí-
Guaraní language of northern Peru. Men use the suffix =kana for plural, and
women use =nu; so the plural of iwira ‘tree’ will be iwira=kana ‘trees’ for men, and
iwira=nu for women; the plural of mesa ‘table’, a recent borrowing from Spanish, will
be mesa=kana for men and mesa=nu for women. Further differences lie in personal
pronouns, demonstratives, and linkers—a selection of forms is in Table 9.2. 6
Male and female speech in Awetí, a language from the Xingu Indigenous Park in
Brazil closely related to Tupí-Guaraní languages, differs in some of the personal
pronouns and prefixes, and all the demonstratives. Table 9.3 illustrates some of the
differences. 7
Just a few lexemes have female and male forms; some male forms tend to have an
initial n and female forms do not, e.g. male nypék, female ypék ‘duck’. The term for
‘duck’ in other Tupí-Guaraní languages does not have an initial n, e.g. Língua Geral
ipêka ‘pato’. The female dialect appears to be more archaic—we return to this in
§9.1.3 (and then in §10.4).
The female dialect of Yanyuwa has six Linguistic Genders (male, female, masculine,
feminine, food, and arboreal). The men’s dialect classifies male humans, male
animates, and a few other items into one Linguistic Gender (marked with the prefix
ki-). The female dialect distinguishes human males (using the ‘male’ class prefix nya-)

140 9 When women and men speak differently
TABLE 9.2. Female versus male forms in Kokama-Kokamilla: a selection (Vallejos
2010: 42)
TRANSLATION FEMALE FORM MALE FORM
First person singular pronoun tsa, etse ta
First person exclusive pronoun (we without you) penu tana
Third person plural pronoun inu rana
Proximal demonstrative ‘this’ ajan ikian
Distal demonstrative ‘that’ yama/yamua rama/ramua
also yay riay
there, then yaepe raepe
TABLE 9.3. Female versus male forms in Awetí: a selection
TRANSLATION FEMALE FORM MALE FORM
First person singular pronoun itó atit
Third person singular pronoun ı̃ nã
Third person plural pronoun ta’i tsã
Proximal demonstrative ‘this’ (close to speaker) ujá jatã
Proximal demonstrative ‘this’ (close to addressee) akyʹj kitã
Distal demonstrative ‘that’ (distant from both speaker
and addressee)
akój kujtã
and male animates (using the ‘masculine’ class prefix ji-). Pronouns, demonstratives,
and verbal prefixes have different forms in the two dialects. 8
Male and female dialects may use different interjections. Among the Gros Ventre
greetings wahei (male speakers) and næhæ (female speakers) ‘hello’ would immediately
tell the Natural Genders apart. When expressing surprise, a man would say
u . tse, and a woman e . ha, and when answering to a hail, a man would say wei', and a
woman o . . Vocal gestures would also be different. As Flannery (1946: 133) put it, ‘in
order to indicate joy and thankfulness, women give a call which may be described as
rattling the tongue, whereas men vocalize and break the sounds, not with the tongue,
but by striking the mouth rhythmically with the palm’.

9.1 ‘Male’ and ‘female’ dialects 141
9.1.2 Speakers and addressees of male and female dialects
The choice between male and female speech may depend on the Natural Gender of
the speaker. In Koasati, Kokama-Kokamilla, Awetí, and Yanyuwa men use the ‘male’
dialect no matter who they talk to. Or the choice of a male or a female dialect may
correlate with the Natural Gender of the person spoken to. ‘Male’ forms in Yana were
used by men speaking to men. ‘Female’ forms were used by women speaking to men
or women, and men speaking to women. 9
Biloxi, an extinct Siouan language, had an elaborate system of male and female
forms of mood particles which would mark a statement, a question, or a command.
There were three sets of forms: the same form is used if a man or a woman is
addressing a woman. Men-to-men and women-to-men forms are all different.
A sample is in Table 9.4.
In addition, special singular and plural forms were used by both sexes to address
children: ki ‘carry it (singular)’ and kitu' ‘carry it (plural)’.
Kũr̩ux, a Dravidian language spoken by more than 100,000 people in north India,
has a strict distinction between male and female dialects whose choice depends on
the sex of speaker and the sex of addressee. The differences are most visible in verbal
paradigms. Male or female forms are also used when talking about a man or a
woman. A selection of forms of the verb bar- ‘come’ in present tense is in Table 9.5.
Tunica, a now extinct isolate, is a rare example of a language where the choice of
male and female speech is based just on the sex of the addressee. The differences are
found only in the pronominal system (including full pronouns, prefixes, and suffixes,
in three numbers: singular, dual, and plural). An example is in Table 9.6.
Ignaciano (or Mojo), an Arawak language from Bolivia, 10 combines Linguistic
Gender with male and female dialectal differences in a rather curious way. Masculine
gender covers all animate males, feminine gender covers all animate females, and
neuter gender is reserved for all inanimates. Female and male distinctions are made
only in masculine forms of personal pronouns, demonstratives, indefinites, and
possessives: one form is used if a man is speaking, and the other one if the speaker
is a woman. This is shown in 9.3, for the demonstrative ‘this’.
TABLE 9.4. Male and female speech determined by speaker and addressee in Biloxi
SEX OF SPEAKER SEX OF ADDRESSEE SINGULAR PLURAL MEANING
Man Man ki-aŋko' ki-takaŋko
Man
Woman
ki-tki' ki'-tatki'
Woman
Woman
Woman Man ki-tate' ki-tatute'
‘Carry it!’

142 9 When women and men speak differently
9.3 Singular Plural
juca ‘this:neuter’ (covers inanimate referents)
suca ‘this:feminine’ (animate)
maca ‘this:masculine (animate)’: male speech
ñica ‘this:masculine (animate)’: female speech
juca-na
naca, náca-na
TABLE 9.5. Male and female speech in Kũr̩ux: present tense of the verb bar- ‘come’
PERSON/NUMBER
SEX OF SPEAKER AND ADDRESSEE
Man to Man Man to Woman Woman to Woman Woman to Man
1sg bar-d-an barʔ-e-n bar-d-an
1pl exclusive bar-d-am barʔ-e-m bar-d-am
1pl inclusive
bar-d-at
2sg bar-d-ay bar-d-i bar-d-in bar-d-ay
2pl bar-d-ar bar-d-ay bar-d-ar
TABLE 9.6. Male and female speech chosen by the sex of the addressee in Tunica
SEX OF ADDRESSEE SINGULAR DUAL PLURAL
Pronominal suffixes on semelfactive verbs
Independent pronouns
Man -ʔa -wi'na -wi'ti
Woman -ʔi -hi'na -hi'ti
Man má wi'nima
Woman h h 'ma hi'nima
Different forms for male and for female speech are there only in the singular. In the
plural, there is a straightforward division into animates and inanimates.
9.1.3 Male and female dialects, and language history
Differences between a male and a female dialect may reflect the history of the
speakers. A striking example comes from a ‘mixed’ language of Arawak-Carib origin
in the Lesser Antilles which came about just before Columbus’ discovery of the West
Indies. The population which originally spoke Iñeri (a dialect of so called Island
Carib), a North Arawak language, were subdued by Carib-speaking invaders. Men
came to speak Carib, with a sprinkling of Iñeri (Arawak) grammatical forms. Women
hung on to the original language, of the Arawak family. 11

9.1 ‘Male’ and ‘female’ dialects 143
The mixture of Carib-speaking men and Arawak-speaking women brought about
a curious distinction between ‘speech of men’ and ‘speech of women’. Women used
morphemes and lexemes of Arawak origin, while men used lexical items of Carib
origin and grammatical morphemes mostly of Arawak origin. The following example
illustrates the female style of the Island Carib mixed language. 12 The morphemes of
Arawak origin are in lower case:
9.4 chile-a-tina t-one —Female style
come-perfective-1st.person 3rd.person.feminine-towards
‘I have come to her’
Examples 9.5 and 9.6 come from the speech style used by men. Lexical roots, and the
postposition ‘towards’, are in capitals: these are of Carib origin. Only the grammatical
morphemes are of Arawak origin (and in lower case):
9.5 NEMBOUI-a-tina t-IBONAM —Male style
come-perfective-1st.person 3rd.person.feminine-towards
‘I have come to her’
9.6 CHALIMAIN ba-a-(n)na —Male style
paddle 2nd.person-AUX-1st.person
‘Take me across’
The norms regulating the use of male and female register were strict: ‘it would be
ridiculous to employ the men’s language in speaking to women and vice-versa’. 13
This implies that the use of male and female speech forms could have been partially
determined by the sex of the addressee, and not just the speaker.
The two speech styles came about as a result of warfare, and coexisted with an
unusual multilingual situation. The Carib speakers were waging wars against the
Iñeri (Arawak)-speaking inhabitants of the Lesser Antilles. The Caribs won, and the
Arawak speakers lost. All able-bodied Arawak-speaking men had been killed. The
Carib men took the place of the Arawak-speaking women’s husbands. But they could
not eliminate the existing networks among women and children who continued
speaking their native Iñeri. At first, they relied on pidgin Carib to speak to their new
husbands, and fathers.
Right after the Caribs’ military victory, three linguistic codes coexisted: Iñeri, an
Arawak language, was spoken exclusively by women and children, the Carib language
by Carib men, and a Pidgin Carib by both parties. As the men began to acquire
some knowledge of the language of their wives, new-born children were growing up
with Iñeri and the Carib Pidgin—which was known to everyone but considered
‘property’ of the men. The society gradually became homogeneous, with everyone
having knowledge of both Iñeri and the Carib pidgin. This, by and by, became a
lexical register in the Iñeri language.

144 9 When women and men speak differently
Its descendant, known as Island Carib, Black Carib, or Garifuna, still survives in
Belize and Honduras in Central America. It bears numerous traces of Carib influence,
both in its lexicon and in its grammar. There is no longer a systematic
distinction between male and female speech—just a few remnants of ‘men-only’
and ‘women-only’ pairs of synonyms. The most common ones include uogori (male
speech), eieri (female speech) ‘man, male’, uori (male speech), hi̹aru (female speech)
‘woman, female’, goui (male speech), bare (female speech) ‘fishhook’ and abo ie (male
speech), bugabu ia (female speech) ‘come here!’. 14
The female dialect of Island Carib reflects the original language of the population,
and is more archaic than the male dialect—a newly formed mixed language. This
supports a frequently voiced assumption that women are more conservative than
men, in their speech patterns. Along similar lines, female forms in Karajá keep the
consonant k of a form, and are thus less innovative than male forms which omit it.
A number of male forms in Awetí appear to be derived from the female forms
(Table 9.3 shows this for forms jatã, kitã, and kujtã derived from female forms úja,
akyj, and akoj with a formative -tã: see Drude 2006: 23; 2011). In many instances,
female speech in Koasati uses older forms than male. In other languages, female
dialects are more innovative than male. Men’s forms in Yana are longer, more
elaborate, and more archaic than women’s. The ways in which male and female
dialects may have come about reflects the history of peoples and their languages of
which not much is known. 15
Phonological differences between male and female dialects can be traced to
women’s practices. Among the Suri of south-western Ethiopia, women had to have
a wooden plate inserted into their lower lip. A plate—made of wood or clay—would
vary in its diameter from one inch to six inches. Having a stretched lip sticking out
makes it difficult for women to articulate bilabial sounds which are replaced by
velars. A man would say mà, and a woman would say ŋà ‘water’. Nowadays, women’s
speech continues to follow the same, innovative, patterns even when they take out
their lip-plates. 16
9.1.4 Male and female speech on the way out
The fate of male and female speech varieties reflects the history of their speakers
in yet another, unfortunate way. Fewer and fewer people across the world continue
speaking small tribal languages. As the languages are spoken less and less, their
linguistic wealth slides into oblivion. And with language attrition, male and female
dialects are in danger.
By the time of Mary Haas’s work in the 1930s, most Muskogean languages other
than Koasati had lost the male and female speech dialects. Koasati speakers with
whom Geoffrey Kimball started working in 1977 retained just a few phonological
differences between male and female speech, including ‘the nasal coloring of men’s
speech contrasted to the clarity of women’s’. The male and female speech differences

9.1 ‘Male’ and ‘female’ dialects 145
documented a few decades earlier by Mary Haas were all but gone. 17 Some obsolescence
and mixture of male and female forms in Lakhota was signalled by Trechter
(1995, 1999). In her comprehensive grammar of a highly endangered Osage language,
the late Carolyn Quintero (2004: 12–14) could only trace some tendencies
which used to differentiate male and female speech when the language used to be
fully spoken. For instance, men would omit a declarative particle after continuative
aspect, and women would tend not to.
Yanyuwa, an Australian language with a rich system of differences for male and
female speech, is on the wane: fewer and fewer people still speak it (Bradley 1988). As
is the case in many communities across the world, Yanyuwa women are more
traditional than men, and are the ones who teach young people the language. No
wonder that ‘on rare occasions when young people attempt to speak it, they use the
women’s variety and get chastised’. The male–female speech distinction is dwindling. 18
The Chukchi language is severely endangered, and hardly being learnt by children—
this is the fate of most minorities in the Russian north. An artificial ‘standard’ language,
promoted throughout the Soviet period, was based on the men’s dialect. Chukchi
teachers and radio announcers use the male variety (female announcers continue using
the female forms in private). Discourse about language preservation is framed entirely
in terms of men’s variety.
The destruction of original cultures and traditional institutions may upset the
status quo of Social Genders, and lead to the obsolescence of male and female ways of
speaking. ‘Male’ speech used to be a sign of social malehood among the Gros Ventre.
But as the language gradually passes into oblivion, the difference between male and
female dialects is on the wane—in favour of the female variety. As Taylor (1982: 305)
puts it, ‘the cause for this is probably to be sought in the advanced state of cultural
obsolescence in which the Gros Ventre find themselves’, and ‘the destruction of the
traditional economy and the disappearance of all of the social institutions which
supported the status of males—war and native religions and their attendant ceremonies,
age-based societies, and a new way of life which tended to enhance traditional
female roles’. The drastic spread of English, and the increasing role of mothers
in boys’ socialization, further contributed to the loss of male and female dialects.
The disintegration of traditional social arrangements speeds up the loss of male
and female dialects. This is so because male and female dialects mark more than just
sex: they are tokens of Social Gender. This is what we turn to now.
9.1.5 Beyond Natural Gender
If a Kokama-Kokamilla man lapses into ‘feminine’ pronouns, everyone laughs:
mixing male and female dialects is a mark of incompetence. A Gros Ventre man
who uses a female form will be looked upon as effeminate. However, switching
dialects is always fine if a man is quoting a woman, or a woman is quoting a man. 19

146 9 When women and men speak differently
A ‘wrong’ form used on purpose may have a special effect. Kũr̩ux men can address
women ‘as if ’ they were men when they ‘treat them as equals’ (Ekka 1972: 26). On
rare occasions, Yanyuwa men and women used to exploit the male/female dialect
differences to draw attention to something unusual or unpleasant. An elderly man in
charge of a funeral ritual was not pleased with the performances, and began orating
his displeasure, in the female dialect. The inappropriate use of female speech register
highlighted the unusual state of affairs, and made people pay attention (Bradley 1988:
131–2). We can recall how Linguistic Gender can also be manipulated, to draw
attention to something out of the ordinary (§3.4).
Lakhota men use female forms to express affection—if the addressee is very much
younger than the speaker, or if the speaker and the addressee are emotionally close
(as a child and a parent would be). One man saw his nephew and said 9.7 to him
using female assertive enclitic: in Sara Trechter’s (1995: 107) words, ‘by imitating
what a woman might say...the speaker is indicating a closeness to the boy’:
9.7 Wa̹lewa̹ hiyu welê
interjection:male:surprise he:came female:assertive
‘Look who’s come!’
This is reminiscent of how men can address each other in Amharic using feminine
Linguistic Gender forms to express familiarity and affection (see §7.2).
Using male speech in Lahkota signals authority. Some men use female forms
because they feel they are not authoritative enough to speak like ‘proper men’.
Some women may use male speech, which ‘implies a certain masculinity’. A female
speaker of Lakhota ‘who used the male clitics regularly was considered by some to
be gay, referred to as a “tomboy” by one speaker, and as a “dyke” or “that woman
with balls” by another. Two women who regularly spoke like men shared several
qualities beyond their speech. They both grew up with a number of brothers and few
or no sisters; their biological mothers died when they were young, and they were in
positions of public authority: one was a judge, the other a police officer.’ 20
In the traditional Chukchi society, a person would behave as a representative of the
opposite sex as part of a shamanic inspiration—changing their Social Gender attributes,
including clothing, social behaviour, work patterns, and also language. 21
Yanyuwa men and women can use the dialect of the opposite sex in joking
situations, especially those relating to sexual encounters. If a male dancer impersonates
a woman, he may start teasing his brothers-in-law as if they were his prospective
wives, using the female dialect. We can recall a similar pattern of Linguistic Gender
switch to refer to humans in joking contexts in Machiguenga and in Manambu
(§7.1). Male and female speech are manipulated to mockingly ‘reclassify’ people, as
something overly grotesque.
The male dialect is normally a correlate of male Social Gender. Among the
Yanyuwa, boys used to be initiated into social manhood through a series of rituals

9.1 ‘Male’ and ‘female’ dialects 147
which culminated in circumcision. Only after that were they considered ‘men’ in
the social sense, and from that time onward they were expected to speak the men’s
dialect.
Male and female dialects can be switched in baby-talk. Karajá men use female
speech forms when talking to their baby daughters. Women use male forms when
talking to baby boys. 22 Early in his fieldwork with the Gros Ventre people, Alan
Taylor was puzzled as to why fluent male speakers insisted on using female forms
when teaching him the language. A breakthrough came when the speaker explained
that he was talking that way to make things ‘easier’ for Alan Taylor to learn: female
speech among the Gros Ventre was used on a par with ‘foreigner talk’. The reason for
this may have to do with child language acquisition. In all likelihood, they were first
taught the female forms, by their mothers, and only later ‘graduated’ to the male form
of speech. In Taylor’s (1982: 304) words, ‘“male” speech is reserved, as it were, for
adult male Gros Ventres, a status acquired only by birth and physical and cultural
maturity’.
The socialization process starts early. Fortune and Fortune (1975: 115) remark that
‘at about age three, mothers start insisting in the Karajá tribe that boys use only men’s
speech. This is the beginning of the socialization process which continues until the
adult male is admitted to the men’s house where no young initiate would ever use
women’s speech.’
Evidence for male and female speech as indexes of Social Gender comes from other
quarters. Traditionally, the Lakhota had at least three Social Gender categories: man,
woman, and wi̹kte (woman-potential), or a ‘would-be woman’: a biological man who
possesses the spirit of a woman (sometimes called ‘berdache’). A wi̹kte would engage
in traditional feminine activities such as beadwork, would dress as a woman, and
might marry a male. There would also be biological females engaged in traditional
male activities (such as hunting), dressed as a man, and married to a woman (referred
to as kos᷈kalaka ‘young man’). A more encompassing term for kos᷈kalaka and wi̹kte
is ‘two-spirit people’. As Trechter (1995: 7) puts it, their ‘gendered language ...in
Lakhota is appropriated from or based upon the men/women’s speech distinction’.
Male and female Social Gender and the use of male and female dialect would go
together. 23
Male speech may be a token of a formal, and more elevated, style. Sapir (1929a:
212) noted that male forms in Yana were more prestigious than ‘clipped’ female
forms. 24 An analysis, by Luthin (1991), of Yana texts collected by Sapir, showed that
the male forms were used in particularly formal and solemn situations (and not just
by men speaking to men). Male forms were used when speaking to a mixed audience.
A man would address his mother-in-law as if she were a man. The male speech forms
in Koasati may have traditionally been part and parcel of the social status of the
speaker and perhaps the stories that only men used to tell. We return to male and
female speech genres in §10.1.

148 9 When women and men speak differently
Male and female forms may be intertwined with politeness, formality, and ultimately
the identity of the speaker. Japanese and Thai, major languages of major
countries in Asia, are the prime examples.
9.1.6 Politeness, identity, and change: male and female dialects
in Japanese and Thai
In Japanese, men and women speak differently. One of the most conspicuous distinctions
lies in personal pronouns. Their choice is partly determined by how the speaker
and the addressee relate to each other. Personal pronouns are part of the well-known
Japanese honorific system which permeates the language. Three levels of politeness—
formal, plain, and informal (or deprecatory)—are not reciprocal: a socially superior
person can use either formal or informal language, and the inferior one is generally
expected to use formal language even if addressed ‘informally’.Table9.7 lists male and
female forms of first and second person pronouns in three politeness (or honorific)
levels. 25
Male and female pronouns differ in their honorific levels. The pronoun watasi ‘I’
belongs to the formal level for men, and the plain level for women. Anata ‘you’ is
formal for men, but could be plain or formal for women. Deprecatory second person
pronouns omae and kisama are not used by women. As Ide (1991: 74) puts it,
‘categorical differences in the repertoire of personal pronouns lead to women’s automatic
expression of deference and demeanor’. Female forms have a politer flavour than
do male.
Further traits contribute to this impression. In Shibatani’s (1990: 374) words, ‘a
conspicuous manifestation of politeness in women’s speech’ is excessive use of prefix
o-, an honorific marker whose meaning is to refer to something with deference. An
TABLE 9.7. Personal pronouns in Japanese: men’s and women’s speech (adapted
from Ide 1991: 73)
PRONOUNS HONORIFIC LEVEL MEN’S SPEECH WOMEN’S SPEECH
First person
Second person
formal
watakusi
watasi
watakusi/atakusi
plain boku watasi/atasi
deprecatory ore Ø
formal anata 26 anata
plain kimi/anta anata/anta
deprecatory
omae
kisama
Ø

9.1 ‘Male’ and ‘female’ dialects 149
object which belongs to a respected person would be marked with o-, e.g. sensei no
o-boosi (teacher genitive HONORIFIC-hat) ‘(respected) teacher’s hat’. Women (and not
men) would use the prefix o- to refer to their own possessions, as in watakusi no
o-saihu (I:female genitive HONORIFIC-purse) ‘my (elegant/lovely) purse’. In its female
usage, o- has gained the name of ‘beautification’,or‘elegant’, prefix. The use of o- and
other honorifics in popular magazines for housewives in Early Modern Japan helped
underscore ‘the stances of elegance and grace, both strongly associated with good
demeanour and femininity’, and taught the reader how to express these virtues, as
Nerida Jarkey (2015: 197) puts it.
Throughout the language, women use more honorific forms than men. Men, but not
women, use words and forms with derogatory overtones. A man, but not a woman, will
say dekee, a derogatory form of dekai ‘big, monstrously huge’. Such ‘male-only’ features
acquire the value of what Peter Trudgill (1975: 102) called ‘covert prestige’: they are
tokens of a speaker’s masculinity and an ‘uninhibited quality’.
Sentence-final particles also reflect the speaker’s sex. The final particle wa has a
broad meaning of softening a statement, or giving the addressee an option to
disagree, or creating empathy. It is exclusively female. Particles ze and zo are only
found in men’s rough speech. A woman, but not a man, will omit the copula da when
it is followed by assertion particle yo. A man will say Kirei da yo (pretty COPULA
PARTICLE), and a woman would say Kirei yo (pretty PARTICLE) ‘It is pretty’. 27
A conventional perception of women in many cultures is that they speak more
politely than men. In Japanese this is a rule, and not just a desirable tendency. Katsue
Akiba Reynolds (1991: 133) comments:
As a woman, I am not allowed to say to anybody, even to my younger sibling, Tot-te-kure ‘Get
(it) for me’, using the Informal-Benefactive-Imperative. I have to say instead Tot-te. [...]Iam
not allowed to say It-ta ka ‘Did you go?’. I must suppress the interrogative marker ka and say
It-ta [...] shifting the rising intonation to the tense marker. Some rules for gender-marking in
Japanese are categorical, while in English rules are variable.
Only under special circumstances can a man use female forms. A salesman who
deals primarily with female customers may use the ‘elegant’ prefix o- ‘as if ’ he were a
woman (Shibatani 1990: 374).
The essence of femininity—as reflected in the traditional norms of Japanese—is
softness, politeness, and deference. A female speaker of Japanese is supposed to talk
unassertively, politely, deferentially—these are the positive attributes of a proper
woman’s behaviour, called onna-rasiku ‘as expected of women’. This behaviour fitted
in well with the Confucian doctrine of ‘men superior, women inferior’—the backbone
of the hierarchical structure of Japanese society, where women were wives and
mothers, or, in Pfarr’s(1984: 224) words, ‘the lowest rung on the entire social ladder,
subordinated within the feudal hierarchy and within the family hierarchy as well’.
When women and men were given equal rights by the new constitution after the end

150 9 When women and men speak differently
of the Second World War, the society transformed. Women were no longer confined
to home. Women as supervisors, administrators, teachers, lawyers, doctors, or
colleagues could now—in theory at least—become equal, or superior to men within
the work environment.
As social roles changed, so did the language. 28 Some women started ‘defeminizing’
their speech, by—perhaps subconsciously—using male forms. But old habits and
traditions die hard. Reynolds (1991: 138–40) describes the difficulties for female
teachers who had to address mixed classes of boys and girls. A 26-year-old female
principal lamented in an interview with a newspaper reporter (the Asahi, 18 July
1984), ‘I tend to speak rough language with an imperative tone in spite of my efforts
not to, perhaps, because I am a teacher. I always think regretfully that this is not good
for me.’ The polite, submissive, and differential overtones of female speech in
Japanese seem hardly compatible with a position of authority. A female principal of
a school or a college cannot display many of the linguistic traits typical of women—
she ‘must avoid giving an impression in her talk that she is indecisive, indirect or
picky. At the same time she cannot be as authoritative as a male counterpart would
be’: this might not be seen as a good example by traditionally minded parents. Every
woman has to face the problem of how feminine she wants to sound according to
her own personality, her role and relationships, and the situation, and make her own
decision. 29
Overall, the strict divide between male and female language appears to be dwindling.
In the late 1970s, junior high school girls in Tokyo were referring to themselves
as boku ‘I:male’ in the school environment. When interviewed, they explained that
they couldn’t compete with boys in classes, games, and fights if they were to call
themselves watasi ‘I:female’. Ryooko Ozawa, a woman politician elected as a city
council member of a large satellite city of Tokyo in 1971, dismissed female speech as
‘a product of the idle life of housewives’. In her own position, when urgent decisions
are to be made, ‘she notes that her own speech tends to become more masculine as
she tries to convey her ideas as precisely and emphatically as possible’. And this is
indeed the impression her way of speaking conveys (Reynolds 1985: 41–2).
In her study of a ‘less-feminine speech’ among younger Japanese women,
Okamoto (1995) shows that younger women (between 18 and 34 years of age) no
longer speak the way textbook principles of Japanese ‘female speech’ tell them to.
Female college students, professional, and self-employed women are no longer averse
to using male (not female) sentence endings, and even expressions commonly
perceived as strongly masculine, or even vulgar—aitsu ‘that guy’, bakayaroo ‘stupid’,
yabai ‘troublesome’, and dekai ‘big, humongous’. Homemakers lag behind, using
forms traditionally classed as ‘female’. Older men lament the changes in female
language—calling it ‘tasteless Japanese’ (the title of Okamoto’s paper).
Changes have gone the other way round, too. Younger men in the 1970s were not
averse to using female sentence endings and subordinators. And young women at

9.1 ‘Male’ and ‘female’ dialects 151
that time were freely using the copula da followed by the particle yo—a traditional
feature of male speech.
Modern Japanese women speak differently depending on their social status,
workplace roles and relationships, surroundings, and the impression they want to
convey. Their self-image and their identity as social beings come through in the ways
they speak. Male and female dialects are indicative of Social Gender, and are
influenced by it. In Reynolds’s (1991: 141) words, ‘the female/male language dichotomy
in Japanese is not a mere differentiation of the two sexes but it reflects the
structure of a society where women were defined as the inferior sex’. Growing
equality between male and female Social Genders is enhancing linguistic equality.
The label ‘Japanese women’s language’ no longer fully reflects the exact ways in
which modern women speak the language. As Okamoto (1995: 317) puts it, ‘Japanese
women strategically use particular speech styles to communicate desired pragmatic
meanings and the images of self.’ And some female forms are penetrating the speech
of younger males, as men ‘continue to adapt to the ongoing changes in Japanese
gender roles and gender ideologies’ (319). Language use is gradually shaped by social
change—we return to this in Chapter 11. 30
Different forms for ‘I’—dichán for women, and phŏm for men—are a notable
feature of female and male speech in Thai, a Tai-Kadai language with a large number
of honorific registers. 31 The pronoun dichán was originally used by royal and highclass
males; it has become restricted to women since the early twentieth century.
Phŏm is a high speech level form used by men when talking to someone older or
higher in status than the speaker, or as a default term, with no honorific implications.
The female form dichán carries overtones of a social distance between the speaker
and the addressee: it is avoided in informal situations. The way urban professional
women are using dichán is gradually becoming parallel to the way men in general use
phŏm to refer to themselves. Social changes affect other means of self-reference.
Traditionally, women would use kinship terms or names to refer to themselves in a
womanly, submissive, and humble way. Now men are also using nicknames to refer
to themselves—something only children and women would do in earlier times. But
changes proceed slowly. If a woman uses dichán to refer to herself in a conversation
with senior colleagues, especially females, she may be classed as ‘arrogant’. A young
woman new in her high-ranking position may find it hard to always refer to herself as
dichán. But if she does not, she may get criticized.
Yingluck Shinawatra, Thailand’s first female Prime Minister, referred to herself
with her nickname pu: rather than with dichán ‘I’, judged more appropriate in a
public speech. She was under fire in the news media—some accused her of mixing up
private life with state affairs, that is, representing herself not as the government leader
but as a ‘normal’ citizen and a woman.
Social changes in women’s place in society affect male and female dialectal form, as
tokens of changes in Social Gender. We now turn to gender-variable ways of speaking.

152 9 When women and men speak differently
9.2 Gender-variable skills: men’s and women’s speech practices
Every language has a pool of linguistic devices which reflect the speaker’s position in
a society (in terms of class, wealth, or education), their attitudes, and the way they
want to be perceived by others. In every society, certain ways of speaking come to be
associated with social status, economic conditions, employment, and educational
opportunities. Each of these may be linked to male and female practices, and ways of
communication. Linguistic forms in gender-variable languages have sets of pragmatic
overtones which can be deployed by men and women depending on the image
they want to construct, inseparable from their societal status. In Sara Mills’s (2003a:
239) words, being male or female ‘cannot simply be correlated with the use of
particular linguistic forms or strategies’—in contrast to the gender-exclusive languages
we have just discussed.
In a classic paper, Elinor Ochs (1992: 342) makes a similar point, especially
important for gender-variable languages:
Knowledge of how language relates to gender is not a catalogue of correlations between
particular linguistic forms and sex of speakers, referents, addressees and the like. Rather,
such knowledge entails tacit understanding of 1) how particular linguistic forms can be used
to perform particular pragmatic work (such as conveying stance and social action) and 2)
norms, preferences and expectations regarding the distribution of this work vis à vis particular
social identities of speakers, referents and addressees.
‘Gender-variable’ ways are particularly visible in traditional societies where women’s
and men’s spheres of labour, socialization, and concomitant stereotypes are
relatively clear-cut.
9.2.1 Social status, and women’s speech
In traditional societies, the way women choose to speak may correlate with their roles
and position. In the Mayan-speaking community of Tenejapa, women display extremes
of positive and negative politeness repertoire in their relationships with men, and with
other women. They use negative questions as commands and offers, sounding mild and
pleading, and frequently stress agreement with the addressee. They employ diminutives
and in-group address forms, expressions like ‘you know’ (yaʔwaʔy) and‘you see’
(yaʔwil), and formulate their requests and offers in a subdued, non-imposing manner.
A woman would inquire in a mild, roundabout manner, ‘You wouldn’t haveany
chickens to sell?’ rather than commanding someone to sell her the chickens. Women’s
way of talking is more deferential. Men’s speechismore‘matter of fact’. One gets the
impression that women are more sensitive than men to what they are saying and modify
their speech accordingly.
Three ethnographic facts provide a reason. First, women are vulnerable to men in a
society where women are likely to be beaten if there is any threat to their reputation,

9.2 Gender-variable skills 153
and women are vulnerable to women as possible sources of damage to their reputation.
Secondly, women may have to be polite to other women in a household because
women traditionally move to live with their husbands, and there is a greater social
distance between women than between men within one family. Thirdly, women
speak more cautiously than men: some patterns of behaviour such as talking to
unrelated males are highly face-threatening. This motivates the choice of some
speech strategies—for instance, the ubiquitous expression denying knowledge or
responsibility, mas᷈kil ‘I don’t know’. In addition, men purposefully exaggerate their
brusqueness and matter-of-factness as a display of tough masculinity, and women
elaborate their ‘polite graciousness as a display of feminine (contrasting to masculine)
values’ (Brown 1980: 131).
Women’s careful, polite, and circumspect discourse in Tenejapa does not mean
that they are entirely powerless. In fact, women make a considerable contribution to
the household, and men depend on them in domestic contexts. Women are seen as
‘indispensable to the order of things’, and ‘in maintaining and guarding society in a
role parallel to men’. The Tenejapan culture is essentially egalitarian, and downplays
all differences in status and power. But a status difference between men and women
remains: husbands directly command wives to do something for them using simple
imperatives. Women do not. While women accord men (especially unrelated males)
‘marked interactional deference, the reverse is not the case. Politically—men hold the
positions that are prestigious and publicly visible, and it is men who make the
decisions affecting the community as a whole. Women’s role in decision making,
while very important domestically, is from a society’s view more or less invisible.’ 32
Malagasy men avoid putting someone else ‘in an uncomfortable or unpleasant
position’, and show ‘respect by avoiding this type of confrontation’. It is men who
tend not to express their sentiments openly, and avoid creating unpleasant face-toface
encounters. Women are the opposite—they tend to speak in a more straightforward
manner, and ‘are associated with the direct and open expression of anger
towards others’, very much unlike Tenejapan women. Again, this difference has an
ethnographic explanation.
In Malagasy, indirectness is a feature of skilful speech associated with men. Directness
and explicitness—a feature of unsophisticated speech—is characteristic of women
(Keenan 1974: 143). This goes with men—not women—being seen as keepers of
proper traditional ways, and women as bringing in destructive innovations. Table 9.8
summarizes the principles of female and male speech use among the Malagasy.
Malagasy women are excluded from the oratory, the major formal speech. Men
shy away from the type of talk associated with women—gossip and accusations. Male
speech is stereotyped as indirect or respectful, and women’s speech as direct or
confrontative. And both men and women agree that men’s speech is superior to
women’s. Even the historical records of the Malagasy language contained warnings
against the potential ‘evil’ of women’s tongues. 33 Elaborate verbal skills are associated

154 9 When women and men speak differently
TABLE 9.8. Traits of men’s and women’s speech among the Malagasy
INDIRECTNESS
MEN
Skilled speech
Traditional speech ways
Malagasy language
DIRECTNESS
WOMEN
Unsophisticated speech
Contemporary speech ways
European languages
with authority and status. The surface difference between men and women has to do
with difference in status, and speech genres associated with each Social Gender. 34
Habits of pronunciation and ways of talking can come to project an image of an
affectionate wife or mother, a submissive and humble being, or a bossy and commanding
man. Features associated with different images create a linguistic stereotype
of a Social Gender. As Cameron (1995: 49) put it, ‘feminine’ and ‘masculine’‘are not
what we ARE, nor traits we have, but effects we produce by particular things we DO’.
We now turn to the resources which help create such stereotypes.
9.2.2 Tokens of men and women in gender-variable languages
Women and men can sound differently due to inherent differences in their physique,
or Natural Gender. As Labov (1972: 303–4) puts it in his classic study of sociolinguistic
variation in the Martha Vineyard variety of American English, ‘there are of
course physical differences between the vocal tracts of men and women to be taken
into account’, but there is enough evidence to believe that ‘the sexual differentiation
of speakers is...not a product of physical factors alone...but rather an expressive
posture which is socially more appropriate for one sex or the other...’.
Women’s voices are higher in pitch than men’s. This feature can be deployed by
women, for a special purpose. A Tenejapan Mayan woman would use exaggerated
falsetto voice to express deference when speaking to her husband and other women,
projecting an image of a circumspect and submissive being compatible with women’s
behaviour in the community. A wider range of pitch differences available to women
is deployed for special effects. Russian-speaking women are reported to use high
pitch for positive feelings, and low pitch for the opposite. In 9.8, the contrast between
the speaker’s delight in finding a pretty mushroom and her disgust on looking at it
more closely (and finding that it is poisonous) is partly encoded in pitch level:
9.8 {Oj kakoj xoros᷈en’kij griboc᷈ek} HIGH PITCH
oh what.masc.sg pretty.masc.sg mushroom.dim.sg
{Fu kakaja gadost’!} LOW PITCH
Yuck what.fem.sg nastiness
‘Oh what a pretty little mushroom! Yuck what a nasty thing!’

9.2 Gender-variable skills 155
Women’s vocal tracts are shorter than men’s. Pulmonic ingressive airstream is easier
to achieve with the smaller female pharynx and larynx. Female speakers of Tohono
O'odham deploy this device in discourse when they take turns in conversation,
clarifications, and repetition, to construct an atmosphere of conversational intimacy. 35
Male speech can sound more harsh and ‘close-mouthed’ than female speech—
contributing to a conventional image of men being tougher and more brusque. On
Martha’s Vineyard, men are more ‘close-mouthed’ than women, and use more
contracted areas of phonological space; conversely, women in New York City and
Philadelphia use wider ranges of phonological space than men with more extreme
lip-spreading, for vowels. In Japanese, vowels sound less harsh in women’s speech
than when pronounced by men. 36
Phonetic differences between men and women are often closely linked to the level
of education and socio-economic level, and sensitivity to the ‘norm’. The pronunciation
of vowels and consonants in Norwich, by men and women, varies depending
on their social background. Trudgill (1974) analysed pronunciation of the velar nasal
[ŋ], and concluded that women tend to use a more prestigious and standard [ŋ] no
matter what their background was. Men—especially, but not exclusively working
class—used the stigmatized [n] more often. In other varieties of English, too, women,
not men, tend to use more prestigious varieties and stick to the linguistic norm. In
Glasgow speech, the glottal stop is the most overtly stigmatized feature—especially
frequent in the speech of working-class groups. Lower-middle-class women used
40 per cent fewer glottal stops than men from the same class. Among the people
interviewed by Macaulay (1977, 1978), a lower-middle-class woman used fewer glottal
stops than any upper-middle-class informant.
Adherence to norm, clearer pronunciation, and prestigious, rather than stigmatized
forms, is a common feature of women’s speech. 37 Is this because women are
traditionally more involved in rearing children and thus reproducing the society
norms? Or that they feel too insecure to deviate from what is accepted and what is
prestigious and eager to maintain the status quo? We return to this in §10.4.
Many scholars have described women’s speech in European languages as more
tentative and emotional than men’s. Women who speak Russian and Brazilian
Portuguese use affectionate diminutives—referring to an object as small and cute
and lovely. A Russian woman will tell a child to dry his ‘little face’ (lic᷈iko, diminutive
of litco ‘face’). 38 Jespersen (1922: 249) quotes Lord Chesterfield’s (The World, 5
December 1754) comments on how women ‘take a word and change it, like a guinea
into shillings for pocket money, to be employed in the several occasional purposes of
the day. For instance, the adjective vast and its adverb vastly mean anything...A fine
woman...is vastly obliged,orvastly offended, vastly glad,orvastly sorry. Large objects
are vastly great, small ones are vastly little.’ According to Jespersen, ‘there is no doubt
that he has here touched on a distinctive trait: the fondness of women for hyperbole
will very often lead the fashion with regard to adverbs of intensity’.

156 9 When women and men speak differently
Excessive use of interjections and emotionally expressive words is often relegated
to the domain of ‘female’ language. Yokoyama (1999: 408) notes that ‘Russian men
do not verbalize emotions that are “unworthy of men”, such as fear, pity, or worry.
Women, on the contrary, do not hesitate to voice them, and they even tend to choose
exaggerated expressions when they do so.’ Only a woman would say Ja z᷈utko
perez᷈ivala (I awfully suffer, worry, go.crazy about+past.fem.sg) ‘I was awfully worried’.
Women are said to swear less than men, and use fewer taboo words. But none
of these features has proved to be exclusively female, on a closer look.
In the English-speaking milieu, men are more likely to interrupt women in
conversations, and to try and ‘silence’ women in online and other debates. Women’s
response to male dominance in different sex-interaction is face-saving and polite—
characterized by hedges, tentative statements, and affective tags. In same sex talk, the
general tendency for women is to be ‘cooperative’, while men-to-men talk tends to be
‘competitive’. Women’s ways of talking are generally perceived as mild, tentative, and
oriented towards cooperation with the addressee, rather than with a ‘man-like’ selfassertion.
Robin Lakoff, a founder of studies of women’s language, put the relative ‘tentativeness’
of women’s speech on the map, as a reflection of their ‘powerless’ talk
(1975). The powerless talk would include intensifiers (‘that’s a really good idea’),
hedges and qualifiers (‘that’s sort of good’), excessive politeness and tag questions
(‘that’s a good idea, isn’t it?’). 39 In Lakoff ’s words, women ‘are socialised to believe
that asserting themselves strongly isn’t nice and ladylike, or even feminine’.
In reality, things are more complex. ‘Powerless’, submissive, and cooperative ways
of speaking are not exclusively associated with women. 40 For one thing, hedges,
qualifiers, and tag questions have many meanings and many functions. Their use is
not a prerogative of either Natural or Social Gender. A study by Holmes (1986)
showed that women tend to use a hedge you know more often than men if the hedge
expresses speaker’s certainty, as in 9.9:
9.9 and that way we’d get rid of exploitation of man by man all that stuff/you know/
you’ve heard it before.
The same you know can express uncertainty—and this is the meaning in which
women use it less than men. Coates (1993: 116–17) shows that women would use
hedges rather frequently when they speak to each other, to make sure the conversation
does not become too face-threatening. The use of a linguistic form depends on
the speaker and the addressee, and their relationships, rather than just the gender.
Hedges and intensifiers are not uniformly ‘male’ or ‘female’. Bradac, Mulac, and
Thompson (1995) interviewed 58 male and 58 female students of the University of
California (aged between 18 and 25). It turned out that hedges—especially kind of and
fairly—were dominant in male speech. Women did use really more often than men;
other intensifiers like pretty much were used by both groups. A general stereotyped

9.2 Gender-variable skills 157
perception of women’s speech being milder and replete with unnecessary exaggeration
and signs of doubt turned out not to be true to life, in one university context.
Tag questions in English are equally problematic. By using a tag question (e.g. you
are coming, aren’t you?), a speaker avoids committing themself and thus may also
give the impression of not being sure of what they are saying. In female speech, a tag
may express a supportive attitude towards the addressee. A statement that women
use tags more often than men under any circumstances has not been fully supported
by hands-on studies. 41
Commands and other directives by women appear to be more tentative than those
by men when women step into social roles traditionally reserved for men. In English,
male doctors will be more likely to use imperatives. Female doctors will use milder
and more roundabout command strategies. A man would say Have my suit cleaned!
and a woman would use longer and politer ways of expression, Would you drop this
by the cleaners on your way? 42 In Russian, an abrupt command in an infinitive
form—a feature of the army language—appears to be more common in male than
in female speech.
The ways men and women talk, and the forms they use, may depend on who they
talk to. Among the Manambu, women—and not men—are involved in child-rearing.
In telling children what to do, women use strong prohibitives and imperative forms.
A dependent clause can be used on its own to express a particularly strong command
which would demand immediate compliance, e.g. Təkərəm da-ku! (chair+LOC sit-
SUBORDINATE.SAME.SUBJECT) ‘Sit on the chair (or else)’ (literally, having sat on the
chair!). Men hardly ever use such forms, and comment that this is the way women talk.
Manambu women sound more tentative than men under other circumstances.
Traditional oratories are the domain of men, and women are less accustomed to
public speaking. If a woman tells a story, she typically would start it with a verbal
form Wau? (speak+1st.person.command), literally ‘let me talk, may I talk’ (with a
rising intonation typical of a polar question). The same form is used in turn-taking in
conversation. Men who are not accustomed to speak at public gatherings do the
same. The man mentioned at 7.3 in §7.2 had violated patrilocal practices, and was
considered something of an outcast with little traditional knowledge or oratory
experience. When he was asked to tell a story of his village, he started it with Wau?
A further feature of Manambu women’s talk points in a similar direction. In the
stories and conversations I collected, the modal adverb waiwai ‘maybe’ was used
about twice as often by women as by men, to mitigate their statements. In many
cases, women sound more realistic than men. A man would say Mival yakna ‘plane
will come’. A woman (or an insecure man who is a poor orator) would qualify that by
using waiwai ‘perhaps’. Manambu men tend to flaunt their knowledge venturing
unqualified statements even if they do not have enough knowledge to support it. The
archetypal difference between men and women using Manambu lies in the images
they automatically project, through words they use.

158 9 When women and men speak differently
Differences between male and female ways of talking go back to the way children
are brought up. The most salient difference between Kaluli girls and boys is to do
with language: girls consistently employ the direct instruction routine, using ɛlɛma
‘say like that’, to their mothers and especially to younger children. They replicate the
ways in which they used to be instructed—as Bambi Schieffelin (1987: 259) puts it,
‘girls engage in verbal routines that are considered within the women’s domain and
boys do not’.
What can be perceived as non-aggressive, reconciliatory, cooperative ‘female’ speech
is also used by those who are on their defensive—for instance, men and women in a
courtroom environment. Linguistic forms and ways of speaking reflect social status
and the ways people—men or women—want to be viewed in different situations. If a
woman in a domestic context uses tag questions to facilitate conversation, she would
create an image of someone deferential or subordinate. And if a woman doctor uses the
same strategy in her consulting rooms, she will be seen as an effective and competent
professional. 43
The differences between what is perceived as male or female language in modern
urban and socially fluid societies in America, Europe, and Australia are best interpreted
as ‘surface reflections of such basic social dimensions as power and solidarity’
(Sherzer 1987: 117). The patterns of Social Gender created on the basis of other social
parameters merge and overlap.
When speakers change their Natural and Social Gender, does their language
change? This is what we turn to next.
9.3 The other genders
Gay, lesbians, and transgender people deploy linguistic features to construe their
male or female identity, and create an image of a female or a male. These traits reflect
the stereotypes of male and female speech in a society.
Phonetic features which are perceived, by listeners, as characteristic of ‘gay men’s’
speech in English include higher pitch—a suprasegmental feature associated with
female speech. Other ‘gay men’s’ features are hyperstandard pronunciation such as
longer duration of sibilant fricatives, or ‘hissy’ s and z, and the release of word-final
stops. In contrast, lesbian speech includes a narrow, male-like, pitch-range and flat
intonation patterns. 44
Standard Brazilian Portuguese requires number agreement on nouns in combination
with a definite article, e.g. a-s coisa-s (the.feminine-plural thing-plural) ‘the
things’. In the informal spoken language, the plural marker on the noun is often
omitted—one hears as coisa ‘the things’. These forms are considered substandard.
Gay men and straight women are less likely to delete the plural marker than straight
men. This is why gay male speakers of Portuguese of São Paulo in Brazil are said to

9.3 The other genders 159
speak a ‘better Portuguese’. 45 This agrees with the idea that women tend to adhere to
more standard varieties than men—we turn to this in §10.4.
Extensive use of diminutives in Brazilian Portuguese is a feature of straight
women, and also of feminine gay men and non-butch lesbians. In contrast, straight
men, masculine gay men, and butch lesbians avoid diminutives, to assert their ‘male’
status. Gay men in Tokyo and Osaka use ‘female’ language patterns to assert their
identity. 46 Transgender male to female speakers of Thai have a broader repertoire of
gendered pronouns than either men or women. They tend to use female-specific
pronouns, and final particles. A sentence final particle hâʔ used to be a feature of
female speech; nowadays it ‘seems to be reserved for feminine males’. 47
Resources of Linguistic Gender can also be deployed. Brigitte Martel described her
own experience as a transsexual male who changed to being female in the 1981
autobiography, with a telling title Né homme, comment je suis devenue femme (Born:
masculine.singular man, how I became (feminine.singular) a woman). French gay
men use feminine forms to address each other and talk to their partners, addressing
them as ma petit-e chéri-e (my.fem.sg little-fem.sg darling-fem.sg) ‘my little darling’.
Lesbians will use masculine forms, to project the image of masculinity. 48 The use of
Linguistic Gender by American gay men has affective overtones: Rudes and Healey
(1979) found that gay men used ‘she’ to refer to someone particularly attractive; ‘he’
was associated with negative attitudes.
Literary fiction reflects stereotypes associated with butch and femme lesbians as
‘manly’ and ‘womanly’ in the ways characters talk. Butch lesbians are portrayed as
curt, avoiding expressions of emotion, using sexually explicit vocabulary. Femme
lesbians are portrayed as more ‘lady-like’. All these features do not necessarily reflect
what gay men and lesbian women do. Rather, they help create recognizable stereotypes
which may then be deployed by gays and lesbians to create a male-like, or a
female-like image in the eyes and ears of the listener. 49
We can recall, from §9.1.5, that ‘two-spirit’, or transgender people, among the
Lakhota used to take on the dialect of their newly acquired Social Gender. Fletcher
and La Fleche (1972: 132) tell a story of an Osage man who became like a woman and
started ‘unconsciously using the feminine terminal of speech’. 50
The hijras of India are perhaps the best described instance of a ‘third sex’. 51 They
are born as males and raised as boys. Their female predilections appear later in life.
Hijras—also referred to as ‘hermaphrodites’ or ‘transvestites’—have a special position
in the social structure, as the ‘third gender’. Their thick male voices are in stark
contrast to their female clothing and woman-like looks and gait.
Hijras deploy the resources of Linguistic Gender to play out their Social Gender
identity and interpersonal relations. Hindi-speaking hijras refer to themselves as
women would—using the feminine (and not the masculine) Linguistic Gender in
pronouns, adjectives, and habitual, progressive, and other verb forms. When a hijra
decides to dress up as a man, they switch to masculine Linguistic Gender. The use of

160 9 When women and men speak differently
feminine Linguistic Gender to talk about themselves is tantamount to the projection
of ‘non-masculine’ identity. Some use masculine forms when they talk to men, and
feminine forms when they talk to women. For a hijra, using a feminine form to
another hijra may be symbolic of solidarity and familiarity. Masculine forms can
mark respect, if addressed to a man. But using a male name, or a masculine form, by a
hijra to a hijra reflects contempt, and is insulting. Hijras manipulate masculine and
feminine Linguistic Gender forms to establish solidarity, power, and social relations
with their peers and the outside world. Linguistic Gender helps construct images and
identities of Social Gender (in the context of partially ambiguous Natural Gender,
and the in-between status of the third ‘sex’).
Male-to-female transsexuals and transvestites in the English-speaking world try
and sound like women—they raise the pitch of their voices and use a wide range of
intonation patterns. They use emotionally coloured words the way women are
perceived to use them, and adopt female names. Transvestites from southern Brazil
use the feminine Linguistic Gender when they talk to each other, and refer to
themselves as bicha (a grammatically feminine noun, roughly translatable as ‘fag’)
rather than travesti ‘transvestite’, a grammatically masculine noun. They would
employ masculine gender to refer to themselves at the time their body transformation
took place and when talking about their family relationships. A transvestite
explained to the researchers that for her mother, she will always remain a male son,
filho homem. Other transvestites they did not identify with were talked about as male
travestis. Masculine and feminine gender forms are manipulated to create an image
of a Social Gender. 52 The French male-to-female transsexual Georgine Noël switches
between masculine and feminine Linguistic Gender throughout her biography.
According to Livia (1997), switches from masculine to feminine convey a sense of
success and triumph in Noël’s newly acquired female self. Her switches back to the
masculine form deploy frustration and failure, in her male body. Linguistic Gender
has acquired a symbolic value of identity mark.
9.4 Male speech, female speech: a summary
In every language and in every society, Social Gender correlates with how men and
women use language. A few languages have, or used to have, ‘gender-exclusive’ male
and female dialects. Men and women would use different phonological systems (as in
Karajá, from Brazil), or even resources from different languages (as in Island Carib).
Male and female dialects would employ different forms of pronouns, various particles,
and other grammatical forms, as in Yanyuwa (from Australia), Kokama-
Kokamilla (from Peru), Japanese, and Thai. The choice of male or female dialect
may depend on whether the speaker, and the addressee, are male or female, as in
Kũrˌux, a Dravidian language, and the now extinct Biloxi and Tunica, from the
USA. The use of ‘gender-exclusive’ dialects could go beyond Natural Gender into

9.4 Male speech, female speech 161
the domain of Social Gender: the male variety will be the way an initiated man would
speak. And a non-initiated man—incomplete in his manhood—will speak ‘like a
woman’. Male and female gender dialects could be manipulated to achieve special
effects—including jokes. This is similar to how Linguistic Genders can be reversed,
for special purposes. As many minority languages and their social structures disappear,
so do the male and female dialects.
Social changes in the status of men and women affect the ways in which male and
female dialects are used. With the rising equality of men and women in modern
Japan, women no longer use the female forms as the textbooks say. To assert
themselves, many employ what used to be male patterns. As Okamoto (1995: 309)
puts it, ‘today’s notion of Japanese women’s language can be seen as a lasting legacy’
of the late nineteenth-century policy when ‘government officials and intellectuals
sought to standardize the language and to discipline women according to the ideal of
ryoosai kenbo “good wife and wise mother”’, based on the speech style of traditional
women in the middle and upper-middle classes in Tokyo. Alongside other symbols—
clothing, bearing, and so on—this ideal was constructed, and it is now changing as
the society and women’s roles transform. Thai-speaking women have started using
pronouns in the same way as men do. Women, and their verbal expression, are a
barometer of change—we return to this in Chapter 11.
Subtle differences between male and female speech in gender-variable
languages—including English—are more difficult to capture. There is no oneto-one
correspondence between a grammatical or lexical feature and women’s or
men’s speech. On the contrary: most features are deployed by women, and by men,
to create an impression and construct social relations (of which Social Gender
relations are only one part). The resources of Natural Gender—such as higher pitch
of women’s voices—can be deployed to express affection or delight. Subdued,
tentative, emotionally charged, affectionate, and hypercorrect speech underlies
a ‘women’s language’ stereotype (see Lakoff 1979: 53). Salient features of this
stereotype—in combination with Linguistic Gender if available—can be manipulated
to assert Social Gender identity in transgender, gay, and lesbian way of
speaking across the world. In Livia’s (1997: 365) words, transgender people reveal
‘resources available in the gender system to which more traditional identities have
scant recourse’. 53 Building one’s Social Gender identity and constructing one’s
Social Gender underlies Linguistic Genders and other speech distinctions identified
as ‘male’ and ‘female’.
Men’s and women’s activities are clearly and distinctly defined in traditional
societies across Africa, New Guinea, and South America. Then, differences in the
verbal genres and associated patterns of speaking clearly reflect the roles, and thus the
achieved status, of men and of women. Differences between men and women lie not
just in how they speak, but in what they talk about, and what is off limits to members
of either Social Gender. This is what Chapter 10 is about.

162 9 When women and men speak differently
NOTES AND SOURCES
1. Competent surveys of male and female dialects are in Bodine (1975), Key (1975),
Günthner (1996), and Trechter (2009); Mithun (1999: 276–80) is a survey of male and
female dialects in North American Indian languages; a brief survey of Amazonian languages
is in Aikhenvald (2012a). Storch (2014: 252–3) discusses male and female dialects
in East African languages. There are many misconceptions about male and female dialects.
Dunn (2014) is a highly flawed and sketchy outline of ‘male’ and ‘female’ dialects, with
numerous mistakes and a peculiar claim that male and female dialects are ‘only attested in
relatively small communities’ (p. 63), backgrounding Japanese and Thai spoken by
millions of people. Male and female dialects are sometimes referred to as ‘genderlects’.
2. Rodrigues (1999: 177), Ribeiro (2009, 2012); Fortune and Fortune (1975, 1987).
3. Flannery (1946: 133–4). What Flannery referred to as tc [ts] was described as c᷈[tʃ] in later
work by Taylor (1982: 101–2).
4. See Bogoras (1922: 665). V. G. (Waldemar) Bogoras was exiled to the Russian north for his
revolutionary activities, and spent a substantial amount of time there. Dunn (2000) isa
brief description of what the author, as a PhD student, could capture of men’s and
women’s dialects during his brief fieldwork in the 1990s.
5. Similarly to Lakhota as described in detail by Sara Trechter, Omaha-Ponca, Quapaw, and
Kansa, extinct members of the same Siouan family, had male and female forms for
declarative and also imperative markers: Trechter (1995: 12); John Koontz, p.c. Differences
between male and female speech have been described for Mayrinax, the most conservative
dialect of Atayal, an Austronesian language of Taiwan: Li (1982, 1983); see Aalto (1959),
Birtalan (2003) for Kalmyk and Oirat (Mongolic), Rankin (1976) for Moldavian; Camden
(1979: 113) for Tangoa; Hollow (1970: 456) and Mithun (1999: 280) for Mandan, and
LeMaster (1997) for the Irish Sign Language.
6. Vallejos (2010: 41–4; 2015), Faust (1963); see also Kroskrity (1983).
7. See Drude (2002, 2006, 2011) on Awetí; Stradelli (1929) for the Língua Geral form; Drude
(2006) for the relationships within Tupí-Guaraní family and the status of Awetí.
8. Kirton (1988: 116).
9. See Haas (1944) for a typology of men’s and women’s speech; Haas (1944: 148–9) on
Koasati, Tunica, and Biloxi; Einaudi (1976: 78–96), Mithun (1999: 280) on Biloxi; Ekka
(1972) onKũr̩ux; Sapir (1929a: 206) on Yana.
10. See Ott and Ott (1983: 26–7), Olza Zubiri, Nuni de Chapi, and Tube (2001: 10); Rose
(2013) discusses only some forms within personal pronominal reference.
11. See Taylor (1977: 89–99); Hoff (1994).
12. Hoff (1994: 162–3).
13. See Hoff (1994: 164), Breton (1978: 16–17, 132), Du Puis (1652: 195–6) on the male and
female registers.
14. Douglas Taylor, a major expert in Island Carib, reported a curious switch in Linguistic
Gender of some nouns depending on the sex of the speaker. The term for an ancestral
ritual dogó belongs to the masculine Linguistic Gender if a woman is speaking, and the
feminine gender if a man is speaking (Taylor 1951: 44–5, 1959: 204; 1977: 24–43, 72–99).
See also Rat (1898: 311). Note that a partial study based on the expatriate community in
the USA by Munro (1997) is inconclusive.
15. See also Rankin (1976) on archaic features of female speech in Moldavian and Drude
(2002) on archaic forms in Awetí female speech. Dunn (2000) argues that Chukchi
women’s dialect may have arisen as a result of conscious dialect mixing.

9.4 Male speech, female speech 163
16. Yigezu (1998: 97–9).
17. Kimball (1991: 54, 224–5).
18. Speakers of Kokama-Kokamilla view the correct use of genderlects as a marker of linguistic
proficiency (which semi-speakers lack: Vallejos forthcoming).
19. See Flannery (1946) on Gros Ventre, Bradley (1988) on Yanyuwa, Vallejos (2010, 2015) and
also Faust (1963) on Kokama-Kokamilla, Sims and Valiquette (1990) and Maring (1975)on
Keresan, Mithun (1999: 278–8) and Günthner (1996: 453), for a general discussion.
20. Trechter (1995: 106–13).
21. See Bogoras (1901: 98–9). Bradley (1988) reports how Yanyuwa men would use female
dialect forms in their traditional song cycles. Such occurrences were classed as ‘It’s just the
dreaming, they are different.’
22. Ribeiro (2012: 149–50).
23. Trechter (1995: 6–7).
24. ‘Possibly the reduced female forms constitute a conventionalized symbolism of the less
considered ceremonious status of women in the community.’
25. See Ide (1991), Shibatani (1990: 371–4), McGloin (1991), Reynolds (1985, 1991). My
discussion here focuses on informal speech in hyoojungo (Standard Japanese) (see
Reynolds 1991, for a brief mention of this issue). In formal speech, while some interesting
differences between male and female speech remain, the difference is much less obvious
(Nerida Jarkey, p.c.).
26. Nerida Jarkey commented that ‘even anata is not formal enough for addressing someone’s
superior. It can be used as a polite form to address a stranger.’
27. See Ide (1991: 75–6), McGloin (1991), and also Uyeno (1971: 61–2), Shibatani (1990: 373).
What Shibatani (1990: 373) calls clausal nominalizers no and koto (roughly corresponding
to the English subordinator that) may end women’s utterances with an overtone of
surprise and mild admiration. The dubitative ending kasira is also ‘female’ only. Men
and women use different interjections. For instance, a woman would say maa ‘Wow!’ or
ara ‘Oh!’, and a man wouldn’t.
28. Traditional sources date the origins of women’s language in Japanese as early as the fourth
century. According to Inoue (2002: 393–4, 403–10), the consolidation of the female
language was due to social changes in modernized Japan in the late nineteenth and early
twentieth century. Exclusively female forms were created and imposed as a ‘law’ within the
context of creating a national language and codifying the overall ‘image’ of an ideal
Japanese woman through novels and other publications. Nakamura (2015) is a comprehensive
history of the Japanese female language, its development, and changes in different
social environments.
29. In addition, Ide and Inoue (1992) show that female supervisors in some Japanese workplaces
speak even ‘more’ politely than secretaries, and suggest that this is a strategy for
constructing their identity as supervisors, and asserting their authority. Female speech
patterns can be used as a group identity marker, on a par with Tokyo language and other
‘in-group’ ways of speaking (see Ide 2003).
30. See Gal (1995: 171) and further references in Okamoto (1995), on the impact of social
changes on Japanese male and female dialects.
31. Anthony Diller (1985) distinguishes at least seven such registers, ranging from a highly
formal ‘official’ to a very informal ‘oral’ one. A further form, kraphom ‘I’, is used just by men,
when talking to a superior. An informal form nu ‘I’ is used just by women in informal
contexts (when talking to older relatives). See Haas (1944), Chirasombutti and Diller (1999),
and Attaviriyanupap (2015: 384, 392–4, 397) on the usage of male and female forms in Thai.

164 9 When women and men speak differently
32. Brown (1980: 133); see also Brown (1993: 159).
33. Ochs (1992: 57), Hill (1987: 159). To say that women are always more, or less, polite than
men is a simplification.
34. Irvine (1978, 1979). See Gal (1989, 1995: 170–4) on the power of silence and verbal skills in
constructing gender stereotypes.
35. See Yokoyama (1999: 403–6), largely based on Zemskaya, Kitajgorodskaja, and Rozanova
(1993); Hill and Zepeda (1999) on Tohono O'odham.
36. See Hagège (2004: 111–12); Labov (1972: 303–4). Yokoyama (1999) gives examples of
vowel lengthening just by women, to express emphasis in Russian.
37. See Coates (1993: 61–86) and Gordon and Heath (1998) for further examples.
38. This may be partly role-based—women, and not so much men tend to take care of
children, and thus use diminutives as ‘baby-talk’. We saw a similar association between
diminution and cuteness in feminine Linguistic Genders in languages where feminine
Linguistic Gender is associated with size (Chapters 6–7). See Zemskaya, Kitajgorodskaja,
and Rozanova (1993), Yokoyama (1999: 406–7) on Russian; Mendes (2014a) on Brazilian
Portuguese. Sadiqi (2003: 154–5) and Gordon and Heath (1998: 437) mention that Arabic
women tend to use more diminutives than men.
39. See Lakoff (1975: 54); Tannen (1990), and a critical survey in Coates (1993: 116–17).
Trechter (1995: 108) reports that Lakhota women do not use enclitics which imply strong
opinion and assertion: this is not part of their repertoire.
40. Coates (1993: 132–4; 2012) offers an in-depth analysis of dominance, difference, and
‘constructing’ one’s discourse using gender tokens.
41. Dubois and Crouch (1975) were among the first ones to challenge the assumption of
women’s overuse of tags as a sign of ‘powerless’ discourse; see Coates (1993: 119–24),
Holmes (1984), and Cameron, McAlinden, and O’Leary (1989).
42. See Key (1975: 76–7); see Coates (1993: 124–6) on female and male commands in
interaction with children and patients.
43. This point was incisively made by Cameron (1992: 24). That women are more verbose
than men also turned out to be untrue, on close examination (Coates 1993). The impression
of female verbosity prominent in what Coates (1993) calls ‘folklinguistics’ comes from
patterns of polite, and over-emotional speech popularly—but not necessarily correctly—
associated with a dubious ideal of ‘standard average women’.
44. See a summary by Barrett (2006), Podesva, Roberts, and Campbell-Kibler (2002), Christ
(1997), Gaudio (1994), Avery and Liss (1996) discuss phonetic features of speech perceived
as ‘gay’; see also Zwicky (1997) and Levon (2006); a useful survey of literature on
gay and lesbian language is in Kulick (2000).
45. Mendes (2014b). See James (1996) for a survey of opinions on how women use more
standard and prestige varieties than do men.
46. See Mendes (2012, 2014b) on diminutives in male and female speech in Brazilian Portuguese
as tokens of female speech and of speech patterns of feminine gay men. The association of
diminutives with ‘female’ or ‘male’ image is also reflected in a reaction straight men have to
those who overuse diminutives. See also Jones (2014) on English-speaking butch lesbians;
Ogawa and Smith (1997) on Japanese-speaking gay men in Tokyo and Osaka.
47. Attaviryanupap (2015: 393–5).
48. Livia (1997: 358–9), Pastre (1997: 372–3).
49. See the detailed and entertaining account in Livia (1995). Gay and lesbian norms of
structuring discourse and conversation reflect specific ‘cooperation’ based on the affirmation
of gay and lesbian identity rather than identifiable ‘male’ and ‘female’ stereotypes: see
Barrett (2006: 319–20 and references there).

9.4 Male speech, female speech 165
50. Trechter (1995: 6–7) alludes to this. Berdaches have been documented for a number of
peoples in the North American Great Plains, including Lakhota, Arapaho, Mandans,
Poncas, Quapaw, Kiowa, and many others. For further features of male and female
Berdaches, see Whitehead (1981: 88–93), Lang (1998), Roscoe (n.d.). No further details
of female language use are available in the literature. The term ‘berdache’ has pejorative
overtones, and is nowadays used to refer to gay people. The traditional berdaches are now
all but gone, as a result of the Christian influence.
51. See Hall and O’Donovan (1996) and references there, especially Nanda (1990). Hijras have
an established position in the society: they are expected to sing and dance at births and
weddings, and receive payments for this. Some view themselves as neither men nor
women, and also as ‘deficiently’ masculine and ‘incompletely’ feminine. Despite their
official recognition in India, Nepal, Pakistan, and Bangladesh, they remain a marginalized
minority.
52. Borba and Ostermann (2007). Kulick (1998: 216) mentions a preference for the feminine
term bicha among Brazilian transvestite prostitutes from the north-east to refer to
themselves, as part of their female identity.
53. Besnier (2003) investigates the ways in which Tongan fakaleiti, or women-like men, subtly
imitate female speech patterns in the language without Linguistic Gender. Third gender
(or male to female transsexuals) have been described for numerous societies, but with little
attention to linguistic practices (e.g. Wikan 1977, 1978 on Omani xanīth). Speech of
female to male transsexuals remains a matter for further studies (cf. Barrett 2006: 320).

10
The rituals of gender
Social Genders in traditional societies tend to be associated with different domains,
and different speech styles. This is where inequality between the male and the female
Social Genders comes to light. Special languages and language registers can come to
be used in male-only rituals. A whole set of terms may be forbidden to women.
Women may have to exercise special caution and use deferential forms to address
their relatives through marriage. Women are often seen as keepers and promoters of
prestigious linguistic norm, and of traditional language. Or they can be viewed as a
dangerous ‘other’ who lead the society in the wrong direction.
10.1 Social Gender, speech genres, and speech practices
In many traditional societies, public speeches, oratories, and important origin myths
are associated with the male domain. 1 We can recall, from §9.2.1, that Malagasy
women are excluded from the oratory—a major formal speech performance. Malagasy
men shy away from gossip and mutual accusations—the type of talk associated
with women. Elaborate verbal skills are associated with authority and status, and thus
with men. Male speech is described as indirect or respectful, and women’s speech as
direct or confrontational.
Manambu men used to be responsible for culturally important rituals including
the mortuary ritual, the Yam Harvest ritual, and initiation. All of them assigned to
the masculine Linguistic Gender and given importance and value (see §3.4.2). Only
some of these are still performed. Traditionally, there used to be (and still are) two
types of songs—laments about foiled marriages and missing or endangered relatives
(namay and sui), sung by men and by women, and mourning songs gra-kudi (crylanguage)
sung exclusively by women after someone’s death and during the mortuary
ritual Keketep which may take place a year later. The Manambu language remains in
active use, but all song genres are on their way out. The totemic knowledge and the
proficiency in clan names associated with composing songs of any genre are rapidly
decreasing as older people pass away. Women under 60 (and even many of the older
women) do not have enough knowledge to perform a mourning song. According to
Kamibau (in her early sixties), they ‘sing’ in men’s fashion: emitting a high-pitched
How Gender Shapes the World. First edition. Alexandra Y. Aikhenvald.
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.

10.1 Social Gender, speech genres, and speech practices 167
whining sound. This is referred to, rather dismissively, as du-gər (man-cry) ‘crying
men’s fashion’, orkapa-gər (just, for nothing-cry) ‘just crying, crying for nothing’.
This was a special legacy of women within the Manambu tradition. Its recent loss is
deplored.
Women-only genres give women ways of expressing themselves. Warao women,
from Eastern Venezuela, are not supposed to speak at public gatherings. The only
important form of social action for them is a ritual wailing genre called sana. When
someone dies, female relatives of the deceased compose and sing sana songs until
they come back from the graveyards. In these improvised wailing songs women use
words and ways of speaking only men use in public gatherings, and sound aggressive
and self-assertive: their status as women gives them the right to ‘chastise’ their
relatives. 2
Male and female speech genres among the Kuna of Panama are largely complementary.
3 Men perform political and ritual genres such as public speeches, tribal
origin myths, and magic and curing chants. Women perform songs which are closely
linked to life-cycles—lullabies and laments. Important political matters are discussed
in exclusively male gatherings (which women attend only if they are accused of
wrongdoing, accuse someone else, or are called as witnesses). At gatherings attended
by women and men together, women are supposed to listen to the chanting and
speaking of male chiefs. In recent years, however, women have started holding
gatherings of their own, with women leading discussion and practising eloquent verbal
skills. Magic and cures can be viewed as predominantly ‘male’ domains; but there are
both men and women who have magic powers to diagnose a disease and cure it.
Kuna puberty rites—held for girls—involve a strict division of labour between men
and women. A girl for whom the ritual is held has her hair cut, by a ritual haircutter
and her assistants (all women), in a special enclosure. Meanwhile, the ritual director
of the puberty rites—a man and his assistants—perform a long (one-, two-, or threeday)
chant to the spirit of the flute in a large centrally located house, which ensures
the proper running of the ritual.
Each of the Kuna genres—the chants of chiefs, the speeches of male political
leaders, the curing chants of medical men and women, and the chants of pubertyrite
directors—have unique linguistic properties. They involve metaphorical language
and special vocabulary which requires special study, and is not intelligible to
lay people. Both men and women have knowledge and understanding of special
properties of these genres.
As Sherzer (1987: 110) puts it,
There is no question that men’s ritual, formal, and public speech is more diversified and
complex than women’s and that men have more access to and control of political authority
through such speaking practices. At the same time, women also have ritual, formal and public
speaking roles and these are always positively valued, by men and women alike, and provide
women with a certain access to and significant involvement in Kuna political life and power.

168 10 The rituals of gender
Kuna men may appear more visible than Kuna women in public life. But women
are equally, if not more, important in social and economic affairs: they are the ones
who make, wear, and sell molar blouses which are primary markers of Kuna ethnicity
and culture (and an attraction for tourists).
Men’s and women’s knowledge reflects different spheres of competence. In traditional
societies, such as Kaluli of the Western Province of Papua New Guinea, girls
learn about gardening, food preparation, and songs—and other expressive genres—
that their mothers teach them. Boys learn about hunting, and the genres which
they will be using in male ceremonies in their later life. Frogs are hunted only by
women of the Nungon tribe in Morobe Province of Papua New Guinea: this food
used to be a taboo for men. As a consequence, women are more proficient than men in
names of frogs and their species. As Schieffelin (1987: 259) aptly puts it, the linguistic
differences between men and women, and boys and girls, concern ‘the different
distribution of linguistic resources’, associated with Natural Gender and Social Gender
stereotypes.
‘Male’ speech can be associated with ceremonial genres—an exclusive property of
older men—and thus signal a high social status. Among the Acoma Keresan speakers in
the American south-west, the differences between male and female speech lie in the
vocabulary, and in pronunciation. The perception of native speakers of Acoma Keresan
that there is a major difference between male and female speech ‘isbasedinpartonthe
fact that only older males use the age-related kiva speech’—a special ceremonial speech
register. 4 We now turn to the registers which only men are allowed to use.
10.2 The languages of manhood
In many traditional societies, Social manhood can only be achieved through initiation.
A number of Aboriginal Australian tribes had (or still have) a special speech
style used only by initiated men on ceremonial occasions and taught to boys at their
initiation.
The most striking example of a secret men-only language is Damin, until recently
used by initiated men of the Lardil tribe, on Mornington island in northern Australia.
Damin had eleven out of the seventeen consonants found in the everyday language,
and at least thirteen additional ones. Among them were nasal clicks, an ingressive
lateral fricative, a bilabial trill, and a glottalized or ejective velar stop. This was the
only language outside Africa to have clicks, and the only one with so much phonetic
complexity in its click sounds. Damin used only about 150 lexemes covering the full
range of Lardil vocabulary—one term for all vegetable food, one for all bony fish, and
one for all liquids. The everyday language has nineteen pronouns. Damin had only
two, ‘I’ and ‘other’—highly unusual for a human language.
Damin used to be taught quite explicitly. The youth who was to be initiated was
simply brought forth, and a Damin speaker would shout the vocabulary at him, while

10.3 Secrets, avoidance, and taboos 169
another repeated the terms in the everyday Lardil. Damin was used openly in the
community: women, children, and uninitiated men had the opportunity of hearing it,
and even understanding it. This is unlike the initiation language among the Warlbiri
of Central Australia—known as the ‘upside-down Warlbiri’ (or tjiliwiri, ‘the antonymous
language’). This used to be ‘spoken by guardians in the presence of junior
novices’, that is, by initiated men in the Warlbiri men’s rituals (Hale 1971: 473), and
was not supposed to be known to women or uninitiated men, or discussed with them.
Numerous Yanomami-speaking groups in Venezuela and the adjacent area of Brazil
use a special archaic variety of Yanomami called the ‘Wayamo’ language. This is a kind
of Yanomami lingua franca, ‘spoken in a formalized and chanted manner between
members of different villages’. All Yanomami men speak Wayamo; women and prepuberty
children can understand it, but are not allowed to use it. The Wayamo language
is particularly useful for communication between villages which speak different (and
hardly mutually intelligible) dialects. It is also employed among villages which speak the
same, or almost the same variety, and on certain occasions even by people from the
same local group. The Wayamo is mostly used at the [reaho] or [yãimowei] gathering,
when surrounding local groups (living from a day’s journey away to as far as fifteen days
away) are invited for a feast. When they arrive, they first camp in a nearby forest, and at
dawn send one or two representatives to the communal house in the host village to
plead not to have food withheld from them. This is done in Wayamo. 5
This is where linguistic inequality between male and female Social Genders comes
to light. Across the world, male secret ritual languages, and languages associated with
male initiation, explicitly exclude women. There have been no reports of secret
languages confined to women’s puberty rites or other ceremonies in Australia, or
elsewhere. Dixon (1980: 68) suggests that the lack of secret language for women may
be explained by differences in men’s and women’s socialization. While ‘there seems
to be a general tendency for the details of male rituals to be taught and explained
simply and explicitly’, ‘women are usually expected to assimilate the structure of
female rituals by exposure and experience’. Anne Storch (2011: 81–5) offers a further
explanation. The Jukun men, from Nigeria, use secret languages ‘to maintain their
exclusive control over shrines and spirits’. Women are excluded for two reasons—
because the men want to dominate them, and also because they are seen ‘as a
potential danger to their communities’ coherence, in the men’s view’. We now turn
to what women are not supposed to know or to mention.
10.3 Secrets, avoidance, and taboos: what women
are not supposed to know
Sacred objects, and associated rituals, which used to be women’s property and had
been appropriated by men, are a taboo for their previous owners. According to Jorge

170 10 The rituals of gender
Muniz (p.c., in 2012), the only able-bodied elder and healer from the Tariana community
of Periquitos, women suffer birth-pains as a punishment for them having tried
to hold on to the Sacred flutes. The Yurupary flutes in north-west Amazonia used to
be women’s property, and were taken from them by force. These instruments, kept
hidden under water, represent a powerful Yurupary spirit, and their sound is his voice.
If a woman catches a glimpse of the flute, she has to be punished by death (typically, by
poison).
Tariana women are not supposed to hear the word piri ‘Yurupary flute’ (let alone
use it), or any other words which contain this root. In the presence of women this
form used to be replaced with another one (not necessarily phonetically similar). The
set of forms for women only was called ina: na-pia-nipe (woman:PL 3pl-hide-NOMIN-
ALIZATION), literally ‘what they hide from women’,orina: na-maña-nipe (women 3plcheat-NOMINALIZATION)
‘their cheating of women’. The original form—off limits to
women—is referred to as mayakani ‘straight one’.
The Tariana language is severely endangered, and a substantial amount of traditional
knowledge has been lost. Catholic missionaries (permanently present in the
area since 1925) did their best to eradicate the Yurupary cult as something belonging
to the devil. As a result, only snippets of the old secret register are still remembered.
The overwhelming majority are names of important locations associated with the
origin myths and travels of the Tariana ancestors. Just a few words relating to the
objects associated with the Yurupary cult (but not the flutes themselves) have a
‘hidden’ counterpart. As is common for secret and avoidance registers across the
world, all the words with a ‘hidden’ counterparts are noun roots. 6 Some examples are
in Table 10.1.
Two names of sacred instruments associated with the Yurupary flute rituals were
used as place names, by men only. They had a counterpart in the ‘hidden’ register.
A stone in the Vaupés river called Iʧiru ‘a sacred musical instrument’ has a ‘hidden’
counterpart Ukara-da (sardine-CL:ROUND), and a river called Urupema, a name for
another sacred instrument, got the name of Karaka-pua (rooster-CL:RIVER)—according
to the late Cândido Brito, ‘then women will not know what they are not supposed to’.
The only documented instances of lexical replacement for place names which do not
contain the root piri are two locations, Ipanoré and Urubuquara. Ipanoré is said by
some Tariana to be the ‘navel’ of the world; the Tariana themselves are said to have
emerged from the rapids of Ipanoré. Even nowadays, these rapids are practically
impassable. The Tariana varieties I have been working on since 1991 are spoken
beyond Ipanoré (all the original Tariana dialects in Ipanoré there are long forgotten).
We travel by boat to Ipanoré, and then hire a truck to take us, the outboard motor, the
boat itself, and all our belongings, to the other side of the rapids—the settlement called
Urubuquara. No one could tell me any stories about the importance of Urubuquara.
The ‘hidden’ register is now largely a thing of the past. Speakers (all of whom
except two were men) were keen on teaching me—a white woman researcher—the

10.3 Secrets, avoidance, and taboos 171
TABLE 10.1. ‘Hidden from women’: tabooed nouns employed in traditional
Tariana
Type
‘Straight’ form not
to be heard by
women
A form used to
‘deceive’ women
Meaning
Place names
containing the
forbidden root piri
Other mythologically
important place
names not containing
the forbidden root
Paraphernalia related
to the Yurupary cult
(containing the
forbidden root piri
‘Yurupary flute’)
Piri-dapana
(Yurupary.flute-CL:
HOUSE)
Piri-na (Yurupary.
flute-CL:VERT)
Piri-pua (Yurupary.
flute-CL:RIVER)
Piri-taku (Yurupary.
flute-CL:POINT)
Piri-ya (Yurupary.
flute-CL:SKIN)
Myaka-dapana
(ancient-THEMATIC-
CL:HOUSE)
Eta-dapana (eagle-
CL:HOUSE)
Mapa-dapana (bee-
CL:HOUSE)
Inari-na (mucura.
rat-CL:VERT)
Inari ite (mucura.
rat POSS+NCL:ANIM)
Inari-taku (mucura.
rat-CL:POINT)
Piawa (?)
Pa-hwa-li-dapana
(IMPERS-LIE-NOM-CL:
HOUSE)
Wayuli-dapana
(vulture-CL:HOUSE)
name of a cave
name of a hill
name of a river
name of a
peninsula
name of a flat
stone
Ipanoré
(a Portuguese
place name)
Urubuquara
(a Portuguese
place name)
piruaɾi yakasolo traditional
necklace; apronlike
clothing
‘straight’ language, to make sure I got everything right. The first ‘hidden’ register
form was spontaneously produced, as an afterthought and a curious fact. After that,
I started consistently asking for a ‘hidden’ register equivalent for each term. The
Tariana men were not inhibited by the presence of two Tariana women: the oldest
speaker’s daughter and my classificatory older sister Olívia and her mother Maria—
who had a special status. We can recall, from §7.2, that a particularly strong woman
in Tariana can be referred to as ‘he’. This was how Maria’s children referred to her.
All the Tariana stories I recorded contained the ‘straight’ form of the name for
Ipanoré. In contrast, the term consistently given for ‘traditional necklace; apron-like
clothing’ was the yakasolo, from the ‘hidden register’ (and not piruari, a word many
younger Tariana did not seem to know).

172 10 The rituals of gender
The term mayakani ‘straight, right, correct’ is used to refer to the ‘original’ form of
nouns and place names available to men, and hidden from women. This fits in with the
general image of women one gets through the Tariana stories and language in general.
We can recall, from §7.2, that the notion of ‘manhood’ in Tariana has positive
overtones. Ricardo Brito, an elder from Santa Rosa, told me a moving story about
how he managed to escape from a jaguar, and had said to himself, at a critical moment:
10.1 nuha-misini tsiãri-naka
I-TOO
‘I am also a man’
man-PRES.VIS
This was a statement of self-encouragement: Ricardo was saying to himself that he
was courageous and ‘manly’ enough to overcome an adverse situation. In another
story, a male otter was teaching his children to be brave and resilient, telling them to
be ‘men’.
Throughout Tariana lore, women are portrayed as wayward and outright dangerous.
They spoil everything (and are referred to as manihta-kadite (NEG+think/reason-NEG
+NCL:ANIM) ‘the one who does not think’). An ancestral woman ate the manioc when she
was told not to—as a result manioc got covered with thick peel, and one has to work
hard at peeling it. Since other ancestral women ‘didn’t listen to advice’ (pa-kalite-ka
mhema-kade-ka (IMPERS-speak-WHEN NEG+listen-NEG-WHEN)), and slept with a smelly
mucura-rat, their bodies lost their pleasant odour and are now smelly. If a man dreams
of a woman before a hunting expedition or another important event, he is in mortal
danger.
Menstruating women are particularly perilous: the late Graciliano Brito explained
to me that the pernicious evil spirit Ñamu likes menstrual blood, and so menstruating
women are an easy target and attract attacks. A menstruating woman can be
referred to as inaru puaya alia-ka (woman different/adverse EXIST-WHEN) ‘when a
woman is in a different, adverse state’. The adverb puale ‘the other way; on a different
side’ (cognate with puaya) has overtones of something adverse. And so does every
form with the meaning ‘different’. That is, women are a different, strange, and
potentially dangerous ‘other’. And there is perhaps a reason for this.
The Tariana language is spoken in the Vaupés River Basin linguistic area known
for its ‘language-based exogamy’. In Tariana society, and throughout all traditional
communities in the Vaupés, women used to come from a language group different
from men. Traditionally, one could only marry someone who speaks a different
language and who belongs to a different tribe. People used to say: ‘My brothers are
those who share a language with me’ and ‘We don’t marry our sisters.’ That is, in the
Vaupés River Basin area ethnic identity used to be inextricably linked to linguistic
allegiance. One would identify one’s father’s language as one’s identity badge. The
area used to be highly multilingual—women would teach their own fathers’ languages
and the languages of their husbands to their children.

10.3 Secrets, avoidance, and taboos 173
Consistent reference to women as being ‘different’ (puaya) rather than ‘straight’
(mayakani) suggests that they may have been looked upon as outsiders. Now that
many indigenous languages of the Brazilian Vaupés—including Tariana—are rapidly
falling into disuse, women are blamed for not teaching it to the children. We return
to this in §10.5.
In a few similar instances, words women are not allowed to pronounce may be
replaced with special synonyms. In Kele, a Bantu language spoken by the Lokele tribe
in the Democratic Republic of Congo, women cannot utter the term libwá ‘nine’
because it is homophonous with the curse-word lilwáa in the male-only initiation
language Libéli. (The female form for ‘nine’ is is h k h , from the stem -s h k- ‘laugh’). 7
This is somewhat similar to what Mary Haas described as ‘interlingual taboo’—
speakers of Creek (a Muskogean language from Oklahoma) and of Thai would avoid
using those words in their language which bear some phonetic similarity to fourletter
words in English. Creek speakers would avoid the word fákki ‘soil, earth, clay’,
and Thai speakers would try and not use fàg ‘sheath’ and fág ‘pumpkin or squash’.
A whole set of terms can be off limits to women. Male body effluvia are kept secret
among the Kwami of north-eastern Nigeria. Kwami women are not allowed to ever
mention male private body parts; they have to pretend not to know that men have
genitals, and have to defecate or urinate (which they do secretly in the areas off limits
to women). A similar belief in Jukun-speaking societies in the same region of Nigeria
has repercussions on the grammar of the language. Taboo words such as ‘sperm’,
‘urine’, ‘penis’, and ‘testicles’, and also all expressions of body effluvia of adult men,
can never appear with a possessive suffix. That is, forms such as *ʒùr-mii ‘my penis’
or *sùm-mii ‘my (masc.) urine’ are not acceptable.
As Anne Storch (2013: 213) puts it,
That men have genitals and body effluvia is a secret and taboo in many Jukun-speaking
societies. This directly relates to the idea of potentially dangerous women. Women were
largely excluded (and often still are) from religious practices and the religious knowledge of
the men, and they were forced to ‘believe’ that men have non-secreting bodies...Consequently,
men are claimed to have no genitals and produce no body effluvia such as sperm and
faeces, so that the male body can be kept as a delicate and protected entity, with all its
secretions and traces. In this sense, it is the vulnerable male body in transformation that is not
to be possessed verbally. 8
In Jukun tales, women are often represented as potential witches. No wonder the
existing taboos serve to protect men from women, and their dangerous powers.
The asymmetrical relations between men and women are reflected in deferential
registers which only women have to employ. In a number of African societies,
women are not allowed to pronounce the names of their in-laws, or any word that
sounds like an in-law name. If a Sidaama woman from Ethiopia has a father-in-law
whose name is Dawasso, she will not be allowed to utter quite a few words starting

174 10 The rituals of gender
with da- (such as daa ‘come’, dancha ‘good’, and darawo ‘peer’); she will have to replace
them with synonyms.
Women-only honorific registers can be viewed as an obligatory mark of respect
towards their male relatives. Among the Big Nambas of North-West Malakula in
Vanuatu, women employ special lexical and grammatical forms when talking to
anyone with a chiefly rank, or to their own first-born son. So, if a woman has an inlaw
called Sənari (literally, trickster), she cannot use this name, nor can she say, in
ordinary speech, i-sənar-i ‘he tricks him’. She will say instead i-palv-i, with a similar
meaning. 9 Many of the forms have been lost, under the pressure of Western impact;
nevertheless, many Big Nambas women continue avoiding a mention of their male
in-law’s name, or any other word that sounds like it.
In traditional Warlbiri, Kaytej, Warumungo, and other communities of the North
Central Desert in Australia, widows are not allowed to speak during a period of
mourning after their husband’s death. This period may last for as long as three years.
An elaborate sign language is used by women only to communicate throughout this
period. 10
To evoke the name of a deceased is to re-evoke the past and the memory of grief.
Since women identify most strongly with immediate kin, they are, in a certain
sense, more closely connected to the deceased and ‘maybemoreindangerof
reminding others of the deceased than men, social withdrawal [and with it avoidance
of spoken language] is an obvious solution...Using sign language, thus,
when, as a widow, one is in a dangerously close association with a dead person,
is a way of making oneself relatively safe for others’ (Kendon 1988: 450). A taboo—
once again, against women using language—is a way of protecting the tribe from
unleashing dangerous forces.
A special avoidance register just for women can be interpreted as a direct
reflection of male supremacy. In Fox’s (1996: 382) words,a‘specialised vocabulary
indicates to what degree the women were required to show subservience to men in
the days when the Big Nambas culture was almost unaffected by external culture
contact. The fact that the sub-dialect was accompanied by bowing, hiding oneself
and the fear of dreadful happenings if the taboos were broken, shows that these
terms were indeed an indication of the lesser position of women, and this is still the
case to the present day.’
There is no evidence that any of the women-only registers and ways of talking (so
far described) are secret from men. In contrast, men-only languages are kept secret
from women. It is as if men had something to hide and to protect themselves. As
Anne Storch (2011: 84) phrased it, male secret languages are used as strategies ‘by
socially or spiritually inferior groups to gain or maintain power they would not
otherwise have’. The linguistic asymmetry in Social Genders reflects fear and need
for protection—and with this, an open oppression and exclusion of women. Along
similar lines, Dr Samuel Johnson (a major figure in English lexicography and

10.4 Men, women, and language change 175
literature) once said: ‘Men know that women are an over-match for them, and
therefore they choose the weakest or most ignorant. If they did not think so, they
never could be afraid of women knowing as much as themselves.’ 11
Secret knowledge does not have to go together with secret language. Throughout
Aboriginal Australia, numerous men-only sacred rituals and sacred sites are off limits
to women. Women-only sites and rituals are off limits to men. 12 There is no information
about any secret linguistic knowledge confined just to women.
10.4 Men, women, and language change
What is the role of men and women in linguistic change? As Jespersen (1922: 242)
put it,
...as a rule women are more conservative than men...they do nothing more than keep to the
traditional language, which they have learnt from their parents and hand on to their children,
while innovations are due to the initiative of men. Thus Cicero, in an often-quoted passage says
that when he hears his mother-in-law Lælia, it is to him as if he heard Plautus or Nævius, for it
is more natural for women to keep the old language uncorrupted, as they do not hear many
people’s way of speaking and thus retain what they have first learnt. This however does not
hold good in every respect and in every people. 13
Men, as ‘chief innovators of language’, 14 lead the change in casual speech in various
English dialects, including that of Norwich. In gender-exclusive languages—where
male speech is easy to tell apart from female speech—women’s forms are often more
archaic than men’s (see §9.1.3). But the reverse can also be the case. Gauchat (1905), in
his study of Charmay French patois, discovered that women used innovative linguistic
forms more than men. For instance, women over 40 palatalized l to y, just like younger
speakers (men over 40 did not show this innovative trait). Women leading change in
the pronunciation of vowels have been described for many areas in the United
States. 15 We can recall, from §9.1.3, that in Yana as described by Edward Sapir female
forms are less archaic than male forms.
The adherence to a standard norm and prestigious forms are often a feature of
women’s speech (as we recall from §9.2.2). Men would use non-standard forms and
stigmatized varieties as tokens of a speaker’s lack of inhibition and tough masculinity,
and also to show solidarity with each other. 16 And women may choose more
prestigious normative forms to escape the bounds of a traditional society where
they are assigned an inferior status.
A salient feature of the variety of Spanish spoken in the Ucieda village in the
province of Cantabria is word-final u rather than o in words like campu/campo ‘field’.
This pronunciation is stigmatized as a symbol of how peasants speak. Holmquist
(1985) discovered that the non-standard pronunciation was in use by younger
men—despite the fact that they were exposed to standard Spanish variety in their

176 10 The rituals of gender
work outside the community and in military service. Younger women’s speechwas
closer to standard Spanish, and sounded more conservative: they would consistently
say campo rather than campu. 17 Through a standard pronunciation form,
women were making an effort to try to break away from the peasant way of life
which offered them few if any benefits. This resonates with Trudgill’s (1975: 91)
suggestion that in Western societies women are more status-conscious than men,
and they feel the necessity to secure their social status by using the normative form.
Language helps women to present themselves as respectable and intellectually
competent people. 18
Women may go with the flow more quickly than men. Many modern English
dialects are losing their specific features, as a result of dialect levelling. The traditional
local forms of Newcastle English are being abandoned in favour of more regionally
widespread—even if non-standard—forms. For instance, a glottalized [tʔ] is shifting
to [ʔ]. Women are leading this change, as communicators with a wider world. 19
Women act as mediators, using a greater variety of forms employed by a majority of
people.
Female speakers of Belizean Creole use the English copula be more frequently than
the original Creole de in its locative meaning, as in We she de? ‘Where is she?’, thus
shifting to English-like ways of speaking. Men use de as a locative verb more often.
Women, rather than men, use English forms of the past tense (rather than other
typical Creole forms). 20 Women tend to adapt to variation more easily than men. But
other forms in Belizean Creole are used by men and women with equal frequency. It
is perhaps the case that with a rise in equality of social position for men and for
women, a wider range of social and linguistic choices becomes available, and people
make personal decisions as to the way they speak. Social Gender is just one of the
variables at play in the choice of forms; many other social parameters, including the
type of community, are equally important.
A number of studies show that women tend to be more conservative than men in
rural settings and more innovative in urban settings. Kera is a Chadic language
spoken in Cameroon and southern Chad. Women who live in Kera-speaking villages
continue using three tones, and can easily recognize them. Women are also more
conservative than men in other respects—they are the keepers of ancestral traditions.
Women who have moved to local towns lose tonal distinctions in their Kera: they are
exposed to French, which does not have phonological tones, and speak Kera less and
less. Men appear to use and recognize tones in rural and urban environment alike.
Urban women promote language change, in all likelihood, under the influence of
prestigious French which represents power and status. 21
The asymmetry between men and women in different social settings tends to reflect
women’s sensitivity to social change, values, and ensuing linguistic stereotypes. How
do men and women fare with regard to language maintenance in communities where
more than one language is spoken?

10.5 Language keepers or language killers? 177
10.5 Language keepers or language killers? Women
and language maintenance
10.5.1 Women as language keepers
In many minority communities women are believed to be the keepers of the
traditional vernacular. Historically, throughout Papua New Guinea, men would
learn fluent Tok Pisin (a lingua franca of the country) during lengthy periods of
contracted labour outside the villages. Women would learn the lingua franca later,
and speak it less well. Women of Gapun, a remote village in the East Sepik region of
Papua New Guinea, speak the local language, Taiap, and lag behind men in their shift
to the lingua franca, Tok Pisin.
All throughout my fieldwork with the Manambu, women were the ones who knew
traditional legends best, and used many fewer words from Tok Pisin than did men.
At a traditional ceremony of ‘name debate’, respectable orators would intersperse
their speeches with Tok Pisin and also English words. There are still some old women
in the villages who speak little Tok Pisin, while all men are full bilinguals in Tok Pisin
and Manambu.
Rural women speakers of Tamazight Berber in Morocco are the linchpins of
language maintenance; they are seen as ‘heritage vessels’ by men within the communities
and also by governments (notwithstanding the fact that urban Berber-speaking
men feature much more prominently in the indigenous rights movements for valorizing
and preserving Tamazight). This is supported by the rhetoric of ‘mother-tongue’
maintenance which accords women a special symbolic status. 22
But in many cases women, not men, abandon the traditional language first, in
search of social status and prestige.
10.5.2 Women leading language shift
About a quarter of the population of the town of Oberwart in eastern Austria are
bilingual in Hungarian and German; the rest are German speakers. Hungarian is
associated with peasant status and peasant way of life, and consequently looked down
upon. The preferred status for young people is a factory worker, not a peasant. The
world of work is German-speaking, and the language itself has come to represent a
higher social status. Young women of Oberwart are especially status-conscious, and
very willing to participate in social and linguistic change. They are the ones who
speak German in most contexts. Young women feel less allegiance than the men to
the traditionally male-dominated system of subsistence farming. They have more to
gain by embracing the new opportunities of industrial employment, and the more
attractive status of a worker and a worker’s life rather than remaining a ‘backward’
peasant. Young women lead in an already advanced language shift to German. Their
language choice reflects facing a social change, breaking away from their traditional

178 10 The rituals of gender
inferior social status. 23 This is reminiscent of how young women in Ucienda tend to
adhere to standard Spanish pronunciation norms, moving away from their ‘peasant’
linguistic roots (see §10.4, and Holmquist 1985).
The Reindeer Saami in Finnish Lapland speak their own language, alongside the
national language, Finnish. A drastic shift to Finnish was accelerated by the evacuation
of the Saami-speaking peoples to Finnish areas during the Second World War,
and the influence of the Finnish-language public school system. The traditional
reindeer herding economy was shattered. Speaking Saami at school and in school
dormitories was forbidden; parents were encouraged to speak Finnish to their
children at all times. As a result, women tend to use the majority language much
more than men—who continue being bilingual in both Saami and Finnish. Men use
Saami in what remains of traditional reindeer herding and ritual life. Women’s
sphere nowadays is the home and child-rearing (back in the old days, both women
and men were involved in reindeer herding). By yielding to the pressure of change,
Reindeer Saami women succumb to the responsibility for their children and future—
which lies with the majority language, and not with the traditional heritage. 24
10.5.3 Women and modernity
The value placed on the local language and appreciation of what women do varies
rather dramatically. Tamazight-speaking women are generally described as ‘lagging
behind men’ in their language shift. Keeping the precious tradition and the language
is seen as a positive thing. Other communities view the traditional vernacular as an
obstacle to the coveted modernization. In the village of Gapun, Tok Pisin has become
a symbol of modernity and highly appreciated ‘knowledge’ and ‘understanding’
(termed save, from Tok Pisin save ‘know’), and is associated with men. The local
language, Taiap, has come to be associated with the female domain, and a traditional,
rather than modern, way of life. Its cultural correlate is the notion of hed (from Tok
Pisin hed ‘head’), encompassing personal will, autonomy, and potentially unruly bigheadedness.
Speakers of Mexicano, a Uto-Aztecan language of Malinche Volcano,
believe that women ‘lag’ linguistically: they are viewed as speaking less Spanish than
men, and thus appearing backward. Sons nag their mothers, ‘Speak Spanish, don’tbe
stupid.’ Men try and control women’s linguistic behaviour, as part of an asymmetry
of Social Genders in the community. 25
In other communities, women can be looked upon as plainly destructive for the
survival of the precious tradition. Malagasy men present themselves as keepers of
proper traditional ways. They are skilled in speech, and employ elaborate ways of
essentially indirect discourse. Women are the opposite—as we can recall from
Table 9.8 in Chapter 9. Women speak more directly and are not averse to openly
criticizing their offenders. Such direct discourse, unsophisticated and contemporary
ways of speaking, and with them European languages, are associated with women as

10.5 Language keepers or language killers? 179
norm breakers. Destructive innovations—brought in by women—involve direct
accusations, and lesser use of proverbs and traditional sayings. These are believed
to be due in part to the influence of European languages, especially French. Accordingly,
in commercial interactions, women (rather than men) are recruited to confront
European buyers as they share directness and matter-of-factness.
10.5.4 Women as language killers
Women can be portrayed as ‘killers’ of an endangered vernacular. Corsican—a
Romance language not mutually intelligible with French—is in decline. Corsican is
associated with predominantly male activities—hunting, shepherding, singing the
traditional song genre of paghjella, and public politics. Women—especially younger
ones—speak decidedly less Corsican than men. In both popular and academic representations
of women’s roles in language shift towards French, women are described
as more attentive to prestige than men. According to Alexandra Jaffe (1999: 107), in the
current climate of nostalgic appreciation for Corsican culture and language, women
are said to be ‘buying in to dominant language and culture’,by‘turning their backs’ on
traditional values and speaking less Corsican than men. Men are viewed more positively,
as preserving important cultural tradition. This is reflected in a poster from Scola
Corsa (Corsican school) which has been in circulation for a few years, where a little
child looks up to their mother beseeching her, ‘O Màe, parlami corsu!’ (‘Oh Mom,
speak Corsican to me!’).
Women are overtly blamed for the demise of the traditional language in Corsica.
This discourse of blame has its roots in traditional attitudes to women. For one thing,
women’s linguistic practices are a priori defined as inferior to men’s. This attitude
underlies ‘scapegoating’ women for the loss of traditional values and the imposition
of new ones. Also, women—who used to be and to a large extent still are excluded
from access to power—are being blamed for having had too much power to influence
the negatively viewed social change. Jaffe (1999: 108) treats this ‘scapegoating’ of the
least powerful members of the society as a way of removing one’s responsibility and
not having to acknowledge one’s powerlessness in the face of French domination
leading to an impending language shift. And powerless people are the easiest to
blame—they are unlikely to strike back.
We recall, from the previous section, that women are portrayed by Tariana men as
wayward and potentially dangerous and destructive. Men are the keepers of the
language as a badge of patrilineal inheritance and identity. And it is women—wives
and mothers, rather than missionaries, or colonial employers—who are uniformly
blamed for not teaching the language to the Tariana children. An anecdote shows
how pervasive this belief is.
Elvia, a teacher in the Tariana school in Iauaretê, is a Tucano. She married a
Tariana man (who no longer speaks the language). When she got married, her
Tucano father said to her: ‘You will now kill his language, this is what women do.’

180 10 The rituals of gender
Many reasons have contributed to the demise of Tariana. More than 200 years of
Portuguese presence in north-west Amazonia, and educational policies of Catholics
of the Salesian order of Dom Bosco, have taken their toll. Just as in many other parts
of the world, priests used to take children away from their parents, so as to educate
them in an appropriate westernized way. While studying with the Salesians, children
were supposed to use just one language—Tucano. The ‘heathen’ multilingualism of
the Vaupés area was looked upon as something to get rid of. Tucano language—
spoken by a numerical majority in the Brazilian Vaupés—was imposed on people as
the major language. Men used to be contracted by white people to work for money
(mostly, on rubber extraction). Lengthy absences of men from the communities may
have contributed to children’s lesser exposure to their fathers’ languages—ironically,
the ones they were to identify with. Unlike many other areas of the world, mother’s
language plays a minor role in the identity of the Vaupés Indians. One’s badge of
identity is the father’s language. Fathers’ languages—Tariana among them—are now
learnt less and less. 26 Just as in the Corsican case, women are an easy scapegoat.
Traditionally, mothers did not aim at maintaining the father’s language. They
would mainly teach children their own languages. However, following the language
etiquette of the area, women would speak the language of their husbands pretty well,
and still continue using it when speaking to their husbands and husbands’ male
relatives including their children (who count as ethnic Tariana). This does not mean
that women or men had to learn their spouses’ languages from scratch as they started
FIGURE 10.1 Elvia, a Tucano woman married to a Tariana man, was supposed to ‘kill’ his
language. Here she is doing a traditional woman’s task: serving manioc beer to Tariana men at
an Assembly of the Tariana in the Tariana school (Aikhenvald 2013b)

10.6 Summary 181
living together. Typically one would marry someone from the same language group
as one’s mother. It is also important to note that the institutionalized multilingualism
in the Vaupés area goes back at least a few hundred years. And the drastic decline of
the language didn’t start until the twentieth century. Just as in the Corsican case,
women are an easy, but not necessarily the right, target for laying blame.
Now the rhetoric of accusation in the Vaupés area is gradually changing. That
children can hardly speak or write Tariana is considered the fault of the school
system, and of the teachers who are not implementing the language programme
properly. Elvia (Figure 10.1) and her colleagues (male and female) will be blamed
again, but in a different context.
10.6 Summary: Social Gender through rituals, genres,
and speech practices
Social Genders are a basis for distinctions between speech genres and speech practices.
In traditional societies across the world, men’s speech is associated with the
public domain, oratories, and origin myths. Women-only genres almost universally
include mourning songs and lullabies—they appear to be more involved with
domestic settings and the natural life-cycle. The division of labour between men
and women is reflected in what they tend to do, and what they tend to know.
However, in many societies male and female domains are not mutually exclusive,
and in Joel Sherzer’s (1987: 120) words,
the diversity of voices speaks through, reminding us of the complexity of language use in
different societies around the world, reminding us...that women’s language, women’s speech
and women’s verbal activities are not everywhere socially and culturally inferior, domestic or
polite.
Secret languages, avoidance, and taboos reveal a striking asymmetry, and a gap,
between male and female Social Genders. In many a traditional society, male
initiation and male cults involve special secret language varieties which women are
not supposed to use or to know. And there are overwhelmingly more restrictions on
what women can hear, or say, than on what men can. The Tariana of north-west
Amazonia used to have a special ‘hidden’ register, to avoid women hearing the name of
the powerful Yurupary flute, women’s former possession which they are now not
allowed to see. Big Nambas women of Vanuatu have to use a special politeness register
when talking to their sons and in-laws. Similar avoidance principles apply among the
Kambaata and a few other groups of Ethiopia and many groups in Southern Africa:
women, and not men, have to avoid pronouncing the names of their in-laws. No such
‘women-specific’ registers or languages have ever been documented. It is as if men try
and purposely exclude women, attempting to protect themselves and their knowledge
from potential intruders who may well show them up.

182 10 The rituals of gender
And indeed, women do—they are the ones who tend to implement prestigious
norms. They develop a greater diversity of interactional skills, and adapt to variation
more easily than men, in many cases thanks to their lower social status which they
are eager to escape from. But a word of caution is in order: in complex industrialized
societies the positions of men and women are more fluid and more complex. The
more social variables—age, education, and socio-economic status—are in play, the
more diffuse the patterns are which differentiate men from women in how normative,
or how innovative, they are.
There is no simple answer to a question of whether women are instrumental in
keeping the traditional languages alive, or in orchestrating their demise. But across
the world, the attitudes to women as agents of change, tradition, modernity, language
maintenance, or language loss reflect the alarming lack of symmetry between the
male and the female Social Gender, often viewed through the lens of male-centred
norms.
As Jane Hill (1987: 159) put it, ‘women may have their own norms, quite
independent from men. Women may find that the symbolic resources emanating
from an urban elite are a good alternative to male usage patterns, offering benefits
that may be more attractive to women, who are excluded from scarce resources in
their own communities, than to men.’ This is when women represent a barometer of
change—cultural and linguistic.
We now turn to the ways in which women and their new—and changing—social
roles affect language use.
NOTES AND SOURCES
1. A useful survey of male and female communicative genres is in Günthner (1996).
2. Briggs (1992).
3. See Sherzer (1987: 112–14) for a discussion of the Kuna, and further examples.
4. Sims and Valiquette (1990: 165).
5. See Dixon (1980: 65–7; 2002: 91–2) on initiation languages in Australia, and references there;
Hale and Nash (1997) on Damin; Migliazza (1978: 568–70) on Wayamo. Numerous examples
of male initiation languages come from Africa. The Lokele tribe (a Bantu-speaking group
of the Democratic Republic of Congo) had a special male initiation language called Libéli.
Many words were markedly different from the day-to-day language. Their grammars were
similar in all but one point. Kele had three demonstratives which agreed in Linguistic
Gender with the noun they modified; Libéli had only one demonstrative which showed no
agreement: see Carrington (1947). Storch (2011: 60–81) offers a comprehensive survey of
secret and initiation languages across Africa. A special male initiation language of the
Gbaya of eastern Cameroon, described by Moñino (1977), was a mixture of several related
languages.
6. See Aikhenvald (2013a); Fleming (2015) on the predominance of nouns in many avoidance
registers. All the Tariana examples come from my original fieldwork and were collected in
1999–2012. A description of the Vaupés Linguistic area (with special attention to Tariana,

10.6 Summary 183
the only Arawak language there) is in Aikhenvald (2002, 2012a: 75–82); recent developments,
including breakdowns of linguistic exogamy, are addressed in Aikhenvald (2013b).
The ‘hidden’ register attested in Tariana has not been described for any other Vaupés
groups; nor has the image of women as ‘wayward’. Many tribes in Amazonia and New
Guinea possess magical objects—especially flutes—which used to be owned by women but
were subsequently appropriated by men and cannot be seen by women under the penalty
of death or humiliation. If a Waujá woman, from an Arawak-speaking group of the Xingu
Indigenous Park in Brazil, saw a magic flute, she would, in former times, have to undergo a
gang-rape. See Villas Bôas and Villas Bôas (1970) on the Xingu mythology and the magic
flutes; see , on the Waujá practices (compiled by
Emilienne Ireland). Ñatabi, one of my teachers of Manambu, from New Guinea, became
blind, on account of her having inadvertently caught a glimpse of a forbidden flute. We
have no information about any special linguistic taboo against women associated with this
sacred knowledge.
7. See note 5 and Carrington (1947: 201) onLibéli; Haas (1951) on interlingual taboo.
8. See Dinslage, Leger, and Storch (2000: 124) on Kwami and Jukun.
9. See Treis (2005) on women’s avoidance register in Kambaata, a Cushitic language, and
similar examples from numerous Cushitic-speaking groups in Ethiopia, including Alaaba,
Sidaama, Libido, and Arsi Oromo, and Teferra (1987: 46) on Sidaama. See Fox (1996: 379)
on Big Nambas. Hlonipha is the name of a similar principle of avoidance among the Zulu,
the Nguni, and the Southern Sotho of Southern Africa: alongside avoiding the names of
the father-in-law and other male affines, women may not mention words with the same
root or root syllables: see Herbert (1990).
10. See Kendon (1988: 449–50). A somewhat similar ban has been described for young
Armenian brides in the Baraninski region of Armenia in the first half of the twentieth
century: the brides were not supposed to talk to their in-laws and used a sign language
instead. See Karbelashvili (1935), and also Fleming (2015), for further examples. In the
Aboriginal communities of Cape York Peninsula, bans on spoken language were not
restricted to women: both widows and widowers were expected to address their nearest
relatives in signed language following the death of their spouses. Signs were also used in
long-distance communication. See Kendon (1988: 44–7).
11. See Henson (1997: 67) for the full reference.
12. See Myers (1986: 64, 85) on the Pintupi of Western Australia; Bell (1983) on the Warlbiri;
Williams (1986: 51) on the Yolngu.
13. Exceptions listed by Jespersen include the Botocudo, a Macro-Jê-speaking group from
Brazil, and Japanese.
14. Jespersen (1922: 247).
15. Gauchat (1905: 205, 211, 218–19, 224–6); Labov (1972: 301–3), Trudgill (1971) on
Norwich.
16. Further examples are in Coates (1993: 61–86) and Gordon and Heath (1998). See Trudgill
(1975, 1998).
17. See James (1996) for a comprehensive survey; Milroy (1999) on the notion of social
networks, their density, and its role in male and female speech varieties.
18. See Eckert (1989).
19. See Coates (1993: 101–4) on women and vernacular speech, and the idea of women as
‘deviants’ in androcentric sociolinguistics. See Milroy (1999) on women and dialect
levelling, and a summary by Brittain (2002: 619).
20. See Escure (2001: 72–81).

184 10 The rituals of gender
21. See Pearce (2013) on Kera. Correlations between the use of more-or-less standard forms,
social genders, and linguistic change are often complicated by parameters such as socioeconomic
status, level of education, and age (see e.g. Eckert 2000), making it difficult to
draw simple conclusions.
22. See Hoffman (2006) on Tamazight Berber; Kulick and Stroud (1990) on the village of
Gapun; Hill and Hill (1977) on Nahuatl, Gal (1979) for language shift in European
communities, and Aikio (1992: 44) on early reports of monolingualism among women
and multilingualism among men.
23. See a detailed discussion in Gal (1978: 159–71).
24. Aikio (1992). Among linguistic minorities in north-east Russia, women tend to obtain a
higher level of education than men; women work outside the communities, while men
stick to the traditional activities of hunting and fishing (Kazakevich 1996).
25. Kulick (1992a: 114, 148–9) on Taiap; Hill (1987: 154–5) on Mexicano.
26. See Aikhenvald (2013b). The situation in the Colombian part of the Vaupés area is
different; but there have never been any Tariana villages there.

11
Gender in grammar and society
Language helps construct the way speakers perceive the world around them. 1 As
Edward Sapir (1929b: 209) put it, ‘The network of cultural patterns of a civilization is
indexed in the language which expresses that civilization . . . Language is a guide to
“social reality” . . . The fact of the matter is that the “real world” is to a large extent
unconsciously built up on the language habits of the group.’
Linguistic Gender is perhaps more of a ‘guide to social reality’ than any other
linguistic category. Linguistic Gender reflects a classification of humans, and is
among the lucky categories which are particularly sensitive to the reality of social
structures. This is what it shares with other means of noun categorization, that is,
classifiers of different types. Linguistic Gender—as assigned to humans—is a repository
of historical asymmetries between Social Genders. Changes in Linguistic Genders
may accompany external changes which impact Social Genders. Language
reforms target Linguistic Genders. These are the topics addressed here.
11.1 What Linguistic Gender can tell us about Social Gender
Meanings of Linguistic Gender can reflect roles and functions of Social Genders. In
Kwami, a West Chadic language with feminine and masculine Linguistic Genders,
nouns which refer to the domestic sphere are typically feminine, and nouns referring
to objects outside the household are usually masculine. We can recall, from §4.1 and
example 4.2, that lòolów-mè (cotton-masculine) is cotton growing on a farm, and
lòolów-jè (cotton-feminine) is cotton which has already been harvested or is at home.
The traditional women’s sphere among the Kwami is the house and the household.
The overwhelming majority of nouns referring to the domestic sphere belong to the
feminine Linguistic Gender (these include gàràn ‘cooking place’, fíní ‘room, hut’, and
shémi ‘entrance hut’). Nouns belonging to the masculine Linguistic Gender tend to
refer to objects outside the household, e.g. máála ‘bush, forest’, zírkì ‘woodland’,
bùnké ‘meeting place in the bush’. 2
The choice of Linguistic Genders for inanimates in Ket, a Yeniseian language from
Siberia, is partly based on size and shape of the object (see §4.1). Terms for large and
economically important species of mammals and fish are masculine, including qàj
How Gender Shapes the World. First edition. Alexandra Y. Aikhenvald.
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.

186 11 Gender in grammar and society
‘elk’, sèl ‘reindeer’, qùr ‘pike’. So are terms for objects of economic or spiritual
importance, including bálbès ‘(Christian) cross’, sùl ‘hook (for holding or fastening)’,
and names for tent parts and snow-sled parts. 3 Men play a major role in fishing and
reindeer herding (the major traditional activities of the Ket)—and this is what we
discern in the Linguistic Gender choice.
The meanings of Linguistic Gender may reflect positive and negative values, and
the relative importance, of Social Genders. Feminine gender choice for non-humans
in Manambu is associated with smaller size, and masculine gender with larger size.
But men’s houses are always masculine, no matter what their size. We saw in §3.4.2
that terms referring to speech and ceremonies in this language are assigned masculine
gender if they are culturally important, and feminine gender if they are less so.
The term jap ‘thing’ can be masculine or feminine, depending on how big the object.
If this same word is used to refer to a traditional valuable object (displayed at malecentred
ceremonies, or used in mortuary payments), it is always masculine. All the
male-oriented rituals and ceremonial objects in Manambu are assigned masculine
Linguistic Gender. So are the names for traditional shell valuables and artefacts made
of them (e.g. yu ‘greensnail shell’).
The feminine Linguistic Gender in Hamar, an Omotic language, is associated with
the importance of an object, and its larger size (see example 3.18). According to Jean
Lydall (1988)—a scholar with an extraordinary knowledge of the Hamar language
and culture—this reflects the importance of women in agricultural and other central
tasks in Hamar life.
The effects of the reversal of masculine and feminine Linguistic Gender choice for
men and for women also point towards attitudes to men and women. We can recall,
from §7.4, that addressing a man with a feminine pronoun in Amharic may imply
belittling him, and is offensive. But addressing a woman with a masculine pronoun is a
way of praising her, by ‘promoting’ her to the status of an honorary male. Similar
examples come from Tariana and Jarawara, two South American languages (a summary
is in Table 7.3).
Asymmetries in the expression, and meanings, of derivational genders can be
indicative of the inequalities in Social Genders.
11.2 Social inequalities through gender asymmetries
The historical asymmetry in male and female Social Genders and their stereotypes
find their expression many a time in derogatory and demeaning overtones of female
terms and functions. We can recall, from §5.3, that in Ndyuka (or the Eastern
Maroon Creole of Suriname) a number of nouns marked with uman ‘woman’ may
have a negative and over-sexualized connotation: waka-man means ‘traveller’, but
waka-uman is not just a ‘travelling woman’: this has strong connotations of a woman
of easy virtue. 4

11.2 Social inequalities through gender asymmetries 187
Conversely, positive connotations of females in a matrilineal society—such as
Ndyuka—may also be reflected in gendered lexical forms. ‘Mother’ is a very prestigious
role in Ndyuka society. It is the mother and her family who have the primary
authority over children. The line of descent of every lineage is determined by
mothers: only the (male) children of the women of elite lineages may carry out the
community’s main socio-political functions. And we find a number of compounds
where mama ‘mother’ means ‘important, main’, asinmama liba ‘the main river’,
mama kinde ‘native village’, mama mofu ‘important or main message’, and a mama
fu a toli ‘the gist of the story’. 5
Specific meanings of male and female terms reflect professions and roles reserved
for men and women. In Oriya, an Indo-Aryan language of India, professions
such as carpenter (kaarigara), constable (daarogaa), or driver (draaibhara) have
only a masculine form. At home women do the day-to-day cooking; but only men
work as professional chefs; so the terms roseiaa and pujaari ‘chef ’ have an
exclusively male reference. A few professions are associated with women—these
include nars ‘nurse’ (from English nurse). The term for ‘male nurse’, aɲɖiraa nars,
contains the noun ‘man’. If one wishes to talk about a female policeman, one needs
to use the term for woman, maakinaa, asinmaakinaa polis. Female counterparts
of many male occupational terms refer to a wife, and not to a female professional: for
instance, casaa means ‘farmer’ and caasuɲi means ‘farmer’s wife’ (and not a female
farmer).
In this male-dominated society, quite a few terms of abuse refer to women only—
these include daari ‘whore’, baanijhuɲi ‘barren’,orbidhabaa ‘widow’. These reflect a
popular belief that misfortunes—such as barrenness, widowhood, divorce, and
socially unacceptable behaviour such as prostitution or extramarital affairs—are
associated with women. 6
Numerous feminine derivations and compounds referring to women in English
have overtones which go beyond what Baron (1986: 115) calls ‘the simple recording
of gender’. A woman appears as distinctly inferior in conventional pairs the weaker
sex versus the stronger sex. Referring to a woman as the fair sex or the gentler sex, and
men as the superior sex indicates disparity and lack of equality between men and
women. In a few instances, feminine forms have no pejorative connotations—the
most oft-quoted examples are aristocratic titles in pairs such as baron/baroness, duke/
duchess, prince/princess, and count/countess. But in an overwhelming majority of
cases compounds and derivations referring to women have negative overtones.
We saw, in §7.5, how a feminine derivational form spinster in English has
somewhat pejorative connotations of an unmarried woman who is unwanted and
‘on the shelf ’. Its masculine counterpart, bachelor, does not. In English, one can talk
about ‘an eligible bachelor’;but‘an eligible spinster’ sounds like a joke. In English, words
denoting women acquire debased or obscene connotations, e.g. madam or mistress.
The equivalent masculine terms—sir or master—do not. The pair mistress–master is a

188 11 Gender in grammar and society
very clear, and oft-quoted, example of disparity in meaning of forms with female and
with male reference.
The term master, ‘a person, predominantly a man, having control and authority’,
was borrowed into Old English (as mægester) about AD 900 from classical Latin
magister. The feminine form mistress was borrowed around 1300, into Middle English,
from Middle French maistresse, first of all in the meaning of ‘woman at the head of a
household’. Mistress, in its pejorative sense of a woman ‘other than his wife with whom
a man has a long-lasting sexual relationship’, was first recorded in 1601. There has
never been a corresponding word for a ‘male extramarital lover’—which, in Cheshire’s
(1985: 22) words, ‘is a telling reflection of the freedom that society has tolerated for
men, but not for women’. However, nowadays, as times are changing, the word mistress
is often replaced by lover, which can refer to people of any sex.
The noun mistress is still used in compounds such as schoolmistress and headmistress,
as female counterpart of schoolmaster and headmaster. An alternative to either
would be schoolteacher and headteacher. However, as Cheshire (2008) reports in her
follow-up to her 1985 paper, in a sample of 500 BBC web-pages, headmaster appears
to be used more frequently than headteacher, and headmistress is lagging behind both
of them. The pair headmaster–headmistress continue being unequal.
Further asymmetry lies in the generic use of the word master: according to the
Oxford English Dictionary Online, master started being used in the second half of the
twentieth century to include women, e.g. In her field, she is a master (from Miller and
Swift 1981: 144). Master is also one of the words with what R. M. W. Dixon (2014)
calls ‘double duty’: it can be used as a verb, and as a modifier. How to master the art of
French cooking refers to a person of any sex. Anyone can become a ‘master landscape
architect’, if they are good enough to master the art. Miller and Swift (1981: 144)
conclude that the word master has ‘outgrown its masculine gender origins’.
In contrast, the noun mistress has retained its female-only reference. A few examples
of a verb mistress in the sense of ‘master something, by a female’ are offered by the
Oxford English Dictionary Online: all are said be ‘used humorously or ironically after
master’ (as a verb). The latest example given is from Fiona Cooper’s 1991 novel Jay
loves Lucy: ‘“What’s this?” said Piggy-wig, straddling a chair with the triumph of one
who has mistressed a Zanussi automatic.’ The term master in its generic and Natural-
Gender-neutral sense appears in numerous derivations and compounds, including
masterful, mastery, masterpiece, mastermind, and mastersinger. 7
Generic masculines—author, actor, poet—cover members of either sex. Their
feminine equivalents may acquire unwanted overtones of lesser, feminine-only,
professionals. 8 The generic uses of masculine terms fit in with the functionally
unmarked character of masculine Linguistic Gender in many languages. Many
feminine forms are derived from, or formed on the basis of, the erstwhile masculine
ones. The exceptions include professions which are considered traditionally female—
such as nurse or prostitute. A male representative of such a profession will then be

11.2 Social inequalities through gender asymmetries 189
marked—we talk about male nurses and male prostitutes. This is how, in Henley’s
(1989: 60–1) words, ‘language stereotypes women’, and also deprecates them—
reflecting traditional practices, even if on their way to obsolescence.
Similar stereotyping is a feature of languages with no Linguistic Gender. The
covert gender bias in Turkish is reflected in implicit information about genderspecific
domains through terms referring to people: çocuk bakɪcɪsɪ ‘nursery school
teacher’ will be understood as a woman, and futbolcu ‘football player’ as a man. Adam
‘man, human’ is the ‘representative member of humanity’ (Braun 2001: 305; see also
§5.3), and women remain on the periphery. In Hungarian, stereotypes of Social
Genders are reflected in professional terms: terms for female occupations which
contain the feminine derivational marker -nő, and do not have a masculine counterpart.
These include varró-nő(sewing-feminine) ‘seamstress’, apoló-nő‘nurse’ or
takaritó-nő‘cleaning woman’. According to Vasvári (2011: 18), one cannot just say
apoló or apoló férfi ‘male nurse’. The marker -nő‘female’ can be added to a formally
unmarked (and traditionally male) professional term to indicate a woman, e.g. orvos
‘doctor’ versus orvos-nő ‘female doctor’, tanár ‘teacher’ versus tanár-nő ‘female
teacher’. However, feminine marked forms reflect a somewhat lower status.
As Vasvári (2011: 19), herself a university lecturer, explains: ‘Tanár is a job
traditionally done for a long time by both sexes, and when I am addressed directly,
I am called tanar-nő[teacher-feminine]. However, if someone were to enquire what
my profession is and I replied tanar-nő[vagyok] “[I am a] female teacher”, they
would assume that I am a secondary school teacher. As a university professor I can
say only egyetemi tanár vagyok, with *egyetemi tanar-nő vagyok impossible. Or
alternately I can be a profeszor [“professor”], but, as one of my former students
reports, whenever she mentions her new yorki profeszor (New York professor) all
listeners always assume . . . that she is speaking of a male professor.’
In Henley’s (1989: 60) words, masculine generics—such as chairman, spokesman,
the working man, the black man, and the man in the street are tokens of how ‘language
ignores women’. They have been a target of struggle against ‘sexist language’—we turn
to this in §11.5.2.
An asymmetry between terms for men and for women may have its roots in a
colonial heritage, and differences in male and female roles for colonizers. In Tariana,
nawiki, the term for ‘Indian’, or indigenous person as a category, is an epicene: it can
refer to a man or to a woman. Male or female reference can be disambiguated with an
article, or by agreement on verb or adjective, e.g. diha nawiki (he/the:masc.sg Indian)
‘the Indian (man)’ and duha nawiki (she/the:fem.sg Indian) ‘the Indian (woman)’.
A number of other generic terms, including the one for evil spirit of the jungle ñamu,
can also have male or female reference. In contrast, there are two separate terms for
‘white (or non-Indian) man’ and ‘white (non-Indian) woman’. The word yalana
‘white man’ is an old and fully integrated borrowing from Língua Geral yara
‘master’. 9 This origin of the term for ‘white man’ iconically reflects the commanding

190 11 Gender in grammar and society
role of non-Indian male masters in the colonial world. Non-Indian women were not
exactly masters on their own; they were defined in terms of men and called yalana
i-sa-do (white.man indefinite-spouse/woman-feminine.singular), literally, ‘white man’s
wife/woman’.
As the relative status of Social Genders changes, so may the composition of
Linguistic Genders and the expression of gender with other means, including classifiers.
This is what we turn to now.
11.3 The value of ‘man’ through gender in lexicon
The asymmetry between male and female Social Genders is especially conspicuous
in the lexicon. The positive overtones of ‘malehood’ in contrast to negative overtones
of ‘femalehood’ are reflected in the lexicon and derivations in many languages. The
ways masculine Linguistic Gender in Hebrew is used as a marker of solidarity,
affection, and protection (discussed in §7.3) may well be based on what Tobin
(2001: 191) calls ‘the inherent desirability of masculine attributes’. In other words,
talking to, or about, a woman as if she were a man has positive connotations. Golda
Meir, during her time in various governments and as Prime Minister of Israel, was
often complimented with male-oriented idioms—such as yesh lah beitsim (exist
to.her balls) ‘she has balls’—and praised as ha-gever ha-yexid ba-memshala (theman
the-only in.the-government) ‘the only man in the government’. This is reminiscent
of the use of the phrasal verb Man up! in English when said to a woman, urging
her to get her act together. 10
Similarly, the noun tsiãri ‘man’ in Tariana has positive overtones of courage,
determination, and reason. The noun inaru ‘woman’ has distinctly opposite overtones.
In order to encourage himself to be brave, a male speaker would say ‘I ama
man’. A turtle who is about to fight a tapir says to himself, as a matter of selfencouragement,
tsiãri-naka inaru-kade-naka nuha (man-PRESENT.VISUAL woman-
NEG-PRESENT.VISUAL I) ‘I am a man, I am not a woman’. This is consistent with a
specially important woman being referred to as ‘he’—as we saw in §7.2. 11 Such
attitudes to men and to women directly reflect the Tariana social reality. In terms
of the relative position of Social Genders, men occupy a distinctly higher status than
women. Among the Tariana, one’s father’s language is the most valuable asset, and is
the badge of one’s identity. Women always come from a different tribe and are held
responsible for language loss (and many other evils). They are a dangerous and
unruly force to beware of—as we saw in §10.5.4.
Schipper’s(2003) study of proverbs across the world reveals a strong association of
women with negative features—weak, unstable, unruly, unreasonable, and better fit
for domestic environment than for dealing with the outside world: this is what we
saw in §8.4. Estonian proverbs (analysed by Hasselblatt 2015: 140–1) illustrate an
asymmetry between the Social Genders: more often than not they portray women as

11.4 How Linguistic Genders reflect social change 191
a negative and disturbing element. Typical examples include Kus on naisi, siel pole
rahu ‘where there are women, there is no peace’, and iga naine on libu oma mehele
‘every women is a bitch to her own husband’. 12
In a few European languages inherently sexist associations between ‘man’ and
‘positive properties’ are embedded in a few derivational forms. Examples of derivations
referring to positive properties of valour and courage associated with ‘man’
include Estonian mehisus (from mees, mehe ‘man’), Russian muzhestvo (from muzh
‘man, husband’), German Mannhaftigkeit, Hungarian ferfiás, all of which mean
‘courage, valour’, literally, ‘manhood’. An Estonian woman-politician can be complimented
as someone who displays ‘statesmanlike’ (riigi-mehe-lik (state-man-AFFIX))
behaviour. The term ‘effeminate’ has a distinctly negative feel to it—as if someone
were coming down in the world, from a status of a ‘manly’ and worthy being to a
decadent one. But to what extent are speakers conscious of the inherently sexist
character of the positive overtones of ‘man’ within established derivations? This is an
issue which has not so far been explored.
Anecdotal evidence indicates that not everyone is aware of any potentially demeaning
connotations of the word man in the idiom man up. A top-level university
administrator told me how she’d complained to her teenage daughter about how
difficult electronic databases are to access. The daughter replied, ‘Man up, Mum, stop
grumbling.’ When I pointed out that such usage of man up could be considered
sexist—as one would not say to anyone *woman up, or*woman down, the administrator
first looked at me with some incredulity, and after a minute or so exclaimed,
‘Oh, I never thought of that.’
We can recall, from §7.3 and §4.1, that Mohawk, an Iroquoian language, offers
two options of referring to a woman—the ‘feminine-indefinite form’ and the
‘feminine-neuter form’. The latter subsumes animals; but, as Mithun (2014: 137)
reports, speakers do not show any awareness of this correlation as ‘demeaning’ for
women. In Lak, a north-east Caucasian language, women have traditionally
belonged to a special feminine gender class II. As the language evolved, respected
women (especially those outside the immediate family) came to be reclassified
as belonging to gender class III (which also subsumes inanimate objects). In this
unusual case, inanimate gender is used as a mark of respect for women. This does
not in any way imply downgrading anyone. On the contrary. Politeness and respect
often involve what Brown and Levinson (1987: 199) call‘conventionalised indirectness’
and distancing. This indirectness is a factor behind the use of a more general
erstwhile inanimate category for talking to, or about, those whom one respects. 13
11.4 How Linguistic Genders reflect social change
Changes in social structures and attitudes may find their reflection in how Linguistic
Genders, and other noun categorization devices, apply to humans. All human beings

192 11 Gender in grammar and society
in Setswana, a Bantu language from Botswana, belong to the human gender (labelled
1/2: see Table 2.2). Names for some groups traditionally treated as strange and
unusual used to be classified within gender 5/6, together with substances, such as
clay or dirt, and abstract nouns, such as foreign institutions. These groups used to
include Bushmen, the indigenous minority of Botswana. In the modern society, it is
no longer correct to refer to Bushmen with this gender—instead, they have to be
referred to like all other humans, with the human gender: when they are referred to
with the non-human class, it is felt that they are being treated like ‘dirt’. 14
We can recall, from §5.1.1, that the choice of numeral classifiers in Maonan, a Tai-
Kadai language spoken in China, reflects the changing place of women within the
society. Similarly to a number of languages of South-East Asia, women used to be
counted with the classifier tɔ 2 , which also subsumes animals and children. Respected
women were typically counted using the human classifier ʔai 1 (NUM.CL:HUMAN). At
present, women who have a professional status—for example teachers in 11.1—are
also counted with the ‘human classifier’.
11.1 ja:u 3 fia:k 8 nde 1 mɛ 2 si 5 ʔai 1
inside school 2pl:inclusive have four NUM.CL:HUMAN
ki 2 sɛŋ la:k 8 ʔbjek 8
teacher woman
‘There are four women teachers in our school’
Respected women—counted with the human classifier ʔai 1 —include ‘women
government and party officials, teachers, women singers’, and elder relatives such
as aunts and grandmothers. They also include female fortune tellers, shamans and
matchmakers, and female monsters: it is believed that counting them with the animal
classifier might offend them and provoke unwanted results. Women who do not hold
any positions associated with a high status are referred to with the classifier tɔ 2
‘animal’.
The choice of the human classifier can also be used as a token of a person’s status,
and their maturity. Children are typically referred to with the animal classifier tɔ 2 .
This applies to primary school pupils, as in sa:m tɔ 2 la:k 8 fia:k 8 (three NUM.CL:ANIMAL
pupil) ‘three pupils’. But college students will be referred to with ʔai 1 , e.g. ŋɔ 4 ʔai 1
ta 4 jɔŋ 5 (five NUM.CL:HUMAN college.student) ‘five college students’. 15 The human
classifier in Maonan has been extended to subsume women of professions and
ranks which they had never occupied before. Derivational gender in Russian offers
similar examples.
One of the consequences of the Bolshevik Revolution in 1917 was, as Rothstein
(1973: 460) put it, ‘the relative liberation of women’ which ‘produced a need for new
terminology to refer to women occupying certain jobs or positions’, for example,
frezerovs᷈c᷈ica ‘female milling machine operator’ (a female equivalent of frezerovs᷈c᷈ik)
or syrovarka ‘female cheesemaker’ (a female equivalent of syrovar). Similarly, women

11.4 How Linguistic Genders reflect social change 193
of minority groups within Russia acquired access to the jobs formerly reserved for
men. In Lak, a north-east Caucasian language from Daghestan, agent nouns such as
xIakin ‘doctor’ used to belong exclusively to the masculine gender marked through
agreement on the verb, e.g. xIakin ur (doctor masc.sg+exist) ‘there is a (male) doctor’.
They can now be used to refer to women, and trigger feminine agreement, e.g. xIakin
dur (doctor fem.sg+exist) ‘there is a (female) doctor’. 16
Old habits die hard—if they die at all. Language change lags behind the changes
in a society. 17 Despite the purported equality of men and women during the Soviet
period, the asymmetry between feminine and masculine derivational forms
remains. 18 Masculine forms of nouns denoting prestigious (and especially traditionally
male) professions continue being used as generics. This is an established feature
of the language—as Roman Jakobson (1971: 213) put it, in his description of Russian
gender in terms of its markedness, ‘the general meaning of the masculine does not
necessarily specify sex’, since the masculine form is used to refer to people in general.
A general term for teacher is the masculine uchitel'; the feminine form uchitelj'-
nica is possible, but it tends to refer to primary school female teachers. The prestigious
title Zasluzhennyj uchitelj' ‘Distinguished teacher’ can be awarded to men and
to women. Making it into Zasluzhennaja uchiteljnica (feminine) would have been
laughable. Along similar lines, awards such as ‘Hero of the Soviet Union’ (Geroj
Sovetskogo Sojuza) or‘Hero of Labour’ (Geroj Truda) contain masculine generic
forms. The title Heroine Mother (matj-geroinja) awarded to mothers with many
children contains the female form heroine. 19 In her Wikipedia entry in Russian, the
first woman cosmonaut, Valentina Tereshkova, is described exclusively with masculine
terms in their generic meanings (kosmonavt ‘cosmonaut’, letchik-kosmonavt
‘pilot-cosmonaut’, general-mayor ‘major-general’, etc.).
Overtones of female derivations can be felt to be deprecatory. We can recall, from
§7.5, how Anna Akhmatova made it clear that she was to be referred to as ‘poet’ and
not as ‘poetess’: the latter term had, for her, somewhat demeaning overtones of notquite-up-to-scratch
feminine writing. Rothstein (1973: 462) mentions the head chef
of a restaurant in St Petersburg (then Leningrad) who burst into tears when an
American tried to compliment her by telling her that she was otlic᷈naja povarixa ‘an
excellent cook’ using the feminine form of ‘cook’. This was demeaning for her—she
considered herself a povar, masculine form for ‘cook’. A female author would refer to
herself as avtor ‘author: masculine’. A female form avtorsha would sound out of place
and somewhat derogatory, and there is no neutral feminine counterpart available.
The late Natalja Y. Shvedova, a leading Russian grammarian and lexicographer,
inscribed a copy of her collected papers to me as ot avtora (from author.masculine.
genitive singular). 20
The meanings of derivational gendered forms reflect social asymmetries in the
position of men and women—in Robin Lakoff ’s (1975: 69) words, ‘linguistic imbalances’
which ‘bring into sharper focus real-world imbalances and inequities’. How

194 11 Gender in grammar and society
can the lack of balance be redressed, by conscious linguistic reform? This is what we
turn to now.
11.5 Thwarting ‘sexist language’
‘Sexist language’ is a term used to describe unequal representation of men and
women across many languages of the world. The following features are seen as its
reflections. 21
First, a cross-linguistically common tendency for the masculine Linguistic Gender
to be the functionally unmarked choice is seen as a way of subsuming women under
‘men’ and making them invisible (see also §7.7). A major bone of contention in the
English ‘sexist language’ has been—and partly continues to be—the generic use of
masculine singular he to refer to people in general (if their sex is not known or
irrelevant). The generic pronoun man ‘person, one’ in German has also been the
target of concern, thanks to its similarity to the generic form Mann ‘man, person’.
Gender agreement and gendered forms of adjectives and verbs have been targeted in
Romance languages (including French and Spanish). The issue does not arise in
English with no gender agreement, nor in German where masculine, feminine, and
neuter Linguistic Genders are not distinguished in the plural. We turn to this in
§11.5.1.
Secondly, a tendency to use terms with male reference as general terms for humans
and for professions has been seen as a sexist way of portraying a man, or the male, as
the main representative of the species, a benchmark for all human beings. The man is
seen as a reference point, or norm. A woman becomes invisible and subsumed under
the category of ‘man’. A term for a woman is often derived from that for a man (e.g.
Hebrew is᷈‘man’, is᷈-a ‘woman’). This markedness is seen as showing woman as a
‘deviation’ or ‘exception’ from an essentially male norm. In newspaper reports, one
often finds a woman singled out from the ‘human’ category—as in the example given
by Cheshire (1985: 25), ‘People would bring their wives, mothers and children’, or
‘Sharing our railway compartment were two Norwegians and their wives’. As we have
seen above, feminine forms and derivations develop pejorative or slighting connotations,
and may sound less prestigious than their male equivalents—used as generics.
In §11.5.2 we look at the attempts to redress the balance.
Thirdly, address practices in the Western world focus on women’s (but not men’s)
marital status—once again putting them in the shadow of a male being. Sexist naming
practices expect women to take on names of their fathers and husbands—making them
invisible and their achievements slighted. This is what we turn to in §11.5.3.
Order of words in coordinations—such as men and women, ladies and gentlemen,
or Bill and Mary—has also come under scrutiny as a potential reflection of gender
bias. However, many factors come into play—among them a tendency to put a
shorter word first. Considerable variation remains in this area. 22

11.5 Thwarting ‘sexist language’ 195
11.5.1 Masculine bias through Linguistic Gender: pronouns and agreement
Personal pronouns in English (she, he, it) are the major means for expressing
Linguistic Gender. What English lacks is a general indefinite form which could
refer to both men and women without specifying their sex. A generic non-gender
specific pronoun monn, man, mann, mane, manne, monne (all derived from an
unstressed form of the noun man) did exist in Old English and in Middle English;
its latest occurrence goes back to 1500. 23 In earlier stages of English, the generic use
of masculine singular pronoun he was an alternative to they covering ‘man’ and
‘woman’ (Curzan 2003: 71). In 11.2, from Old English, generic he refers to a person in
general.
11.2 Swyche a persone ys ful slogh, Be he hygh, or be he logh . . .
‘Such a person is very lazy, be he high or be he low’
(Robert Mannyng of Brunne’s ‘Handlyng Synne’ 161)
And in 11.3, from Old English Alfred’s Introduction to Laws, hie ‘they’ has a singular
reference and can refer to a woman or to a man:
11.3 Gif oxa ofhnite wer oððe wif, Þæt hie dead sien,
sie he mid stanum ofworpod . . .
‘If an ox gores a man or a woman, so that they be dead, may he [the ox] be
killed with stones’ (Alfred’s Introduction to Laws 32)
Similar examples are attested in Middle English, as in 11.4:
11.4 Yff man or woman take sekenes that day, thei schuld sone recouer . . .
‘If a man or woman take sickness that day, they should soon recover . . .’
(Metham, Physiognomy 149)
That is, the option of using a generic form they for singular reference has been
available in the written language since the earliest times. The generic use of he can be
accounted for by a purely linguistic reason: the general word for person mann in Old
English belonged to the masculine Linguistic Gender. The early spread of the generic
masculine form in Old English and then in Middle English—notwithstanding the
demise of agreement—may have had another, social, reason: men, more than
women, were literate in their majority, and thus were the targets of the written texts.
The ‘masculinization’ of language had its roots in the grammatical system of the
language where masculine Linguistic Gender is the functionally unmarked choice.
This is a feature English shares with a number of other languages, including many
ancient members of the Indo-European family. To avoid a plethora of wordy
alternatives, the generic he was given the support of a law by the English Parliament
in 1850: ‘An Act for shortening the language used in acts of Parliament’ ordered ‘that
in all acts words importing the masculine gender shall be deemed and taken to

196 11 Gender in grammar and society
include females, and the singular to include plural, and the plural the singular, unless
the contrary as to gender and number is expressly provided.’ 24
The lack of a proper generic pronoun continued to be felt as a ‘gap’—the generic
‘he’ was seen as mostly ‘masculine’. Attempts at trying to introduce a new epicene
pronoun into English started in earnest in the early nineteenth century. In 1884,
Charles Converse, an American lawyer and hymn-writer, created perhaps the widest
known epicene pronoun thon (and possessive thons) as a contraction, or a blend, of
‘that one’. The form never got to be widely used—though it did make its way into a
number of dictionaries, among them Webster’s Second New International Dictionary.
Feminists were not the first to feel that the generic use of he reflected sexual bias
within the language. Otto Jespersen (1894: 27–9) believed that introducing a common
gender pronoun would make sexual bias less prominent. In his own words, ‘ladies
would be spared the disparaging implication that the leading poets are men’,ifthon as a
generic third person pronoun replaced he in the sentence: It would be interesting if each
of the leading poets would tell us what he considers his best work. He saw no future for
thon, instead favouring the singular they. Many other alternatives sprang up, among
them borrowings (French on, le, anden), blends (he'er, shem), clippings (e, per), and
new creations (na, ae, ip). 25 Most of them attracted criticism or were neglected, and
none of them made its way into the language used by any of the English-speaking
communities. The title of Dennis Baron’s (1981) extensive account of epicene pronouns—‘the
word that failed’—captures the sad fate of the new coinages.
The most recent ones are the attempts to create special pronouns for the transgender
community—gender-neutral ze (and its possessive hir) was introduced by
students of Wesleyan University. We are yet to see whether and when it will get
established: the article on ze in The Harvard Crimson (Caputo, 2 November 2005)
ends with a quote from a Wesleyan student saying that ze isn’t ‘exactly ubiquitous’
even in Wesleyan; ‘it is more frequently invoked with irony’.
This is somewhat similar to the gap in second person in English: you covers
singular and plural, and may create difficulties in communication. The literary
language offers no solution. But speakers of English come up with various alternatives,
among them y’all or you all, yous, andyou guys—to which we will return
shortly.
The generic use of he came under attack as a direct reflection of male dominance in
the second half of the twentieth century with the rise of the feminist movement (the
major question being whether women were to be included in the scope of he as a
generic pronoun or excluded from it: in the latter case, he is seen as ‘pseudo-generic’). 26
Experimental studies showed a tendency to identify ‘he’ as a male. 27
A linguistic issue—using one pronoun as a functionally unmarked option—soon
became political. In November 1971 a number of women students at Harvard
Divinity School called for a ban on the use of ‘man’, ‘men’, and masculine pronouns
to refer to God or to people in general. A group of leaders in the field of linguistics

11.5 Thwarting ‘sexist language’ 197
reacted to this proposal ‘to recast part of the grammar of the English language’, ina
letter to The Harvard Crimson (the Harvard University periodical) of 16 November
1971. The letter was signed by the departmental secretary and sixteen linguists—
many of them household names in linguistics, including Ives Goddard, Michael
Silverstein, Calvert Watkins, Einar Haugen, Jay Jasanoff, Stephen Anderson, Sandra
Chung, Alan Timberlake, and Robert Underhill.
The authors appealed to the notion of functional markedness (see §7.7), as ‘one of
the fundamental principles which govern the organization of the internal economics
of all human languages’. In their own words,
in the matter of gender, in some cases the feminine is unmarked, in other cases the masculine.
The feminine goose is unmarked—geese can be all male, all female, or of mixed sex, but
ganders are all male. On the other hand the masculine lion is unmarked—contrast the possible
ranges of meaning of lions and lionesses.
Turning to pronouns and terms for people, they continued:
the masculine is the unmarked and hence is used as a neutral or unspecified term. This
reflects the ancient pattern of the Indo-European languages, seen also, for example, in
French: hommes et femmes heureux ‘happy men and women’ (with the masculine form of
the adjective). Thus we say: All men are created equal. Each student shall discuss his paper
topic with his section man. Madam Chairman, I object. The fact that the masculine is the
unmarked gender in English (or that the feminine is unmarked in the language of the
Tunica Indians) is simply a feature of grammar. It is unlikely to be an impediment to any
change in the patterns of the sexual division of labor towards which our society may wish to
evolve. There is really no cause for anxiety or pronoun-envy on the part of those seeking
such changes.
The idea of ‘pronoun-envy’ can be seen in another way: R. M. W. Dixon pointed
out that English-speaking women have a better linguistic deal than do men, as they
do have a pronoun of their own while men have to share their pronoun with the rest
of the world. He experienced the reverse situation first-hand, during his fieldwork
among the Jarawara of southern Amazonia. Feminine gender in Jarawara is the
functionally unmarked option (see §2.3 and §7.7). All pronouns are marked on the
verb with a feminine agreement suffix: one of the difficulties R. M. W. Dixon
encountered in using the language was that he had to talk about himself using the
feminine form (e.g. I walked: feminine). This in no way implies a privileged or
elevated position of women in the Jarawara society (which continues to be maledominated).
As we saw in §7.2, a respected Jarawara woman can be referred to with
masculine gender, ‘promoting’ her to the status of an honorary male.
This is not the way feminist linguists see it—for many, having a feminine pronoun
is an impediment rather than a gain. But the lack of a specifically feminine
pronoun can be felt as a ‘gap’ in the language: we can recall, from §5.3, how a new
partly jocular form temake has been recently introduced into Modern Estonian.

198 11 Gender in grammar and society
Having this form helps translate expressions like ‘his and hers’ into the language—
and we find it used this way in adverts and the like.
And yet the trend against generic, or pseudo-generic, he—and its sexist overtones,
especially for the vociferous milieu of feminists somewhat naive with regard to
general linguistic issues—is winning. Robin Lakoff—writing in 1973 and 1975—
was pessimistic about changing people’s attitudes to the pseudo-generic he which
she used throughout her 1975 classic. And now, forty years on, he as a generic is on
the wane. 28
Any person starting on a new job in a university in an English-speaking country is
given a set of instructions to use ‘gender-neutral’,or‘gender-inclusive’ language. Staff
and students at the University of Melbourne are encouraged to use a gender-neutral
pronoun they (and not he), or reword a sentence to avoid personal pronouns, or
replace he with s/he, she/he, he/she,orhe or she. A dispreferred alternative in 11.4ais
rephrased as 11.4b:
11.4a Where the scholar has been pursuing research he shall present to the Council
a report embodying the results of his work
11.4b Where the scholar has been pursuing research they shall present to the
Council a report embodying the results of their work
And 11.5a should be abandoned in favour of 11.5b:
11.5a The student must present his research to the examiner
11.5b The student must present the research to the examiner 29
As Jenny Cheshire reports (2008: 9–10), ‘it is encouraging that in 2007, in British
university circles at least, it has become unthinkable to use he with an intended
generic sense. Plural subjects are used, or the compound he or she. In my own
university [Queen Mary College, University of London], he or she is de rigueur in
speaking as well as in writing. None of my colleagues would dare to use a generic he
pronoun; if one should slip out accidentally, it is soon corrected, by someone else.
This is surely a sign of change.’
The guidelines for European international organizations (such as the Council of
Europe) follow similar principles. And so do most publishers. In The Cambridge
Australian English style guide, Peters (1995: 332) states that ‘in ordinary usage he/
his/him seems to be losing its capacity to be common and generic’. The generic
‘unsex’ they is becoming the preferred option (with the singular reflexive themself no
longer rejected as ‘incorrect’). In her analysis of a variety of sources in British
English, Wales (1996: 125–33) shows how they comes to be used more and more
frequently to refer to nouns whose gender is not specified, in current usage by native
speakers. Similar guidelines for ‘non-discriminatory language use’ are at work for job
advertisements in Welsh: if an advertisement is phrased in the singular masculine

11.5 Thwarting ‘sexist language’ 199
and feminine pronoun are used together (e.g. iddo ef/iddi hi (for.3sg.masc him/for.3.
fem her) ‘for him/for her’), or a plural form ‘they’ (which does not distinguish
genders) is used. 30
Over the last decades we—as lingusts—have witnessed a rather remarkable change
in the grammar of English which has affected the closed class of pronouns. The
sphere of use of the generic he is now drastically diminished. But this is not to say that
the generic he is fully obsolete; one hears this use in casual conversations, and in the
literature. In a recent novel by Ian McEwan Solar (2011: 71), the main character,
Michael Beard, walks into a change room on a ship in the Arctic and discovers that
‘someone has put his gear on Bear’s station’. The group of people on the ship is mixed
sex: ‘someone’ may have been a man or a woman. In Jenny Cheshire’s experience
(2008: 10), ‘several students still write he in their written work’, so much so that ‘there
is no guarantee that, if change has occurred amongst some people, it will persist’. For
some, this is still an ideological issue: as Michael Silverstein (1985: 253) put it, the use
of generic he (or lack thereof) ‘is turned into an index of a certain absence or presence
of ideological solidarity with the reformers’.
Despite all this, the masculine gender forms in English have not lost their functionally
unmarked status. They remain, in Goddard et al.’s words (1971), as ‘the
unmarked and hence [ ...]used as a neutral or unspecified term’. This can be seen in
the new plural form for second person in English you guys which contains the form
guy with an erstwhile male reference. You guys can be directed to men, women, or to
a mixed group, and is used by women and by men. (I first heard it used by a highprofile
female financial analyst talking to her teenage daughters.) You guys can hardly
be considered ‘sexist language’: there are certain limits beyond which even the most
ardent reformers cannot go.
Unlike English, German has a generic pronoun man ‘one, person’ which is
perfectly gender-neutral. However, it sounds similar to Mann ‘man, person’ (from
which it comes historically). To avoid potential masculine bias, a new pronoun frau
‘generic (woman)’ based on Frau ‘woman’ was introduced. According to a study in
Storjohann (2004: 313) itfirst appeared in the Duden dictionary of German in
1997. 31 This indefinite feminine form has partly made it into general usage (though
it does not appear in many dictionaries). It differs from man in that its reference is
exclusively feminine. The pronoun differs from the generic man in its meaning, and
in its syntactic behaviour. For instance, in complex sentences, frau ‘she: generic’ can
be anaphorically replaced by sie ‘she’,asin11.6. In contrast, man ‘generic person’ can
only be replaced by man, and not by er ‘he’ or sie ‘she’—as shown in 11.7.
11.6 Wenn frau sich bemüht, kann sie es schaffen
‘If one:feminine makes an effort, she can make it’
11.7 Wenn man sich bemüht, kann man/*er es schaffen
‘If one:generic makes an effort, one (*he) can make it’

200 11 Gender in grammar and society
That is, the pronoun man has a much less gender-specific meaning than frau. In a
further attempt to redress the balance between indefinite pronouns, feminist linguists
have coined another indefinite mann ‘generic male person’ as a counterpart to frau,
in addition to a number of female-specific alternatives to established generic pronouns
which contain man ‘man’, e.g. jedefrau ‘every woman’ (a counterpart of the
established form jederman ‘everybody, “every man”’) and jefraud ‘some women’ (as a
counterpart of jemand ‘someone’). It remains to be seen whether or not these will
become fully established in the language beyond feminist literature.
A generic gender-neutral pronoun in Norwegian has been an issue for the Norwegian
language council since the 1980s. The feminine pronoun ho or hun ‘she’ was
suggested as a replacement for the generic masculine han (alongside the indefinite
pronoun ein/en ‘one’, replacing man ‘one’ which is obviously derived from mann
‘man’, the noun folk ‘people’, or a plural form which does distinguish genders). The
feminine form was finally accepted in children’s schoolbooks in 1989—this was
considered a major victory for feminists. 32
In the official guidelines for language use in Swedish, masculine pronoun han is
used for generic or gender-neutral reference. The Language Council of Sweden, a
major authority on language planning, also supports den ‘it’ or han eller hon ‘he or
she’ as a way of avoiding the generic masculine. A further innovation is a new genderneutral
pronoun hen increasingly used among lesbian, gay, bisexual, and transgender
people. The form may have originated in a blend of three pronouns: han ‘he’, hon
‘she’, and den ‘it’; an alternative hypothesis is that it comes from Finnish hän ‘he/she/
it’. The new form has not been supported by the Language Council; nevertheless it is
making its way into the language, especially for feminists and members of the queer
community. It is as yet hard to predict whether it will spread beyond these. 33
The issue of masculine gender as a functionally unmarked choice in agreement
on adjectives and verbs does not arise in English or in German. 34 But it continues to
be an issue in French and Spanish. Following the recommendations of The French
Council of Europe Guidelines, French occupational terms have to always be given in
a feminine and a masculine form. But agreement follows the established grammatical
pattern: only a masculine form will be used with two conjoined nouns, such as un
expert et une expert sera nommé ‘an expert (masc.sg) and an expert (fem.sg) will be
named (masc.sg)’. 35
Similarly, in Spanish a mixed group has to be referred to with masculine gender:
‘one single man in an audience will be sufficient for the speaker to address the
listeners by masculine terms, e.g. vosotros (masculine plural) “you” or los presentes
(masculine plural) “the ones present”. To use feminine forms in such cases would be
tantamount to breaking the rules of grammar. The same is true with respect to
agreement.’ For instance, one would say Juan (m) y Juana (f) son campesinos
extraordinarios ‘Juan and Juana are extraordinary (masculine pl) peasants (masculine
plural)’, despite the fact that Juana on her own will be referred to as campesina

11.5 Thwarting ‘sexist language’ 201
extraordinaria. 36 The European Parliament guidelines for Portuguese reflect such
usage: they explicitly recommend the use of generic masculines whenever a group of
different sexes is referred to:
11.8 os antigos primeiros-ministros
the:masc.plural former.masc.pl prime.masc.pl-minister.masc.pl
Maria Pintassilgo, Mário Soares, Margaret Thatcher e Tony Blair
Maria Pintassilgo, Mário Soares, Margaret Thatcher and Tony Blair
‘the former prime-ministers Maria Pintassilgo, Mário Soares, Margaret Thatcher,
and Tony Blair’
This takes us back to Goddard et al.’s (1971) point: the use of masculine forms as
functionally unmarked generics is a fact of grammar which lies beyond the reach of
language reformers.
11.5.2 Fighting the ‘generic masculine’ throughout the language
As women started entering professions previously considered ‘male only’, the question
of ‘feminization’ of job titles became ripe. In §7.6, we briefly mentioned the issue
of feminine forms for job titles and positions. For the first time in the history of
France, women constituted more than one-third of ministers in Lionel Jospin’s
government (elected in 1997). The six women ministers explicitly insisted on being
addressed as Madame la ministre. The meeting of the Council of Ministers on 17
December 1997 decided that in the future women should be addressed with feminine
titles. The proposal was immediately put into practice, and in March 1998 the Prime
Minister legalized it by producing a decree Circulaire du 6 mars relative à la
féminisation des noms de métier, fonction, grade ou titre (Jospin 1998). The official
recognition was met with a protest from the Académie française, the main authority
on French language and culture.
In 1999, the guidelines of the Institut National de la Langue Française offered an
extensive list of feminized job titles. Feminization of titles and professions in modern
French has achieved a certain measure of success. And yet, as Pastre (1997: 376) put
it, ‘The honor and prestige of masculine forms fascinate many women, who are
happy to accede to honors that have until now been reserved for men . . . When the
writer Hélène Carrère d’Encausse was accepted into the Académie française, for
example, her daughters made it clear that they did not consider her une écrivaine
(a woman writer) and did not approve of the feminisation of job titles.’ This author,
now the Permanent Secretary of the Académie, is referred to with the masculine form
as le secrétaire (and not as a feminine la secrétaire) ().
Marking both genders—especially in the written language—is an uncanny way of
creating compromise structures. For instance, a feminine marker in upper-case letters

202 11 Gender in grammar and society
can be added to a masculine form, to create an all-embracing form which would
subsume men and women, e.g. manifestant-E (demonstrator(masc.sg)-fem.sg ‘demonstrator
(masculine and feminine)’. 37
In other Romance-speaking countries, the situation is different. The only female
minister in the ill-fated Collor government in Brazil (1990–2), Zélia Maria Cardoso
de Mello, was always referred to as ministra (feminine form of ministro, minister)
and mentora (feminine of mentor) of Collor himself. Most job titles in Brazil are used
in a masculine and a feminine version with no negative overtones attached to either,
e.g. juiz ‘male judge’ versus juiza ‘female judge’, professor ‘male professor, teacher’,
professora ‘female professor, teacher’, orientador ‘male supervisor’ and orientadora
‘female supervisor’, chefe ‘male chief, head (of an organization)’, chefa ‘female chief,
head (of an organization)’. The feminine form of presidente (presidenta) was reintroduced
by law in 2012, and is used to specifically refer to the first female president of
Brazil, Dilma Rousseff (see §7.5).
The first official guidelines for non-sexist language use in Spanish (1988) and the
subsequent UNESCO guidelines explicitly stated the ‘propagation of the feminisation
of words explicitly referring to women’, and ‘replacement of the masculine form for
generic reference by epicene words or by splitted [sic!] constructions’. The sign @ has
been suggested as a way of graphically combining masculine and feminine endings—
so l@s alumn@s will be a blend of los alumnos (masc.pl students.masc.pl) and las
alumnas (fem.pl students.fem.pl). Bengoechea’s (2011: 48) results show that such
blends, and feminine forms of professions, are being more and more accepted. This
echoes Nissen’s (2013) conclusions—that speakers of Spanish in 2005 showed much
less sexist bias in interpretation of generic masculine forms than they did in 1995.
Women have become more visible in public life—and the language appears to have
become more inclusive, and perhaps more receptive to feminine forms. 38
Guidelines for non-discriminatory usage in English warn against ‘false generics’—
that is, man and derivations which include -man. This has indeed resulted in general
avoidance of gender-specific forms such as Chairman: we talk about Chair, or
Chairperson. What Goddard et al. (1971) phrased as ‘Madam Chairman, I object’
sounds distinctly obsolete: one will now say ‘Madam Chair’. Guidelines for nongender-biased
use in German, under the title ‘More women in language’ (Braun
2000b), focus on gender equality in derivations. The guidelines are being followed—
at least in official contexts. 39
Dictionaries tend to lag behind the times in how they represent women. Hampares
(1976) looked at 105 job denominations in The Spanish Royal Academy dictionary
and two Spanish–English dictionaries, and concluded that only a masculine form was
given for high-ranking jobs, such as científico ‘scientist’ and diplomático ‘diplomat’.
Lower-ranking jobs such as sirvienta ‘servant’ or asistenta appeared in the feminine
form. Quite a few feminine forms were in the dictionary alongside their masculine
counterpart but were interpreted as ‘the wife of . . . ’, e.g. embajador ‘ambassador’,

11.5 Thwarting ‘sexist language’ 203
embajadora ‘wife of ambassador’. There were no feminine forms for jobs such as
mecánico ‘mechanic’ or ingeniero ‘engineer’. Nissen (2002: 266) reports that the latest
edition of the Academy’s prescriptive dictionary did not show much improvement
on the earlier version.
French dictionaries have not fared better. In her critique of English–French
dictionaries (Collins, Harrap, Larousse, and Oxford-Hachette, 1978–2001), Élisabeth
Campbell (2004) pointed out a strong masculine bias: mostly masculine singular
forms are given in French translations, and the generic he is prevalent in the English
part. Within dictionary entries, there is a strong masculine bias; examples involving
women are focused on domestic activities. Feminine equivalents are given for some
professions but not for others: Oxford-Hachette contains the word ministre as
masculine and feminine, and the feminine form of député ‘delegate’ (députée ‘female
delegate’); feminine forms of other nouns, including avocate ‘female lawyer’, éditrice
‘female editor’, jurée ‘female member of a jury’, were not included. This in itself may
have nothing to do with being anti-feminist. Dictionaries are prescriptive and
notoriously conservative, reluctant to embrace new terms, and to keep pace with
language change. Indirectly, their conservatism serves to reinforce the masculine bias
and is likely to cause offence. 40
The ‘generic masculine’ is not perceived as sexist in mainstream Hebrew-speaking
society. We can recall, from §2.3 and also §7.7, that in Hebrew and in Arabic, masculine
forms (in singular and in plural) have generic reference. In Hebrew, the masculine
singular pronoun ata can be used in general, to talk about a woman or a man, much like
English one or you. A Hebrew-speaking Israel fashion designer explained why she had
become self-employed: ‘One day you (masc.sg) suddenly realize that you (masc.sg) are
(masc.sg) a mother and all you (masc.sg) wish (masc.sg) to do is stay at home with the
children.’ 41 Using generic masculine is an unconscious choice. At a meeting of the
Israeli Parliament (Knesset) committee for the promotion of the status of women in
March 2004, dedicated to the address of a mixed-gender audience, the representative of
the Hebrew Language Academy—a major regulatory entity for the language—opposed
all attempts to introduce feminine forms in their generic reference. She argued that this
would go against the internal structure of the language. The point she made was
reminiscent of Goddard et al. (1971) (masculine and feminine endings are in bold
face): ‘When I talk about giving birth, I say “when giving birth (masc.pl) (ksheyoldim)”.
Ican’tsay“when giving birth (fem.pl) (ksheyoldot)”. You will agree with me that these
forms are not masculine, they are unmarked. It is true that historically these were
masculine forms, but essentially they are forms that include the female. When I say
‘writing.fem.pl’ (kotvot) I exclude you [the addressee was a man], but when I say ‘writing
(masc.pl) (kotvim) [masc.pl], I don’t exclude myself. I am in there. I am telling you as a
woman that it doesn’t disturb me.’
There appears to be little support for the feminist language reform in Israeli
society. Jacobs (2004: 237) tells a cautionary tale about Na'ama, an elected official

204 11 Gender in grammar and society
on the governing council of a major Israeli city. Her term in office was marked by
struggle against institutional practices that limited women’s access to political and
economic power, and her individual attempt to reform the language. She attempted
to get city council stationery that would use the female title for ‘member of city
council’—xavrat mo'etsa rather than the masculine xaver mo'etsa which used to be
used for males and female alike before her. At first, her request was denied, and only
after repeated struggle, honoured: after two years, the municipality ended up printing
official council stationery with the feminine title—but Na'ama was the only one to
use it. Other female members of the council chose not to. As Jacobs (2004: 245) puts
it, this example underscores ‘the marginal position of feminists in the mainstream
Israeli socio-political context’. The resistance, or rather, indifference, to the generic
masculine forms in Hebrew does not in any way imply that women have a lower
status than men. 42 It is just that the feminist movement has not impacted the
language—social and linguistic change do not have to proceed simultaneously.
11.5.3 Bias in address terms and naming patterns
In most European languages, courtesy titles and honorifics treat women differently
from men. Women are addressed on the basis of their marital status: a married
woman will be traditionally addressed as a Mrs in English, Mevrouw in Dutch,
Madame in French, Senhora in Portuguese, Señora in Spanish, Signora in Italian,
Fru in Danish, and Frue in Norwegian. An unmarried woman will be addressed as
Miss in English, Juffrouw or Mejuffer in Dutch, Mademoiselle in French, Fräulein in
German, Senhorita in Portuguese, Señorita in Spanish, Signorina in Italian, Fruken in
Danish, and Froken in Norwegian. In contrast, there is just one way of addressing a
man; special courtesy titles for young unmarried men, such as Master in English and
Jonge Heer in Dutch, are not often used and more markers of age than of marital
status. Surnames may reflect the marital status of a woman. In Lithuanian, an
unmarried girl will be given her father’s surname with the suffix -aitė, -ytė, -utėʼ or
-tė. A married woman’s surname will be derived from her husband’s name with the
suffix -ene. There are no markers which would reflect the marital status of a man.
The marital status of women has traditionally been flagged—in contrast to that of
men—signalling the lack of equality between Social Genders. As Silverstein (2015)
put it, the private becomes political (echoing a slogan of the second wave feminist
movement). The term Ms in English was registered in the Oxford English Dictionary
Online as early as 1901 and was later adopted as an alternative to avoid marking
marital status (though it is not accepted by everyone). 43 The reform in German went
even further. Since the 1970s, the form Frau is the only option to address a grown-up
woman (no matter whether married or not). The term Fräulein ‘miss, unmarried
woman’ is all but obsolete. 44 This is evidence of progress in eliminating male–female
asymmetry in address terms.

11.5 Thwarting ‘sexist language’ 205
Western naming practices have—traditionally—ensured that women remain
invisible. Children in Western societies are given the surname of the father. In
Spanish- and Portuguese-speaking countries, children are given the mother’s and
father’s name, but are typically referred to by the surname of the father. The Spanish
and the Portuguese traditions differ in the order of names: for Spanish speakers, the
mother’s name comes last. For instance, Gabriela García Salazar has a father with the
surname García, and a mother with the surname Salazar. She would be referred to as
García—which is her main surname in her native Peru. In Portuguese, the order is
the opposite: the mother of Rafael da Silva Brito has the surname da Silva, and his
father’s surname is Brito. He is referred to as Rafael Brito, for short. During her
studies in Brazil, Gabriela Garcia Salazar was known as Gabriela Salazar, following
the Brazilian custom. Patronymics in Slavic-speaking traditions reflect the name of
the father—but not of the mother. The maiden name of my father’s paternal
grandmother was Nina Kirillovna Tomasheva (1868–1943). I thus know the first
name of her father, Kirill, and his surname (Tomashev, the masculine version of
Tomasheva). But I have no idea what her mother’s name was. By keeping this secret,
and by adopting her husband’s name, Nina Kirillovna Aikhenvald hid information
about her aristocratic background—which was by no means an asset under the Soviet
regime. The language came to her aid. The woman’s identity, name, and achievements
were lost and obscured. So far, no attempts have been made to thwart the
practice of patronymics. 45
Traditionally, Western women were expected to adopt their husbands’ surnames.
At present, this is no longer the case; and many professional women keep their own
name on marriage. As a consequence of the liberation of women in the Soviet
Union, many educated professional women kept their maiden names—for some,
giving up their name was like giving up their own identity. As my mother said to
me, ‘if I had changed my surname (Gerlin) to that of my husband’s (Aikhenvald),
I would have stopped being me’. The guidelines for non-discriminatory language
from Melbourne University overtly state that ‘it has become increasingly common
for women to keep their birth names’. ButPauwels(1998: 59) remarks that ‘women
who do not conform to this practice still face many practical and attitudinal
obstacles’. Mills (2003b) undertook a study of naming practices among professional
women in Britain. Over half of thirty-six subjects did take their husband’s nameon
marriage—but they gave different reasons for the choice. One woman took a
husband’s name because of lack of loyalty to her father. Another one liked the
husband’s name better. Naming practices have become negotiable, at least in middleclass
Britain.
A dangerous reverse trend is at work in developing countries which embrace the
influence of Christianity and Western naming practices. In many societies of New
Guinea, women and children have recently been obliged to take on their husbands’
and fathers’ names, following the Western colonial practice.

206 11 Gender in grammar and society
Personal names are a major asset of the traditional Manambu people. In creating a
person’s identity, both maternal and paternal name still play a role. The Manambu
personal names (associated with a totem or a totemic ancestor) are ‘owned’ by
subclans, so that every person carries names of his or her mythological forebears. 46
The first, and the main name bestowed upon a person, would be a name belonging to
their paternal clan. The name is termed ap-a-sə (bone-LINKER-name), literally, ‘the
bone name’ (we can recall, from §3.4.2, that ‘bone’ is associated with masculinity and
patrilineal inheritance). Numerous paternal names bestowed upon a person by their
paternal relatives throughout their lives indicate where people belong. Names
bestowed upon a person by their maternal relatives protect them from evil spirits
while they were alive. This naming practice is still alive—but only barely so. Many
younger people and children can hardly remember their traditional names.
Nowadays, children who go to school, and adults who have identity cards and
passports, have to go under their father’s name. Married women are obliged to take
their husbands’ paternal names (something unheard of in traditional times). The late
Gaialiwag whose husband’s name was Yirvi was a very traditional speaker of Manambu,
with an impressive knowledge of totemic lore and names. She never referred
to herself with her husband’s name until her various stints in the hospital in Wewak
(the provincial capital) where she was made to follow this now standard practice
for Papua New Guinea. In the last story she told me in 2013 she introduced herself
as Wun Gaia, Yirvi du sawa (I Gaia, Yirvi man name+COMITATIVE) ‘I am Gaia, Yirvi
by man’s name’. This was accompanied by an ironic raise of eyebrows and shoulder
shrug—later explained to me as her dismissive reaction to this ‘new’ fashion of
referring to a woman by her husband’s name. Angela Filer, a local politician and a
highly proficient speaker of Kwoma, a neighbouring language, was scathing about the
westernized imposition of husbands’ names on married women, which background
and ultimately destroy their identity. 47
Names are just one aspect of a dangerous trend termed ‘repatriarchialization’.
Abbink (2015: 294) comments on how traditional division of labour between men
and women, and special female practices in Suri communities in Ethiopia, weaken
under the newly introduced patriarchic practices of Evangelical Christianity; ‘the
gender egalitarianism that existed ...is undermined’. Colonial influences and the
destruction of traditional balance between male and female Social Genders among
the Igbo of Nigeria led to marginalization of women (Amadiume 2015: 147–61).
11.6 Expression of Linguistic Gender and social change: a summary
The meanings of Linguistic Gender may reflect the status, and the occupations, of men
and women. So do other noun categorization devices applied to humans, including
numeral classifiers. Linguistic Genders, especially the use of pronouns and derivational
forms, may mirror social changes through language reforms. The most remarkable

11.6 Expression of Linguistic Gender and social change 207
success story for feminist reformers is that of a gradual demise of generic man and
generic he in English, as a way of removing the masculine bias. A number of earlier
attempts to create a generic sex-neutral (or ‘epicene’) pronoun, variation in anaphoric
agreement, and other options—such as a generic they, with singular reference—
prepared a fertile ground for this change. The demise of generic he was not an
innovation: rather, this was an engineered enhancement of already existing patterns.
If the language is a guide to social reality, the way Linguistic Gender is manipulated
is the most salient part of it. Incidentally, these features of Linguistic Gender and
the ways in which they reflect societal stereotypes and ingrained attitudes to male
and female practices fully debunk the myth of gender as a ‘useless’ decoration in a
language.
Scholars of gender should be careful in making their associations too straightforward.
Having feminine gender as the functionally unmarked choice does not mean
that men and women are socially equal, or that women occupy a privileged position.
A conventionalized generic use of an erstwhile feminine pronoun does not in any
way imply a privileged status of women. In Modern Italian, two polite address forms,
Lei and Ella, are identical to third person feminine pronouns, even though they are
used equally in addressing respected males and females. The reason for this rests
with the history of the language: honorific expressions used to indicate deference,
respect, and admiration—similar to English ‘your honour’, ‘your Excellency’, and
‘your worship’—tended to comprise nouns of feminine Linguistic Gender, e.g. la
vostra signoria ‘your lordship’, la vostra eccellenza ‘your Excellency’. These nouns
could also be referred to, anaphorically, with third person feminine pronouns ella
and lei. They gradually became general forms of polite address. 48
Linguistic change never happens quickly enough. Despite numerous substantial
efforts in formulating, and imposing, guidelines for eliminating gender-biased language,
some analysts remain pessimistic. At the end of their comprehensive analysis
of European international organizations and language reform, Teso and Crolley
(2013: 155) remark on ‘a disturbing gap between language policy initiatives and
practice’, leading to ‘the lack of success in the feminisation of language’, even in the
languages of Western Europe. A full ‘feminization’ of any language is an impossible
task: it would be linguistically naive to try and override grammatical rules—such as
masculine agreement forms for mixed sex groups—for the sake of political struggle to
make men and women equal. And using a formation you-guys to refer to a mixed
group of men and women can hardly be seen as a mark of male supremacy. Linguistic
reforms can only go so far: echoing Goddard et al. (1971), it is important to
remember that grammatical features, such as markedness, can hardly be changed
by guidelines or decrees—no matter how much feminist linguists may lament the
‘sexist’ nature of masculine agreement for mixed groups.
Using Robin Lakoff ’s (2004: 171) words, ‘language discrepancy makes manifest
social inequality’ in yet another way. We can recall how in gender-exclusive

208 11 Gender in grammar and society
languages (discussed in §9.1), male and female dialects may correlate with the relative
status of the Social Genders (see §9.1.5). Along similar lines, ‘male-only’ genres are
often seen as more prestigious and showing more knowledge (§§10.1–3). A covert
male bias may become detrimental to the survival of female genderlects. We can
recall, from §9.1.4, that the Chukchi language is severely endangered, and hardly
learnt by children. A partly artificial ‘standard’ language, promoted throughout the
Soviet period, was based on the men’s dialect—women were made invisible and not
heard. Chukchi teachers and radio announcers use the male variety (female announcers
continue using the female forms in private). Discourse about language preservation
is framed entirely in terms of men’s language. The male bias is creeping in.
Social change and growing equality affect the use of genderlects in westernized
developed societies. The partial levelling of genderlects in Japanese and in Thai goes
together with women getting access to leading positions in society, and phasing out
social asymmetries in Social Genders.
More than just language forms are at stake. We can recall, from §8.4 and §11.2,
how in many cultures of patriarchic type women are stereotyped in predominantly
negative terms. More often than not women, but not men, are portrayed as
sexualized objects. As Cheshire (1985: 24) puts it, ‘newspapers give personal details
about women, but not about men, so that we read about “blonde, shapely, motherof-three
Mrs Smith”, but not about “bronzed, muscular Mr Jones”’. As demonstrated
by Baker (2014) in his analysis of adjectives used with ‘man’ and ‘woman’ in
the Historical Corpus of American English, the imbalance appears to be slowly
changing (with men’s bodies being sexualized as much as women’s). As Jenny
Cheshire (2008: 10) puts it in her reappraisal of linguistic sexism, ‘there have been
changes in the use of English in the last twenty-three years that suggest to me that
our language is freer than before of masculine bias, even if the reform is far from
complete. What needs to be determined now is the extent to which our thinking is
equally free of masculine bias.’
NOTES AND SOURCES
1. A tight relationship between language and reality has been captured by the Sapir–Whorf
hypothesis: see Pauwels (1998: 82–3) and Talbot (2010: 15–16), for feminist approaches.
Excellent discussion is in Salzmann (2004: 42–69).
2. See Leger (1998: 206–7) and Dinslage, Leger, and Storch (2000: 125). Human nouns are
assigned to Linguistic Genders based on their sex.
3. Vajda (2004: 17–18); Werner (1997: 92–4).
4. Note that, so far, sexist ‘positive’ associations of ‘man’ and negative associations of ‘woman’
have been documented exclusively for patriarchic societies. We lack any information on the
images of men and women in traditional matriarchies, where women traditionally play a

11.6 Expression of Linguistic Gender and social change 209
leading role in inheritance and governance. In her study of matriarchies, Goettner-
Abendroth (2012) does not mention the issue of names, naming, or language.
5. Migge (2001: 99–100) on this and other pairs. Unfortunately, little information is available
on overtones of gendered derivations in other languages spoken in matrilineal societies.
6. Sahoo (2003: 252–3); see Allan and Burridge for this general trend (2006: 81), and
Gregersen (1979) for feminine terms and insults cross-linguistically.
7. See Barnhart (1988: 639, 668), the Oxford English Dictionary Online (entry mistress, sense
7 for an earlier date). See also Lakoff (1975: 59–60), Henley (1989: 60), Cheshire (1985: 22).
8. See Baron (1986: 121–2) for the history of these feminine forms in English.
9. See Aikhenvald (2002: 227).
10. In the Oxford English Dictionary Online, the meaning of man up is given as ‘to demonstrate
manliness, toughness, or courage when faced with a difficult situation; to take
responsibility; to own up’.
11. See also Aikhenvald (2013a). My corpus of Tariana contains several similar examples of
the word ‘man’.
12. Hasselblatt (2015: 141–2).
13. Khaidakov (1963: 49–50) on Lak; Mithun (2014: 137–8) on Mohawk.
14. Anderson and Janson (1997: 34–5), Joe Tsonope p.c.
15. Lu(2012: 83–4, 101–2, 115, 119–21) on classifiers in Maonan and the correlations between
prestige and the use of human classifiers with women. Raised numbers indicate tones.
16. Khaidakov (1963: 50).
17. See also Vinogradov (1947: 62).
18. See Yokoyama (1999: 420–1), Fesenko and Fesenko (1955: 161).
19. In his state-of-the-art description of word formation in twentieth-century Russian,
Lopatin (1970: 125) stresses that many female forms are only used if the sex of the person
needs to be emphasized.
20. See a similar example in note 14 in Rothstein (1973: 465); Yokoyama (1999: 418–20), and
Martynyuk (1990).
21. See Pauwels (1998: 34–80), Cheshire (1985, 2008), Henley (1989: 60), Teso and Crolley
(2013). Investigations of the ways women speak tie in with the history of feminism (see
Walters 2005, for a brief history; and Mills (2003b) for a comparison of Second-wave
feminism, from the 1960s until the 1990s, and the subsequent Third-wave feminism. See
Lakoff (2004: 205) for a brief reappraisal of methodology used by feminist linguists, and
warnings against the limitations of any one methodology—be it the Conversation Analysis
or the Critical Discourse Analysis.
22. See Hegardty (2014); and further discussion in Key (1975).
23. See the Oxford English Dictionary Online; the form gradually weakened to men and to me.
One of the reasons for its demise may have been its similarity to the first person form me.
24. Baron (1981: 84), Evans and Evans (1957: 221).
25. See Baron (1981: 88–96) and (1986: 198ff.; 205–9) for their chronology. Stotko and Troyer
(2007) reported a newly coined generic third person singular pronoun yo in Baltimore
public schools.
26. A history of generic masculine ‘he’ and attempts to replace it with ‘they’ and other forms is
in Newman (1997), Baron (1986: 191–7), and Curzan (2003: 70–9).
27. Echoing a sweeping statement by Penelope (1990: 94), that ‘in English all persons are
assumed to be male unless otherwise specified’. See §7.6, and also Henley and Abueg
(2003), Henley (1989), Cheshire (1985) and Miller and Swift (1981: 44–60).

210 11 Gender in grammar and society
28. In her guidelines to changing sexist language, Pauwels (1998: 127) outlines a number of
alternatives—recasting a sentence in the plural, using passive, repeating a generic noun,
replacing he with singular they or he or she, and many further variants, such as s/he, she/he,
he/she.Wefind similar guidelines in the Handbook of non-sexist writing for writers, editors
and speakers by Miller and Swift (1981).
29. Watch your language: guidelines for non-discriminatory language. University of Melbourne.
30. See Teso and Crolley (2013: 144–5) for details on institutional guidelines. The singular
themself goes back to early Middle English (the Oxford English Dictionary Online);
nevertheless, in the late 1990s I encountered resistance from some publishers. This echoes
Greenbaum’s(1991: 172) appraisal of the singular they (and themself )as‘faulty’. In a later
edition of his English grammar, Greenbaum (1996) predicted that the singular they would
become more acceptable. See Awbery, Jones, and Morris (2002: 326–7) for Welsh.
31. See Kochskämper (1999: 464) on masculine overtones of the pronoun man. The reader
will notice that nouns in German are written with the capital initial letter. Pronouns
(derived from nouns) are written with a small letter.
32. The two forms of the feminine pronoun represent two varieties of Norwegian: Nynorsk
ho/Bokmål hun; see Bull and Swan (2002: 244–6).
33. See Hornscheidt (2003: 360–1), Hornscheidt (2006) for attempts to use the feminine
pronoun hon as a generic; and Milles (2011: 22–3, 27–8) for a history of reactions to the
feminist language planning and the new generic pronoun.
34. We can recall that gender agreement in German is based on formal assignment of
Linguistic Gender in the singular; and there is one plural form without any gender
distinctions.
35. Teso and Crolley (2013: 145) lament that ‘the female remains invisible in structures with
plural nouns, adjectives and verbs which include both male and female’, in such examples
in French.
36. See Nissen (2002: 257–8) on Spanish. Similar rules apply in Italian (Marcato and Thüne
2002) and Portuguese (Endruschat 2015).
37. See Burr (2003: 128–34) for the history of feminization of titles in France; Abbou (2011)on
‘double gender marking’ in the written language. See Yaguello (1991: 118–38).
38. The situation in Italian appears to be unstable: see Marcato and Thüne (2002: 206), on
general preference for generic masculine forms.
39. See also Bussmann and Hellinger (2003). See also Vasvári (2015) for Hungarian and
Hasselblatt (2015) for Estonian.
40. The feminine form la soldate ‘female soldier’ (alongside the masculine noun le soldat) and
la professeure ‘female professor’ (alongside masculine le professeur) were added to Oxford
Language Dictionaries in 2008. Other dictionaries, including The Collins French Dictionary
and Grammar (2014), contain a negligible number of feminized forms. Lynda Mugglestone
(2011: 74–6) comments on the reluctance of dictionaries to embrace innovations, and
include feminized titles. As Cheris Kramarae (1992: 146) put it, dictionaries can be seen
as ‘instruments of social control’ which focus on the language of men and male authors
backgrounding women’s voices.
41. Example from Sa'ar (2007: 413). See Sa'ar (2007) on Hebrew and Palestinian Arabic, and
on ata ‘you masculine’ as a generic form translatable with English ‘one’, and Hachimi
(2001) on Arabic.
42. Israeli society does not exclude women from high administrative positions (at least twelve
women have served in Israeli governments since the 1970s; Golda Meir has so far been the
only female Prime Minister (1969–74)).

11.6 Expression of Linguistic Gender and social change 211
43. See Pauwels (1998: 58–63), Braun (2000a), and Spender (1980: 27) for the criticism of such
practices from a feminist perspective. See Miller and Swift (1981: 128–9) onMs.
44. Duden Dictionary online.
45. Just a handful of peoples have an established tradition of matronyms—or surnames based
on the mother’s name: see Haviland et al. (2014: 130).
46. See Harrison (1990) and Aikhenvald (2008).
47. See Sini (2011: 195–6) on the gradual replacement of matrilineally inherited names by
surnames based on fathers’ names in traditionally matriarchic communities in Malabar
in India.
48. See Maiden and Robustelli (2007: 459); Maiden (1998: 178ff.). The spread of these forms
may have been facilitated by Spanish influence, and copying of Spanish forms like Vuestra
Merced ‘your honour’. Jacqueline Brunet (2003) offers an entertaining description of how
Italian linguists and lay people (including members of the National Fascist party in the
1930s) tried to eradicate the feminine pronoun as a formal term of address. In their view,
this was a lamentable trace of a ‘damned’ Spanish influence. Needless to say, these attempts
have failed. The form Vuestra Merced in Spanish gave rise to the polite second person
Usted.

12
The heart of the matter: envoi
Gender is a wondrous construct, multifaceted in its functions and meanings. It
permeates and shapes the world we live in. We can now recapitulate the most salient
points discussed so far. The notion of gender has three sides to it.
I. The three faces of Gender
• First, NATURAL GENDER (or sex) reflects the dichotomy between women and men,
or females and males.
• Secondly, SOCIAL GENDER reflects the social implications, expectations, and norms of
being a woman or a man. Social Gender is a status to be achieved within a society.
• And thirdly, the nouns in a language may divide into LINGUISTIC GENDERS. Then
one class will be marked in one way, and another class in another way. The class
which includes words referring to females will be called ‘feminine’. The one
which includes words for males will be ‘masculine’. Linguistic genders are
defined by their male and female members, but often extend beyond those.
And this is where the trouble begins. Linguistic Genders for objects without
Natural Gender can be shrouded in mystery. Linguistic Gender of living beings—
animals, birds, insects, and so on—and of natural phenomena may reflect their role
in legends and myths. In Tunica, the sun is believed to be a mythical man, and
belongs to the masculine gender. In Dyirbal, it is a mythical woman, and so is
classified as feminine. Every noun in a language will belong to a linguistic gender,
but not every instance will have an explanation. This has earned linguistic genders a
reputation of being a hard nut to crack, among some impatient linguists and hasty
language learners.
We now recapitulate a few home truths about Linguistic Gender.
II. Linguistic Gender in a nutshell
Linguistic Gender is typically expressed on adjectives agreeing with a noun, on
pronouns, and sometimes also on the noun itself. We know that the noun menina
How Gender Shapes the World. First edition. Alexandra Y. Aikhenvald.
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.

12 The heart of the matter 213
‘girl’ in Portuguese is feminine because it requires the feminine form of an article
(uma ‘a’) and the feminine form of an adjective, and also because it ends in ‐a. Every
noun in a language will be assigned a Linguistic Gender. There will always be a
meaning to a gender. But not every instance of it will be explainable. Chapters 2, 3,
and 4 focus on what Linguistic Genders mean and what they are good for.
Linguistic Genders can have complex histories. They may develop out of words for
man or woman, or out of other sources. Children acquire them at an early age. And if
languages are in contact, the chances are that their Linguistic Genders will influence
each other. Language engineering, and reform, play a role in reshaping Linguistic
Genders. More on this in Chapters 6 and 11, and also in section VI below.
Linguistic Gender of women and men is special: it is where the three faces of
Gender interact. This takes us to the next point.
III. How the three faces of gender relate to each other
The three faces of Gender—Natural, Social, and Linguistic—are inextricably linked.
They are like a chain of elements feeding into each other: this is what we see in
Scheme 12.1.
Biological men and women acquire their male and female social status as they
reach maturity. Linguistic Genders reflect biological malehood and femalehood.
They also mirror the social status of men and of women. Through Linguistic
Genders, objects can be endowed with features of natural and of social genders.
This is how Linguistic Genders and gender-based metaphors make the world come
alive. Heinrich Heine, one of the greatest German poets, alluded to a pine-tree (of the
masculine Linguistic Gender) longing for a palm-tree (feminine): the Linguistic
Gender of plants allowed him to treat trees like human beings (more details and
examples are in §8.2).
Linguistic Genders can be associated with typical shapes—elongated and thin for
males and masculine gender, round for females and feminine gender. Choosing
Linguistic Genders by shape is a way to go for objects which do not have ‘sex’. But
classifying humans by shape can be offensive. Among the Manambu of New Guinea,
Natural Gender
Social Gender for humans
Linguistic Gender of humans
Social Gender features projected onto non-humans
SCHEME 12.1 How the three faces of Gender relate to each other

214 12 The heart of the matter
talking about a man as if he were ‘round’ and feminine means downgrading him to
the status of a mere ‘thing’.
Linguistic Genders interact with social values and may reflect them. In many
languages from New Guinea, highly valued objects are masculine. Objects of lesser
value are feminine. Linguistic Genders mirror the values of the societies where men,
and male cults, are most appreciated. Referring to men as if they were women, and to
women as if they were men, is incongruous. A ‘feminine’ man fails to conform to
Social Gender expectations. A ‘masculine woman’ may be too big for her boots.
Alternatively, in some languages talking about a woman as if she were a man implies
promoting her to proper human status. Among the Tariana of north-western Brazil,
an important woman is a ‘he’. This is akin to the English expression ‘man up!’, in the
meaning of ‘get your act together’. A man (and not a woman) is an epitome of
positive features. The Social Gender finds its direct match in Linguistic Gender. More
on this in Chapter 7.
Aspects of language other than Linguistic Gender set men and women apart. This
brings us to our fourth point.
IV. Social Genders in speech practices
In a number of societies, men and women have to use different words. A Karajá
woman, from Brazil, will say bicikreta, for ‘bicycle’ (a borrowing from Portuguese
bicicleta). A man will say bicileta, for the same thing: men’s speech lacks the sound k.
Special forms for men and women (known as genderlects) define Social Genders:
mature men are expected to use the male language as a feature of their social
malehood. The same applies to women, and female language. The trouble is that
fewer and fewer people across the world continue speaking minority languages with
fascinating gender differences. The systems become endangered, and harder and
harder to understand. Japanese and Thai are examples of larger communities with
pretty well-defined male and female speech. A Japanese-speaking woman will use a
final particle wa, and a man will not. Recent social changes and growing equality of
men and women have affected genderlects. As Japanese and Thai-speaking women
get access to leading positions in the society, asymmetries in speech practices become
less obvious. More on this in §9.1.
In most other languages, men and women may (but don’t have to) speak somewhat
differently. Speakers have an option of making their language sound more
male-like, or more female-like, in agreement with the stereotypes one has in mind,
and the image one wishes to project. So, in Brazilian Portuguese and in Moroccan
Arabic women use many more diminutives than men. Female speech is felt to be
more polite. These are tendencies, and not steadfast rules. The fascinating thing is
that these trends turn into stereotypes of Social Genders. They also play out in
projecting and defining the identity of genders other than plain male or plain female.

12 The heart of the matter 215
Gays, lesbians, and transsexuals project the image of men, women, or something inbetween,
using the gender-variable language resources. More on this in §§9.2–3.
Differences between men and women lie not just in how they speak, but in what
they talk about. Some words, concepts and genres may be plainly off limits to one
Social Gender. This takes us to our next point.
V. Social Genders, speech genres, and taboos
In many traditional societies, the language reflects a division of labour between Social
Genders. Special male registers and rituals are off limits to women. Men have traditional
secret languages which women are not allowed to know. In the Tariana society
of north-western Brazil, women were not supposed to see the magic flutes—a woman
who might catch a glimpse of those by pure chance would have to die. Nor was a
woman allowed to hear, let alone use, the term piri for magic flute. More on these in
Chapter 10.
Interestingly enough, no women-only registers which would be forbidden to men
have so far been documented. Is it indeed the case, as Anne Storch (2011: 84) phrased
it, that male secret languages are strategies ‘by socially or spiritually inferior groups to
gain or maintain power they would not otherwise have’, and all they reflect is fear and
need for protection against women—a potentially powerful competitor?
And this brings up our next point: how can the social status of women be seen
through the prism of Linguistic Gender and ways of saying things? And how does the
Linguistic Gender change as the societies transform?
VI. Women through the prism of language
In terms of numbers, women are hardly a minority. Yet they have been backgrounded,
downplayed, relegated to the domestic sphere, and generally treated as
‘second class citizens’ throughout the history of many societies and cultures. Linguistic
Genders clearly reflect the historical inequality between men and women, and
the asymmetry between Social Genders. In many languages, the masculine gender is
used to talk about ‘valuable’ and important objects. A woman can be ‘promoted’ to
social manhood and appreciatively spoken of as ‘aman’. The formidable Golda Meir,
during her time as Israeli Prime Minister, was praised as ha-gever ha-yexid bamemshala
(the-man the-only in.the-government) ‘the only man in the government’.
In many languages the word ‘man’ has overtones of courage and bravery. More on
this in §§11.1.–3.
In contrast, women can be treated as a substandard and somewhat suspicious
‘other’. Tariana women are dangerous outsiders—they are to blame for many things,
including the demise of the Tariana language. So are Corsican women, as was mentioned
in §10.5. The historical asymmetry in male and female Social Genders—and

216 12 The heart of the matter
suspicious attitude against women—find their expression in proverbs across the world.
Never marry a woman with big feet, says the title of Mineke Schipper’s(2003) study of
proverbs all over the world: a woman with big feet may try and overrule the dominant
man. An Arabic proverb from Lebanon reminds us that ‘a man, even a man of small
size, will be called great in comparison to women’ (Schipper 2003: 20). ‘Where there are
women, there is no peace’, says an Estonian proverb.
Derogatory and demeaning overtones of female terms and functions tell a similar
story. The feminine form spinster in English has unwanted connotations, as was
discussed in §7.5. Its masculine counterpart, bachelor, does not. For centuries and in
many languages, women were subsumed under ‘man’ as a general word for ‘human
being’. Some dictionaries defined woman as ‘subtype of man’. The traditional practice
in English was to use ‘man’ as a general term for human and ‘he’ as a generic pronoun,
to refer to people of either sex (though other options have been available since Old
English times). With the rise of feminism and general rise in the status of women in
Western societies, the question of the inherent sexism of generic masculine came to a
head. It became the target for linguistic reforms. The political importance of Linguistic
Gender was brought to the forefront of the antisexist struggle. Incidentally, what better
way of proving that Linguistic Gender is central to language use, and not an ‘irrational’
excrescence? Linguistic Gender has come to be seen by feminists as a way of redressing
the balance between Social Genders.
Different languages followed different paths, depending on what resources are
available, and how the systems work. The main issue in English has been the avoidance
of generic masculine pronoun he and terms involving ‘man’. And a victory has been
achieved: it is no longer advisable to use ‘he’ as a generic. We can recall Jenny
Cheshire’s (2008: 9–10) encouraging report, that none of her colleagues would now
‘dare to use a generic he pronoun’. Addressing a woman chair as Madam Chairman
was fine in 1971 (see Goddard et al. 1971). Nowadays, we will say Madam Chair.
German has gone a different way. Feminists and language reformers targeted
generic forms which involve ‘man’ or masculine forms. Thanks to their efforts,
feminine and masculine derivational forms are now mostly used in a parallel fashion:
one talks and writes about Kollegen and Kolleginnen (male colleagues and female
colleagues), Studenten and Studentinnen (male students and female students), rather
than subsuming women under the masculine Kollegen or Studenten. Along similar
lines, French feminists and reformers have concentrated on equality in derivations.
Since 1998, using feminine forms for professions and titles has become officially
recognized in France (though it is not yet followed by everyone). Incidentally, the
feminine forms of many professions and titles in English often have slighting or even
pejorative overtones: a manageress is not equal to a manager, and an authoress is seen
as not quite as prestigious as an author. (Other nouns like colleague or student simply
have no female form.) In all likelihood, this is a reason why English-speaking
feminists never focused on equality in derivations (a point aptly made by Anna

12 The heart of the matter 217
Livia in her 2001 book Pronoun envy). The choice of gendered pronouns in German
and in French is determined by the Linguistic Gender of the noun, and is hard to
manipulate. There have been attempts to get rid of the German generic pronoun man
‘one, person’ because of its uncanny similarity to the masculine Mann ‘man’; but the
success of this remains to be seen. The strife to redress the imbalance in ‘sexist
language’ by feminists across the Western world is a living proof of the centrality of
Linguistic and Social Gender working together, with their uniting and divisive
powers. This is what we saw in §11.5.
The criticism of Linguistic Genders comes from other quarters. Some feminists
complain that having an obligatory Linguistic Gender, and having to always make a
choice between ‘he’ and ‘she’ and masculine and feminine forms is an obstacle for
Social Gender equality: there is no way one can escape the ‘prison-house’ of Gender.
One thing to remember is that the presence of a linguistic category hardly makes a
society more or less sexist. In no way does having no Linguistic Gender in a language
correlate with gender equality or gender neutrality. Turkish, a language with no
Linguistic Gender, does not offer an equal treatment of men and women. Neither
does Hungarian, nor Estonian. The demeaning stereotypes associated with Social
Gender, and the male bias, will be reflected in any language—with or without
Linguistic Gender—as long as they are relevant for the society which speaks it. The
lack of Linguistic Gender does not address the problem in the slightest.
Where to from now? The three faces of Gender are the life-blood of human
interaction and history. Societies evolve, and language change follows suit. We expect
the growing equality of women to find more and more reflection in the languages of
the world. The linguistic expression of genders other than plain male or plain female
is another fascinating issue to further explore, as new studies come to light. We hope
that new developments, and new descriptive studies, will help unravel the intricacies
of human cognition and social life.

References
Aalto, Pentti. 1959. ‘Über die kalmückische Frauensprache’. Studia Mongolica 1: 1–8.
Abbink, Jon. 2015. ‘Menstrual synchrony claims among Suri girls (Southwest Ethiopia):
between gender and biology’. Cahiers d’études africaines (2), 218: 279–302.
Abbott, Clifford. 1984. ‘Two feminine genders in Oneida’. Anthropological Linguistics 26:
125–37.
Abbou, Julie. 2011. ‘Double gender marking in French: a linguistic practice of antisexism’.
Current Issues in Language Planning 12: 55–75.
Abondolo, Daniel. 1998. ‘Finnish’, pp. 149–83 of The Uralic languages, edited by Daniel
Abondolo. London and New York: Routledge.
Adams, K. L. 1989. Systems of numeral classification in the Mon-Khmer, Nicobarese and Aslian
subfamilies of Austroasiatic. Canberra: Pacific Linguistics.
Adelaar, Willem F. H. 2006. ‘The Quechua impact in Amuesha’, pp. 290–311 of Grammars in
contact: a crosslinguistic typology. Latrobe, Australia: Research Centre for Linguistic Typology,
edited by Alexandra Y. Aikhenvald and R. M. W. Dixon. Oxford: Oxford University Press.
Ahland, Colleen Anne. 2012. ‘A grammar of northern and southern Gumuz’. PhD dissertation,
University of Oregon.
Aikhenvald, Alexandra Y. 2000. Classifiers: a typology of noun categorization devices. Oxford:
Oxford University Press.
Aikhenvald, Alexandra Y. 2002. Language contact in Amazonia. Oxford: Oxford University Press.
Aikhenvald, Alexandra Y. 2003. A grammar of Tariana, from north-west Amazonia. Cambridge:
Cambridge University Press.
Aikhenvald, Alexandra Y. 2004. ‘Gender’, Article 98, pp. 1031–45 of Morfologie/Morphology.
Ein Handbuch zur Flexion und Wortbildung/A handbook on inflection and word formation,
2 Halbband, edited by Geert E. Booij, Christian Lehmann, Joachim Mugdan, and Stavros
Skopetas. Berlin: Mouton de Gruyter.
Aikhenvald, Alexandra Y. 2006. ‘Classifiers and noun classes, semantics’, pp. 463–70 of
Volume 1 (article 1111) ofEncyclopedia of language and linguistics, 2nd edition, edited by
Keith Brown. Oxford: Elsevier.
Aikhenvald, Alexandra Y. 2007. ‘Classifiers in multiple environments: Baniwa of Içana/Kurripako:
a North Arawak perspective’. International Journal of American Linguistics 27:
475–500.
Aikhenvald, Alexandra Y. 2008. The Manambu language, from East Sepik, Papua New Guinea.
Oxford: Oxford University Press.
Aikhenvald, Alexandra Y. 2009. ‘Language contact along the Sepik River’. Anthropological
Linguistics 50: 1–66 (dated 2008, published 2009).
Aikhenvald, Alexandra Y. 2010. Imperatives and commands. Oxford: Oxford University Press.
Aikhenvald, Alexandra Y. 2012a. The languages of the Amazon. Oxford: Oxford University
Press.

References 219
Aikhenvald, Alexandra Y. 2012b. ‘Round women and long men: shape and size in gender
choice in Papua New Guinea and beyond’. Anthropological Linguistics 54 (1): 33–86.
Aikhenvald, Alexandra Y. 2013a. ‘The language of value and the value of language’. Hau: a
journal of ethnographic theory 3.2: 55–73.
Aikhenvald, Alexandra Y. 2013b. ‘Shifting language attitudes in north-west Amazonia’. International
Journal of the Sociology of Language 222: 195–216.
Aikhenvald, Alexandra Y. 2013c. ‘Possession and ownership: a typological perspective’,
pp. 1–64 of Aikhenvald and Dixon 2013.
Aikhenvald, Alexandra Y. 2014a. The art of grammar. Oxford: Oxford University Press.
Aikhenvald, Alexandra Y. 2014b. ‘Double talk: parallel structures in Manambu songs, and their
origin’. Language and Linguistics in Melanesia 32.2: 86–109.
Aikhenvald, Alexandra Y. 2014c. ‘The grammar of knowledge in typological perspective’,
pp. 1–51 of The grammar of knowledge, edited by A. Y. Aikhenvald and R. M. W. Dixon.
Oxford: Oxford University Press.
Aikhenvald, Alexandra Y. 2015a. ‘Body, mind and spirit: what makes up a person in Manambu’.
Studies in Language 39: 85–117.
Aikhenvald, Alexandra Y. 2015b. Oxford Bibliography Online: Classifiers (refereed updatable
resource with summaries and evaluation for each entry). General Editor: Mark Aronoff.
Aikhenvald, Alexandra Y. Forthcoming. ‘A typology of noun categorization devices’, to appear
in The Cambridge handbook of linguistic typology, edited by Alexandra Y. Aikhenvald and
R. M. W. Dixon. Cambridge: Cambridge University Press.
Aikhenvald, Alexandra Y. and R. M. W. Dixon. 2011a. Language at large: essays in syntax and
semantics. Empirical approaches to linguistic theory, 2. Leiden: Brill.
Aikhenvald, Alexandra Y. and R. M. W. Dixon. 2011b. ‘Dependencies between grammatical
systems’, pp. 170–204 of Language at large: essays on syntax and semantics, by Alexandra
Y. Aikhenvald and R. M. W. Dixon. Leiden: Brill.
Aikhenvald, Alexandra Y. and R. M. W. Dixon (eds.). 2013. Possession and ownership. Oxford:
Oxford University Press.
Aikhenvald, Alexandra Y. and Diana Green. 1998. ‘Palikur and the typology of classifiers’.
Anthropological Linguistics 40: 429–80.
Aikhenvald, Alexandra Y. and Diana Green. 2011. ‘Palikur and the typology of classifiers’.
pp. 394–450 of Language at large: essays on syntax and semantics, by Alexandra Y. Aikhenvald
and R. M. W. Dixon. Leiden: Brill.
Aikio, Marjut. 1992. ‘Are women innovators in the shift to a second language? A case study of
Reindeer Sámi women and men’. International Journal of the Sociology of Language 94: 43–61.
Alexeyev, M. 1985. Voprosy sravniteljno-istoricheskoj grammatiki lezginskih jazykov. Morfologija.
Sintaksis. (Problems of comparative and historical grammar of Lezghic language.
Morphology. Syntax.) Moscow: Nauka.
Allan, Keith and Kate Burridge. 2006. Forbidden words: taboo and the censoring of language.
Cambridge: Cambridge University Press.
Alpher, Barry. 1987. ‘Feminine as the unmarked gender: Buffalo girls are no fools’. Australian
Journal of Linguistics 7: 169–87.
Alvanoudi, Angeliki. 2014. Grammatical gender in interaction: cultural and cognitive aspects.
Leiden: Brill.

220 References
Alvanoudi, Angeliki. 2015. ‘The routine achievement of gender in Greek conversation’. Gender
and Language 9: 11–31.
Amadiume, Ifi. 2015. Male daughters and female husbands: gender and sex in an African
society. London: Zed books.
Ameka, Felix K. 2012. Ewe: its grammatical constructions and illocutionary devices. Munich:
Lincom Europa.
Amha, Azeb. 2001. The Maale language. Amsterdam: Research School of Asian, African, and
Amerindian Studies, Universiteit Leiden.
Amha, Azeb. 2012. ‘Omotic’, pp. 423–504 of Frajzyngier and Shay 2012.
Amha, Azeb. 2013. ‘Male drums and female drums: natural gender and inanimate nouns in
Omotic languages’. Paper presented at the International Workshop on Gender. LCRC, JCU,
November 2013.
Amha, Azeb. Forthcoming. ‘Omotic languages’, to appear in The Cambridge handbook of
linguistic typology, edited by Alexandra Y. Aikhenvald and R. M. W. Dixon. Cambridge:
Cambridge University Press.
Anderson, L.-G. and T. Janson. 1997. Languages in Botswana: language ecology in Southern
Africa. Botswana: Longman.
Armon-Lotem, Sharon and Orit Amiram. 2014. ‘The assignment of gender in L2 Hebrew: the
role of the L1 gender system’, pp. 230–49 of The acquisition of Hebrew phonology and
morphology, edited by Outi Bar-El. Leiden: Brill.
Asher, R. E. 1985. Tamil. London: Croom Helm.
Atanga, Lilian Lem, Sibonile Edith Ellece, Lia Litosseliti, and Jane Sunderland. 2013. Gender and
language in Sub-Saharan Africa: tradition, struggle and change. Amsterdam: John Benjamins.
Attaviriyanupap, Korakoch. 2015. ‘The linguistic representation of gender in Thai’, pp. 369–400
of Hellinger and Motschenbacher 2015.
Audring, Jenny. 2013. ‘A pronominal view of gender agreement’. Language Sciences 35: 32–46.
Avery, Jack and Julie Liss. 1996. ‘Acoustic characteristics of less-masculine-sounding male
speech’. Journal of the Acoustic Society of America 99: 3738–48.
Awbery, Gwenllian, Kathryn Jones, and Delyth Morris. 2002. ‘The politics of language and
gender in Wales’, pp. 313–30 of Hellinger and Bussmann 2002, Volume 2.
Axenov, Serge. 2006. The Balochi language of Turkmenistan: a corpus-based grammatical
description. Uppsala: Department of Linguistic and Philology.
Baker, Paul. 2014. ‘Two hundred years of American man’, pp. 34–52 of Milani 2014.
Bani, Ephraim. 1987. ‘Masculine and feminine grammatical gender in Kala Lagaw Ya’.
Australian Journal of Linguistics 7: 189–201.
Barnhart, Robert K. 1988. Chambers dictionary of etymology. Edinburgh: Chambers.
Baron, Dennis. 1981. ‘The epicene pronoun: the word that failed’. American Speech 56: 83–97.
Baron, Dennis. 1986. Language and gender. New Haven: Yale University Press.
Baron, Naomi S. 1971. ‘A reanalysis of English grammatical gender’. Lingua 27: 113–40.
Barrett, R. 2006. ‘Queer talk’, pp. 316–23 of Encyclopedia of languages and linguistics, edited by
Keith Brown. Oxford: Elsevier.
Bateson, Gregory. 1958. Naven. 2nd edition. Stanford, Calif.: Stanford University Press.
Bauman, R. and J. Sherzer (eds.). 1974. Explorations in the ethnography of speaking. Cambridge:
Cambridge University Press.

References 221
Beachy, Marvin Dean. 2005. ‘An overview of Central Dizin phonology and morphology’.
Masters thesis in linguistics, University of Texas at Arlington.
Beauvoir, Simone de. 1949. Le deuxième sexe. Paris: Gallimard.
Becker, A. L. 1975. ‘A linguistic image of nature: the Burmese numerative classifier system’.
Linguistics 165: 109–21.
Bell, D. 1983. Daughters of the dreaming. Melbourne: McPhee Gribble.
Benamara, Hassane. 2013. Dictionnaire amazighe–français. Parler de Figuig et ses régions.
Rabat: IRCAM.
Bengoechea, Mercedes. 2011. ‘Non-sexist Spanish policies: an attempt bound to fail?’ Current
Issues in Language Planning 12: 35–53.
Bergen, John J. 1980. ‘The semantics of gender contrasts in Spanish’. Hispania 63: 48–57.
Berman, Ruth A. 1981. ‘Regularity vs anomaly: the acquisition of Hebrew inflectional morphology’.
Journal of Child Language 8: 265–82.
Berman, Ruth A. 1985. ‘The acquisition of Hebrew’, pp. 255–71 of The crosslinguistic study of
language acquisition. Hillsdale, NJ: Lawrence Erlbaum.
Berman, Ruth A. and Dan I. Slobin. 1994. Relating events in narrative: a crosslinguistic
developmental study. Hillsdale, NJ: Lawrence Erlbaum Associates.
Besnier, Niko. 2003. ‘Crossing genders, mixing languages: the linguistic construction of
transgenderism in Tonga’, pp. 279–301 of Holmes and Meyerhoff 2003.
Bhatia, Tej K. 1993. Punjabi. London: Routledge.
Birtalan, Ágnes. 2003. ‘Oirat’, pp. 210–28 of Mongolic languages, edited by Juha Janhunen.
London: Routledge.
Bisang, W. 1993. ‘Classifiers, quantifiersandclassnounsinHmong’. Studies in Language 17: 1–51.
Bloomfield, Leonard. 1933. Language. New York: Holt, Rinehart and Winston.
Bloomfield, Morton W. 1963. ‘A grammatical approach to personification allegory’. Modern
Philology 60: 161–71.
Boas, Franz. 1911. ‘Introduction’. Handbook of American Indian languages 1: 5–83.
Boas, Franz. 1938. ‘Language’, pp. 124–45 of General anthropology, edited by Franz Boas.
Boston and New York: D. C. Heath and Company.
Bodine, Ann. 1975. ‘Sex differentiation in language’, pp. 130–51 of Language and sex: difference
and domination, edited by Barrie Thorne and N. Henley. Rowley, Mass.: Newbury House.
Bogoras, Waldemar. 1901. ‘The Chukchi of Northeastern Asia’. American Anthropologist 3:
80–108.
Bogoras, Waldemar. 1922. ‘Chukchee’, pp. 631–903 of Handbook of American Indian languages,
Volume 2, edited by Franz Boas. Washington, DC: Government Printing Office.
Bolinger, Dwight. 1991. ‘Reference and inference: inceptiveness in the Spanish preterit’,
pp. 319–34 of his Essays on Spanish: words and grammar. Newark, Del.: Juan de la Cuesta.
Bonfante, G. 1946. ‘Semantics, language’, pp. 847–51 of Encyclopedia of psychology, edited by
P. L. Harriman. New York: ADD.
Bonvillain, Nancy. 1973. A grammar of Akwesasne Mohawk. Mercury Series, Ethnology
Division, Paper 8. Ottawa: National Museum of Man.
Borba, Rodrigo and Ana Christina Ostermann. 2007. ‘Do bodies matter? Travestis’ embodiment
of (trans)gender identity through the manipulation of the Brazilian Portuguese
grammatical gender system’. Gender and Language 1: 131–47.

222 References
Boroditsky, Lera and Lauren A. Schmidt 2000. ‘Sex, syntax, and semantics’. Proceedings of the
Cognitive Science Society 22: 42–7.
Boroditsky, Lera, Lauren A. Schmidt, and Webb Phillips. 2003. ‘Sex, syntax and semantics’,
pp. 61–80 of Language in mind: advances in the study of language and cognition, edited by
Dedre Gentner and Susan Goldin-Meadow. Cambridge, Mass.: MIT Press.
Bossong, Georg. 1985. Empirische Universalienforschung. Differentielle Objekstmarkierung in
den neuiranischen Sprachen. Tübingen: Gunter Narr.
Bossong, Georg. 1991. ‘Differential object marking in Romance and beyond’, pp. 143–70 of
New analyses in Romance linguistics, edited by D. Wanner and D. Kibbee. Amsterdam: John
Benjamins.
Bowden, Ross. 1997. A dictionary of Kwoma, a Papuan language of north-east New Guinea.
Canberra: Pacific Linguistics.
Bradac, J. J., A. Mulac, and S. A. Thompson. 1995. ‘Men’s and women’s use of intensifiers and
hedges in problem-solving interaction: molar and molecular analyses’. Research on Language
and Social Interaction 28.2: 93–116.
Bradley, John. 1988. ‘Yanyuwa: “men speak one way, women speak another”’. Aboriginal
Linguistics 1: 126–34.
Brauer, Marcus and Michaël Landry. 2008. ‘Un ministre peut-il tomber enceinte? L’impact du
générique masculin sur les représentations mentales’. Année psychologique 108.2: 243–72.
Braun, Friederike. 1997a. ‘Genuslose Sprache—egalitäre Sprache? Die Behandlung der Geschlechter
in Sprachen mit und ohne Genus’, pp. 39–49 of Jahrbuch für finnisch-deutsche
Literaturbeziehungen 29, edited by Hans Fromm et al. Helsinki: Deutsche Bibliotek.
Braun, Friederike. 1997b. ‘Making men out of people: the MAN principle in translating
genderless forms’, pp. 3–30 of Communicating gender in context, edited by H. Kotthof and
R. Wodak. Amsterdam: John Benjamins.
Braun, Friederike. 2000a. Geschlecht in Türkischen. Untersuchungen zum sprachlichen Umgang
mit einer sozialen Kategorie. Wiesbaden: Harrassowitz Verlag.
Braun, Friederike. 2000b. Mehr Frauen in die Sprache. Leitfaden zur geschlechtergerechten
Formulierung. Kiel: Ministerium für Justiz, Frauen, Jugend und Familie des Landes Schleswig-
Holstein.
Braun, Friederike. 2001. ‘The communication of gender in Turkish’, pp. 283–310 of Hellinger
and Bussmann 2001, Volume I.
Braun, Friederike and Geoffrey Haig. 2010. ‘Where are German “girls” feminine? How the
semantics of age influences the grammar of gender agreement’, pp. 68–84 of Language in its
socio-cultural context: new explorations in gendered, global and media uses, edited by Markus
Bieswanger, Heiko Motschenbacher, and Susanne Mühleisen. Frankfurt, M. u.a.: Lang.
Braun, Friederike, Sabine Sczesny, and Dagmar Stahlberg. 2005. ‘Cognitive effects of masculine
generics in German: an overview of empirical findings’. Communications 30: 1–21.
Breen, G. 1976a. ‘Wangkumara’, pp. 336–9 of Grammatical categories in Australian languages,
edited by R. M. W. Dixon. Canberra: Australian Institute of Aboriginal Studies.
Breen, G. 1976b. ‘Wagaya’, pp. 340–2 and 590–4 of Grammatical categories in Australian
languages, edited by R. M. W. Dixon. Canberra: Australian Institute of Aboriginal Studies.
Breton, Raymond. 1978. Relations de l’isle de la Guadeloupe, Volume I. Basse-Terre: Société
d’histoire de la Guadeloupe. (Bibliothèque d’histoire antillaise 3.)

References 223
Briggs, C. L. 1992. ‘“Since I am a woman, I will chastise my relatives”: gender, reported speech
and the (re)production of social relations in Warao ritual wailing’. American Ethnologist 19:
337–61.
Brittain, David. 2002. ‘Space and spatial diffusion’, pp. 603–37 of The handbook of variation and
change, edited by J. K. Chambers, Peter Trudgill, and Natalie Schilling-Estes. London: Blackwell.
Brosman, P. W. Jr. 1979. ‘The semantics of Hittite gender system’. Journal of Indo-European
Studies 7: 227–36.
Brown, Penelope. 1980. ‘How and why are women more polite: some evidence from a Mayan
community’, pp. 111–36 of Women and language in literature and society, edited by
S. McConnell-Ginet, R. Borker, and N. Furman. New York: Praeger.
Brown, Penelope. 1993. ‘Gender, politeness and confrontation in Tenejapa’,pp.144–61 of Gender
and conversational interaction, edited by Deborah Tannen. New York: Oxford University Press.
Brown, Penelope and S. C. Levinson. 1987. Politeness: some universals in language usage.
Cambridge: Cambridge University Press.
Brown, R. 1981. ‘Semantic aspects of some Waris predications’, pp. 93–123 of Syntax and
semantics in Papua New Guinea languages, edited by K. J. Franklin. Ukarumpa: Summer
Institute of Linguistics.
Bruce, Les. 1984. ‘The Alamblak language of Papua New Guinea, East Sepik’. Canberra: Pacific
Linguistics.
Brunet, Jacqueline. 2003. ‘La troisième personne de politesse en italien: fait de langage, fait de
culture’. Franco-British Studies 33/34: 22–34.
Bucholtz, Mary and Kira Hall (eds.). 1995a. Gender articulated: language and the socially
constructed self. New York: Routledge.
Bucholtz, Mary and Kira Hall. 1995b. ‘Introduction: twenty years after Language and woman’s
place’, pp. 1–24 of Bucholtz and Hall 1995a.
Bucholtz, Mary, Anita C. Liang, and Laurel A. Sutton (eds.). 1999. Reinventing identities: the
gendered self in discourse. New York: Oxford University Press.
Bull, Tove and Toril Swan. 2002. ‘The representation of gender in Norwegian’, pp. 219–50 of
Hellinger and Bussmann 2002, Volume 2.
Bullock, Alan, Oliver Stallybrass, and Stephen Trombley. 1977. The Fontana dictionary of
modern thought. London: Fontana/Collins.
Bulygina, T. V. and A. D. Shmelev. 1996. ‘Nespecifirovannyj pol i soglasovanie pri anafore’
(‘Unspecified gender and anaphoric agreement’). Moscow Linguistic Journal 2: 98–103.
Burr, Elisabeth. 2003. ‘Gender and language politics in France’, pp. 119–40 of Hellinger and
Bussmann 2003, Volume 3.
Bussmann, Hadumod and Marlis Hellinger. 2003. ‘Engendering female visibility in German’,
pp. 141–74 of Hellinger and Bussmann 2003, Volume 3.
Butt, John and Carmen Benjamin. 2004. A new reference grammar of Modern Spanish. 4th
edition. London: Hodder Arnold.
Caldas-Coulthard, Carmen Rosa and Rosamund Moon. 2010. ‘“Curvy, hunky, kinky”: using
corpora as tools for critical analysis’. Discourse and Society 21: 99–133.
Camden, W. G. 1979. ‘Parallels in structure of lexicon and syntax between New Hebrides
Bislama and the South Santo language as spoken at Tangoa’, pp. 51–117 of Papers in Pidgin
and Creole Linguistics, 2, edited by Peter Mühlhäusler et al. Canberra: Pacific Linguistics.

224 References
Cameron, Deborah. 1992. ‘Not gender difference but the difference gender makes—
explanation in research on sex and language’. International Journal of the Sociology of
Language 94: 13–26.
Cameron, Deborah. 1995. Verbal hygiene. London: Routledge.
Cameron, Deborah. 1997. ‘Theoretical debates in feminist linguistics: questions of sex and
gender’,pp.21–36 of Gender and discourse, edited by Ruth Wodak. London: Sage publications.
Cameron, Deborah. 2000. ‘Review of Anne Pauwels, Women changing language’. Language in
Society 29: 269–72.
Cameron, Deborah. 2011. ‘Gender and language’, pp. 331–3 of The Cambridge encyclopedia of
the language sciences, edited by Patrick Hogan. Cambridge: Cambridge University Press.
Cameron, Deborah and Don Kulick. 2003. Language and sexuality. Cambridge: Cambridge
University Press.
Cameron, Deborah, Fiona McAlinden, and Kathy O’Leary. 1989. ‘Lakoff in context: the social
and linguistic functions of tag question’, pp. 74–93 of Women in their speech communities,
edited by Jennifer Coates and Deborah Cameron. London: Longman.
Campbell, Élisabeth. 2004. ‘La représentation des femmes dans les dictionnaires bilingues’.
French Studies 57: 61–76.
Capell, A. and H. E. Hinch. 1970. Maung grammar, texts and vocabulary. The Hague: Mouton.
Caputo, B. Britt. 2005. ‘Gender neutrality hits Wesleyan’. Harvard Crimson 2 November.
Carrington, J. F. 1947. ‘The initiation language: Lokele tribe’. African Studies 6: 196–207.
Castellino, G. R. 1975. ‘Gender in Cushitic’, pp. 333–59 of Hamito-Semitica, edited by James
and Theodora Bynon. Jauna Linguarum, Series Practica, 200. The Hague: Mouton.
Chafe, Wallace L. 1967. Seneca morphology and dictionary. Washington, DC: Smithsonian Press.
Chafe, Wallace L. 1977. ‘The evolution of third person verb agreement in the Iroquoian
languages’, pp. 493–524 of Mechanisms of syntactic change, edited by Charles Li. Austin:
University of Texas.
Chafe, Wallace. 2004. ‘Masculine and feminine in the Northern Iroquoian languages’,
pp. 99–109 of Ethnosyntax: explorations in grammar and culture, edited by N. J. Enfield.
Oxford: Oxford University Press.
Cheshire, J. 1985. ‘A question of masculine bias’. English Today 1: 22–6.
Cheshire, J. 2002. ‘Sex and gender in variationist research’, pp. 423–43 of The handbook of
language variation and change, edited by J. K. Chambers, P. Trudgill, and N. Schilling-Estes.
London: Blackwell.
Cheshire, J. 2008. ‘Still a gender-biased language? Updates on gender inequalities and the
English language’. English Today 93: 7–10.
Chirasombutti, Voravudhi and Anthony Diller. 1999. ‘Who am “I” in Thai?—The Thai first
person: self-reference’, pp. 114–33 of Genders and sexualities in modern Thailand, edited by
Peter A. Jackson and Nerida M. Cook. Chiang Mai: Silkworm Books.
Christ, S. 1997. ‘Duration of onset consonants in gay male stereotypes speech’. University of
Pennsylvania Working Papers in Linguistics 4.3: 53–70.
Clackson, James. 2007. Indo-European languages. Cambridge: Cambridge University Press.
Clamons, Cynthia Robb. 1993. ‘Gender assignment in Oromo’, pp. 269–80 of Principles and
prediction: the analysis of natural language, edited by Mushira Eid and Gregory Iverson.
Amsterdam: John Benjamins.

References 225
Claudi, U. 1985. Zur Entstehung von Genussystemen. Hamburg: Buske.
Coates, Jennifer. 1993. Women, men and language. London and New York: Longman.
Coates, Jennifer. 2012. ‘Gender and discourse analysis’, pp. 90–103 of The Routledge handbook
of discourse analysis, edited by James Paul Gee and Michael Handford. London: Routledge.
Coates, Jennifer and Pia Pichler (eds.). 2011. Language and gender: a reader. Oxford: Wiley
Blackwell.
Cobbinah, Alexander. 2010. ‘The Casamance as an area of intense language contact: the case of
Baïnounk Gubaher’. Journal of Language Contact 3: 175–201.
Collinder, Björn. 1965. An introduction to the Uralic languages. Berkeley: University of
California Press.
Collins French dictionary and grammar. 2014. London: Harper Collins.
Comrie, Bernard and Greville G. Corbett. 1993. ‘Introduction’, pp. 1–19 of The Slavonic
languages, edited by Bernard Comrie and Greville G. Corbett. London: Routledge.
Connors, Kathleen. 1971. ‘Studies in feminine agentives in selected European languages’.
Romance Philology 24: 573–98.
Contini-Morava, Ellen and Marcin Kilarski. 2013. ‘Functions of nominal classification’. Language
Sciences 40: 263–99.
Cooper, Fiona. 1991. Jay loves Lucy. London: Serpent’s tail.
Corbett, Greville G. 1991. Gender. Cambridge: Cambridge University Press.
Corbett, Greville (ed.). 2014a. The expression of gender. Berlin: Mouton de Gruyter.
Corbett, Greville. 2014b. ‘Gender typology’, pp. 87–130 of Corbett 2014a.
Cowley, Roger, Marvin L. Bender, Charles A. Ferguson, Hailu Fulass, and Getatchew Haile.
1976. ‘The Amharic language’, pp. 77–98 of Language in Ethiopia, edited by M. L. Bender,
J. D. Bowen, R. L. Cooper, and C. A. Ferguson. London: Oxford University Press.
Craig, C. G. 1986a. ‘Jacaltec noun classifiers’. Lingua 70: 241–84.
Craig, C. G. (ed.). 1986b. Noun classes and categorisation. Amsterdam: John Benjamins.
Craik, Brian. 1982. ‘The animate in Cree language and ideology’, pp. 29–35 of Papers of the
Thirteenth Algonquian Conference, edited by William Cowan. Ottawa: Carleton University.
Creed, Gerald W. 1984. ‘Sexual subordination: institutionalized homosexuality and social
control in Melanesia’. Ethnology 23: 157–76.
Curzan, Anne. 2003. Gender shifts in the history of English. Cambridge: Cambridge University
Press.
Dahlstrom, Amy. 1995. ‘Motivation vs. predictability in Algonquian gender’, pp. 52–66 of
Papers of the Twenty-Sixth Algonquian Conference, edited by David H. Pentland. Winnipeg:
University of Manitoba.
Daley, Karen Ann. 1998. Vietnamese classifiers in narrative texts. University of Texas at
Arlington: Summer Institute of Linguistics.
Darnell, Regna and Anthony L. Vanek. 1976. ‘The semantic basis of the animate/inanimate
distinction in Cree’. Papers in Linguistics 9: 159–80.
Dawkins, Richard M. 1916. Modern Greek in Asia Minor: a study of the dialects of Silli,
Cappadocia and Phárasa with grammar, texts, translations and glossary. Cambridge: Cambridge
University Press.
de Courtenay, Baudouin. 1929. ‘Einfluss der Sprache auf Weltanschauung und Stimmung’.
Prace filologiczne 14: 185–256.

226 References
De León, Maria de Lourdes Pasquel. 1987. ‘Noun and numeral classifiers in Mixtec and Tzotzil:
a referential view’. PhD thesis, University of Sussex.
Demuth, K. A. 1988. ‘Noun classes and agreement in Sesotho acquisition’, pp. 305–22 of
Agreement in natural language: approaches, theories, descriptions, edited by M. Barlow and
C. Ferguson. Stanford, Calif.: Center for the Study of Language and Literature.
Demuth, K. A., N. Faraclas, and L. Marchese. 1986. ‘Niger-Congo noun class and agreement
systems in language acquisition and historical change’, pp. 453–71 of Noun classes and
categorisation, edited by C. G. Craig. Amsterdam: John Benjamins.
Denny, J. P. 1976. ‘What are noun classifiers good for?’ Papers from the Annual Regional
Meeting of the Chicago Linguistic Society 12: 122–32.
Deutscher, Guy. 2010. Through the language glass: why the world looks different in other
languages. London: Heineman.
Diakonoff, I. M. 1990. Afrasian languages. Moscow: Nauka.
Diller, A. V. N. 1985. ‘High and low Thai: views from within’, pp. 51–76 of Papers in South-
East Asian Linguistics No. 9: Language policy, language planning and sociolinguistics in
South-East Asia, edited by David Bradley. Canberra: Pacific Linguistics.
Dimmendaal, Gerrit J. 1983. The Turkana language. Dordrecht: Foris Publications.
Dimmendaal, Gerrit J. 2000. ‘Number marking and noun categorization in Nilo-Saharan
languages.’ Anthropological Linguistics 42: 214–61.
Dinslage, Sabine, Rudolf Leger, and Anne Storch. 2000. ‘Space and gender: cultural limitation
of space in two communities of northeastern Nigeria’. Anthropos 95: 121–7.
Dixon, R. M. W. 1972. The Dyirbal language of North Queensland. Cambridge: Cambridge
University Press.
Dixon, R. M. W. 1980. The languages of Australia. Cambridge: Cambridge University Press.
Dixon, R. M. W. 1982. Where have all the adjectives gone? And other essays in semantics and
syntax. Berlin: Mouton.
Dixon, R. M. W. 2002. Australian languages: their nature and development. Cambridge:
Cambridge University Press.
Dixon, R. M. W. 2004. The Jarawara language of southern Amazonia. Oxford: Oxford
University Press.
Dixon, R. M. W. 2010. Basic linguistic theory. Volume 1. Methodology. Volume 2. Grammatical
topics. Oxford: Oxford University Press.
Dixon, R. M. W. 2012. Basic linguistic theory. Volume 3. Further grammatical topics. Oxford:
Oxford University Press.
Dixon, R. M. W. 2014. Making new words: morphological derivation in English. Oxford: Oxford
University Press.
Dixon R. M. W. 2015. Edible gender, mother-in-law style, and other grammatical wonders:
studies in Dyirbal, Yidiñ, and Warrgamay. Oxford: Oxford University Press.
Dixon R. M. W. 2016. Are some languages better than others? Oxford: Oxford University Press.
Dobrin, Lise M. 2012. Concreteness in grammar: the noun class systems of the Arapesh
languages. Stanford, Calif.: Stanford Studies in Morphology and the Lexicon.
Doleschal, Ursula and Sonja Schmid. 2001. ‘Doing gender in Russian: structure and perspective’,
pp. 253–82 of Hellinger and Bussmann 2001, Volume 1.

References 227
Dorian, Nancy C. 1981. Language death: the life cycle of a Scottish Gaelic dialect. Philadelphia:
University of Pennsylvania Press.
Drapeau, Lynn and Renée Lambert-Brétière. 2011. ‘Verbal classifiers in Innu’. Anthropological
Linguistics 53: 293–322.
Drude, Sebastian. 2002. ‘Fala masculina e fala feminina em Awetí’, pp. 177–90 of Línguas
indígenas brasileiras: fonologia, gramática e história, edited by Aryon Dall’Igna Rodrigues
and Ana Suelly Arruda Câmara Cabral. Belém: EDUFPA.
Drude, Sebastian. 2006. ‘On the position of the Awetí language in the Tupí family’, pp. 11–45
of Guarani y ‘Maweti-Tupi-Guarani’. Estudios históricos y descriptivos sobre una familia
lingüistica de America del Sur, edited by W. Dietrich and H. Symeonidis. Berlin: LIT Verlag.
Drude, Sebastian. 2011. ‘Aweti in relation with Kamayurá: the two Tupian languages of the
Upper Xingu’, pp. 155–92 of Alto Xingu: uma sociedade multilíngue, edited by Bruna
Franchetto. Rio de Janeiro: Museu do Índio/FUNAI.
Dubois, Betty and Isabel Crouch. 1975. The question of tag questions in women’s speech: they
really don’t use more of them’. Language in Society 4: 289–94.
Dul'son, A. P. 1968. Ketskij jazyk (The Ket language). Tomsk: Izdateljstvo Tomskogo
Universiteta.
Dunn, Michael. 2000. ‘Chukchi women’s language: a historical-comparative perspective’.
Anthropological Linguistics 42: 305–28.
Dunn, Michael. 2014. ‘Gender determined dialect variation’, pp. 39–68 of Corbett 2014a.
Du Puis, Mathias. 1652. Relation de l’establissement d’une colonie française dans la Guadeloupe
isle de l’Amérique et des moeurs des Sauvages. Caen: Yvon. (Reissue in 1972, Bibliothèque
d’histoire antillaise 1, Basse Terre: Société d’histoire de la Guadeloupe.)
Eckert, Penelope. 1989. ‘The whole woman: sex and gender differences in variation’. Language
Variation and Change 1: 245–67.
Eckert, Penelope. 2000. Linguistic variation as social practice: the linguistic construction of
identity in Belten High. Oxford: Blackwell.
Eckert, Penelope and Sally McConnell-Ginet. 2003. Language and gender. Cambridge: Cambridge
University Press.
Eckert, Penelope and Sally McConnell-Ginet. 2006. ‘Putting communities of practice in their
place’. Gender and Language 1.1: 27–38.
Edelman, D. I. 1966. ‘Yazgulamskij jazyk (Yazgulami language)’, pp. 436–54 of Jazyki narodov
SSSR. Volume 1. Indoevropejskie jazyki (Languages of the USSR. Indo-European languages),
edited by V. V. Vinogradov et al. Moscow: Nauka.
Edelman, D. I. 1999. ‘Garvi, or Bashkarik’, pp. 56–8 of Yazyki mira. Dardskie i nuristanskie
jazyki (Languages of the world. Dardic and Nuristani languages). Moscow: Indrik.
Edenmyr, Niklas. 2004. ‘The semantics of Hadza gender assignment: a few notes from the
field’. Africa and Asia 4: 3–19.
Einaudi, Paula. 1976. A grammar of Biloxi. New York: Garland.
Ekka, Francis. 1972. ‘Men’s and women’s speech in Kũrˌux’. Linguistics 81: 25–31.
Emeneau, Murray. 1980. ‘India and linguistic areas’, pp. 126–66 of Language and linguistic
area. Essays by Murray B. Emeneau. Selected and introduced by Anwar S. Dil. Stanford,
Calif.: Stanford University Press.

228 References
Endruschat, Annette. 2015. ‘Gender in Portuguese’, pp. 303–32 of Hellinger and Motschenbacher
2015.
Enfield, N. J. 2004. ‘Nominal classification in Lao: a sketch’, pp. 117–43 of Nominal classification,
Special issue of Sprachtypologie und Universalienforschung 57: 2/3, edited by Alexandra
Y. Aikhenvald.
Enfield, N. J. 2007. A grammar of Lao. Berlin: Mouton de Gruyter.
England, N. 1983. A grammar of Mam, a Mayan language. Austin: University of Texas Press.
Ervin, Susan M. 1962. ‘The connotations of gender’. Word 18: 249–61.
Escure, Geneviève. 2001. ‘Belizean Creole: gender, creole, and the role of women in language
change’, pp. 53–84 of Hellinger and Bussmann 2001, Volume 1.
Evans, Bergen and Cornelia Evans. 1957. Dictionary of contemporary American usage. New York:
Random House.
Faust, Norma, 1963. ‘El lenguaje de los hombres y mujeres en Cocama’. Perú Indígena 10.22–3:
115–17.
Ferguson, C. A. 1964. ‘Baby talk in six languages’. American Anthropologist 66: 103–14.
Ferrari, F. 2005. ‘A syntactic analysis of the nominal system of Italian and Luganda: how nouns
can be formed in syntax’. PhD dissertation, New York University.
Fesenko, Andrej and Tatiana Fesenko. 1955. Russkij jazyk pri sovetakh (Russian language
under the Soviets). New York: Rausen Bors.
Flannery, Regina. 1946. ‘Men’s and women’s speech in Gros Ventre’. International Journal of
American Linguistics 12: 133–5.
Fleming, Luke. 2012. ‘Gender indexicality in the Native Americas: contributions to the
typology of social indexicality’. Language in Society 41: 295–320.
Fleming, Luke. 2015. ‘Australian exceptionalism in the typology of affinal avoidance registers’.
Anthropological Linguistics 56: 155–8.
Fletcher, Alice and Frances La Fleche. 1972. The Omaha tribe. Lincoln: University of Nebraska
Press.
Fodor, I. 1959. ‘The origin of grammatical gender I, II’. Lingua 8: 1–41, 186–214.
Foley, W. A. 1991. The Yimas language of New Guinea. Stanford, Calif.: Stanford University
Press.
Fortune, David and Gretchen Fortune. 1975. ‘Karajá’s men’s-women’s speech differences with
social correlates’. Arquivos de anatomia e antropologia (Instituto de Antropologia Professor
Souza Marques, Rio de Janeiro) 1: 111–34.
Fortune, David and Gretchen Fortune. 1987. ‘Karajá literary acquisition and sociocultural
effects on a rapidly changing culture’. Journal of Multilingual and Multicultural Development
8: 469–91.
Fox, Helen. 1996. ‘An honorific sub-dialect used among Big Nambas women’, pp. 375–82 of
Oceanic studies: proceedings of the first international conference on Oceanic linguistics, edited
by John Lynch and Fa’afo Pat. Canberra: Pacific Linguistics.
Frajzyngier, Zygmunt. 1993. A grammar of Mupun. Berlin: Dietrich Reimer.
Fraizyngier, Zygmunt. 2012. ‘Typological outline of the Afroasiatic phylum’, pp. 505–624 of
Frajzyngier and Shay 2012.
Fraizyngier, Zygmunt and Erin Shay (eds.). 2012. Afroasiatic languages. Cambridge: Cambridge
University Press.

References 229
Franklin, K. 1981. ‘Existential and pro-verbs in Kewa’, pp. 151–72 of Syntax and semantics in
Papua New Guinea languages, edited by K. J. Franklin. Ukarumpa: Summer Institute of
Linguistics.
Fraurud, Kari. 2000. ‘Proper names and gender in Swedish’, pp. 167–219 of Unterbeck and
Rissanen 2000.
Friedman, Victor. 1993. ‘Macedonian’, pp. 249–305 of The Slavonic languages, edited by
Bernard Comrie and Greville G. Corbett. London: Routledge.
Gal, Susan. 1978. ‘Peasant men can’t get wives: language change and sex roles in a bilingual
community’. Language in Society 7: 1–16.
Gal, Susan. 1979. Language shift. New York: Academic Press.
Gal, Susan. 1989. ‘Between speech and silence: the problematics of research on language and
gender’. IPrA Papers in Pragmatics 3: 1–38.
Gal, Susan. 1991. ‘Between speech and silence: the problematics of research on language and
gender’, pp. 175–203 of Gender at the crossroads of knowledge: toward a new anthropology of
gender, ed. M. DiLeonardo. Berkeley: University of California Press.
Gal, Susan. 1995. ‘Language, gender and power: an anthropological review’, pp. 169–82 of
Bucholtz and Hall 1995a.
Garibian, A. S. 1976. ‘Armjanskij jazyk’ (Armenian language), pp. 94–109 of Jazyki Azii i
Afriki. Indo-evropejskie jazyki. Moscow: Nauka.
Gauchat, L. 1905. ‘L’unité phonétique dans le patois d’une commune’, pp. 175–232 of Aus
romanischen Sprachen und Literaturen: Festschrift Heinrich Mort. Halle: Max Niemeyer.
Gaudio, R. 1994. ‘Sounding gay: pitch properties in the speech of gay and straight men’.
American Speech 69: 30–7.
Georg, Stefan. 2007. A descriptive grammar of Ket. Part 1: Introduction, phonology and
morphology. Folkestone: Global Oriental Ltd.
Gerdts, Donna and Mercedes Q. Hinkson. 2004. ‘Salish numeral classifiers: a lexical means to a
grammatical end’, pp. 247–79 of Nominal classification, Special issue of Sprachtypologie und
Universalienforschung 57 2/3, edited by Alexandra Y. Aikhenvald.
Goddard, C. 1985. A grammar of Yankunytjatjara. Alice Springs: Institute for Aboriginal
Development.
Goddard, Ives. 2002. ‘Grammatical gender in Algonquian’, pp. 195–231 of Papers of the Thirty-
Third Algonquian Conference, edited by H. Christoph Wolfart. Winnipeg: University of
Manitoba.
Goddard, Ives, Michael Silverstein, Calvert Watkins, Einar Haugen, Sandra L. Chung,
T. L. Markey, Robert Underhill, Jay Jasanoff, Engin Sezer, Toeio Sato, Ariene Berman,
Alan H. Timberlake, Michael Szamosl, Anthony Ariotto, Stephen Anderson, Gregory Nagy,
and Janet F. Hinkley. 1971. ‘Pronoun envy’. Harvard Crimson, 16 November.
Goettner-Abendroth, Heide. 2012. Matriarchal societies: studies on indigenous cultures across
the globe. New York: Peter Lang.
Gordon, Matthew and Jeffrey Heath. 1998. ‘Sex, sound symbolism, and sociolinguistics’.
Current Anthropology 39.4: 421–49.
Grandi, Nicola. 2002. Morfologie in contatto: le costruzioni valutative nelle lingue del Mediterraneo.
Milan: Franco Angeli.
Green, Ian. 1997. ‘Nominal classification in Marrithiyel’, pp. 229–53 of Harvey and Reid 1997.

230 References
Greenbaum, Sidney. 1991. An introduction to English grammar. London: Longman.
Greenbaum, Sidney. 1996. The Oxford English grammar. Oxford: Oxford University Press.
Greenberg, J. H. 1963. ‘Some universals of grammar with particular reference to the order of
meaningful elements’, pp. 58–90 of Universals of language, edited by J. H. Greenberg.
Cambridge, Mass.: MIT Press.
Greenberg, J. H. 1978. ‘How does language acquire gender markers’, pp. 241–70 of Universals
of human languages, Volume 3. Word structure, edited by J. H. Greenberg, C. A. Ferguson,
and E. A. Moravcsik. Stanford, Calif.: Stanford University Press.
Gregersen, Edgar A. 1979. ‘Sexual linguistics’, pp. 3–19 of Language, sex and gender: does ‘la
différence’ make a difference?, edited by J. Orasanu, M. K. Slater, and L. L. Adler. New York:
New York Academy of Sciences.
Grünberg, A. L. 1999. ‘Gawar’, pp. 91–5 of Yazyki mira. Dardskie i nuristanskie jazyki
(Languages of the world. Dardic and Nuristani languages). Moscow: Indrik.
Gumperz, J. J. and R. Wilson. 1971. ‘Convergence and creolization: a case from the Indo-
Aryan/Dravidian border in India’, pp. 151–68 of Pidginization and creolization of languages,
edited by D. Hymes. Cambridge: Cambridge University Press.
Günthner, Susanne. 1996. ‘Male–female speaking practices across cultures’, pp. 447–73 of
Contrastive sociolinguistics, edited by Marlis Hellinger and Ulrich Ammon. Berlin: Mouton
de Gruyter.
Haas, Mary. 1941. Tunica. Extract from HAIL 4: 1–143. New York: J. J. Augustin.
Haas, Mary. 1944. ‘Men’s and women’s speech in Koasati’. Language 20: 142–9.
Haas, Mary. 1951. ‘Interlingual word taboos’. American Anthropologist NS 53: 338–44.
Hachimi, Atiqa. 2001. ‘Shifting sands: language and gender in Moroccan Arabic’, pp. 27–51 of
Hellinger and Bussman 2001, Volume 1.
Hagège, C. 2004. ‘On the part played by human conscious choice in language structure and
language evolution,’ pp. 105–17 of Linguistic diversity and language theories, edited by
Z. Frajzyngier, A. Hodges, and D. Rood. Amsterdam: John Benjamins.
Hagman, Roy S. 1977. Nama Hottentot grammar. Bloomington: Indian University Press.
Haiman, J. 1980. Hua: a Papuan language of the Eastern Highlands of New Guinea. Amsterdam:
John Benjamins.
Hale, Kenneth. 1971. ‘A note on a Warlbiri tradition of antonymy’, pp. 472–82 of An
interdisciplinary reader in philosophy, linguistics and psychology, edited by Danny
D. Steinberg and Leon A. Jakobovits. Cambridge: Cambridge University Press.
Hale, Kenneth and David Nash. 1997. ‘Damin and Lardil phonotactics’, pp. 247–59 of Boundary
rider: essays in honour of Geoffrey O’Grady, edited by D. T. Tryon and M. Walsh. Canberra:
Pacific Linguistics.
Hall, Kira. 2002. ‘“Unnatural” gender in Hindi’, pp. 133–62 of Gender across languages: the
linguistic representation of women and men, edited by Marlis Hellinger and Hadumod
Bussman. Amsterdam: John Benjamins.
Hall, Kira and Veronica O’Donovan. 1996. ‘Shifting gender positions among Hindi-speaking
hijras’, pp. 228–66 of Rethinking language and gender research: theory and practice, edited by
V. Bergvall et al. London: Longman.
Hamp, Eric. 1965. ‘The Albanian dialect of Mandres’. Die Sprache 11: 137–54.
Hampares, Katherine. 1976. ‘Sexism in Spanish lexicography?’. Hispania 59: 100–9.

References 231
Harbert, Wayne. 2007. Germanic languages. Cambridge: Cambridge University Press.
Harrison, Simon J. 1990. Stealing people’s names: history and politics in a Sepik river cosmology.
Cambridge: Cambridge University Press.
Harvey, Mark. 1987. ‘The Warray language from Adelaide River’. MA thesis, Australian
National University.
Harvey, Mark. 1992. The Gaagudju people and their language. PhD thesis, University of Sydney.
Harvey, Mark. 1997. ‘Nominal classification and gender in aboriginal Australia’, pp. 17–62 of
Nominal classification in Aboriginal Australia, edited by Mark Harvey and Nicholas Reid.
Amsterdam: John Benjamins.
Harvey, Mark. 2002. A grammar of Gaagudju. Berlin: Mouton de Gruyter.
Harvey, Mark and Nicholas Reid (eds.). 1997. Nominal classification in Aboriginal Australia.
Amsterdam: John Benjamins.
Haspelmath, M. 1993. A grammar of Lezgian. Berlin: Mouton de Gruyter.
Hasselblatt, Cornelius. 2015. ‘The representation of gender in Estonian’, pp. 125–52 of
Hellinger and Motschenbacher 2015.
Haugen, Einar Ingvald. 1976. The Scandinavian languages: an introduction to their history.
Cambridge, Mass.: Harvard University Press.
Hauser-Schäublin, Brigitta. 1977. Frauen in Kararau. Zur Rolle der Frau bei den Iatmul am
Mittelsepik, Papua New Guinea. Basler Beiträge zur Ethnologie, Volume 18. Basel: Wepf.
Hauser-Schäublin, Brigitta. 1995. ‘Puberty rites, women’s Naven, and initiation: women’s
rituals of transition in Abelam and Iatmul culture’, pp. 55–82 of Gender rituals: female
initiation in Melanesia, edited by Nancy C. Lutkehaus and Paul B. Roscoe. London:
Routledge.
Haviland, William A., Harald E. L. Prins, Bunny McBride, and Dana Walrath. 2014. Cultural
anthropology: the human challenge, 14th edition. Belmont, Calif.: Wadsworth/Cengage
Learning.
Heath, Jeffrey. 1984. Functional grammar of Nunggubuyu. Canberra: Australian Institute of
Aboriginal Studies.
Hegardty, Peter. 2014. ‘Ladies and gentlemen: word order and gender in English’, pp. 69–86 of
Corbett 2014a.
Heine, Bernd. 1982a. ‘African noun class systems’, pp. 189–216 of Apprehension: Das sprachliche
Erfassen von Gegenständen, Teil II: Die Techniken und ihr Zusammenhang in
Einzelsprachen, edited by Hansjakob Seiler and F. J. Stachowiak. Tübingen: Narr Language
Universals Series I/II.
Heine, Bernd. 1982b. The Nubi language of Kibera: an Arabic Creole. Berlin: Dietrich Reimer
Verlag.
Heine, Bernd and Tania Kuteva. 2002. World lexicon of grammaticalization. Cambridge:
Cambridge University Press.
Heine, Bernd and Mechthild Reh. 1984. Grammaticalization and reanalysis in African languages.
Hamburg: Helmut Buske Verlag.
Heine, Bernd and Rainer Vossen. 1983. ‘On the origin of gender in Eastern Nilotic’, pp. 255–68
of Nilotic studies: proceedings of the International Symposium on Languages and History of
the Nilotic Peoples, edited by R. Vossen and M. Bechchaus-Gerst. Berlin: Reimer.
Hellenthal, Anne-Christie. 2010. A grammar of Sheko. Amsterdam: LOT.

232 References
Hellinger, Marlis and Hadumod Bussmann (eds.). 2001–3. Gender across languages, Volumes
1–3. Amsterdam: John Benjamins.
Hellinger, Marlis and Heiko Motschenbacher. 2015. Gender across languages, Volume 4.
Amsterdam: John Benjamins.
Hellinger, Marlis and Anne Pauwels. 2007. ‘Language and sexism’, pp. 651–81 of Handbook of
language and communication: diversity and change, edited by Marlis Hellinger and
Anne Pauwels. Berlin: Mouton de Gruyter.
Henley, Nancy M. 1989. ‘Molehill or mountain? What we know and don’t know about sex bias
in language’, pp. 59–78 of Gender and thought: psychological perspectives, edited by Mary
Crawford and Margaret Gentry. New York: Springer Verlag.
Henley, Nancy M. and Joselito Abueg. 2003. ‘A review and synthesis of research on comprehension
of the masculine as a generic form in English’. Estudios de Sociolingüística 4:
427–54.
Henson, Eithne. 1997. ‘Johnson and the condition of women’, pp. 67–85 of The Cambridge
companion to Samuel Johnson, edited by Greg Clingham. Cambridge: Cambridge University
Press.
Herbert, R. K. 1990. ‘Hlonipha and the ambiguous woman’. Anthropos 85: 455–73.
Herbert, R. K. 1991. ‘Patterns in language change, acquisition and dissolution: noun prefixes
and concords in Bantu’. Anthropological Linguistics 33.2: 103–34.
Herdt, Gilbert. 1987. The Sambia: ritual and gender in New Guinea. New York: Holt, Rinehart
and Winston Inc.
Hill, Jane H. 1987. ‘Women’s speech in modern Mexicano’, pp. 121–60 of Philips, Steele, and
Tanz 1987.
Hill, Jane and Kenneth Hill. 1977. ‘Language death and relexification in Tlaxcalan Nahuatl’.
International Journal of the Sociology of Language 12: 55–67.
Hill, Jane and Ofelia Zepeda. 1999. ‘Language, gender, and biology: pulmonic ingressive
airstream in women’s speech in Tohono O'odham’. Southwest Journal of Linguistics 18:
15–40.
Hoben, Susan J. 1976. ‘The meaning of the second-person pronouns in Amharic’, pp. 281–8 of
Language in Ethiopia, edited by M. L. Bender, J. D. Bowen, R. L. Cooper, and C. A. Ferguson.
London: Oxford University Press.
Hodge, Carleton T. 1958. ‘Non-native Hausa’, pp. 57–69 of Report on the Ninth Annual Round
Table Meeting on Linguistics and Language Studies, edited by William M. Austin. Washington,
DC: Georgetown University Press.
Hoff, Berend. 1994. ‘Island Carib, an Arawakan language which incorporated a lexical register
of Cariban origin, used to address men’, pp. 161–8 of Mixed languages: 15 case studies in
language intertwining, edited by P. Bakker and M. Mous. Amsterdam: IFOTT.
Hoffman, Katherine E. 2006. ‘Berber language ideologies, maintenance and contraction:
gendered varieties in the indigenous margins of Morocco’. Language and Communication
26: 144–67.
Hollow, Robert. 1970. A Mandan dictionary. PhD dissertation, University of California,
Berkeley.
Holm, John. 2000. An introduction to pidgins and creoles. Cambridge: Cambridge University
Press.

References 233
Holmes, Janet. 1984. ‘Women’s language: a functional approach’. General Linguistics 24.3:
149–78.
Holmes, Janet. 1986. ‘Functions of you know in women’s and men’s speech’. Language in
Society 15: 1–21.
Holmes, Janet and Miriam Meyerhoff (eds.). 2003. The handbook of language and gender.
Oxford: Blackwell.
Holmquist, Jonathan C. 1985. ‘Social correlates of a linguistic variable: a study in a Spanish
village’. Language in Society 14: 191–203.
Holmquist, Jonathan C. 1991. ‘Semantic features and gender dynamics in Cantabrian Spanish’.
Anthropological Linguistics 33: 57–81.
Hornscheidt, Antje. 2003. ‘Linguistic and public attitudes towards gender in Swedish’,
pp. 339–68 of Hellinger and Bussmann 2003, Volume 3.
Hornscheidt, Antje. 2006. Die sprachliche Benennung von Personen aus konstruktivistischer
Sicht. Genderspezifizierung und ihre diskursive Verhandlung im heutigen Schwedisch. Berlin:
de Gruyter.
Houseman, Michael and Carlo Severi. 1998. Naven, or the Other Self: a relational approach to
ritual action. Leiden: Brill.
Huntley, David. 1980. ‘The evolution of genitive-accusative animate and personal nouns
in Slavic dialects’, pp. 189–212 of Historical morphology, edited by J. Fisiak. The Hague:
Mouton.
Ibrahim, Muhammad Hasan. 1973. Grammatical gender: its origin and development. The
Hague: Mouton.
Ide, Sachiko. 1991. ‘How and why do women speak more politely in Japanese’, pp. 63–80 of Ide
and McGloin 1991.
Ide, Sachiko. 2003. ‘Women’s language as a group identity marker in Japanese’, pp. 227–39 of
Gender across languages, Volume 3, edited by Marlis Hellinger and Hadumod Bussmann.
Amsterdam: John Benjamins.
Ide, Sachiko and Miyako Inoue. 1992. ‘Josei kotoba ni miru aidentitii’ (Identities in women’s
language). Gengo 21.10: 46–7.
Ide, Sachiko and Naomi Hanaoka McGloin (eds.). 1991. Aspects of Japanese women’s language.
Tokyo: Kurosio Publishers.
Ikoro, S. M. 1996. The Kana language. Leiden: University of Leiden.
Inoue, Miyako. 2002. ‘Gender, language, and modernity: towards an effective history of
Japanese women’s language’. American Ethnologist 29: 392–422.
Irvine, Judith T. 1978. ‘Wolof noun classification: the social setting of divergent change’.
Language in Society 7: 37–64.
Irvine, Judith T. 1979. ‘Formality and informality in communicative events’. American
Anthropologist 81: 773–90.
Isac᷈enko, A. V. 1939. Narec᷈je vasi Sele na Roz᷈u. Ljubljana: Uc᷈iteljska tiskarna.
Jacobs, Andrea Michele. 2004. ‘Language reform as language ideology: an examination of
Israeli feminist language practice’. PhD dissertation, University of Texas at Austin.
Jaffe, Alexandra. 1999. Language politics on Corsica. Berlin: Mouton de Gruyter.
Jakobson, Roman O. 1959. ‘On linguistic aspects of translation’, pp. 113–18 of On Translation,
edited by R. A. Brower. Cambridge, Mass.: Harvard University Press.

234 References
Jakobson, Roman O. 1971. ‘Signe zéro’, pp. 211–19 of Roman Jakobson: selected writings,
Volume 2. Word and language. The Hague and Paris: Mouton.
Jakobson, Roman O. 1984. ‘The gender pattern of Russian’, pp. 141–3 of Russian and Slavic
grammar: studies 1931–1981, edited by L. R. Waugh and M. Halle. Berlin: Mouton.
James, Deborah. 1996. ‘Women, men and prestige speech forms: a critical review’, pp. 98–125
of Rethinking language and gender research: theory and practice, edited by Victoria
L. Bergvall, Janet M. Bing, and Alice F. Freed. London: Longman.
Janda, Laura A. 1999. ‘From number to gender, from dual to virile: bridging cognitive
categories’, pp. 73–86 of Between grammar and lexicon, edited by Ellen Contini-Morava
and Yishai Tobin. Amsterdam: John Benjamins.
Janse, Mark. 2002. ‘Aspects of bilingualism in the history of the Greek language’, pp. 332–90 of
Bilingualism in ancient society: language contact and the written text, edited by J. N. Adams,
Mark Janse, and Simon Swain. Oxford: Oxford University Press.
Jarkey, Nerida. 2015. ‘The Housewife’s Companion: identity construction in a Japanese
women’s magazine’, pp. 179–201 of Language and identity across modes of communication,
edited by Dwi Noverini Djenar, Ahmar Mahboob, and Ken Cruickshank. Berlin: De
Gruyter.
Jasanoff, Jay. 1989. ‘Language and gender in the Tarim Basin: the Tocharian 1sg pronoun’.
Tocharian and Indo-European Studies 3: 125–48.
Jespersen, Otto. 1894. Progress in language with special reference to English. London: Swan
Sonnenschein.
Jespersen, Otto. 1922. Language, its nature, development and origin. London: Allen and Unwin.
Jespersen, Otto. 1949. A modern English grammar on historical principles. Part VII. Syntax.
London: George Allen and Unwin Ltd.
Jespersen, Otto. 1972. Growth and structure of the English language. Oxford: Basil Blackwell
(reprint of 1912 edition, Stuttgart: Teubner Verlag).
Jones, Lucy. 2014. ‘Masculinity in lesbian discourse’, pp. 175–196 of Milani 2014.
Joseph, Brian D. and Irene Philippaki-Warburton. 1987. Modern Greek. London: Routledge.
Jospin, Lionel. 1998. ‘Circulaire du 6 mars 1998 relative à la féminisation des noms de
métier, fonction, grade ou titre, NOR: PRMX9802741C’. Journal Officiel 58, 8 March
1998: 3565.
Jungraithmayr, H. 1995. ‘Was ist am Tangale noch tschadisch/hamitosemitisch?,’ pp. 197–205
of Sprachkulturelle und historische Forschungen in Afrika, edited by A. Fleisch and D. Otten.
Cologne: Köppe.
Kahane, Henry and René Kahane. 1948–9. ‘The augmentative feminine in the Romance
languages’. Romance Philology 2: 135–75.
Karatsareas, Petros. 2009. ‘The loss of grammatical gender in Cappadocian Greek’. Transactions
of the Philological Society 107: 196–230.
Karatsareas, Petros. 2013. ‘On the diachrony of gender in Asia Minor Greek: the development
of semantic agreement in Pontic’. Language Sciences 43: 77–101.
Karbelashvili, D. P. 1935. Ruchnaia rech na Kavkaze: issledovanie po materialam Baraninskogo
raiona SSR Armenii. Izd. Nauchno-issledovatelʹskogo instituta kavkazovedeniia.
Karmiloff-Smith, A. 1979. A functional approach to child language: a study of determiners and
reference. Cambridge: Cambridge University Press.

References 235
Kastovsky, Dieter. 2000. ‘Inflectional classes, morphological restructuring, and the dissolution
of Old English grammatical gender’, pp. 709–27 of Unterbeck and Rissanen 2000.
Kazakevich, Olga A. 1996. ‘The education of the Selkups in Russia: teaching the mother tongue
as a foreign language’. International Review of Education 42.4: 388–9.
Keenan, Elinor. 1974. ‘Norm-makers, norm-breakers: uses of speech by men and women in a
Malagasy community’, pp. 125–43 of Bauman and Sherzer 1974.
Kendon, Adam. 1988. Sign languages in Aboriginal Australia: cultural, semiotic, and communicative
perspectives. Cambridge: Cambridge University Press.
Key, Mary Ritchie. 1975. Male/female language, with a comprehensive bibliography. Metuchen,
NJ: The Scarecrow Press, Inc.
Khaidakov, S. M. 1963. ‘The principles of non-class division in Lak’ (in Russian, with a
summary in English). Studia Caucasica 1: 48–55.
Kibrik, A. E. (ed.). 1996. Godoberi. Munich: Lincom Europa.
Kilarski, Marcin. 2007. ‘Algonquian and Indo-European gender in a historiographic perspective’.
Historiographia Linguistica 34: 333–49.
Kilarski, Marcin. 2014. Nominal classification: a history of its study from the classical period to
the present. Amsterdam: John Benjamins.
Kilian-Hatz, Christa. 2008. A grammar of Modern Khwe (Central Khoisan). Cologne: Rüdiger
Köppe.
Kimball, Geoffrey D. 1987. ‘Men’s and women’s speech in Koasati: a reappraisal’. International
Journal of American Linguistics 53: 30–8.
Kimball, Geoffrey D. 1991. Koasati grammar. Lincoln: University of Nebraska Press.
Kirton, Jean F. 1988. ‘Men’s and women’s dialects’. Aboriginal Linguistics 1: 111–25.
Klenin, E. 1983. Animacy in Russian: a new interpretation. Columbus, Oh.: Slavica.
Klumpp, D. 1990. Piapoco grammar. Colombia: SIL.
Kochskämper, Birgit. 1999. ‘Frau’ und ‘Mann’ im Althochdeutschen. Frankfurt: Peter Lang.
Köhler, O. 1962. ‘Studien zum Genussystem und Verbalbau der zentralen Khoisan-Sprachen’.
Anthropos 57: 529–46.
Konishi, T. 1993. ‘The semantics of grammatical gender: a cross-cultural study’. Journal of
Psycholinguistic Research 22: 519–34.
Kooyers, O. 1974. Washkuk grammar sketch. Workpapers in Papua New Guinea Languages 6:
5–74.
Kossmann, Maarten. 2014. ‘Derivational gender in Moroccan Berber: examples from Ayt
Seghrushen’. Sprachtypologie und Universalienforschung 67.1: 21–33.
Kramarae, Cheris. 1992. ‘Punctuating the dictionary’. International Journal of the Sociology of
Language 94: 135–54.
Kramarae, Cheris and Paula A. Treichler. 1992. Amazons, bluestockings and crones: a feminist
dictionary. A woman’s companion to words and ideas. London: Pandora Press.
Krishnamurti, Bhadriraju. 2003. The Dravidian languages. Cambridge: Cambridge University
Press.
Krishnamurti, Bhadriraju and John Peter Lucius Gwynn. 1985. A grammar of modern Telugu.
New Delhi: Oxford University Press.
Kroskrity, Paul V. 1983. ‘On the male and female speech in the Pueblo Southwest’. International
Journal of American Linguistics 49: 88–91.

236 References
Kryk-Kastovsky, Barbara. 2000. ‘Norm vs. use: on gender in Polish’, pp. 729–47 of Gender in
grammar and cognition. Part 2: Manifestations of gender, edited by Barbara Unterbeck and
Matti Rissanen. Berlin: Mouton de Gruyter.
Kulick, Don. 1992a. Language shift and cultural reproduction: socialisation, self, and syncretism
in a Papua New Guinean village. Cambridge and New York: Cambridge University Press.
Kulick, Don. 1992b. ‘Anger, gender, language shift and the politics of revelation in a Papua
New Guinean village’. Pragmatics 2: 281–96.
Kulick, Don. 1998. Sex, gender, and culture among Brazilian transgendered prostitutes. Chicago:
University of Chicago Press.
Kulick, Don. 2000. ‘Gay and lesbian language’. Annual Review of Anthropology 29: 243–85.
Kulick, Don and C. Stroud. 1990. ‘Code-switching in Gapun: social and linguistic aspects of
language used in a language shifting community’, pp. 205–34 of Melanesian Pidgin and Tok
Pisin, edited by J. W. M. Verhaar. Amsterdam: John Benjamins.
Kuntjara, Esther. 2001. ‘Gender in Javanese Indonesian’. Gender across Languages 1: 199–227.
Kwee, Jihn B. 1965. Indonesian. London: The English Universities Press Ltd.
Labov, William. 1972. Sociolinguistic patterns. Oxford: Blackwell.
Lakoff, George. 1987. Women, fire and dangerous things: what categories reveal about the mind.
Chicago: University of Chicago Press.
Lakoff, Robin Tolmach. 1973. ‘Language and woman’s place’. Language in Society 2: 45–80.
Lakoff, Robin Tolmach. 1975. Language and woman’s place. New York: Harper and Row.
Lakoff, Robin Tolmach. 1979. ‘Stylistic strategies within a grammar of style’, pp. 53–78 of
Language, sex and gender: does ‘la différence’ make a difference?, edited by J. Orasanu,
M. K. Slater, and L. L. Adler. New York: New York Academy of Sciences.
Lakoff, Robin Tolmach. 2004. Language and woman’s place: text and commentaries. Oxford:
Oxford University Press.
Landor, Roland. 2012. ‘Grammatical categories and cognition across five languages: the case of
grammatical gender and its potential effects on the conceptualisation of objects’. PhD thesis,
Griffith University.
Lang, Adrienne. 1975. The semantics of classificatory verbs in Enga and other Papua New
Guinea languages. Canberra: Pacific Linguistics.
Lang, Sabine. 1998. Men as women, women as men: changing gender in Native American
cultures. Austin: University of Texas Press.
Laycock, D. C. 1965. The Ndu language family (Sepik District, New Guinea). Canberra:
Linguistic Circle of Canberra Publications.
Leap, William. 2012. ‘Queer linguistics, sexuality and discourse analysis’, pp. 558–71 of The
Routledge handbook of discourse analysis, edited by James Paul Gee and Michael Hanford.
London: Routledge.
Lee, J. 1987. Tiwi today: a study of language change in a contact situation. Canberra: Pacific
Linguistics.
Lee, Yun-seok. 2014. Classifiers in Korean. Munich: Lincom Europa.
Leger, Rudolf. 1998. ‘Grammatical gender in some southern Bole-Tangale languages—Kwami,
Kupto, Kushi and Piya’, pp. 204–16 of Africa: society, culture, and languages. Institute of
African Studies. Moscow: University Press.

References 237
Lehmann, Thomas. 1993. A grammar of Modern Tamil. Pondicherry: Pondicherry Institute of
Linguistics and Culture.
LeMaster, Barbara. 1997. ‘Sex difference in Irish Sign Language’, pp. 67–86 of The life of
language: papers in linguistics in honor of William Bright, edited by Jane H. Hill, P. J. Mistry,
and Lyle Campbell. Berlin: Mouton de Gruyter.
Levon, Erez. 2006. ‘Hearing “gay”: prosody, interpretation and the affective judgement of
men’s speech’. American Speech 81: 56–78.
Levy, Y. 1983a. ‘The acquisition of Hebrew plurals: the case of the missing gender category’.
Journal of Child Language 10: 107–21.
Levy, Y. 1983b. ‘It’s frogs all the way down’. Cognition 15: 75–93.
Lewis, Geoffrey. 2000. Turkish grammar. Oxford: Oxford University Press.
Li, Paul Jen-kuei. 1982. ‘Male and female forms of speech in the Atayalic group’. Bulletin of the
Institute of History and Philology 53.2: 265–304.
Li, Paul Jen-kuei. 1983. ‘Types of lexical derivation of men’s speech in Mayrinax’. Bulletin of
the Institute of History and Philology 54.3: 1–18.
Livia, Anna. 1995. ‘“I ought to throw a Buick at you”: functional representations of butch/
femme speech’, pp. 245–78 of Bucholtz and Hall 1995a.
Livia, Anna. 1997. ‘The creation of coherence in coming-out stories’, pp. 287–309 of Livia and
Hall 1997.
Livia, Anna. 2001. Pronoun envy: literary uses of linguistic gender. Oxford: Oxford University
Press.
Livia, Anna and Kira Hall (eds.). 1997. Queerly phrased: language, gender, and sexuality. New
York: Oxford University Press.
Lock, Arnold (Arjen) Hugo. 2011. Abau grammar. Data Papers on Papua New Guinea
languages 57. Ukarumpa: SIL-PNG Academic Publishers.
Lopatin, V. V. 1970. ‘Word formation of nouns’ (Slovoobrazovanie imen suschestviteljnykh),
pp. 46–176 of Grammatika sovremennogo russkogo literaturnogo jazyka, edited by
N. Y. Shvedova. Moscow: Nauka.
Loporcaro, Michele, Vincenzo Faraoni, and Francesco Gardani. 2014. ‘The third gender of Old
Italian’. Diachronica 31: 1–22.
Lu, Tian-Qiao. 2012. Classifiers in Kam-Tai languages: a cognitive and cultural perspective.
Boca Raton, Fla.: Universal Publishers.
Luraghi, Silvia. 2009. ‘The origin of the feminine gender in PIE: an old problem in a new
perspective’, pp. 1–13 of Grammatical change in Indo-European languages, edited by Vit
Bubenik, John Hewson, and Sarah Rose. Amsterdam: John Benjamins.
Luraghi, Silvia. 2011. ‘The origin of the Proto-Indo-European gender system: typological
considerations’. Folia Linguistica 45.2: 435–63.
Luraghi, Silvia. 2014. ‘Gender and word formation: the PIE gender system in cross-linguistic
perspective’, pp. 199–232 of Studies on the collective and feminine in Indo-European from a
diachronic and typological perspective, edited by Sergio Neri and Roland Schuhmann.
Leiden: Brill.
Luthin, Herbert. 1991. ‘Restoring the voice of Yanan traditional narrative: prosody, performance,
and presentational form’. PhD dissertation, University of California, Berkeley.

238 References
Lydall, Jean. 1988. ‘Gender, number and size in Hamar’, pp. 77–99 of Cushitic and Omotic:
papers from the International Symposium on Cushitic and Omotic Languages, Cologne,
January 6–9, 1986, edited by Marianne Bechhaus-Gerst and Fritz Serzisko. Hamburg:
Helmut Buske Verlag.
Macaulay, R. K. S. 1977. Language, social class and education. Edinburgh: Edinburgh University
Press.
Macaulay, R. K. S. 1978. ‘The myth of female superiority in language’. Journal of Child
Language 5: 353–63.
McConnell-Ginet, Sally. 1979. ‘Prototypes, pronouns and persons’, pp. 63–83 of Ethnology:
Boas, Sapir and Whorf revisited, edited by Madeleine Mathiot. The Hague: Mouton.
McConnell-Ginet, Sally. 2003. ‘“What’s in a name?” Social labeling and gender practices’,
pp. 69–97 of Holmes and Meyerhoff 2003.
McConnell-Ginet, Sally. 2014. ‘Gender and its relation to sex: the myth of “natural gender”’,
pp. 3–38 of Corbett 2014a.
McEwan, Ian. 2011. Solar. London: Vintage Books.
McGloin, Naomi Hanaoka. 1991. ‘Sex difference and sentence-final particles’, pp. 23–42 of Ide
and McGloin 1991.
McGregor, Donald E. and Aileen R. F. McGregor. 1982. Olo language materials. Canberra:
Pacific Linguistics.
McWhorter, John. 2014. The language hoax: why the world looks the same in any language.
Oxford: Oxford University Press.
Mahapatra, B. P. 1979. Malto: an ethnosemantic study. Manasagangotro, Mysore: Central
Institute of Indian Languages.
Maiden, Martin. 1998. A linguistic history of Italian. London: Routledge.
Maiden, Martin and Cecilia Robustelli. 2007. A reference grammar of Modern Italian. 2nd
edition. New York: MacGraw Hill.
Mallinson, Graham. 1986. Rumanian. London: Croom Helm.
Marcato, Gianna and Eva-Maria Thüne. 2002. ‘Gender and female visibility in Italian’,
pp. 187–18 of Hellinger and Bussmann 2002, Volume 2.
Maring, J. 1975. ‘Speech variation in Acoma Keresan’, pp. 471–85 of Linguistics and anthropology
in honor of C. F. Voegelin. Lisse: de Ridder.
Martel, Brigitte. 1981. Né homme, comment je suis devenue femme. Montréal: Québecor.
Martin, Samuel E. 1975. A reference grammar of Japanese. Tokyo: Charles E. Tuttle Company.
Martin, Samuel E. 1992. A reference grammar of Korean: a complete guide to the grammar and
history of the Korean language. Tokyo: Tuttle Publishing.
Martynyuk, A. 1990. ‘A contrastive study of male and female occupational terms in English
and Russian’. Papers and Studies in Contrastive Linguistics 26: 103–10.
Masica, Colin. 1991. The Indo-Aryan languages. Cambridge: Cambridge University Press.
Matasović, Ranko. 2004. Gender in Indo-European. Heidelberg: Winter.
Mathiot, Madeleine (ed.). 1979a. Ethnology: Boas, Sapir and Whorf revisit(ed). The Hague:
Mouton.
Mathiot, Madeleine. 1979b (assisted by Marjorie Roberts). ‘Sex roles as revealed through
referential gender in American English’, pp. 1–47 of Ethnology: Boas, Sapir and Whorf
revisited, edited by Madeleine Mathiot. The Hague: Mouton.

References 239
Matras, Yaron and Jeannette Sakel (eds.). 2007. Grammatical borrowing in cross-linguistic
perspective. Berlin: Mouton de Gruyter.
Matthews, P. H. 2010. Oxford concise dictionary of linguistics. Oxford: Oxford University Press.
Matthews, W. K. 1956. ‘The Latvian element in modern Livonian’, pp. 307–18 of Festschrift für
Max Vasmer zum 70 Geburtstag, edited by Margarete Woltner und Herbert Bräuer. Berlin:
Osteuropa Institut an der freien Universität Berlin.
Meillet, Antoine. 1931. Essai de chronologie des langues indo-européennes. Paris: Librairie
Ancienne Honoré Champion.
Meillet, Antoine. 1964. ‘The feminine gender in the Indo-European languages’, p.124 of
Language in culture and society: a reader in linguistics and anthropology, edited by Dell
Hymes. New York: Harper and Row.
Melchert, H. C. 1994. ‘The feminine gender in Anatolian’, pp. 231–44 of Früh-, Mittel-,
Spätindogermanisch: Akten der IX Fachtagung der indogermanischen Gesellschaft. Wiesbaden:
Dr. Ludwig Reichert Verlag.
Mendes, Ronald Beline. 2012. ‘Diminutivos como marcadoes de sexo/gênero’. Revista Linguística
(PGL UFRJ) 8: 113–24.
Mendes, Ronald Beline. 2014a. ‘Diminutives and masculinity in Brazilian Portuguese’,
pp. 117–32 of Milani 2014.
Mendes, Ronald Beline. 2014b. ‘Gendered perceptions of noun agreement in Brazilian Portuguese’.
Revista Internacional de Lingüística Iberoamericana 17: 93–108.
Menn, Lise, Michael P. O’Connor, and Audrey Holland. 1995. Studies in speech pathology and
clinical linguistics: non-fluent aphasia in a multilingual world. Amsterdam: John Benjamins.
Merlan, F. 1994. A grammar of Wardaman: a language of the Northern Territory of Australia.
Mouton Grammar Library 11. Berlin: Mouton de Gruyter.
Merlan, F., S. P. Roberts, and A. Rumsey. 1997. ‘New Guinea “classificatory verbs” and
Australian noun classification: a typological comparison’, pp. 63–103 of Nominal classification
in Aboriginal Australia, edited by M. Harvey and N. Reid. Amsterdam: John Benjamins.
Migge, Bettina. 2001. ‘Communicating gender in the Eastern Maroon Creole of Suriname’,
pp. 84–104 of Hellinger and Bussmann 2001, Volume 1.
Migliazza, Ernest C. 1978. ‘Yanomama diglossia’, pp. 561–80 of Approaches to language, edited
by William McCormack and Stephen A. Wurm. The Hague: Mouton.
Milani, Tommaso N. (ed.). 2014. Language and masculinities: performances, intersections,
dislocations. Hoboken: Taylor and Francis.
Miller, Casey and Kate Swift. 1981. The handbook of non-sexist writing for writers, editors and
speakers. London: The women’s press.
Milles, Karin. 2011. ‘Feminist language planning in Sweden’. Current Issues in Language
Planning 12: 21–33.
Mills, A. E. 1986. The acquisition of gender: a study of English and German. Berlin: Springer.
Mills, Sara. 2003a. Gender and politeness. Cambridge: Cambridge University Press.
Mills, Sara. 2003b. ‘Third Wave feminist linguistics and the analysis of sexism’. Discourse
analysis online.
Mills, Sara. 2003c. ‘Caught between sexism, anti-sexism and “political correctness”: feminist
women’s negotiations with naming practices’. Discourse and Society 14.1: 87–110.
Mills, Sara. 2008. Language and sexism. Cambridge: Cambridge University Press.

240 References
Milroy, L. 1999. ‘Women as innovators and norm-creators: the sociolinguistics of dialect
leveling in a northern English city’, pp. 361–76 of Engendering communication: proceedings
of the 5th Berkeley women and language conference, edited by S. Wertheim, A. C. Bailey, and
M. Corston-Oliver.
Miranda, Rocky V. 1975. ‘Indo-European gender: a study in semantic and syntactic change’.
Journal of Indo-European Studies 3: 199–215.
Mithun, Marianne. 1999. The languages of Native North America. Cambridge: Cambridge
University Press.
Mithun, Marianne. 2014. ‘Gender and culture’, pp. 131–60 of Corbett 2014a.
Mladenova, Olga M. 2001. ‘Neuter designations of humans and norms of social interactions in
the Balkans’. Anthropological Linguistics 43: 18–53.
Moñino, Yves. 1977. ‘Conceptions du monde et langue d’initiation lá'bì des Gbaya-Kara’,
pp. 115–47 of Langages et cultures africaines: essais d’etnolinguistique, edited by Geneviève
Calame-Griaule. Paris: Maspero.
Moore, Henrietta L. 1994. ‘Understanding sex and gender’, pp. 813–30 of Companion encyclopedia
of anthropology: humanity, culture and social life, edited by Tim Ingold. London:
Routledge.
Morris, Lori. 1999. ‘The grammar of English gender’, pp. 185–203 of Between grammar and
lexicon, edited by Ellen Contini-Morava and Yishai Tobin. Amsterdam: John Benjamins.
Motschenbacher, Heiko. 2010. ‘Redefining genderlect’, pp. 45–59 of Language, gender and
sexual identity, by Heiko Motschenbacher. Amsterdam: John Benjamins.
Mous, Maarten. 2012. ‘Cushitic’, pp. 342–422 of Frajzyngier and Shay 2012.
Mugglestone, Lynda. 2011. Dictionaries: a very short introduction. Oxford: Oxford University
Press.
Mullen, M. K. 1990. ‘Children’s classification of nature and artifact pictures into female and
male categories’. Sex Roles 23: 577–87.
Munro, Pamela. 1997. ‘The Garifuna gender system’, pp. 443–61 of The life of language: papers
in linguistics in honor of William Bright, edited by Jane H. Hill, P. J. Mistry, and Lyle
Campbell. Berlin: Mouton de Gruyter.
Myers, Fred R. 1986. Pintupi country, Pintupi self: sentiment, place, and politics among Western
Desert Aborigines. Washington, DC: Smithsonian Institution; Canberra: Australian Institute
of Aboriginal Studies.
Nakamura, Momoko. 2015. Gender, language and ideology: a genealogy of Japanese women’s
language. Amsterdam: John Benjamins.
Nanda, Serena. 1990. Neither man nor woman: the hijras of India. Belmont, Calif.: Wadsworth
Publishing Co.
Nekitel, Otto. 1986. ‘A sketch of nominal concord in Abu' (an Arapesh language)’, pp. 177–205
of Papers in New Guinea linguistics 24, edited by D. C. Laycock et al. Canberra: Pacific
Linguistics.
Newman, Michael. 1997. Epicene pronouns: the linguistics of a prescriptive problem. New York:
Garland.
Newman, Paul. 2000. The Hausa language: an encyclopedic reference grammar. New Haven
and London: Yale University Press.
Nissen, Uwe Kjær. 2002. ‘Aspects of translation gender’. Linguistik online 11.2/2: 25–37.

References 241
Nissen, Uwe Kjær. 2013. ‘Is Spanish becoming more gender fair? A historical perspective
on the interpretation of gender-specific and gender-neutral expressions’. Linguistik online
58.1/13: 99–117.
Ochs, Elinor. 1992. ‘Indexing gender’, pp. 335–58 of Rethinking context: language and an
interactive phenomenon, edited by Alessandro Duranti and Charles Goodwin. Cambridge:
Cambridge University Press.
Ogawa, Naoko and Janet S. Smith (Shibamoto). 1997. ‘The gendering of the gay male sex class
in Japan: a case study based on Rasen no Sobyo’, pp. 402–15 of Livia and Hall 1997.
Okamoto, Shigeko. 1995. ‘“Tasteless” Japanese: less “feminine” speech among young Japanese
women’, pp. 297–325 of Bucholtz and Hall 1995a.
Olza Zubiri, Jesús, Conchita Nuni de Chapi, and Juan Tube. 2001. Gramática Moja Ignaciana
(Morfosintaxis). San Cristóbal: Universidad Católica del Táchira.
Onishi, M. 1994. A grammar of Motuna (Bougainville, Papua New Guinea). PhD thesis,
Australian National University.
Onishi, M. 1996. ‘A grammatical summary of Bengali’. Materials for the Research Project
‘Universals of Human Languages’, Australian National University/James Cook University.
Ortner, Sherry B. 1972. ‘Is female to male as nature to culture?’, pp. 57–87 of Woman, culture
and society, edited by Michelle Zimbalist Rosaldo and Louise Lamphere. Stanford, Calif.:
Stanford University Press.
Ortner, Sherry B. and Harriet Whitehead. 1981. ‘Introduction: accounting for sexual meanings’,
pp. 1–28 of Sexual meanings: the cultural construction of gender and sexuality, edited
by S. Ortner and H. Whitehead. Cambridge: Cambridge University Press.
Osborne, C. R. 1974. The Tiwi language: grammar, myths and dictionary of the Tiwi language
spoken on Melville and Bathurst Islands, northern Australia. Canberra: Australian Institute
of Aboriginal Studies.
Osgood, Charles E. 1960. ‘The cross-cultural generality of visual-verbal synesthetic tendencies’.
Behavioral Science 5: 146–69.
Ott, W. and R. Ott. 1983. Diccionario ignaciano y castellano con apuntes gramaticales.
Cochabamba: SIL.
Oussikoum, Bennaser. 2013. Dictionnaire amazighe–français. Le parler des Ayt Wirra. Moyen
Atlas–Maroc. Rabat : IRCAM.
Owens, Jonathan. 2001. ‘Creole Arabic: the orphan of all orphans’. Anthropological Linguistics
43: 348–78.
Owens, Jonathan. 2014. ‘The morphologization of an Arabic Creole’. Journal of Pidgin and
Creole Languages 29: 232–98.
Palmer, W. D. 2009. Kokota grammar. Honolulu: University of Hawaii Press.
Pandharipande, Rajeshwari V. 1997. Marathi. London: Routledge.
Pankhurst, Helen. 1992. Gender, development and identity: an Ethiopian study. London: Zed Books.
Pastre, Geneviève. 1997. ‘Linguistic gender play among French gays and lesbians’, pp. 369–79
of Livia and Hall 1997.
Pauwels, Anne. 1998. Women changing language. London: Longman.
Pawley, Andrew K. 2004. ‘Using He and She for inanimate referents in English: questions of
grammar and world view’, pp. 110–37 of Ethnosyntax: explorations in grammar and culture,
edited by N. J. Enfield. Oxford: Oxford University Press.

242 References
Payne, Judith. 1989. Lecciones para el aprendizaje del idioma Asheninca. Serie Lingüística
Peruana 28, Pucalpa, Peru.
Pearce, Mary. 2013. ‘Variation with gender in the tonal speech varieties of Kera (Chadic)’,
pp. 79–93 of Atanga et al. 2013.
Penelope, Julia. 1990. Speaking freely. New York: Pergamon Press.
Pensalfini, R. 2003. A grammar of Jingulu, an Aboriginal language of the Northern Territory.
Canberra: Pacific Linguistics.
Pérez-Pereira, M. 1991. ‘The acquisition of gender: what Spanish children tell us’. Journal of
Child Language 18: 571–90.
Perlak, D. and G. Jarema. 2003. ‘The recognition of gender-marked nouns and verbs in Polishspeaking
aphasic patients’. PubMed 39: 383–403.
Pet, W. J. A. 1987. Lokono Dian, the Arawak language of Suriname: a grammatical sketch of its
grammatical structure and lexicon. PhD thesis, Cornell University.
Peters, Pam. 1995. The Cambridge Australian style guide. Cambridge: Cambridge University
Press.
Pfarr, S. J. 1984. ‘Status conflict: the rebellion of the tea pourers’, pp. 214–40 of Conflict in
Japan, edited by E. S. Kraus, T. P. Rohlen, and P. G. Steinhoff. Honolulu: University of
Hawaii Press.
Philips, S. U. 1980. ‘Sex differences in language’. Annual Review of Anthropology 9: 523–44.
Philips, S. U., S. Steele, and C. Tanz (eds.). 1987. Language, gender, and sex in comparative
perspective. Cambridge: Cambridge University Press.
Plank, F. and W. Schellinger. 1997. ‘The uneven distribution of genders over numbers:
Greenberg Nos. 37 and 45’. Linguistic Typology 1: 53–101.
Podesva, R., S. Roberts, and K. Campbell-Kibler. 2002. ‘Sharing resources and indexing
meanings in the production of gay styles’, pp. 175–90 of Language and sexuality: contesting
meaning in theory and practice, edited by K. Campbell-Kibler, R. Podesva, S. Roberts, and
A. Wong. Stanford, Calif.: CSLI.
Pope,M.K.1934. FromLatintoModernFrench. Manchester: Manchester University
Press.
Popova, M. I. 1958. ‘Grammatic᷈eskie elementy jazyka v rec᷈i detej preddos᷈kol'nogo vozrasta’
(‘Grammatical elements of language in the speech of children of pre-school age’). Voprosy
psixologii 4.3: 106–17. (Russian version of Popova 1973.)
Popova, M. I. 1973. ‘Grammatical elements of language in the speech of pre-school children’,
pp. 269–80 of Studies of child language development, edited by C. A. Ferguson and
D. I. Slobin. New York: Holt, Rinehart and Winston.
Poser, William J. 2005. ‘Noun classification in Carrier’. Anthropological Linguistics 47: 143–68.
Posner, Rebecca. 1966. The Romance languages: a linguistic introduction. New York: Anchor
books.
Pountain, C. J. 2005. ‘Gender without sex: the semantic exploitation of the masculine/feminine
opposition in the history of Spanish’, pp. 329–48 of Studies on Ibero-Romance linguistics
dedicated to Ralph Penny, edited by R. Wright and R. Ricketts. Newark, Del.: Juan de la
Cuesta.
Priestly, Tom M. S. 1983. ‘On “drift” in Indo-European gender systems’. Journal of Indo-
European Studies 11.3/4: 339–63.

References 243
Puşcariu, Sextil. 1943. Die rumänische Sprache. Ihr Wesen und ihre volkliche Prägung. Leipzig:
Otto Harrassowitz.
Quinn, Conor. 2001. ‘A preliminary survey of animacy categories in Penobscot’, pp. 395–426
of Actes du trente-deuxième congrès des Algonquinistes, edited by John D. Nichols. Winnipeg:
University of Manitoba.
Quintero, Carolyn. 2004. Osage grammar. Lincoln and London: University of Nebraska Press.
Quirk, Randolph and C. L. Wrenn. 1957. An Old English grammar. 2nd edition. London:
Methuen.
Raffelsiefen, Renate. 2011. ‘Reference to body parts in attributive adjective constructions’.
Paper presented at a Conference ‘Body parts’, University of Warsaw.
Raga, Amanuel and Hirut Woldemariam. 2011. ‘Gender bias as manifested in the grammar of
Afan Oromo’. Journal of Language and Culture 2: 162–73.
Rankin, Robert L. 1976. ‘Latin kw, gw > Rumanian p, b: an explanation’, pp. 14–26 of Current
studies in romance linguistics. Washington, DC: Georgetown University Press.
Rapold, Christian J. 2006. ‘Towards a grammar of Benchnon’. PhD thesis, University of Leiden.
Rat, Joseph Numa. 1898. ‘The Carib language as now spoken in Dominica’. Journal of the
Anthropological Institute of Great Britain and Ireland 27: 293–315.
Reid, Nicholas. 1997. ‘Class and classifier in Ngan.'gityemerri’, pp. 165–228 of Harvey and
Reid 1997.
Repp, Hanna. 1996. ‘Die Genustausch in der Kindersprache des Arabischen und das Komplementärwesen
“Qarîna”—Konzepte des Volksglaubens zur Erklärung eines sprachlichen
Phänomens’. Zeitschrift für Arabische Linguistik 32: 25–56.
Reynolds, Katsue Akiba. 1985. ‘Female speakers of Japanese’. Feminist Issues 5: 13–46.
Reynolds, Katsue Akiba. 1991. ‘Female speakers of Japanese in transition’, pp. 129–46 of Ide
and McGloin 1991.
Reynolds, Katsue Akiba. 2011. ‘Female speakers of Japanese in transition’, pp. 291–9 of Coates
and Pichler 2011.
Ribeiro, Eduardo Rivail. 2009. ‘Macro-Jê’, pp. 665–9 of Concise encyclopedia of languages of the
world, edited by Keith Brown and Sarah Ogilvie. Oxford: Elsevier.
Ribeiro, Eduardo Rivail. 2012. ‘A grammar of Karajá’. PhD dissertation, University of Chicago.
Richards, Cara B. 1974. ‘Among the liberated’, pp. 401–19 of Many sisters: women in crosscultural
perspective, edited by Carolyn J. Matthiasson. New York: Free Press.
Rodina, Yulia. 2014. ‘Variation in the input: child and caregiver in the acquisition of grammatical
gender in Russian’. Language Sciences 43: 116–32.
Rodrigues, Aryon Dall’Igna. 1999. ‘Macro-Jê’, pp. 165–206 of The Amazonian languages,
edited by R. M. W. Dixon and A. Y. Aikhenvald. Cambridge: Cambridge University Press.
Romaine, Suzanne. 1999. Communicating gender. London: Lawrence Erlbaum Associates.
Romani Miranda, Maggie M. 2004. ‘Toponimia en el Gran Pajonal con especial atención a los
topónimos de afiliación Ashaninka’. PhD dissertation, Universidad Nacional Mayor de San
Marcos, Lima.
Roscoe, Paul B. 2001. ‘“Strength” and “sexuality”: sexual avoidance and masculinity in New
Guinea and Amazonia’, pp. 279–308 of Gender in Amazonia and Melanesia: an exploration
of the comparative method, edited by Thomas A. Gregor and Donald Tuzin. Berkeley and
Los Angeles: University of California Press.

244 References
Roscoe, Will. n.d. ‘Berdaches’. Encyclopedia of the Great Plains. .
Rose, Françoise. 2013. ‘Los generolectos del mojeño’. LIAMES 13: 115–34.
Rosenhouse, J. and N. Dbayyat. 2006. ‘Gender switch in female speech of an urbanized Arabic
dialect in Israel’. Anthropological Linguistics 48: 169–86.
Ross, Kristiina. 1996. ‘Naine eesti keele grammatikas’ [The woman in Estonian grammar].
Vikerkaar 11/12: 102–6.
Ross, Kristiina. 2012. ‘Sõbranna—sõbralaadne toode. Kultuurlisi representatsioone meilt ja
mujalt’. [Female friend—a friend resembling product: cultural representations from us and
from elsewhere]. Vikerkaar 1.2: 166–73.
Rothstein, Robert A. 1973. ‘Sex, gender, and the October Revolution’, pp. 460–6 of A Festschrift for
Morris Halle, edited by S. R. Anderson and P. Kiparsky. New York: Holt, Rinehart and Winston.
Rubino, C. R. G. 1997. A reference grammar of Ilocano. PhD thesis, University of California
(Santa Barbara).
Rudes, B. and B. Healey. 1979. ‘Is she for real? The concepts of femaleness and maleness in the
gay world’, pp. 49–61 of Mathiot 1979a.
Rushforth, S. 1991. ‘Uses of Bearlake and Mescalero (Athapaskan) classificatory verbs’. International
Journal of American Linguistics 57: 251–66.
Sa'ar, Amalia. 2007. ‘Masculine talk: on the subconscious use of masculine linguistic forms
among Hebrew- and Arabic-speaking women in Israel’. Signs 32: 405–29.
Sadiqi, F. 2003. Language and gender in Morocco. Leiden: Brill.
Sahoo, Kalyanamalini. 2003. ‘Linguistic and socio-culural implications of gendered structures
in Oriya’, pp. 239–58 of Hellinger and Bussmann 2003, Volume 3.
Salzmann, Zdenek. 2004. Language, culture and society: an introduction to linguistic anthropology.
Boulder, Colo.: Westview Press.
Sandström, Caroline. 2000. ‘The changing system of grammatical gender in the Swedish
dialects of Nyland, Finland’, pp. 793–806 of Unterbeck and Rissanen 2000.
Sapir, Edward. 1929a. ‘Male and female forms of speech in Yana’, pp. 79–85 of Donum
Natalicium Schrijnen, edited by St. W. J. Teeuwen. Nijmegen-Utrecht: N. V. Dekker and
van de Vegt.
Sapir, Edward. 1929b. ‘Status of linguistics as a science’. Language 5: 207–14.
Sarvasy, Hannah. 2016. ‘Sexless babies, sexed grandparents. Nungon gendered person forms’.
International Journal of Language and Culture.
Sasse, Hans-Jürgen. 1974. ‘Notes on the structure of Galab’. BSOAS 37: 407–38.
Sasse, Hans-Jürgen. 1985. ‘Sprachkontakt und Sprachwandel: die Gräzisierung der albanischen
Mundarten Griechenlands’. Papiere zur Linguistik 32: 37–95.
Saville-Troike, M. 1988. ‘A note on men’s and women’s speech in Koasati’. International
Journal of American Linguistics 54: 241–2.
Sayers, Dorothy L. 1995. Lord Peter. New York: Harper paperback.
Schane, S. A. 1970. ‘Phonological and morphological markedness’, pp. 286–94 of Progress in
linguistics, edited by M. Bierwisch and K. E. Heidolph. The Hague: Mouton.
Schapper, Antoinette. 2010a. ‘Bunaq, a Papuan language of central Timor’. PhD thesis,
Australian National University.

References 245
Schapper, Antoinette. 2010b. ‘Neuter gender in Eastern Indonesia’. Oceanic Linguistics 49:
407–35.
Schaub, W. 1985. Babungo. Croom Helm Descriptive Grammars. London: Croom Helm.
Schenker, Alexander M. 1993. ‘Proto-Slavonic’, pp. 60–121 of The Slavonic languages, edited
by Bernard Comrie and Greville G. Corbett. London: Routledge.
Schieffelin, Bambi B. 1987. ‘Do different worlds mean different words? An example from
Papua New Guinea’, pp. 249–60 of Philips, Steele, and Tanz 1987.
Schieffelin, Edward L. 1977. The sorrow of the lonely and the burning of the dancers. St Lucia:
University of Queensland Press.
Schipper, Mineke. 2003. Never marry a woman with big feet: women in proverbs from around
the world. New Haven and London: Yale University Press.
Schmidt, A. 1985. Young People’s Dyirbal: an example of language death from Australia.
Cambridge: Cambridge University Press.
Senft, Gunter. 1986. Kilivila: the language of Trobriand Islanders. Berlin: Mouton de Gruyter.
Senft, Gunter. 1996. Classificatory particles in Kilivila. New York: Oxford University Press.
Sera, M., C. Berge, and J. del Castillo. 1994. ‘Grammatical and conceptual forces in the
attribution of gender by English and Spanish speakers’. Cognitive Development 9: 261–92.
Seyoum, Mulugeta. 2008. A grammar of Dime. Amsterdam: LOT.
Shepard Jr., G. 1997. ‘Noun classification and ethnozoological classification in Machiguenga,
an Arawakan language of the Peruvian Amazon’. Journal of Amazonian Languages 1: 29–57.
Sherzer, Joel. 1987. ‘The diversity of voices: men’s and women’s speech in ethnographic
perspective’, pp. 95–120 of Philips, Steele, and Tanz 1987.
Shibatani, Masayoshi. 1990. The languages of Japan. Cambridge: Cambridge University Press.
Silverman, Eric K. 2001. Masculinity, motherhood and mockery: psychoanalyzing culture and
the Iatmul Naven Rite in New Guinea. Ann Arbor: The University of Michigan Press.
Silverstein, Michael. 1972a. ‘Chinook Jargon: language contact and the problem of multi-level
generative systems I’. Language 48: 378–406.
Silverstein, Michael. 1972b. ‘Chinook Jargon: language contact and the problem of multi-level
generative systems II’. Language 48: 596–625.
Silverstein, Michael. 1985. ‘Language and the culture of gender: at the intersection of structure,
usage and ideology’, pp. 219–59 of Semiotic meditation: sociocultural and psychological
perspectives, edited by Elizabeth Mertz and Richard J. Parmentier. Orlando, Fla.: Academic
Press.
Silverstein, Michael. 2015. ‘Presupposing demographic sex, entailing sociocultural gender’.
Paper presented at the symposium ‘Indexing gender revisited: on the non-referential aspects
of gendering’, 14th International Pragmatics Conference, International Pragmatics Association,
26–31 July, Antwerp, Belgium.
Sims, C. and H. Valiquette. 1990. ‘More on male and female speech in (Acoma and Laguna)
Keresan’. International Journal of American Linguistics 56: 162–6.
Singer, Ruth. 2010. ‘Creativity in the use of gender agreement in Mawng: how the discourse
functions of a gender system can approach those of a classifier system’. Studies in Language
34: 382–416.
Sini, M. 2011. ‘Emergence of patrilineal system and its impact on North Malabar—a study of
Chrakkal village in Kannur district’. PhD dissertation, Kannur University.

246 References
Slobin, Dan I. 1996. ‘From “thought and language” to “thinking for speaking”’, pp. 70–96 of
Rethinking linguistic relativity, edited by John J. Gumperz and Stephen G. Levinson. Cambridge:
Cambridge University Press.
Smith-Stark, S. 1974. ‘The plurality split’. Papers from the Annual Regional Meeting of the
Chicago Linguistic Society 10: 657–71.
Sneddon, James N. 1996. Indonesian reference grammar. Sydney: Allen and Unwin.
Sohn, Ho-Min. 1994. Korean. London: Routledge.
Sontag, Susan. 1973. ‘The third world of women’. Partisan Review 40.2: 181–206.
Southerland, Ronald H. and Francis Katamba. 1996. ‘Language in social contexts’, pp. 540–90
of Contemporary linguistics: an introduction, edited by William O’Grady, Michael Dobrovolsky,
and Francis Katamba. London: Longman.
Souza, L. 2011. ‘Fala masculina e fala feminina no kadiwéu.’. Anais do VII Congresso Internacional
da Abralin 2488–96.
Speece, R. n.d. Angave grammar. Ukarumpa: Summer Institute of Linguistics.
Spender, Dale. 1980. Man made language. London: Routledge and Kegan Paul.
Spitulnik, D. 1989. ‘Levels of semantic restructuring in Bantu noun classification’, pp. 207–20
of Current approaches to African linguistics, Volume 5, edited by P. Newman and R. D. Botne.
Dordrecht: Foris Publications.
Stebbins, Tonya N. 2005. ‘Nominal classification in Mali’. Anthropological Linguistics 47:
77–131.
Steinhauer, H. 1986. ‘Number in Biak: counterevidence to two alleged language universals a
summary’, pp. 171–3 of FOCAL I: papers from the Fourth International Conference on
Austronesian Linguistics, edited by P. Geraghty, L. Carrington, and S. A. Wurm. Canberra:
Pacific Linguistics.
Storch, Anne. 2006. ‘How long do linguistic areas last? Western Nilotic grammars in contact’,
pp. 94–113 of Grammars in contact: a cross-linguistic typology, edited by A. Y. Aikhenvald
and R. M. W. Dixon. Oxford: Oxford University Press.
Storch, Anne. 2011. Secret manipulations. New York: Oxford University Press.
Storch, Anne. 2013. ‘Possession in Hone’, pp. 208–23 of Aikhenvald and Dixon 2013.
Storch, Anne. 2014. A grammar of Luwo: an anthropological approach. Amsterdam: John
Benjamins.
Storjohann, Petra. 2004. ‘frau auf dem linguistischen Prüfstand: eine korpusgestützte Gebrauchsanalyse
feministischer Indefinitpronomen’. German Life and Letters 57.3: 309–26.
Stotko, Elaine M. and Margaret Troyer. 2007. ‘A new gender-neutral pronoun in Baltimore,
Maryland: a preliminary study’. American Speech 82: 262–79.
Stradelli, Ermano. 1929. Vocabulários da Língua Geral Portuguez-Nheêngatú e Nheêngatú-
Portuguez. Rio de Janeiro: Revista do Instituto Historico-Geográfico.
Straus, Anne Terry and Robert Brightman. 1982. ‘The implacable raspberry’. Papers in
Linguistics 15: 97–137.
Studzińska, Joanna. 2012. ‘Translating personification across gendered languages’, pp. 259–68
of Cognitive processes in language, edited by K. Kosecki and J. Badio. Bern: Peter Lang.
Sumbuk, Kenneth Memson. 1999. ‘Morphosyntax of Sare’. PhD thesis, University of Waikato.
Svartengren, T. Hilding. 1927. ‘The feminine gender for inanimate things in Anglo-American’.
American Speech 3: 83–113.

References 247
Svartengren, T. Hilding. 1954. ‘The use of feminine gender for inanimate things’. Moderna
Sprak 48: 261–92.
Tadmor, Uri. 2007. ‘Grammatical borrowing in Indonesian’, pp. 301–28 of Matras and Sakel
2007.
Taïfi, Miloud. 1991. Dictionnaire tamazight–français. Parlers du Maroc central. Paris:
L’Harmattan.
Talbot, M. 2010. Language and gender. Cambridge: Polity Press.
Tannen, Deborah. 1990. You just don’t understand: women and men in conversation. New
York: William Morrow.
Taylor, Allan R. 1982. ‘“Male” and “female” speech in Gros Ventre’. Anthropological Linguistics
24: 301–7.
Taylor, Douglas M. 1951. ‘Sex gender in Central American Carib’. International Journal of
American Linguistics 17: 102–4.
Taylor, Douglas M. 1959. ‘La catégorie du genre en caraïbe insulaire: un cas d’interférence
linguistique?’. Bulletin de la Société de Linguistique de Paris 54: 201–7.
Taylor, Douglas M. 1977. Languages of the West Indies. Baltimore: Johns Hopkins University
Press.
Teferra, Anbessa. 1987. ‘Balishsha: women’s speech among the Sidaama’. Journal of Ethiopian
Studies 20: 44–59.
Teso, Elena and Liz Crolley. 2013. ‘Gender-based linguistic reform in international organizations’.
Language Policy 12: 139–58.
Thomason, S. G. and T. Kaufman. 1988. Language contact, creolization and genetic linguistics.
Berkeley and Los Angeles: University of California Press.
Thompson, L. C. 1987. A Vietnamese grammar. Seattle: University of Washington Press.
Tobin, Yishai. 2001. ‘Gender switch in Modern Hebrew’, pp. 177–98 of Hellinger and Bussmann
2001, Volume 1.
Trask, L. 2000. The dictionary of historical and comparative linguistics. Edinburgh: Edinburgh
University Press.
Trechter, Sara. 1995. ‘The pragmatic functions of gender deixis in Lakhota’. PhD dissertation,
University of Kansas.
Trechter, Sara. 1999. ‘Contextualizing the exotic few: gender dichotomies in Lakhota’,
pp. 101–22 of Bucholtz, Liang, and Sutton 1999.
Trechter, Sara. 2009. ‘Native languages of North America, variation in’, pp. 753–60 of Concise
encyclopedia of languages of the world, edited by Keith Brown and Sarah Ogilvie. Oxford:
Elsevier.
Treis, Yvonne. 2005. ‘Avoiding their names, avoiding their eyes: how Kambaata women respect
their in-laws’. Anthropological Linguistics 47: 292–320.
Trudgill, Peter. 1971. ‘The social differentiation of English in Norwich’. PhD dissertation,
University of Edinburgh.
Trudgill, Peter. 1974. The social differentiation of English in Norwich. Cambridge: Cambridge
University Press.
Trudgill, Peter. 1975. ‘Sex, covert prestige and linguistic change in the urban British English of
Norwich’, pp. 88–104 of Language and sex: difference and dominance, edited by B. Thorne
and N. Henley. Rowley, Mass.: Newbury House.

248 References
Trudgill, Peter. 1977. ‘Creolization in reverse: reduction and simplification in the Albanian
dialects of Greece’. Transactions of the Philological Society 77: 32–50.
Trudgill, Peter. 1998. ‘Sex and covert prestige’, pp. 20–6 of Language and gender: a reader. 2nd
edition, edited by Jennifer Coates and Pia Picher. London: Wiley-Blackwell.
Trudgill, Peter. 2007. ‘On grammatical gender as an arbitrary and redundant category’,
pp. 24–36 of History of Linguistics 2005: selected papers from the Third International
Conference on the History of Language Sciences, Urbana-Champaign, Illinois, 1–5 September
2005, edited by Douglas Kibbee. Amsterdam: John Benjamins.
Tsegaye, Mulugeta T., Maarten Mous, and Niels O. Schiller. 2014. ‘Plural as a value in the
investigation of grammatical gender’, pp. 161–90 of Corbett 2014a.
Unterbeck, Barbara and Matti Rissanen (eds.). 2000. Gender in grammar and cognition,
I. Approaches to gender. II. Manifestations of gender. Berlin: Mouton de Gruyter.
Upton, Rebecca. 2012. ‘Gender’. Oxford bibliographies online. Oxford: Oxford University Press.
Uyeno, T. 1971. ‘A study of Japanese modality: a performative analysis of sentence particles’.
PhD dissertation, University of Michigan.
Vajda, Edward J. 2004. Ket. Munich: Lincom Europa.
Valentine, J. Randolph. 2001. Nishnaabemwin reference grammar. Toronto: University of
Toronto Press.
Vallejos, Rosa Yopán. 2010. ‘A grammar of Kokama-Kokamilla’. PhD dissertation, University
of Oregon.
Vallejos, Rosa Yopán. 2015. ‘La indexalidad de género en kukama-kukamiria desde una
perspective tipológica’, to appear in Estudios de lenguas ameríndias 3. Escenarios de diversidad
lingüística, edited by Zarina Estrada Fernández, Ana V. Fennándex Garay, and Albert
Alvarez Gonzalez. Hermosillo, Son.: Editorial UNISON.
Vallejos, Rosa Yopán. Forthcoming. ‘Structural outcomes of obsolescence and revitalization:
documenting variation among the Kukama-Kukamiria’, to appear in Language documentation
and revitalization in Latin America, edited by Gabriela Perez-Baez, Chris Rogers, and
Jorge Rosés-Labrada. Berlin: De Gruyter Mouton.
van Baarle, Peter and M. A. Sabajo. 1997. Manuel de la langue Arawak. Paris: Éditions du Saule.
van den Berg, René. 2015. ‘The loss of clusivity and the rise of gender in West Oceanic
pronominals’. Language and Linguistics in Melanesia 33: 10–47.
Vasvári, Louise O. 2011. ‘Grammatical gender trouble and Hungarian gender[lessness]: Part
I. Comparative linguistic gender.’ AHEA E-Journal 4. .
Vasvári, Louise O. 2015. ‘Gender trouble in a grammatically genderless language: Hungarian’,
pp. 203–25 of Hellinger and Motschenbacker 2015.
Villas Bôas, Cláudio and Orlando Villas Bôas. 1970. Xingu: Os índios, seus mitos. Rio de
Janeiro: Zahar.
Vinogradov, V. V. 1947. Russkij Jazyk. (The Russian language.) Moscow: Uchpedgiz.
Wales, Katie. 1996. Personal pronouns in present-day English. Cambridge: Cambridge University
Press.
Walsh, M. 1976. The Murinypata language of north-west Australia. PhD thesis, Australian
National University.

References 249
Walsh, M. 1997. ‘Nominal classification and generics in Murrinhpatha’, pp. 255–92 of Harvey
and Reid 1997.
Walters, Margaret. 2005. Feminism: a very short introduction. Oxford: Oxford University Press.
Watkins, L. 1984. A grammar of Kiowa. Lincoln: University of Nebraska Press.
Weber, Doris. 2000. ‘On the function of gender’, pp. 495–509 of Gender in grammar and
cognition, I. Approaches to gender, edited by Barbara Unterbeck and Matti Rissanen. Berlin:
Mouton de Gruyter.
Wegener, Claudia. 2012. Grammar of Savosavo. Berlin: Mouton de Gruyter.
Welmers, Wm. E. 1973. African language structures. Berkeley: University of California Press.
Werner, Heinrich. 1994. Das Klassensystem in den Jenissei-Sprachen. Wiesbaden: Harrassowitz.
Werner, Heinrich. 1997. Die ketische Sprache. Wiesbaden: Harrassowitz.
Wertz, C. A. 1977. ‘The number of genders in Polish’. Canadian Slavonic Papers 19: 50–63.
Westermann, D. 1947. ‘Pluralbildung und Nominalklassen in einigen afrikanischen Sprachen’,
Abhandlungen der Deutschen Akademie der Wissenschaften zu Berlin 1945/46. Phil.-hist.
Klasse. Nr 1. Berlin.
Whitehead, Harriet. 1981. ‘The bow and the burden strap: a new look at institutionalized
homosexuality in native North America’, pp. 80–115 of Sexual meanings: the cultural
construction of gender and sexuality, edited by Sherry B. Ortner and Harriet Whitehead.
Cambridge: Cambridge University Press.
Wiedemann, Ferdinand Johann. 1875. Grammatik der ehstnischen Sprache, zunächst wie sie in
Mittelehstland gesprochen wird, mit Berücksichtigung der anderen Dialekte. St-Petersbourg.
Wikan, Unni. 1977. ‘Man becomes woman: transsexualism in Oman as a key to gender roles’.
Man 12: 304–19.
Wikan, Unni. 1978. ‘The Omani Xanith: a third gender role?’ Man 13: 473–5.
Williams, Nancy M. 1986. The Yolngu and their land: a system of land tenure and the fight for
its recognition. Canberra: Australian Institute of Aboriginal Studies.
Winford, Donald. 2005. ‘Contact-induced changes: classification and processes’. Diachronica
22: 373–427.
Woidich, Manfred. 2005. ‘Kindersprache in ilBas᷈andi. Ein text aus der Oase Dakhla’,
pp. 411–31 of Alltagsleben und materielle Kultur in der arabischen Sprache und Literatur.
Festschrift für Heinz Grotzfeld. Abhandlungen für die Kunde des Morgenlandes, Band LVI.
Wolfart, H. Christoph. 1973. Plains Cree: a grammatical study. Transactions of the American
Philosophical Society, NS 63–5. Philadelphia: The American Philological Society.
Wołk, Ewa. 2009. ‘Positive and negative emotions encoded in Amharic forms of address’,
pp. 128–36 of Coding and rituals of emotions in Asian and African cultures, edited by Nina
Pawlak. Warsaw: Dom Wydawniczy ELIPSA.
Yadav, Ramawatar. 1996. A reference grammar of Maithili. Berlin: Mouton de Gruyter.
Yaguello, Marina. 1979. Les mots et les femmes. Paris: Payot.
Yaguello, Marina. 1991. Petits faits de langue. Paris: Seuil.
Yigezu, Moges. 1998. ‘Women in society and female speech among the Suri of Southwestern
Ethiopia’, pp. 83–101 of Surmic languages and cultures, edited by Gerrit J. Dimmendaal and
Marco Last. Cologne: Rüdiger Köppe Verlag.
Yokoyama, Olga T. 1999. ‘Russian genderlects and referential expressions’. Language in Society
28: 401–29.

250 References
Zavala, Roberto. 1992. El Kanjobal de San Miguel Acatán. Volume 6. Universidad nacional
autónoma de México.
Zavala, Roberto. 2000. ‘Multiple classifier systems in Akatek (Mayan)’, pp. 114–46 of Systems
of nominal classification, edited by Bunter Senft. Cambridge: Cambridge University Press.
Zemskaya, Elena A., Margarita V. Kitajgorodskaja, and Nona N. Rozanova. 1993. ‘Osobennosti
muz᷈skoj i z᷈enskoj rec᷈i’ (Features of male and female speech), pp. 90–157 of Russkij jazyk v
ego funkcionirovanii: Kommunikativno-pragmatic᷈eskij aspekt, edited by Elena A. Zemskaya
and Dmitri N. Shmelev. Moscow: Nauka.
Zgusta, Ladislav. 1971. Manual of lexicography. The Hague: Mouton.
Zhang, Sihong. 2014. ‘A grammar of Ersu’. PhD thesis, James Cook University.
Zimman, Lal and Kira Hall. Forthcoming. ‘Language, embodiments, and the “third sex”’,to
appear in Language and identities, edited by Dominic Watt and Carmen Llamas. Edinburgh:
Edinburgh University Press.
Zubin, David A. and Klaus-Michael Köpcke. 1984. ‘Affect classification in the German gender
system’. Lingua 63: 41–86.
Zubin, David A. and Klaus-Michael Köpcke. 1986. ‘Gender and folk taxonomy: the indexical
relation between grammatical and lexical categorization’, pp. 139–80 of Noun classes and
categorisation, edited by C. G. Craig. Amsterdam: John Benjamins.
Zwicky, Arnold M. 1997. ‘Two lavender issues for linguists’, pp. 21–34 of Livia and Hall 1997.

Index of languages, linguistic families,
and peoples
For each language, the linguistic family or area to which it belongs is noted. Established
subgroups have been added for a number of well-researched families.
Abau (isolate, Papuan area) 121
Abu Arapesh (Arapesh, Papuan area) 23, 31
Acoma Keresan (isolate) 168, see also Keresan
Adamawa (Adamawa-Ubangi) 87
Afan Oromo, see Oromo
African languages 13, 31
Afroasiatic languages 13, 38–9, 44–5, 49–50
Aït Mguild (Berber, Afroasiatic) 103, 109
Aït Wirra (Berber, Afroasiatic) 103, 109
Akatek (Mayan) 78, 95
Alaaba (Cushitic, Afroasiatic) 183
Alamblak (Sepik Hill, Papuan area) 35, 44,
62, 120
Albanian (Indo-European) 23, 92
Albanian, dialect of Mandres 87, 97
Algonquian languages 8, 13, 20–2, 31, 55,
67, 118, 137
Alor-Pantar, Austronesian languages of 97
Amazonian languages 53, 162
American English (Germanic, Indo-European)
109–10, 188, 129–30, 154, 208, see also
English
Amharic (Semitic, Afroasiatic) 44–5, 51, 100,
104–5, 108–9, 117, 128, 146, 186
Amuesha (Arawak) 87
Anatolian languages (Indo-European) 81
Ancient Greek (Indo-European) 13, 89
Angan languages 51
Angave (Angan) 51
Arabic (Semitic, Afroasiatic) 38, 46, 105,
108–9, 117, 164, 203, 210, 216, see also
Egyptian Arabic, Moroccan Arabic,
Omani Arabic, Palestinian Arabic
Arabic Creole (Creole) 75
Arapaho (Algonquian) 165
Arapesh languages (Papuan area) 23, 31
Arawá languages 27, 32
Arawak, see Lokono Dian
Arawak languages 18, 40, 46, 67, 87, 91,
102–4, 121, 141–2, 183
Arawak-Carib mixed language 142–3
Archi (North-east Caucasian) 53
Armenian (North-east Caucasian) 183
Arsi Oromo (Cushitic, Afroasiatic) 183
Arvanitika (Albanian, Indo-European) 92
Ashaninka (Arawak) 121
Ashanti (Kwa) 103
Asheninca (Arawak) 121
Asia Minor Greek (Indo-European) 225,
see also Greek
Assamese (Indo-Aryan, Indo-European) 84
Atayal (Formosan, Austronesian) 162
Athabaskan languages 30, 67–8
Atsina (Algonquian) 8
Australian English (Germanic, Indo-European)
109, 198, see also English
Australian languages 3, 18, 21, 30–2, 41, 58,
62, 66, 69, 75, 80–9, 91, 120, 136, 145
Austroasiatic languages 74
Austronesian languages 18, 55, 63, 71, 97, 127
Avestan (Iranian, Indo-European) 82
Awetí (Tupí) 139–41, 144, 162
Ayt Seghrushen (Berber, Afroasiatic) 50–1
Babungo (Bantu) 21
Baining languages (Papuan area) 41, 44

252 Index of languages, linguistic families, and peoples
Baïnouk (Atlantic) 31
Balochi (Iranian, Indo-European) 82, 87
Baltic languages (Indo-European) 84
Balto-Finnic languages (Finno-Ugric) 87
Baluchi (Iranian, Indo-European) 82,
see also Balochi
Bandjalang (Australian area) 95
Baniwa of Içana (Arawak) 31
Bantu languages 12, 16, 21, 56, 71, 78, 92, 95,
98, 173, 182, 192
Bantu-based Creole, see Fanagalo
Baré (Arawak) 92
Bari (Nilotic) 77
Belizean Creole (Creole) 176
Benchnon (Omotic, Afroasiatic) 38–9,
46–7, 50
Bengali (Indo-Aryan, Indo-European) 83–4
Berber languages (Afroasiatic) 16, 38, 45,
103–4, 109, 117, 177, 184
Biak (Austronesian) 30
Big Nambas (Oceanic, Austronesian) 174,
181, 183
Biloxi (Siouan) 141, 160–2
Black Carib (Arawak) 144
Boiken (Ndu, Papuan area) 44, 49–50, 128
Botocudo (Macro-Jê) 183
Brahui (Dravidian) 87, 97
Brazilian Portuguese 51, 112, 118–19, 155,
158–9, 164, 214, see also Portuguese
Breton (Celtic, Indo-European) 84
Bulgarian (Slavic, Indo-European) 90, 106, 118
Bunaq (Austronesian) 55–6
Burmese (Tibeto-Burman) 64
Camus (Nilotic) 40
Cantabrian Spanish (Romance, Indo-
European) 42–7, 56, 61, 175–6, see also
Spanish
Carib languages 15, 18, 71, 142–3, 160, 162
Carrier (Athabaskan) 67–8, 74
Catalan (Romance, Indo-European) 84
Celtic languages (Indo-European) 84
Central Khoisan 40, 44, 56
Chadic languages 16, 50, 54, 79, 87, 176, 185
Chamicuro (Arawak) 87
Charmay French patois (Romance,
Indo-European) 175, see also French
Chehalis (Salish) 71
Chinese (Sino-Tibetan) 71, 98
Chinook Jargon (a creolized variety of
Chinook) 71, 75
Chinookan languages 71
Chukchi (Chukotko-Kamchatkan) 139,
145–6, 162, 208
Chukotko-Kamchatkan languages 8, 139
Clackamas Chinook (Chinookan) 88
Coatzoquitengo Mixtec (Mixtecan) 79
Cocama, see Kokama-Kokamilla
Cora (Uto-Aztecan) 68
Corsican (Romance, Indo-European)
179–81, 215
Cree (Algonquian) 20, 31
Creek (Muskogean) 173
Creole Arabic, see Arabic Creole
Creole languages 41, 71, see also Arabic
Creole, Belizean Creole, Fanagalo,
French Creole, Guyanais Creole French,
Maroon Creole (or Ndyuka),
Negerhollands Creole Dutch,
Papiamentu, Tok Pisin
Cross-River languages (Benue-Congo) 65,
87, 97
Cushitic languages 16, 22, 31, 39, 44, 50–1, 183
Czech (Slavic, Indo-European) 98, 123
Damin (initiation language of Lardil,
Australian area) 168–9, 182
Dangbon (Australian area) 90
Danish (Germanic, Indo-European) 93,
96–8, 204
Dardic languages (Indo-European) 87
Dasenech (Cushitic, Afroasiatic) 50
Dizin (Omotic, Afroasiatic) 129
Dolakha Newar (Tibeto-Burman) 4
Dravidian languages 13, 18, 31, 68, 83, 87,
90, 141, 160
Dutch (Germanic, Indo-European) 61, 72,
96, 204

Index of languages, linguistic families, and peoples 253
Dyirbal (Australian area) 91, 97
development of gender in 80–1
gender in 7, 13, 57, 117
obsolescence of the language 91
semantic basis of gender choice 19–22,
29–31, 35, 59, 69, 120–1, 212
variable gender 53–4
Eastern Maroon Creole of Suriname, see
Maroon Creole
Egyptian Arabic (Semitic, Afroasiatic) 98,
117, see also Arabic
Enga (Engan, Papuan area) 51, 67, 75
English (Germanic, Indo-European) 11–12,
53, 59, 69, 73–4, 81, 91, 94, 105, 123–35,
145, 173, see also American English,
Australian English, English of
Tasmania, Middle English, Old English
acquistion of gender in 92–3
female and male speech patterns in 138,
149, 155–61, 164
gender in 7, 11, 15, 30, 53, 72, 94–7,
108–12, 118–19, 187–90
generic masculine pronoun in 4, 26, 29,
86, 112–16, 174, 195–200, 203–16
history of gender in 6, 85–90, 96–7
language change in 175–7
loans from 12, 24, 71, 187
loss of gender agreement in 6, 15, 85–90
term for gender in 11–12
English of Tasmania 110, see also English
English-based Creoles 41, 72
Ersu (Tibeto-Burman) 67, 74
Estonian (Balto-Finnic, Uralic) 30, 72–5,
88–9, 93, 97, 122–3, 129, 134, 190–1,
197, 210, 216–17
Ewe (Kwa) 69, 75
Fanagalo (a Bantu-based Creole) 71, 75
Farizandi (Iranian, Indo-European) 86
Figuig (Berber, Afroasiatic) 104, 108–9
Finnish (Balto-Finnic, Uralic) 15, 30, 72, 75,
88, 97, 200
French (Romance, Indo-European)
gender in 5, 7, 11, 13–15, 23, 26, 84,
89, 97–8, 103, 113, 118–19, 156–60,
176, 217
loans from 23, 46, 71–2, 179
markedness in gender 26
sexist language in 194–204, 210
translation into 124–6
French Creole 23
Frisian (Germanic, Indo-European) 96
Gaagudju (Australian area) 18, 31
Gala (Ndu, Papuan area) 16, 44, 50, 99
Galab, see Dasenech
Gapun, see Taiap
Garifuna (Arawak) 144
Garvi (Dardic, Indo-European) 87
Gavar (Dardic, Indo-European) 87
Gbaya-Kara (Atlantic-Congo) 182
German (Germanic, Indo-European) 11
acquisition of 92–3, 97
assignment of gender, principles of 13–15,
22–5, 53, 57–8, 210
gender in 6–7, 13–16, 58–9, 85, 113,
122–35, 177, 213, 217
gender resolution in 27
loans from 88
neuter gender in 7, 52–3, 60
sexist language in 191, 194, 199–204,
210, 216
Germanic languages (Indo-European) 72,
88–9, 96–8
Godoberi (North-east Caucasian) 18
Gola (West Atlantic) 58
Greek (Indo-European) 11, 22–3, 52, 81, 88,
93, 113, 118, 135
Greek, Modern, see Greek
Greek dialects (Indo-European) 88
Gros Ventre (Algonquian) 137–40, 145–7, 163
Guyanais Creole French (Creole) 72
Hadza (isolate, Tanzania) 42, 44
Hamar (Omotic, Afroasiatic) 42–4, 49, 186
Hausa (Chadic, Afroasiatic) 13, 22, 71, 83
Hawaiian (Polynesian, Austronesian) 71

254 Index of languages, linguistic families, and peoples
Hebrew (Semitic, Afroasiatic)
acquisition of 92–3
gender in 13, 52, 98, 100, 132, 194
gender reversal in 106, 109, 190
markedness in 26–7, 31–2, 114
masculine generic in 203–4, 210
Hindi-Urdu (Indo-Aryan, Indo-
European) 90
Hittite (Anatolian, Indo-European) 81
Hixkaryana (Carib) 15
Hmong (Hmong-Mien) 68
Hmong-Mien languages 68
Hunde (Bantu) 95
Hungarian (Ugric, Uralic) 30, 73, 122, 132,
177, 189, 191, 210, 217
Huron (Iroquoian) 115
Iatmul (Ndu, Papuan area) 44, 50, 101–3
Igbo (Atlantic-Congo) 61–2, 73, 206
Ignaciano (Arawak) 141–2, see also Mojo
Ilocano (Philippine, Austronesian) 127
Indo-Aryan languages (Indo-European) 15,
22, 83–90, 96, 187
Indo-European languages 6, 10, 13, 16, 24–6,
31, 70, 81, 84, 95–6, 114, 195–6
Indonesian (Western Austronesian,
Austronesian) 30, 64, 71–5, 89, 97
Iñeri (Arawak) 142–3
Innu (Algonquian) 67, 74
Iranian languages (Indo-European) 82,
86–7, 96
Irish sign language 162
Iroquoian languages 32, 56, 107,
115–18, 191
Island Carib (Arawak) 142–44, 160, 162
Island Carib mixed language 142–3
Italian (Romance, Indo-European) 13, 51,
61, 84, 93, 131, 204, 207, 210
Italian dialects 96
Jacaltec (Mayan) 64–6, 68, 74, 95
Japanese (isolate) 67, 74, 131
female and male speech in 8, 137, 148–51,
155, 160–4, 183, 208, 214
Jarawara (Arawá) 4, 27–8, 32, 53, 104, 109,
114–16, 186, 197
Jê languages 15, 18
Jingulu (Australian area) 21
Jukun (Jukunoid, East Benue-Congo) 169,
173, 183
Kaingang (Jê) 15, 18
Kala Lagaw Ya (Australian area) 32
Kalapuyan languages 88
Kalmyk (Tungusic) 129
Kaluli (Kaluli-Kasua, Bosavi) 130, 133,
158, 168
Kambaata (Cushitic, Afroasiatic) 181, 183
Kana (Cross-River, Benue Congo) 86
Kannada (Dravidian) 90
Kansa (Siouan) 162
Kapriman, see Sare
Karajá (Macro-Jê) 138, 144, 147, 160, 214
Katcha (Kadugli-Krongo) 22, 40, 51
Kayardild (Australian area) 69
Kaytej (Australian area) 174
Kele (Bantu) 173
Kera (Chadic, Afroasiatic) 176, 184
Keresan (isolate)163, see also Acoma Keresan
Ket (Yeniseian) 56, 61, 128, 185–6
Kewa (Engan, Papuan area) 74
Kilivila (Oceanic, Austronesian) 68, 75, 78
Kiowa (Kiowa-Tanoan) 82, 165
Kiowa-Tanoan languages 82
Koasati (Muskogean) 8, 12, 69, 139–44,
147, 162
Kokama-Kokamilla (Tupí-Guaraní,
Tupí) 139–40, 145, 160, 163
Kokota (Oceanic, Austronesian) 77
Kolami (Dravidian) 18
Konkani (Indo-Aryan, Indo-European) 83
Korean (isolate) 64, 74, 78, 95
Kristang (Creole) 72
Kru languages 97
Ku Waru (Chimbu-Wahgi, Papuan area) 74
Kuna (Chibchan) 167–8, 182
Kupto (Chadic, Afroasiatic) 10
Kupwar village, languages of 90

Index of languages, linguistic families, and peoples 255
Kurdish (Iranian, Indo-European) 82
Kũr̩ux (Dravidian) 141–2, 146, 162
Kushi (Chadic, Afroasiatic) 10
Kwa languages 69, 70
Kwami (southern Bole-Tangale, Chadic,
Afroasiatic) 10
Kwoma (Kwoma-Nukuma, Papuan area) 44,
50, 91, 206
Labwor (Nilotic) 91
Lak (North-east Caucasian) 56, 118,
191–3, 209
Lakhota (Siouan) 136–9, 145–7, 159, 162–5
Lao (Tai-Kadai) 64–5, 68, 74
Lardil (Australian area) 168–9
Latin (Italic, Indo-European) 13, 51, 84, 89, 95
Latvian (Baltic, Indo-European) 84
Lezgian (North-east Caucasian) 70
Libido (Cushitic, Afroasiatic) 183
Língua Geral (Tupí-Guaraní, Tupí) 91–2,
139, 162, 189
Livonian (Balto-Finnic, Uralic) 87
Lokele (Bantu) 173, 182
Lokono Dian (Arawak) 46–7, 104, 108–9, 117
Lower Sepik languages (Papuan area) 23
Lower Chinook (Chinookan) 71
Luwo (Nilotic) 91
Maa (Ubangi, Benue-Congo) 18
Maale (Omotic, Afroasiatic) 38, 50
Maasai (Nilotic) 51
Macedonian (Slavic, Indo-European) 70, 75
Machiguenga (Arawak) 102, 108
Macro-Jê languages 138, 183
Maithili (Indo-Aryan, Indo-European) 15, 17
Malagasy (Western Austronesian,
Austronesian) 153–4, 166, 178
Mali (Baining, Papuan area) 41, 44
Malto (Dravidian) 18, 68
Mam (Mayan) 66, 74, 78, 91
Manambu (Ndu, Papuan area)
cultural associations of gender in 29,
47–50, 113–14, 117, 186
female and male speech in 146–57
gender and physical properties in 33–7,
44, 48, 50
gender in 16, 27, 53, 91–3, 99–104, 213
markedness in gender in 27–8
names in 206
Mandan (Siouan) 162
Marathi (Indo-Aryan, Indo-European) 90,
97, 108–9, 117
Maroon Creole (Creole) 72, 186
Marra (Australian area) 69
Maung (Australian area) 54–5
Mawayana (Arawak) 87
Mayali dialect chain (Australian area) 90
Mayan languages 13, 64–6, 78, 152–4
Mayan-speaking community of
Tenejapa 152–4
Mayrinax dialect of Atayal (Formosan,
Austronesian) 162
Mescalero Apache (Athabaskan) 67
Mexican Spanish (Romance, Indo-European)
131, see also Spanish
Mexicano (Uto-Aztecan) 178, 184
Middle English 86–8, 89, 94, 97, 188, 195,
210, see also English
Middle Irish (Celtic, Indo-European) 84
Mixtecan languages 71, 79, see also
Coatzoquitengo Mixtec
Modern Khwe (Central Khoisan) 40, 44, 51,
56, 61
Mohawk (Iroquioan) 56, 107, 115,
191, 209
Mojo (Arawak) 141–2, see also Ignaciano
Moldavian (Romance, Indo-European) 162
Mongolic languages 162
Moroccan Arabic (Semitic, Afroasiatic) 32,
214, see also Arabic
Motuna (Bougainville, Papuan area) 57–8
Mupun (Chadic, Afroasiatic) 79, 95
Murrinhpatha (Australian area) 32, 66, 74
Muskogean languages 8, 69, 139, 144, 173
Nahuatl (Uto-Aztecan) 184, see also Mexicano
Nama (Khoekhoe) 61
Navajo (Athapaskan) 131

256 Index of languages, linguistic families, and peoples
Ndu languages (Papuan area) 16, 33, 50, 91
Ndyuka, see Maroon Creole
Negerhollands Creole Dutch (Creole) 72
Ngala (Ubangi) 78
Ngandi (Australian area) 69
Ngan.gityemerri (Australian area) 79–81
Nguni (Bantu) 183
Nheêngatú, see Língua Geral
Nilotic languages 40, 44, 58, 77, 91, 95–7
Nootka (Wakashan) 71
Norman French (Romance,
Indo-European) 97
North America, languages of 4, 8, 13, 18, 27,
64, 118, 136, 162, 165
North-east Caucasian languages 18, 53, 56,
70, 191–3
Northern Mao (Omotic, Afroasiatic) 87
Norwegian (Germanic, Indo-European) 93,
96–8, 200, 204, 210
Nubi (Creole) 72, 75
Nunggubuyu (Australian area) 58, 62
Nynorsk 93, 210, see Norwegian
Oceanic languages (Austronesian) 68, 77–8, 97
Oirat (Mongolic) 162
Ojibwe (Algonquian) 55
Old English (Germanic, Indo-European) 6,
15, 85–6, 90, 95–6, 188, 195, 216
Old French (Romance, Indo-European) 84, 90
Old Irish (Celtic, Indo-European) 84
Old Norse (Germanic, Indo-European) 93, 97
Old Persian (Iranian, Indo-European) 82
Ollari (Dravidian) 18
Omaha-Ponca (Siouan) 162
Omani Arabic (Semitic, Afroasiatic) 165,
see also Arabic
Omotic languages (Afroasiatic) 38–9, 42–6,
49, 87, 129, 186
Oneida (Iroquoian) 56, 107
Onondaga (Iroquioan) 107, 115
Oriya (Indo-Aryan, Indo-European) 84, 96, 187
Oromo (Cushitic, Afroasiatic) 39, 45–7, 105,
118, 128–9, 183
Osage (Siouan) 145, 159
Palestinian Arabic (Semitic, Afroasiatic) 27,
31, 100, 108, 117, 210, see also Arabic
Palikur (Arawak) 18, 23, 31, 45–7, 67–9,
74, 121
Papago (Uto-Aztecan) 2, 68, 155, 164,
see also Tohono O’odham
Papiamentu (Creole) 72
Papua New Guinea, languages of 23, 71–2,
75, 81, 133, 135
Papuan languages 15, 23, 38, 40, 44, 56–7,
67, 74, 93, 128–30, see also Papua New
Guinea, languages of
Parji (Dravidian) 18
Parthian (Iranian, Indo-European) 82
Persian (Iranian, Indo-European) 82
Pidgin languages 71, 133
Pintupi (Australian area) 183
Piya (Chadic, Afroasiatic) 10
Plains Cree (Algonquian) 31
Polish (Slavic, Indo-European) 70, 82, 92–3,
96–8, 117–18, 122
Portuguese 13–15, 23–6, 45, 52, 59, 84, 113,
118–19, 122, 201, 204–5, 210–13
female and male speech patterns 155,
158–9, 164, 214
loans from 138, 171
Proto-Arawak 31
Proto-Bantu 21
Proto-Germanic 85
Proto-Indo-European 81, 83
Proto-Mixtec 79
Proto-Uralic 69
Punjabi (Indo-Aryan, Indo-European) 22,
87, 97
Quapaw (Siouan) 162
Romance languages (Indo-European) 17, 31,
83–4, 90, 96, 179, 194
Romanian (Romance, Indo-European) 17,
31, 90, 96
Russian 13–16, 26, 31, 51, 54–6, 59, 111–13,
117, 121–5, 134–5, 191–3, 209
acquisition of gender in 92–3, 97–8

Index of languages, linguistic families, and peoples 257
animate declension in 83
endearment and gender 105, 109
female and male speech patterns 154–7, 164
gender, term for 11–12
Saami (Finno-Ugric, Uralic) 178
Salish languages 64
Sanskrit (Indo-Aryan, Indo-European) 72,
84–5, 89
Sardinian (Romance, Indo-European) 84
Sare (Sepik Hill, Papuan area) 36–7, 44,
97, 129
Savosavo (Papuan area) 38, 44, 56
Scottish Gaelic (Celtic, Indo-European) 91
Sele Fara dialect of Slovene (Slavic,
Indo-European) 84
Seneca (Iroquioian) 115
Sepik Hill languages (Papuan area)
35–6
Serbo-Croatian (Slavic, Indo-European) 84
Sesotho (Bantu) 92
Setswana (Bantu) 192
Sheko (Omotic) 38, 50
Sidaama (Cushitic, Afroasiatic) 173, 183
Sinhalese (Indo-Aryan, Indo-European)
84, 96
Siouan languages 18, 136, 141, 162
Slavic languages (Indo-European) 70, 82–4,
89–90, 96, 205
Slovene (Slavic, Indo-European) 84
South Dravidian (Dravidian) 83
Southern Sotho (Bantu) 183
Spanish (Romance, Indo-European) 46–7,
51, 61, 84, 98, 113, 126–7, 131–4, 175,
178, 205, see also Cantabrian Spanish,
Mexican Spanish
differential object marking in 70, 82, 97
loans from 133, 211
markedness in gender 25–6, 200–4
sexist language in 113–14, 194, 200
Suri (Surmic) 206
Surmic languages 87
Swedish (Germanic, Indo-European) 93,
96–8, 200
Taiap (Papuan area) 135, 177–8, 184
Tai-Kadai languages 64, 141, 192
Tamazight (Berber, Afroasiatic) 174,
177–8, 184
Tamil (Dravidian) 13, 18, 96
Tangale (Chadic, Afroasiatic) 87, 97
Tariana (Arawak) 105, 109, 128–9, 170–3,
179–86, 189–90, 209, 215
Tat (Iranian, Indo-European) 82
Talysh (Iranian, Indo-European) 82
Telugu (Dravidian) 31
Tenejapa, see Mayan-speaking community
of Tenejapa
Thai (Tai-Kadai) 64–5
female and male spech 8, 137, 148, 151,
159–60, 172, 208, 214
Tiwi (Australian area) 41, 44, 97
Tocharian A (Tocharian, Indo-European)
31, 96
Tohono O’odham (Uto-Aztecan) 2, 155,
164, see also Papago
Tok Pisin (Creole) 41, 72, 177–8
Tongan (Polynesian, Austronesian) 165
Toposa (Nilotic) 77
Tualatin (Kalapuyan) 88
Tucano (East-Tucanoan, Tucanoan) 179–80
Tucanoan languages 105
Tungusic languages 71
Tunica (isolate) 18, 20, 120, 141–2, 160–2,
197, 212
Tupí languages 71
Tupí-Guaraní languages (Tupí) 91, 139, 150,
162, 178, 183
Turkana (Nilotic) 40, 51, 58, 62
Turkic languages 71
Turkish (Turkic) 5, 71–5, 79, 87–9, 95, 97,
189, 217
Uralic languages 69, 71–2, 75
Vietnamese (Austroasiatic) 74
Wagaya (Australian area) 32
Wangkumara (Australian area) 32

258 Index of languages, linguistic families, and peoples
Wära (isolate, Papuan area) 37, 44, 50
Warao (isolate, Venezuela) 167
Wardaman (Australian area) 58, 62
Waris (Papuan area) 51
Warlbiri (Australian area) 169, 174, 183
Warndarran (Australian area) 81
Warray (Australian area) 58, 62
Warumungo (Australian area) 174
Waujá (Arawak) 183
Welsh (Celtic, Indo-European) 84, 198
West Chadic (Afroasiatic) 79
Wolaitta (Omotic, Afroasiatic) 79
Wolof (West-Atlanic) 38, 50
Wosera (Ndu, Papuan area) 44, 50
Yalaku (Ndu, Papuan area) 44, 50
Yana (isolate) 139–41, 144, 162, 147, 175
Yangoru Boiken, see Boiken
Yankunytjatjara (Australian area)
66, 74
Yanomami (Yanomami) 169
Yanyuwa (Australian area) 91, 136, 139–41,
145–6, 160, 163
Yazgulami (Iranian, Indo-European) 86
Yeniseian languages 128, 185
Yidiñ (Australian area) 65, 69, 79
Yimas (Lower Sepik, Papuan area) 23, 31
Yolngu (Australian area) 183
Young People’s Dyirbal (Australian area) 91,
97, see also Dyirbal
Zande (Ubangi) 18, 77–8
Zazaki (Iranian, Indo-European) 82
Zulu (Bantu) 183
!Xun (Central Khoisan) 95

Index of authors
Aalto, Pentti 162
Abbink, Jon 206
Abbott, Clifford 61
Abbou, Julie 210
Abondolo, Daniel 97
Abueg, Joselito 209
Adams, K. L. 74
Adelaar, Willem F. H. 97
Ahland, Colleen Anne 97
Aikhenvald, Yuri A. 111
Aikio, Marjut 184
Akhmatova, Anna A. 111, 193
Alexeyev, M. 75
Allan, Keith 209
Alpher, Barry 32, 112, 134
Alvanoudi, Angeliki 11, 135
Amadiume, Ifi 61–2, 73, 206
Ameka, Felix K. 75
Amha, Azeb 49–50, 97
Amiram, Orit 93
Anderson, L-G. 209
Armon-Lotem, Sharon 93
Asher, R. E. 30
Attaviriyanupap, Korakoch 163
Audring, Jenny 61
Avery, Jack 164
Awbery, Gwenllian 210
Axenov, Serge 95
Baker, Paul 130, 208
Bani, Ephraim 31
Barnhart, Robert K. 209
Baron, Dennis 11, 27, 111–12, 118, 187,
196, 209
Baron, Naomi S. 96–7
Barrett, R. 11, 164–5
Bateson, Gregory 3, 11, 101
Baudelaire, Charles 135
Beachy, Marvin Dean 129
Beauvoir, Simone de 2
Becker, A. L. 74
Bell, D. 183
Benamara, Hassane 117
Bengoechea, Mercedes 202
Benjamin, Carmen 51
Bergen, John J. 51
Berman, Ruth A. 97
Besnier, Niko 165
Bhatia, Tej K. 31, 97
Birtalan, Ágnes 162
Bisang, W. 75
Bloomfield, Leonard 12
Bloomfield, Morton W. 126
Boas, Franz 4, 138
Bodine, Ann 162
Bogoras, Waldemar 139, 162–3
Bolinger, Dwight 10
Bonfante, G. 51
Bonvillain, Nancy 61, 107
Borba, Rodrigo 165
Boroditzky, Lera 126–7, 131–2
Bossong, Georg 75
Bowden, Ross 50
Bradac, J. J. 156
Bradley, John 145–6, 163
Brauer, Marcus 119
Braun, Friederike 11, 16, 73–5, 95–7, 113,
189, 202, 211
Breen, G. 32
Breton, Raymond 162
Briggs, C. L. 182
Brightman, Robert 162
Brittain, David 183
Brosman, P. W. Jr. 95
Brown, Penelope 153, 164, 191
Brown, R. 51

260 Index of authors
Bruce, Les 50, 62, 120
Brunet, Jacqueline 211
Bull, Tove 210
Bullock, Alan 11
Bulygina, T. V. 31
Burridge, Kate 209
Butt, John 51
Caldas-Coulthard, Carmen Rosa 130
Camden, W. G. 162
Cameron, Deborah 154, 164
Campbell, Élisabeth 203
Campbell-Kibler, K. 164
Capell, A. 54
Caputo, B. Britt 196
Carrington, J. F. 182–3
Castellino, G. R. 50
Chafe, Wallace L. 52, 96, 115–19
Cheshire, J. 11, 188, 194, 198–9, 208–9, 216
Chirasombutti, Voravudhi 163
Christ, S. 154
Chung, Sandra L. 197
Clackson, James 95–6
Clamons, Cynthia Robb 39, 48, 118
Claudi, U. 95
Coates, Jennifer 11, 119, 156, 164, 183
Cobbinah, Alexander 31
Collinder, Björn 75
Comrie, Bernard 75
Connors, Kathleen 118
Contini-Morava, Ellen 61
Cooper, Fiona 118
Corbett, Greville G. 11, 30–1, 61, 75, 96–7
Cowley, Roger 45, 51
Craig, C. G. 74, 95
Craik, Brian 31
Creed, Gerald W. 11
Crolley, Liz 207, 209–10
Crouch, Isabel 164
Curzan, Anne 31, 85–6, 96–7, 195, 209
Dahlstrom, Amy 31, 61
Daley, Karen Ann 74
Darnell, Regna 31
Dawkins, Richard M. 97
de Courtenay, Baudouin 122
De León, Maria de Lourdes Pasquel 95
del Castillo, J. 126
Demuth, K. A. 97
Denny, J. P. 31
Deutscher, Guy 62, 124, 132–5
Diakonoff, I. M. 50
Diller, Anthony 74, 163
Dimmendaal, Gerrit J. 30, 51, 58, 62
Dinslage, Sabine 61, 183, 208
Dixon, R. M. W. 30–1, 59, 120, 135, 169
on Australian languages 30, 75, 90–1, 95,
169, 182
on Dyirbal 22, 29, 31, 54, 69, 97, 117, 121
on English 97, 118, 188, 197
on Jarawara 27, 32, 104, 114–15, 197
Doleschal, Ursula 117
Dorian, Nancy C. 91, 97
Drapeau, Lynn 74
Drude, Sebastian 144, 162
Du Puis, Mathias 162
Dubois, Betty 164
Dul’son, A. P. 61, 135
Dunn, Michael 162
Eckert, Penelope 119, 183–4
Edelman, D. I. 86, 97
Edenmyr, Niklas 51
Einaudi, Paula 162
Ekka, Francis 146, 162
Emeneau, Murray 97
Endruschat, Annette 210
Enfield, N. J. 74
England, N. 74, 95
Ernst, Amélie 124
Ervin, Susan M. 51
Escure, Geneviève 183
Evans, Bergen 209
Evans, Cornelia 209
Faraclas, N. 97
Faraoni, Vincenzo 96
Faust, Norma 162–3

Index of authors 261
Ferguson, Charles A. 117
Ferrari, F. 51, 61
Fesenko, Andrej 209
Fesenko, Tatiana 209
Flannery, Regina 137, 140, 162–3
Fleming, Luke 182–3
Fletcher, Alice 159
Fodor, I. 95, 125
Foley, W. A. 31
Fortune, David 147, 162
Fortune, Gretchen 147, 162
Fox, Helen 174, 183
Frajzyngier, Zygmunt 95
Franklin, K. 74
Friedman, Victor 75
Gal, Susan 163–4, 184
Gardani, Francesco 96
Garibian, A. S. 96
Gauchat, L. 175, 183
Gaudio, R. 164
Georg, Stefan 128
Gerdts, Donna 74
Glazkov, Nikolaj 134
Goddard, C. 74
Goddard, Ives 31, 55, 61, 118, 197, 199,
201–3, 207, 216
Goettner-Abendroth, Heide 209
Gogol, Nikolaj 134
Gordon, Matthew 164, 183
Grandi, Nicola 51
Green, Diana 31, 51, 71
Green, Ian 95
Greenbaum, Sidney 210
Greenberg, J. H. 30, 81, 97
Gregersen, Edgar A. 93, 98, 209
Grünberg, A. L. 97
Gumperz, J. J. 90, 97
Günthner, Susanne 162–3, 182
Gwynn, John Peter Lucius 31
Haas, Mary 31, 134, 139, 144–5, 162–3,
173, 183
Hachimi, Atiqa 210
Hagège, C. 98, 164
Hagman, Roy S. 61
Haig, Geoffrey 16, 60
Hale, Kenneth L. 169, 182
Hall, Kira 11, 165
Hamp, Eric 97, 230
Hampares, Katherine 202
Harbert, Wayne 96–7
Harrison, Simon J. 48, 117, 211
Harvey, Mark 31, 62
Haspelmath, M. 75
Hasselblatt, Cornelius 75, 97, 190,
209–10
Haugen, Einar 97, 197
Hauser-Schäublin, Brigitta 101, 104
Haviland, William A. 211
Healey, B. 159
Heath, Jeffrey 62, 164, 173, 183
Hegardty, Peter 209
Heine, Bernd 30–1, 51, 62, 75, 95, 97
Heine, Heinrich 123–4, 134, 213
Hellenthal, Anne-Christie 50
Hellinger, Marlis 119, 210
Henley, Nancy M. 189, 209
Henson, Eithne 183
Herbert, R. K. 98, 183
Herdt, Gilbert 11
Hill, Jane H. 2, 164, 182, 184
Hill, Kenneth 184
Hinch, H. E. 54
Hinkson, Mercedes Q. 74
Hoben, Susan J. 51, 104–5, 117
Hodge, Carleton T. 75
Hoff, Berend 162
Hoffman, Katherine E. 184
Holland, Audrey 98
Hollow, Robert 162
Holm, John 75
Holmes, Janet 11, 156, 164
Holmquist, Jonathan C. 46, 51, 61,
175, 178
Hornscheidt, Antje 210
Houseman, Michael 117
Huntley, David 96

262 Index of authors
Ibrahim, Muhammad Hasan 96
Ide, Sachiko 148, 163
Ikoro, S. M. 74
Inoue, Miyako 163
Irvine, Judith T. 164
Isačenko, A. V. 84
Jacobs, Andrea Michele 203–4
Jaffe, Alexandra 179
Jakobson, Roman O. 31, 59, 118, 121–3,
132, 193
James, Deborah 164, 183
Janda, Laura A. 95
Janse, Mark 97
Janson, T. 75, 209
Jarkey, Nerida 75, 149, 163
Jasanoff, Jay 31, 96, 197
Jespersen, Otto 2, 6, 155, 175, 183, 196
Jones, Kathryn 210
Jones, Lucy 164
Joseph, Brian D. 23
Jospin, Lionel 201
Jungraithmayr, H. 97
Kahane, Henry 51
Kahane, René 51
Karatsareas, Petros 97
Karbelashvili, D. P. 183
Karmiloff-Smith, A. 97
Kastovsky, Dieter 97
Kazakevich, Olga A. 138
Keenan, Elinor 184
Kendon, Adam 174, 183
Key, Mary Ritchie 162, 164, 209
Khaidakov, S. M. 56, 118, 209
Kibrik, A. E. 31
Kilarski, Marcin 30–1, 61
Kilian-Hatz, Christa 51, 61
Kimball, Geoffrey D. 144, 163
Kirton, Jean F. 137, 162
Kitajgorodksaja, Margarita V. 111, 117–18, 164
Klenin, E. 95
Klumpp, D. 95
Kochskämper, Birgit 210
Köhler, O. 51
Konishi, T. 126
Kooyers, O. 50
Köpcke, Klaus-Michael 24–5, 52, 62, 123
Kossmann, Maarten 50–1, 117
Kramarae, Cheris 11, 111, 210
Krishnamurti, Bhadriraju 31, 96
Kroskrity, Paul V. 162
Kryk-Kastovsky, Barbara 96
Kulick, Don 135, 164–5, 184
Kuntjara, Esther 73
Kuteva, Tania 95
Kwee, John B. 75
La Fleche, Frances 159
Labov, William 2, 154, 164, 183
Lakoff, George 31
Lakoff, Robin Tolmach 9, 156, 161, 164, 193,
198, 207–8
Lambert-Brétière, Renée 74
Landor, Roland 132, 135
Landry, Michaël 119
Lang, Adrienne 51, 74
Lang, Sabine 165
Laycock, D. C. 117
Lazarus, Emma 124–5
Leap, William 11
Lee, J. 97
Lee, Yun-seok 74, 95
Leger, Rudolf 50, 61, 183, 208
Lehmann, Thomas 30
LeMaster, Barbara 162
Lermontov, Mikhail 124
Levon, Erez 164
Levy, Y. 97
Lewis, Geoffrey 97
Li, Paul Jen-kuei 162
Liss, Julie 164, 220
Livia, Anna 160–1, 164, 217
Lock, Arnold (Arjen) Hugo 50
Lopatin, V. V. 209
Loporcaro, Michele 96
Lu, Tian-Qiao 65, 74
Luraghi, Silvia 95–6

Index of authors 263
Luthin, Herbert 147
Lydall, Jean 42, 49–51, 186
McAlinden, Fiona 164
Macaulay, R. K. S. 155
McConnell-Ginet, Sally 11, 113, 119, 122
McEwan, Ian 199
McGloin, Naomi 163
McGregor, Aileen R. F. 50
McGregor, Donald E. 50
McWhorter, John 135
Mahapatra, B. P. 69
Maiden, Martin 211
Maikov, Apollon 125
Mallinson, Graham 31
Mann, Heinrich 16
Marcato, Gianna 210
Marchese, L. 97
Maring, J. 163
Martel, Brigitte 3, 108, 159
Martin, Samuel E. 74, 95
Martynyuk, A. 209
Masica, Colin 84, 96
Matasovic, Ranko 31, 75, 96–7
Mathiot, Madeleine 110, 118
Matthews, P. H. 61
Matthews, W. K. 97
Mänd, Heljo 134
Meillet, Antoine 6, 95
Melchert, H. C. 95
Mendes, Ronald Beline 51, 164
Menn, Lise 98
Merlan, F. 62, 74
Migge, Bettina 72, 209
Migliazza, Ernest C. 182
Milani, Tommaso N. 220
Miller, Casey 188
Milles, Karin 210
Mills, A. E. 92
Mills, Sara 11, 119, 152, 205, 209
Milroy, L. 183
Miranda, Rocky V. 96
Mithun, Marianne 31, 49, 56, 61, 96–7, 107,
115–18, 134, 162–3, 191, 209
Mladenova, Olga M. 118
Moñino, Yves 182
Moon, Rosamund 130
Moore, Henrietta L. 11
Morris, Delyth 210
Morris, Lori 118
Motschenbacher, Heiko 30, 220
Mous, Maarten 31
Mugglestone, Lynda 210
Mulac, A. 156
Mullen, M. K. 128–30, 135
Munro, Pamela 162
Myers, Fred R. 183
Nakamura, Momoko 163
Nanda, Serena 165
Nash, David 182
Nekitel, Otto 24, 31
Neruda, Pablo 135
Newman, Michael 31, 209
Newman, Paul 96
Nissen, Uwe Kjær 113, 134,
202–3, 210
Nuni de Chapi, Conchita 162
O’Connor, Michael P. 98
O’Donovan, Veronica 11, 165
O’Leary, Kathy 164
Ochs, Elinor 152, 164
Ogawa, Naoko 164
Okamoto, Shigeko 150–1, 161, 163
Okudzhava, Bulat 125
Olza Zubiri, Jesús 162
Onishi, M. 62, 96
Ortner, Sherry B. 3, 130
Osborne, C. R. 51
Osgood, Charles E. 131
Ostermann, Ana Christina 165
Ott, R. 162
Ott, W. 162
Oussikoum, Bennaser 117
Owens, Jonathan 75
Oxford English Dictionary Online 11, 97,
111, 188, 204, 209–10

264 Index of authors
Palmer, W. D. 95
Pandharipande, Rajeshwari V. 97
Pankhurst, Helen 51, 100, 104–5, 128
Pastre, Geneviève 164, 201
Pauwels, Anne 119, 205, 208–11
Pawley, Andrew K. 110
Payne, Judith 134
Pearce, Mary 184
Penelope, Julia 209
Pensalfini, R. 31
Pérez-Pereira, M. 97
Perlak, D. 98
Pet, W. J. A. 117
Peters, Pam 214
Pfarr, S. J. 149
Philippaki-Warburton, Irene 23
Philipps, Webb 126, 131
Pichler, Pia 11
Plank, F. 30
Podesva, R. 164
Pope, M. K. 84, 96
Popova, M. I. 97
Poser, William J. 74–5
Posner, Rebecca 96
Pountain, C. J. 51, 61
Priestly, Tom M. S. 51, 84, 96–7
Puşcariu, Sextil 90
Quinn, Conor 31
Quintero, Carolyn 145
Quirk, Randolph 96
Raffelsiefen, Renate 135
Raga, Amanuel 135
Rankin, Robert L. 162
Rapold, Christian J. 50
Rat, Joseph Numa 162
Reh, Mechthild 75, 95
Reid, Nicholas 95
Repp, Hanna 117
Reynolds, Katsue Akiba 149–51, 163
Ribeiro, Eduardo Rivail 122
Richards, Cara B. 115
Ringelnatz, Joachim 134
Roberts, S. 164
Roberts, S. P. 62, 74
Robustelli, Cecilia 211
Rodina, Yulia 92, 97
Rodrigues, Aryon Dall’Igna 162
Romaine, Suzanne 11
Romani Miranda, Maggie M. 134
Roscoe, Paul B. 128
Roscoe, Will 165
Rose, Françoise 162
Rosenhouse, J. 117
Ross, Kristiina 72, 75
Rothstein, Robert A. 31, 117–18, 192–3, 209
Rozanova, Nona N. 111, 117–18, 164
Rubino, C. R. G. 127, 135
Rudes, B. 159
Rumsey, A. 62, 74
Rushforth, S. 74
Sa’ar, Amalia 31, 117, 210
Sabajo, M. A. 46
Sadiqi, F. 32, 164
Sahoo, Kalyanamalini 209
Salzmann, Zdenek 208
Sapir, Edward 139, 147, 162, 175, 185
Sarvasy, Hannah 40
Sasse, Hans-Jürgen 50–1, 97
Saville-Troike, M. 12
Sayers, Dorothy L. 61
Särsa, Riisto 50
Schane, S. A. 31
Schapper, Antoinette 30, 61, 97
Schaub, W. 31
Schellinger, W. 30
Schenker, Alexander M. 75
Schieffelin, Bambi B. 130, 158, 168
Schieffelin, Edward L. 130
Schipper, Mineke 103, 129, 135, 190, 216
Schmid, Sonja 117
Schmidt, A. 97
Schmidt, Lauren 126, 131
Sczesny, Sabine 113
Senft, Gunter 75, 95
Sera, M. 126

Index of authors 265
Severi, Carlo 117
Seyoum, Mulugeta 50
Shepard Jr., G. 102
Sherzer, Joel 102, 158, 167, 181–2
Shibatani, Masayoshi 148–9, 163
Shmelev, A. D. 31
Shvedova, Natalja Y. 193
Silverman, Eric K. 11
Silverstein, Michael 11–12, 75, 197, 188, 204
Sims, C. 163, 182
Singer, Ruth 55
Sini, M. 211
Slobin, Dan I. 133
Smith, Janet S. (Shibamoto) 164
Smith-Stark, S. 75
Sneddon, James N. 64, 74
Sontag, Susan 113
Southerland, Ronald H. 138
Speece, R. 51
Spender, Dale 211
Spitulnik, D. 31
Stahlberg, Dagmar 113
Stallybrass, Oliver 11
Stebbins, Tonya N. 51
Steinhauer, H. 30
Storch, Anne 61, 97, 162, 169, 173–4, 182–3,
208, 215
Storjohann, Petra 199
Stotko, Elaine M. 209
Stradelli, Ermano 162
Straus, Anne Terry 20, 31
Stroud, C. 184
Studzińska, Joanna 122
Sumbuk, Kenneth Memson 37, 50, 129
Surikov, Ivan 134
Svartengren, T. Hilding 118
Swan, Toril 210
Swift, Kate 188, 209–11
Tadmor, Uri 97
Taïfi, Miloud 117
Talbot, M. 11, 208
Tannen, Deborah 164
Taylor, Allan R. 12, 145–7, 162
Taylor, Douglas M. 162
Teferra, Anbessa 183
Teso, Elena 207–10
Thompson, James 123
Thompson, L. C. 74
Thompson, S. A. 156
Thüne, Eva-Maria 210
Timberlake, Alan H. 197
Tjutschev, Fjodor 124
Tobin, Yishai 27, 31, 106, 190
Trask, L. 95
Trechter, Sara 11, 136–7, 145–7,
162–5
Treichler, Paula A. 11, 111
Treis, Yvonne 183
Trombley, Stephen 11
Troyer, Margaret 209
Trudgill, Peter 97, 149, 155, 176, 183
Tsegaye, Mulugeta T. 31
Tube, Juan 162
Underhill, Robert 197
Upton, Rebecca 11
Uyeno, T. 163
Vajda, Edward J. 128
Valentine, J. Randolph 208
Valiquette, H. 163, 182
Vallejos, Rosa Yopán 140, 163
van Baarle, Peter 46
van den Berg, René 95–7
Vanek, Anthony L. 31
Vasvári, Louise O. 75, 189, 210
Villas Bôas, Cláudio 183
Villas Bôas, Orlando 183
Vinogradov, V. V. 113, 134, 209
Vossen, Rainer 95
Wales, Katie 31, 109, 118, 198
Walsh, M. 32, 74
Walters, Margaret 209
Watkins, Calvert 197

266 Index of authors
Watkins, L. 82
Weber, Doris 61
Wegener, Claudia 50, 61
Welmers, Wm. E. 16
Werner, Heinrich 135, 208
Wertz, C. A. 96
Westermann, D. 62
Whitehead, Harriet 3, 165
Wiedemann, Ferdinand Johann 75, 97
Wikan, Unni 165
Williams, Nancy M. 183
Wilson, R. 90, 97
Winford, Donald 97
Woidich, Manfred 117
Woldemariam, Hirut 135
Wolfart, H. Christoph 31
Wołk, Ewa 104–5, 117
Wrenn, C. L. 96
Yadav, Ramawatar 31
Yaguello, Marina 118, 210
Yigezu, Moges 163
Yokoyama, Olga T. 117–18, 156, 164, 209
Zavala, Roberto 95
Zemskaya, Elena A. 111, 117–18, 164
Zepeda, Ofelia 2, 164
Zgusta, Ladislav 62
Zhang, Sihong 74
Zimman, Lal 11
Zubin, David A. 24–5, 52, 62, 123
Zwicky, Arnold M. 164

Index of subjects
abstract noun 6, 18–19, 21, 23, 67, 81, 122, 191
acquisition of linguistic gender 7, 92–3,
97, 147
address terms and practices 56, 104, 106,
108, 141, 159, 163, 186, 204–9
adjective 14, 15–18
adverb 14, 71, 78
agglutinating language 16
agreement 14–18, 30–1, see also gender,
linguistic
alliterative 23, 31
anaphoric 14–16
hierarchy 30
loss 76–98
origin of 76–98
variability in 33–51, 53–7
agreement gender 53, 70–2, 76–82, 86–8,
91–4, 97, 113, 126
alliterative agreement, see agreement,
alliterative
anaphoric gender, see gender, anaphoric
animacy 5, 13–25, 29–30, 34–51, 55–6, 72–5,
82–7, 118, 121–35, 185–6, 191
and classifiers 63–9
and grammatical relations 69–71
and number 69–71, 82
and pronouns 69–71
animate declension 83–4, 96
aphasia 93–4
areal diffusion 87–92, see also language
contact
article 14–19
assignment of gender 18–25, see also gender
assignment
augmentative 22–3, 38, 51
avoidance style 9, 57, 169–75, 181–3, 202,
215, see also Hlonipha, Mother-in-law
style, taboo
baby talk 72, 108–9, 147, 164
beliefs 19–20, 120–1, 132–5, see also
myth-and-belief principle in linguistic
gender choice
Berdache 147, 165
bilingualism 131, 177–8, see also
multilingualism
borrowing, see language contact, loans
case 7, 63, 69–71, 82–3, 86, 92–6, see also
differential case marking
child language acquisition, see acquisition of
linguistic gender
classificatory verbs, see verbal classifier
classifier 63, 68
noun classifier 52, 63, 65, 68–9, 94
numeral classifier 52, 63–5, 68–9, 94,
190–2
possessive classifier 61, 68, see also
possession
relational classifier 61
verbal classifier 52, 67–9
cognition 8, 10, 32, 87–8, 126–32
collective 20–1, 51, 81
command 4, 136–7, 141, 152–4, 157, 164, 189
common gender 61, 93, 96, 98, 196
compound 25, 81, 88, 113, 187–8
convergence, see areal diffusion, language
contact
coordination 26–9, 194, 200–1
covert marking of linguistic gender 14–15,
58, 81, see also gender, linguistic
creolization 97
declension 13, 69–70, 83–6, 94, 96
default choice in linguistic gender 28,
114–15, 151, see also markedness,
functional

268 Index of subjects
deferential register 173–4, see also politeness
definiteness and gender 57–9
demonstrative 14–20, 27, 31–3, 54–6, 63,
67–8, 78, 81–5, 88, 95–6, 138–41, 182
deprecatory meanings of linguistic
gender 46, 102–5, 111–12, 193
derivation 22–3, see also derivational gender
derivational gender 5, 13–14, 76, 79–83,
88–90, 94, 104, 112, 117, 186, 192–3,
see also gender, linguistic
differential case marking 70, 83, see also case
diminutives
and linguistic gender 7, 15, 22–3, 38, 41,
45, 51, 60, 73, 78, 90–2, 117
in female speech 152–5, 159, 164, 214
in gay and lesbian speech 159
discourse role of linguistic gender 52, 57–9
dissolution of linguistic gender 7, 92–3
double duty words 188
edible gender 19–20, 57, 80–1, 159
elegant prefix 149, see also honorific
endearment 15, 21, 35, 44–7, 92, 100,
105–9, 117
epicene 53, 189, 196, 202, 207
evidential 35
experimental studies of gender 126–35
female social gender 1–9, see also gender, social
female speech 2, 8, 136–65, 214, see also male
speech
phonetic features of 2, 8, 138–41, 144,
154–5, 158, 161–3, 166
feminine as an unmarked linguistic
gender 27–9, 32
feminine gender 1–9, 13–29, 34–61, 63–75,
83–4, 91–135, 185–211, see also gender,
linguistic
feminine-indefinite in Iroquoian
languages 107, 191
feminization of job titles 112, 118, 120, 201,
207, 210
feminist language reform 93–4, 114, 134,
186, 203–8, 211, 216–17
feminist scholars 9, 59–60, 93, 111–13,
196–7, 200–1, 204–11, 216
formal markedness, see markedness, formal
functional markedness, see markedness,
functional
fusional language 16
gay men’s speech 3, 9, 11, 31–2, 80, 130, 146,
158–9, 161, 164–5, 200, 215
gender as term 1–14, 212–17
gender, biological, see gender, natural
gender, linguistic 1–9, 13–32, 214–17
anaphoric 7, 14–16, 30–1, 60, 72, 76–7,
81–8, 91–6, 124–6
and other categories 16–17
and social stereotypes 2, 44–58, 71–5,
109–19, 185–206
assignment of 19–25, 83–7, 185–6
covert marking of linguistic gender 14–15,
58, 81
derivational, see derivational gender
functions of 52–61
gender agreement 14–18, 30–1, see also
agreement gender
in Athabaskan linguistic tradition 30
loss of 83–8, 94–7, 126
meanings of 19–22, 52–61, see also
semantic opacity in linguistic gender,
semantic transparency in linguistic
gender
origin of 66, 76–91
variable assignment of 53–7
gender, natural 1–9, 185–217, see also sex
gender, social 1–9, 25, 29–32, 44–51, 71–5,
100–2, 137–84, see also gender
reversal
and linguistic gender 185–217
gender bias 194–208, see also male bias,
masculine bias, sexist language
gender etiquette 2, 180
gender reversal 99–118
gender stereotype 4–5, 44–52, 56, 65, 69–74,
109, 130, 164, 185–211, 214–17
gender switch, see gender reversal

Index of subjects 269
gender-exclusive languages 136–51, 161–2,
see also genderlect, gender-variable
languages
gender-inclusive language, see gender-neutral
language
gender-less languages 5, 7, 30, 60–3, 71–2,
87, 92, 123, 127–8
gender-neutral language 5, 61, 73, 188,
198, 217
gender-neutral pronoun 196–200
gender-variable languages 137, 152–62
genderlect 162–3, 208, 214, see also genderexclusive
languages, gender-variable
languages
generic he 6, 86, 195–9, 203–7
generic masculine 12–14, 27–9, 86, 89,
112–16, 188–204, 207–17, see also male
bias, masculine bias
generic noun 4–6, 15, 24–8, 58, 65–6, 72–81,
112–14, 123, see also generic
masculine
generic pronoun 4, 15, see also generic he,
generic masculine, pseudogeneric,
singular they
gesture 8, 140
grammaticalization 76–81, 88, 95
greetings 100, 140
hedges 156
hijra 3, 8, 159–60, 165
Hlonipha 183
honorific 4, 15, 148–51, 174, 204, 207
humanness 5, 13–14, 19–22, 29–30, 63–75,
90–1, 191–2
hybrid nouns 127
inanimate, see animacy
individuation as a function of gender 52,
58, 61
initiation 2–3, 47–8, 103–4, 166–9, 173,
181–2
insults 102–5
interrogatives 26–7, 47, 69, 81
isolating language 15
joking and gender 8, 54, 72–3, 100–2, 106–8,
128, 146, 161, 187, 197
kinship terms 53, 66, 68, 151
language change 76–98, 175–84
language contact 87–97, see also linguistic
area, loans
language endangerment 92, 139, 145, 166,
170, 179, 208, see also language
obsolescence
language loss, see language endangerment,
language obsolescence
language obsolescence 7, 47, 91–4, 137, 145,
189, 199, 202–4
language planning 4, 93–5, see also feminist
language reform, language reform
language reform 88, 93–5, 114, 134,
201–204, 217
lesbian language 3, 9, 11, 159–61, 164, 200, 215
lexicon 52–3
linguistic area 172, 182, see also language
contact
loans 12, 23–5, 28, 38, 41, 46, 56–7, 72,
88–90, 93–7, 138–9, 188–92, 196, 214
male bias 15, 73–4, 208, 217, see also
masculine bias, sexist language
male social gender 1–9, see also gender,
social
male speech 2, 8, 136–65, 214
markedness, formal 25–29, 31–2
markedness, functional 25–9, 31–2, 114–16,
see also generic he, generic noun, generic
pronoun
masculine bias 15, 73–4, 112–14, 127, 189,
194–208
masculine gender 1–9, 13–29, 34–61, 63–75,
83–4, 91–35, 185–211, see also gender,
linguistic
mass noun 35–6
matrilineal societies 115–16, 187, 209–11
metaphor 2, 5, 7–8, 60, 74, 88, 103, 121–5,
132–6, 167, 213

270 Index of subjects
mixed language, Arawak-Carib 142–4
Mother-in-law style 57, see also avoidance
style
multilingualism 143, 172, 180–1, 184,
see also bilingualism, language contact,
linguistic area
multiple classifier system 63, 68, 78
myth 19–20, 59, 120–1, 166–8
myth-and-belief principle in linguistic gender
choice 19–20, 91, 120–1
myths about gender in English 11
naming practices 137, 194, 204–6, 209
natural gender 29–30, 56–7, 128–30, 133–51,
see also sex
Naven 3, 101, 117, 220
neuter gender 5, 13–16, 22–5, 28–9, 52–3, 60,
72, 83–4, 91–3, 96, 106, 128, 141
nominal hierarchy 75
non-feminine gender, see gender, linguistic
non-rational gender 13, 18
non-verbal communication, see gesture
noun class 13, 16–17, see gender, linguistic
noun classifier, see classifier
noun compounding, see compound
number 15–18, 26–30, 51, 69–71, 82–3, 141
number agreement 16, 29–30
numeral classifier, see classifier
overt marking of linguistic gender 14–16, 54,
58, 81, see also covert marking of
linguistic gender
patrilineal societies 35, 101–2, 133, 179, 206
patronymic 205
pejorative meanings of linguistic
gender 45–51, 111, see also deprecatory
meanings of linguistic gender
personal name, see naming practices
personal pronoun 14–19, 26–30, 38, 71–8,
83–7, 92–6, 99, 102, 110, 115, 137–9,
141, 148, 195, 198, see also generic he,
generic pronoun, singular they
phonological assignment of linguistic
gender 22–5, 31
pitch 2, 8, 154, 158, 161, 166
poetry, gender in 5, 60, 109–10, 120–35
politeness 4, 56, 118, 137, 152–3, 156–7,
163–4, 181, 191, 207, 214, see also
honorific, respect in Japanese 148–51
possession 4, 27, 54, 61, 64, 68, 71, 101, 110,
135, 141, 173, 183, 196
power and gender 8–9, 21, 29, 48, 55, 61, 68,
105, 109, 114–19, 128, 131–3, 153,
156–60, 164, 173–6, 179–81, 204
pragmatic functions of linguistic
gender 57–61
proper name, see naming practices
proverbs, women in 73, 103, 179, 190, 216
pseudogeneric 196–8
queer linguistics 11, 200
questions, see interrogatives, tag questions
rational gender 13, 18
reanalysis 76–7, 81–2, 88–9, 94–5
reinterpretation 51, 63, 76–7, 80–5, 88, 95–6,
107
residual gender 19, 24–5, 30
respect 4, 27, 64–8, 78, 104–7, 116, 122, 137,
149, 153, 160, 166, 174–7, 191–2, 197,
207, see also honorific
ritual language 146, 166–9, 181–4, see also
initiation
Sapir–Whorf hypothesis 208
secret language 169–75, 181, 215, see also
avoidance style, taboo
semantic functions of linguistic
gender 18–22, 51–61
semantic opacity in linguistic gender 5–6,
18–25
semantic transparency in linguistic
gender 5–6, 18–25, 185–6
sex 3, 69–75, 109–18, see also natural gender
in classifier choice 63–9

Index of subjects 271
in linguistic gender choice 1, 13, 18–19,
25, 91–118
sexism 60, 112–14, 208, 216, see also male
bias, masculine bias, sexist language
sexist language 93–4, 100, 112, 114, 189,
194–205, see also male bias, masculine bias
shape in linguistic gender choice 13–14, 18,
34–51, 185–6
sign languages 162, 174, 183
singular they 196–8
size in linguistic gender choice 13–14, 18,
34–51, 97, 102–7, 128–9, 185–6
social change 191–211
societal hierarchies 4, see also honorific,
respect
socio-cultural change 9, 47–51
socio-cultural gender, see gender, social
solidarity and gender 8, 100, 104–9, 158–60,
175, 190, 199
songs 163, 166–8, 179, 181
speech register, see female speech, male speech
Sprachbund, see linguistic area
taboo 169–75, 215, see also
avoidance style
tag questions 156–8
target gender 31
topicality and linguistic gender 57–8
transgender 3, 11, 108, 158–61,
196, 200
transsexual 3, 7, 159–60, 165, 215
transvestite 3, 9, 101, 159–60, 165
two-spirit people 147, 165
value as a semantic parameter in linguistic
gender 7, 44–51, 56, 118, 128–9, 133,
166–7, 190–1, 213
warfare 143
word order 194
writing system 43
Yurupary ritual 105, 169–71, 181
zero marking, see markedness, formal