You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
<strong>How</strong> <strong>Gender</strong> <strong>Shapes</strong> <strong>the</strong> <strong>World</strong>
<strong>How</strong> <strong>Gender</strong> <strong>Shapes</strong><br />
<strong>the</strong> <strong>World</strong><br />
ALEXANDRA Y. AIKHENVALD<br />
Language and Culture Research Centre<br />
James Cook University<br />
1
3<br />
Great Clarendon Street, Oxford, OX2 6DP,<br />
United Kingdom<br />
Oxford University Press is a department of <strong>the</strong> University of Oxford.<br />
It fur<strong>the</strong>rs <strong>the</strong> University’s objective of excellence in research, scholarship,<br />
and education by publishing worldwide. Oxford is a registered trade mark of<br />
Oxford University Press in <strong>the</strong> UK and in certain o<strong>the</strong>r countries<br />
© Alexandra Y. Aikhenvald 2016<br />
The moral rights of <strong>the</strong> author have been asserted<br />
First Edition published in 2016<br />
Impression: 1<br />
All rights reserved. No part of this publication may be reproduced, stored in<br />
a retrieval system, or transmitted, in any form or by any means, without <strong>the</strong><br />
prior permission in writing of Oxford University Press, or as expressly permitted<br />
by law, by licence or under terms agreed with <strong>the</strong> appropriate reprographics<br />
rights organization. Enquiries concerning reproduction outside <strong>the</strong> scope of <strong>the</strong><br />
above should be sent to <strong>the</strong> Rights Department, Oxford University Press, at <strong>the</strong><br />
address above<br />
You must not circulate this work in any o<strong>the</strong>r form<br />
and you must impose this same condition on any acquirer<br />
Published in <strong>the</strong> United States of America by Oxford University Press<br />
198 Madison Avenue, New York, NY 10016, United States of America<br />
British Library Cataloguing in Publication Data<br />
Data available<br />
Library of Congress Control Number: 2016931923<br />
ISBN 978–0–19–872375–2<br />
Printed in Great Britain by<br />
Clays Ltd, St Ives plc<br />
Links to third party websites are provided by Oxford in good faith and<br />
for information only. Oxford disclaims any responsibility for <strong>the</strong> materials<br />
contained in any third party website referenced in this work.
In loving memory of remarkable and independent women in whose<br />
shadow I grew up—my great-aunt Frania S. Rosman who showed<br />
what a woman can achieve against all odds, my grandmo<strong>the</strong>r Maria<br />
S. Bonné who educated me concerning <strong>the</strong> nature of <strong>the</strong> status of<br />
women, and my great-grandmo<strong>the</strong>r Nina K. Aikhenvald whose<br />
indomitable spirit and strength inspired me.
Contents<br />
Acknowledgements xi<br />
Abbreviations and conventions xiii<br />
List of boxes, figures, schemes, and tables<br />
xv<br />
1. The multifaceted <strong>Gender</strong> 1<br />
1.1 Disentangling ‘<strong>Gender</strong>’ 1<br />
1.2 What is special about Linguistic <strong>Gender</strong> 4<br />
1.3 <strong>How</strong> this book is organized 6<br />
1.4 The empirical basis, and a note on conventions 9<br />
2. Linguistic <strong>Gender</strong> and its expression 13<br />
2.1 Linguistic <strong>Gender</strong> in a nutshell 13<br />
2.1.1 Marking Linguistic <strong>Gender</strong> 14<br />
2.1.2 <strong>Gender</strong> agreement and anaphoric gender 15<br />
2.1.3 Linguistic <strong>Gender</strong> and o<strong>the</strong>r linguistic categories 16<br />
2.1.4 <strong>How</strong> many genders? 17<br />
2.2 <strong>How</strong> to choose a Linguistic <strong>Gender</strong> 18<br />
2.2.1 <strong>Gender</strong> choice by meaning 18<br />
2.2.2 <strong>Gender</strong> choice by form 22<br />
2.2.3 Meaning meets form: mixed principles of gender assignment 23<br />
2.3 Markedness and Linguistic <strong>Gender</strong>s 25<br />
2.4 Linguistic <strong>Gender</strong>s and <strong>the</strong>ir labels: envoi 29<br />
3. Round women and long men: physical properties in Linguistic <strong>Gender</strong> 33<br />
3.1 Small round women and long slender men 33<br />
3.2 When ‘women’ are larger than ‘men’ 41<br />
3.3 Physical properties in Linguistic <strong>Gender</strong> choice: contrasting<br />
<strong>the</strong> two scenarios 43<br />
3.4 Beyond mere physique: attitude, value, and importance<br />
in Linguistic <strong>Gender</strong> 44<br />
3.4.1 Endearment and disdain through Linguistic <strong>Gender</strong> 45<br />
3.4.2 Value and importance in Linguistic <strong>Gender</strong> 47<br />
4. What are Linguistic <strong>Gender</strong>s good for? 52<br />
4.1 Variable choice of Linguistic <strong>Gender</strong> 53<br />
4.2 Linguistic <strong>Gender</strong> in discourse 57<br />
4.3 The utility of Linguistic <strong>Gender</strong> 59
viii<br />
Contents<br />
5. <strong>Gender</strong> meanings in grammar and lexicon 63<br />
5.1 Sex, humanness, and animacy in classifier systems 63<br />
5.1.1 Numeral classifiers 64<br />
5.1.2 Noun classifiers 65<br />
5.1.3 Verbal classifiers and o<strong>the</strong>r classifier types 67<br />
5.1.4 Linguistic <strong>Gender</strong>s and classifiers as noun categorization devices:<br />
commonalities and differences 68<br />
5.2 Sex, humanness, and animacy in noun categories 69<br />
5.3 <strong>Gender</strong> in gender-less languages 71<br />
6. The rise and fall of Linguistic <strong>Gender</strong>s 76<br />
6.1 Developing Linguistic <strong>Gender</strong> 76<br />
6.1.1 From generic nouns to Linguistic <strong>Gender</strong>s 77<br />
6.1.2 From generic nouns to noun classifiers and <strong>the</strong>n to<br />
Linguistic <strong>Gender</strong>s 79<br />
6.1.3 From anaphoric gender to agreement gender 81<br />
6.1.4 From derivational gender to agreement gender 81<br />
6.1.5 Linguistic <strong>Gender</strong> from o<strong>the</strong>r nominal categories 82<br />
6.2 Reshaping Linguistic <strong>Gender</strong>s 83<br />
6.3 Partial loss and reinterpretation of gender: <strong>the</strong> story of English 85<br />
6.4 Linguistic <strong>Gender</strong> in language contact 87<br />
6.4.1 To lose a Linguistic <strong>Gender</strong> 87<br />
6.4.2 Evolving a Linguistic <strong>Gender</strong> 88<br />
6.4.3 Readjusting Linguistic <strong>Gender</strong> 90<br />
6.4.4 Linguistic <strong>Gender</strong> in language obsolescence 91<br />
6.5 Linguistic <strong>Gender</strong> in language acquisition and language<br />
dissolution 92<br />
6.6 Linguistic <strong>Gender</strong> and language reforms 93<br />
6.7 Linguistic <strong>Gender</strong>s, <strong>the</strong>ir development, demise,<br />
and transformations 94<br />
7. Manly women and womanly men: <strong>the</strong> effects of gender reversal 99<br />
7.1 Reversing Linguistic <strong>Gender</strong>s with jocular effects 100<br />
7.2 Offence and praise in Linguistic <strong>Gender</strong> reversals 102<br />
7.3 Linguistic <strong>Gender</strong> reversal: endearment and solidarity 105<br />
7.4 Men as women, women as men: a summary of Linguistic<br />
<strong>Gender</strong> reversals 108<br />
7.5 Attitudes to Social <strong>Gender</strong>s through Linguistic <strong>Gender</strong> reversals 109<br />
7.6 ‘Women’ as a subtype of ‘men’? The overtones of masculine generics 112<br />
7.7 Markedness, status, and power in Linguistic <strong>Gender</strong> choice 114
Contents<br />
ix<br />
8. The images of gender 120<br />
8.1 Myth-and-belief in <strong>the</strong> choice of Linguistic <strong>Gender</strong> 120<br />
8.2 The metaphors of Linguistic <strong>Gender</strong> 121<br />
8.3 Does Linguistic <strong>Gender</strong> affect cognition? 126<br />
8.4 What men and women look like 127<br />
8.5 ‘<strong>Gender</strong>ing’ <strong>the</strong> world: images, metaphors, and cognition 132<br />
9. When women and men speak differently 136<br />
9.1 ‘Male’ and ‘female’ dialects 136<br />
9.1.1 <strong>How</strong> male and female dialects differ 138<br />
9.1.2 Speakers and addressees of male and female dialects 141<br />
9.1.3 Male and female dialects, and language history 142<br />
9.1.4 Male and female speech on <strong>the</strong> way out 144<br />
9.1.5 Beyond Natural <strong>Gender</strong> 145<br />
9.1.6 Politeness, identity, and change: male and female dialects<br />
in Japanese and Thai 148<br />
9.2 <strong>Gender</strong>-variable skills: men’s and women’s speech practices 152<br />
9.2.1 Social status, and women’s speech 152<br />
9.2.2 Tokens of men and women in gender-variable languages 154<br />
9.3 The o<strong>the</strong>r genders 158<br />
9.4 Male speech, female speech: a summary 160<br />
10. The rituals of gender 166<br />
10.1 Social <strong>Gender</strong>, speech genres, and speech practices 166<br />
10.2 The languages of manhood 168<br />
10.3 Secrets, avoidance, and taboos: what women are not supposed<br />
to know 169<br />
10.4 Men, women, and language change 175<br />
10.5 Language keepers or language killers? Women and language<br />
maintenance 177<br />
10.5.1 Women as language keepers 177<br />
10.5.2 Women leading language shift 177<br />
10.5.3 Women and modernity 178<br />
10.5.4 Women as language killers 179<br />
10.6 Summary: Social <strong>Gender</strong> through rituals, genres,<br />
and speech practices 181<br />
11. <strong>Gender</strong> in grammar and society 185<br />
11.1 What Linguistic <strong>Gender</strong> can tell us about Social <strong>Gender</strong> 185<br />
11.2 Social inequalities through gender asymmetries 186<br />
11.3 The value of ‘man’ through gender in lexicon 190<br />
11.4 <strong>How</strong> Linguistic <strong>Gender</strong>s reflect social change 191
x<br />
Contents<br />
11.5 Thwarting ‘sexist language’ 194<br />
11.5.1 Masculine bias through Linguistic <strong>Gender</strong>: pronouns<br />
and agreement 195<br />
11.5.2 Fighting <strong>the</strong> ‘generic masculine’ throughout <strong>the</strong> language 201<br />
11.5.3 Bias in address terms and naming patterns 204<br />
11.6 Expression of Linguistic <strong>Gender</strong> and social change: a summary 206<br />
12. The heart of <strong>the</strong> matter: envoi 212<br />
References 218<br />
Index of languages, linguistic families, and peoples 251<br />
Index of authors 259<br />
Index of subjects 267
Acknowledgements<br />
I have been working and publishing on gender for over two decades now, and am<br />
indebted to many people, of different continents, backgrounds, and walks of life. My<br />
gratitude goes to native speakers of Amazonian languages who taught me <strong>the</strong>ir<br />
remarkable languages, especially my Tariana family—José, Jovino, Olívia, Rafael,<br />
Leo, Maria, Diká, Emílio, Juvenal, <strong>the</strong> late Graciliano, Ismael, and Cândido Brito,<br />
toge<strong>the</strong>r with <strong>the</strong> Muniz family, and Afonso Fontes, Ilda Cardoso, and <strong>the</strong> late<br />
Marcília Rodrigues from whom I learnt Baniwa.<br />
I am immensely grateful to my adopted family in <strong>the</strong> Manambu-speaking Avatip<br />
village (East Sepik Province, Papua New Guinea)—especially Jacklyn Yuamali Ala,<br />
Pauline Yuaneng Agnes Luma Laki, James Sesu Laki, Dameliway, Jenny Kudapakw,<br />
Motuway, <strong>the</strong> late Gaialiwag, Yuawalup, and John Sepaywus Angi. Special thanks go<br />
to my new family in <strong>the</strong> Yalaku village of <strong>the</strong> East Sepik Province, especially Joel<br />
Ukaia and his wife Rethi, Ritha Saun, Elsa Kasandemi, yafa Mark, yafa Solomon, and<br />
David Kwaibori. Angela Filer, a Kwoma speaker from <strong>the</strong> East Sepik Province, was<br />
<strong>the</strong> first one to draw my attention to <strong>the</strong> ways in which Papuan women lose <strong>the</strong>ir<br />
identity by having to take <strong>the</strong>ir husband’s name.<br />
I am grateful to my friends who taught me Estonian—and from whom I learnt that<br />
speaking a language without a Linguistic <strong>Gender</strong> is not as boring as it may seem—Aet<br />
Lees, Reet Bergman, Reet Vallak, and Krista Gardener in Australia, and Sana<br />
Valliulina, toge<strong>the</strong>r with <strong>the</strong> late Elsa Endemann, Maimu Endemann, and Lembit<br />
Oiari back in Estonia. I owe bursts of revelation concerning Linguistic <strong>Gender</strong> in<br />
Modern Hebrew to my dear cousin Lia Shaked (née Rosman). A debt of love and<br />
gratitude goes to Emma Aikhenvald (née Breger), Tata Baeva, Ella Vainermann, and<br />
indefatigable Tamara Margolina, for patiently answering my linguistic and o<strong>the</strong>r<br />
questions concerning Russian as she is spoken today.<br />
I am grateful for <strong>the</strong> support and comments of many colleagues, students, and<br />
friends who allowed me to learn from <strong>the</strong>ir work, provided references, additional<br />
sources, and patiently answered my questions concerning <strong>the</strong>ir areas of expertise—<br />
especially Willem Adelaar, Angeliki Alvanoudi, Azeb Amha, Juliane Böttger, Nancy<br />
Dorian, Sebastian Drude, Luke Fleming, Diana Forker, Valérie Guérin, Jenia<br />
Gutova, Emi Ireland, Olga Kazakevitch, Pablo Kirtchuk, John Koontz, Maarten<br />
Kossmann, Iwona Kraska-Szlenk, Maxim Kronhaus, Aet Lees, Mike Lu, Silvia<br />
Luraghi, Lev Michael, Elena Mihas, Marianne Mithun, Edith Moravcsik, Heiko<br />
Narrog, Simon Overall, Chia-jung Pan, Nick Piper, Vera Podlesskaya, Natasha<br />
Pokrovsky, Renate Raffelsiefen, Nick Reid, David Rood, Hannah Sarvasy, Dineke<br />
Schokkin, Glenn Shepard, Cácio and Elisângela Silva, Anne Storch, Marilena
xii<br />
Acknowledgements<br />
Thanassoula, Yishai Tobin, Rosa Vallejos, Louise Vasvári, Mary Ruth Wise, Kasia<br />
Wojtylak, Sihong Zhang, and <strong>the</strong> late Bob Rankin.<br />
Invaluable comments on just about every page came from R. M. W. Dixon,<br />
without whose incisive criticism and ideas, and constant encouragement and support<br />
this book would not have appeared. Special thanks go to Angeliki Alvanoudi for her<br />
comments on <strong>the</strong> manuscript, to Nerida Jarkey for looking over <strong>the</strong> discusssion of<br />
Japanese, to Sebastian Drude for materials and comments on Awetí, and to Rosa<br />
Vallejos for materials and comments on Kokama-Kokamilla.<br />
While working on this book, I made extensive use of <strong>the</strong> Oxford English Dictionary<br />
Online—a wonderful and comprehensive resource on English and its history. I owe a<br />
debt of gratitude to JCU Library who provided us with this resource, and especially to<br />
Bronwyn Forster and Caroline Tredrea. The efficient interlibrary loan system worked<br />
like clockwork—particular thanks go to Lynn Clarke, Janine Meakins, and Bridie<br />
Soo, also at JCU Library. I owe a considerable debt to Brigitta Flick and Jolene Overall<br />
for carefully reading through drafts of this book and making corrections, and to<br />
Amanda Parsonage for looking after all things administrative while I was doing <strong>the</strong><br />
writing. This book was supported by <strong>the</strong> Australian Laureate Fellowship (from <strong>the</strong><br />
Australian Research Council) ‘<strong>How</strong> gender shapes <strong>the</strong> world: a linguistic perspective’.<br />
This volume would have never been brought to fruition without encouragement<br />
from Julia Steer, <strong>the</strong> Linguistics Editor of Oxford University Press. Her constant<br />
support makes <strong>the</strong> whole enterprise worthwhile.
Abbreviations and conventions<br />
1 first person (I)<br />
2 second person (you)<br />
3 third person (she, he, it, <strong>the</strong>y)<br />
A<br />
transitive subject<br />
ABS absolutive<br />
ACC accusative<br />
ADJ adjective<br />
ANIM animate<br />
ART article<br />
AUX auxiliary<br />
CL<br />
classifier<br />
dim<br />
diminutive<br />
DOM differential object marking<br />
ERG ergative<br />
fem, FEM feminine<br />
GEN genitive<br />
IMPERS impersonal<br />
IMPV imperative<br />
INANIM inanimate<br />
inanim inanimate<br />
INDEF indefinite<br />
LK<br />
linker<br />
LOC locative<br />
masc, MASC masculine<br />
NCL noun class<br />
NEG negation<br />
NOM nominalization<br />
NUM.CL numeral classifier<br />
O<br />
object<br />
pl<br />
plural<br />
POSS possessive
xiv<br />
Abbreviations and conventions<br />
PRES<br />
S<br />
sg<br />
VERT<br />
VIS<br />
present<br />
intransitive subject<br />
singular<br />
vertical<br />
visual<br />
Numbers of examples, tables, boxes, figures, and schemes consist of <strong>the</strong> chapter number and<br />
<strong>the</strong>n are numbered consecutively. So, Table 6.1 is <strong>the</strong> first table in Chapter 6. The majority of<br />
examples from different languages are glossed and <strong>the</strong>n translated into English. I keep <strong>the</strong><br />
original orthography and also <strong>the</strong> glossing of <strong>the</strong> quoted sources.
List of boxes, figures, schemes,<br />
and tables<br />
Box 2.1 General properties of linguistic genders 14<br />
Box 6.1 <strong>How</strong> to gain a linguistic gender: pathways of development 76<br />
Figure 3.1 <strong>Gender</strong> assignment in Manambu 36<br />
Figure 10.1 Elvia, a Tucano woman married to a Tariana man,<br />
was supposed to ‘kill’ his language. Here she is doing a traditional woman’s<br />
task: serving manioc beer to Tariana men at an Assembly of <strong>the</strong> Tariana<br />
in <strong>the</strong> Tariana school (Aikhenvald 2013b) 180<br />
Scheme 6.1 <strong>Gender</strong> and number in Zande personal pronouns 77<br />
Scheme 12.1 <strong>How</strong> <strong>the</strong> three faces of <strong>Gender</strong> relate to each o<strong>the</strong>r 213<br />
Table 2.1 <strong>Gender</strong>s in Romanian and <strong>the</strong>ir marking 17<br />
Table 2.2 <strong>Gender</strong>–number pairs in Bantu 21<br />
Table 2.3 Semantic basis of gender choice in German: an illustration 24<br />
Table 3.1 Semantic features in gender choice for nouns in Cantabrian Spanish 43<br />
Table 3.2 Physical properties in Linguistic <strong>Gender</strong> choice 44<br />
Table 3.3 Endearment and disdain through Linguistic <strong>Gender</strong>s 47<br />
Table 4.1 Variable <strong>Gender</strong> assignment in Maung 55<br />
Table 5.1 Animate classifiers in Burmese 64<br />
Table 5.2 Noun classifiers for humans and deities in Jacaltec (a selection) 66<br />
Table 5.3 Mescalero Apache classificatory verb categories 67<br />
Table 6.1 Noun classifiers for humans in Mam and nouns <strong>the</strong>y originated from 78<br />
Table 7.1 Personal pronouns in Gala compared with Manambu 99<br />
Table 7.2 Treating ‘men’ as ‘women’: masculine to feminine Linguistic<br />
<strong>Gender</strong> reversal 108<br />
Table 7.3 Treating ‘women’ as ‘men’: feminine to masculine Linguistic<br />
<strong>Gender</strong> reversal 109<br />
Table 7.4 Meanings of ‘she’ vs ‘he’ in American English (Mathiot 1979b) 110<br />
Table 9.1 Male and female forms of a selection of enclitics in Lakhota<br />
(Trechter 1995: 57) 137<br />
Table 9.2 Female versus male forms in Kokama-Kokamilla: a selection<br />
(Vallejos 2010: 42) 140<br />
Table 9.3 Female versus male forms in Awetí: a selection 140<br />
Table 9.4 Male and female speech determined by speaker and addressee in Biloxi 141<br />
Table 9.5 Male and female speech in Kũr̩ux: present tense of <strong>the</strong> verb bar- ‘come’ 142
xvi<br />
List of boxes, figures, schemes, and tables<br />
Table 9.6 Male and female speech chosen by <strong>the</strong> sex of <strong>the</strong> addressee in Tunica 142<br />
Table 9.7 Personal pronouns in Japanese: men’s and women’s speech<br />
(adapted from Ide 1991: 73) 148<br />
Table 9.8 Traits of men’s and women’s speech among <strong>the</strong> Malagasy 154<br />
Table 10.1 ‘Hidden from women’: tabooed nouns employed in traditional Tariana 171
1<br />
The multifaceted <strong>Gender</strong><br />
The multifaceted notion of ‘gender’ pervades every aspect of life and of living. <strong>Gender</strong><br />
differences form <strong>the</strong> basis for family life, patterns of socialization, distribution of tasks,<br />
spheres of responsibility, and occupational predilections. Understanding <strong>the</strong> nature<br />
of ‘gender’ is central to many disciplines—anthropology, sociology (and, of course,<br />
women’s studies), criminology, linguistics, and biology, to name a few. The way<br />
gender is articulated shapes <strong>the</strong> world of individuals, and of <strong>the</strong> societies <strong>the</strong>y live in.<br />
To different people, <strong>the</strong> word ‘gender’ means different things. For a grammarian<br />
and a linguist concerned with <strong>the</strong> structure of languages, ‘gender’ is a linguistic way of<br />
categorizing nouns reflected in <strong>the</strong>ir form, <strong>the</strong> form of an adjective or a verb which<br />
would agree with <strong>the</strong> noun, or a personal pronoun.<br />
For a sociolinguist, a psychologist, and an anthropologist, ‘gender’ is a set of<br />
norms, attitudes, feelings, and behaviours that a given culture or society associates<br />
with <strong>the</strong> person’s biological sex (male or female). A philosopher defines gender as<br />
‘social construction of male/female identity’ distinct from ‘sex, <strong>the</strong> biologically-based<br />
distinction between men and women’. 1 <strong>Gender</strong> is also defined as a set of ideas about<br />
relations and behaviours, and principles of social organization, to be understood<br />
within a social context. 2 For some, ‘gender’ reflects a social and cultural elaboration<br />
of sex differences, ‘a process that restricts our social roles, opportunities, and expectations’,<br />
and also determines some ways in which we speak. 3 And when we fill in a<br />
customs declaration, we need to state which gender we belong to—male or female.<br />
That is, in day-to-day usage, <strong>the</strong> term ‘gender’ has expanded at <strong>the</strong> expense of ‘sex’:<br />
<strong>the</strong>n gender is a physiological distinction between men and women.<br />
<strong>How</strong> to reconcile all <strong>the</strong> different meanings packaged into one word?<br />
1.1 Disentangling ‘<strong>Gender</strong>’<br />
The multifaceted concept of <strong>Gender</strong> has three faces.<br />
• LINGUISTIC GENDER. This is <strong>the</strong> original sense of ‘gender’ as a linguistic term. One<br />
class of nouns may be marked in a particular way, ano<strong>the</strong>r class in ano<strong>the</strong>r way.<br />
That class which includes most words referring to females is called ‘feminine’,<br />
<strong>How</strong> <strong>Gender</strong> <strong>Shapes</strong> <strong>the</strong> <strong>World</strong>. First edition. Alexandra Y. Aikhenvald.<br />
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.
2 1 The multifaceted <strong>Gender</strong><br />
similarly for males and ‘masculine’. <strong>Gender</strong> classes are defined by <strong>the</strong>ir male and<br />
female members but may extend beyond those. The ways in which animals,<br />
birds, insects, plants, and natural phenomena (such as thunder and wind) are<br />
assigned to genders may reflect <strong>the</strong>ir roles in legends and metaphors, and reveal<br />
folk taxonomies. Linguistic <strong>Gender</strong> is integrated in <strong>the</strong> grammar, and is one of<br />
<strong>the</strong> ways languages use to categorize nouns.<br />
• NATURAL GENDER. This is what was until recently simply called ‘sex’—male versus<br />
female. A female is able to bear children, a male is not. Natural <strong>Gender</strong> entails<br />
anatomical and hormonal differences, linked to concomitant physiological and<br />
psychological traits. In <strong>the</strong> day-to-day use, ‘gender’ has just about displaced <strong>the</strong><br />
term ‘sex’—perhaps felt to be too blunt and rude.<br />
• SOCIAL GENDER. This reflects <strong>the</strong> social implications, and norms, of being a man<br />
or a woman (or perhaps something in between). In Simone de Beauvoir’s(1949:<br />
267) adage: ‘One is not born, but ra<strong>the</strong>r becomes, a woman.’ In many traditional<br />
societies of New Guinea, social manhood is achieved, and defined, through male<br />
initiation. Similarly, in o<strong>the</strong>r traditional societies, social womanhood used to be<br />
achieved through female initiation. Social <strong>Gender</strong> relates to contrasting social<br />
roles of <strong>the</strong> sexes, and how <strong>the</strong>se are embodied in cultural practices and public<br />
ritualized behaviour. (These patterns include conventions for <strong>the</strong> behaviour of men<br />
and women, known as ‘gender etiquette’, social stereotypes associated with males<br />
and females, and a traditional complex of knowledge and beliefs about mythical<br />
women and mythical men. 4 )<br />
The three faces of <strong>Gender</strong> interact. Investigations of Natural <strong>Gender</strong> focus on<br />
innate biological differences between men and women. They are also played out in<br />
<strong>the</strong> ways men and women communicate, within <strong>the</strong>ir Social <strong>Gender</strong>s. As Labov<br />
(1972: 304) puts it, ‘<strong>the</strong> sexual differentiation of speakers is . . . not a product of<br />
physical factors alone’, but ‘ra<strong>the</strong>r an expressive posture which is socially more<br />
appropriate for one sex or ano<strong>the</strong>r’. In a ground-breaking study of physical features<br />
of ‘women’s’ speech among <strong>the</strong> Tohono O'odham (a Uto-Aztecan group from<br />
Mexico), Hill and Zepeda (1999) show how women (not men) use a pulmonic<br />
ingressive airstream in order to construct a special atmosphere of conversational<br />
intimacy, taking advantage of size differences between male and female vocal tracts.<br />
Such sound production is easier to achieve with <strong>the</strong> smaller female larynx and<br />
pharynx. Physical attributes—including high pitched voice—typical of female Natural<br />
<strong>Gender</strong> come to be associated with ‘female talk’, and redeployed as tokens of<br />
Social <strong>Gender</strong> and associated attitudes.<br />
The difference between Natural <strong>Gender</strong>, or sex, and Linguistic <strong>Gender</strong> was captured<br />
by Jespersen (1949: 174):<br />
Sex is a natural quality shown primarily in <strong>the</strong> productive organs . . . <strong>Gender</strong> is a grammatical<br />
category. Many languages have class distinctions of different characters. <strong>Gender</strong> in primary
1.1 Disentangling ‘<strong>Gender</strong>’ 3<br />
words (chiefly substantives [nouns]) is not always shown by <strong>the</strong> form of <strong>the</strong> word itself, but it<br />
may influence <strong>the</strong> form of o<strong>the</strong>r words (secondary) and is thus chiefly a syntactic category.<br />
Languages vary very much with regard to <strong>the</strong> number of classes distinguished, also with regard<br />
to <strong>the</strong> correspondence forums between <strong>the</strong>se grammatical classes and natural distinctions such<br />
as those of sex, between big and small, between living and lifeless, etc. <strong>Gender</strong> thus cannot be<br />
defined as <strong>the</strong> grammatical expression of sex, but may relate to o<strong>the</strong>r things.<br />
Natural <strong>Gender</strong> and Social <strong>Gender</strong> work toge<strong>the</strong>r creating stereotypes of behaviour<br />
in each society and culture. Ortner and Whitehead (1981: 1) put this as follows:<br />
Natural features of gender, and natural processes of sex and reproduction, furnish only a<br />
suggestive and ambiguous backdrop to <strong>the</strong> cultural organization of gender and sexuality. What<br />
gender, what men and women are, what sorts of relations do or should obtain between <strong>the</strong>m—<br />
all of <strong>the</strong>se notions do not simply reflect or elaborate upon biological ‘givens’, but are largely<br />
products of social and cultural processes. 5<br />
The division of biological, or ‘Natural’ <strong>Gender</strong>s, goes beyond a male and female<br />
dichotomy. Numerous traditional and modern societies have ‘groups whose gender<br />
identities and enactments fall outside of sociocultural norms for women and men’—<br />
<strong>the</strong>se are <strong>the</strong> ones described as ‘a third sex’, ora‘third (Natural) <strong>Gender</strong>’. 6 ‘Transgender’<br />
is ano<strong>the</strong>r umbrella term which encompasses those whose behaviour shows<br />
patterns associated with <strong>the</strong> opposite sex—including transsexuals, transvestites, and<br />
drag queens and kings. Transsexuals are those whose Social <strong>Gender</strong> identity does not<br />
correspond to <strong>the</strong> male or female Natural <strong>Gender</strong> characteristics <strong>the</strong>y were born with.<br />
Some undergo sex-reassignment surgery to change <strong>the</strong>ir biological features, so that it<br />
should match <strong>the</strong>ir gender identity. Transvestites are men and women in terms of<br />
<strong>the</strong>ir Natural <strong>Gender</strong> who dress and behave as members of <strong>the</strong> opposite gender. The<br />
way gays, lesbians, and transgender people speak reflects <strong>the</strong>ir identity as special<br />
groups, and highlights linguistic features perceived as characteristic of being ‘male’ or<br />
‘female’. One of <strong>the</strong>se features is manipulating Linguistic <strong>Gender</strong>. Brigitte Martel, a<br />
transsexual male who became female, aptly captured this in <strong>the</strong> title of her autobiography<br />
by changing Linguistic <strong>Gender</strong>s which accompanied <strong>the</strong> change of her Natural<br />
and Social <strong>Gender</strong> from male to female: Né homme, comment je suis devenue femme—<br />
‘Born (masculine) a man, how I became (feminine) woman’. Hijras—womanly men in<br />
India and Nepal—talk about <strong>the</strong>mselves using feminine or masculine Linguistic<br />
<strong>Gender</strong> depending on circumstances and attitudes. 7<br />
Traditionally, transgender practices play a role in initiation and o<strong>the</strong>r rituals across<br />
<strong>the</strong> world. These include Naven, made famous by Gregory Bateson, and fur<strong>the</strong>r<br />
explored in <strong>the</strong> literature on Sepik cultures, and across New Guinea. 8 Cross-dressing<br />
of men as women during initiation ceremonies in <strong>the</strong> Sepik area and <strong>the</strong> Highlands of<br />
New Guinea is thought to be a way of getting initiates to acquire a proper male social<br />
identity—or masculine Social <strong>Gender</strong>—and rationalize sexual roles (usually characterized<br />
by male dominance). Male and female transgender people have been documented
4 1 The multifaceted <strong>Gender</strong><br />
for numerous groups of <strong>the</strong> North American Plains. Just a few correlations with<br />
language have so far been recorded for <strong>the</strong>se cultural practices. 9<br />
From a linguist’s perspective, Linguistic <strong>Gender</strong> occupies a central position in<br />
shaping <strong>the</strong> role and <strong>the</strong> meanings of <strong>Gender</strong> in its three faces. Linguistic <strong>Gender</strong> as<br />
a way of categorizing entities through language—and o<strong>the</strong>r realizations of Social<br />
and Natural <strong>Gender</strong> in language—shapes <strong>the</strong> world we live in and <strong>the</strong> ways we<br />
perceive and construct it. O<strong>the</strong>r aspects of language use also set women and men apart.<br />
These include ways of speaking, speech genres and speech practices, and often politeness<br />
forms.<br />
The multifaceted concept of ‘gender’ spans a linguistic category, a complex of<br />
social norms, and a set of biological features. This book is about <strong>the</strong> ways in which<br />
gender is reflected in language—and more specifically, <strong>the</strong> role of Linguistic <strong>Gender</strong><br />
in <strong>the</strong> expression of Social <strong>Gender</strong> and Natural <strong>Gender</strong>, <strong>the</strong>ir manipulations and<br />
development.<br />
1.2 What is special about Linguistic <strong>Gender</strong><br />
Some linguistic categories show strong correlations with cultural values, social<br />
hierarchies, and <strong>the</strong>ir conceptualization. Imperatives and commands reflect relationships<br />
between people: for instance, if a speaker of Dolakha Newar is considerably<br />
younger than <strong>the</strong> addressee, or is talking to someone <strong>the</strong>y particularly respect, <strong>the</strong>y<br />
will use special honorific imperatives. Simple imperatives will be reserved for <strong>the</strong>ir<br />
equals. 10 Meanings encoded within possessive structures often reflect relationships<br />
within a society, and change if <strong>the</strong> society changes.<br />
Toge<strong>the</strong>r with o<strong>the</strong>r ways of categorizing nouns, Linguistic <strong>Gender</strong> tends to mirror<br />
social and cultural stereotypes and patterns of human perception. Linguistic <strong>Gender</strong><br />
is a repository of beliefs about what men and women are like and how <strong>the</strong>y behave, and<br />
features which are ‘male’-like or ‘female’-like. 11 Language planning, political correctness,<br />
and societal changes shape various aspects of Linguistic <strong>Gender</strong>—especially with<br />
regard to how humans are categorized. The ways in which people use Linguistic<br />
<strong>Gender</strong> may mirror <strong>the</strong> status of Social <strong>Gender</strong>s. For instance, in Jarawara, a small<br />
Arawá language from sou<strong>the</strong>rn Amazonia, a particularly respected woman can be<br />
referred to with masculine gender, as if she were being ‘promoted’ to <strong>the</strong> male gender<br />
status. <strong>Gender</strong>s reflect <strong>the</strong> history of ideas and attitudes. The recent trend against <strong>the</strong><br />
generic masculine pronoun in English reflects <strong>the</strong> ways in which established stereotypes<br />
can be gradually remoulded.<br />
More than half <strong>the</strong> world’s languages have Linguistic <strong>Gender</strong> in <strong>the</strong>ir grammar. 12<br />
As Franz Boas (a founding fa<strong>the</strong>r of modern linguistics) put it, languages differ not in<br />
what one can say but in what kind of information must be stated: ‘grammar . . .<br />
determines those aspects of each experience that must be expressed’ (Boas 1938:<br />
132). Having to be always conscious of which Linguistic <strong>Gender</strong> to apply—especially
1.2 What is special about Linguistic <strong>Gender</strong> 5<br />
to humans—makes one alert to o<strong>the</strong>r faces of <strong>Gender</strong>, as a natural division of people<br />
into male, female, and perhaps ‘o<strong>the</strong>r’, and as a cultural and social construct.<br />
Linguistic <strong>Gender</strong> comes in many guises and serves many masters. It helps follow<br />
<strong>the</strong> thread of communication, and figure out who or what is being talked about. Having<br />
an obligatory Linguistic <strong>Gender</strong> allows for rich and expansive imagery, and makes <strong>the</strong><br />
language we speak more colourful and versatile. Linguistic gender, its choice and<br />
associations, is something speakers are ready to discuss and argue about. This metalinguistic<br />
perception of gender makes it central for metaphors—especially in poetry.<br />
No term in linguistics is fully straightforward; <strong>the</strong>re is some approximation in each.<br />
Of all linguistic terms, ‘gender’ is uniquely polysemous in its coverage—it subsumes<br />
‘gender’ in a strict grammatical meaning, and extends to a biological division of<br />
humans into males and females, and to conventionalized differences in <strong>the</strong>ir social<br />
status and linguistic behaviour. The diffuse nature of <strong>the</strong> term is sometimes<br />
irritating—as every mention of <strong>the</strong> term has to be unpacked and explained. Turkish<br />
can be considered ‘gender-less’—in terms of <strong>the</strong> absence of grammatical Linguistic<br />
<strong>Gender</strong>. Yet meanings of Linguistic <strong>Gender</strong> can be expressed through some derivational<br />
affixes on nouns, e.g. hoc-anIm ‘female teacher’. And <strong>the</strong> language is not<br />
‘gender-neutral’ in <strong>the</strong> sense that Social <strong>Gender</strong>s find <strong>the</strong>ir expression through o<strong>the</strong>r<br />
means. A nurse is likely to be a woman, and a taxi driver a man. 13<br />
On <strong>the</strong> o<strong>the</strong>r hand, <strong>the</strong> advantage of having a general term encompassing every<br />
aspect of gender classification—linguistic, biological, and social—helps bring <strong>the</strong><br />
three toge<strong>the</strong>r, and highlight <strong>the</strong>ir commonalities and <strong>the</strong> ways in which <strong>the</strong>y may<br />
influence each o<strong>the</strong>r. The ambiguity of <strong>the</strong> term ‘<strong>Gender</strong>’ alerts us to <strong>the</strong> existence of<br />
an overarching concept behind it, spanning linguistic expression, social aspects, and<br />
biological features. Linguistic <strong>Gender</strong> in its various guises and <strong>the</strong> expression of<br />
<strong>Gender</strong> in language reflect and shape Social <strong>Gender</strong> stereotypes, associations, and<br />
attitudes, in <strong>the</strong>ir relationship to Natural—or biological—<strong>Gender</strong> distinctions. 14 This<br />
is what this book is about.<br />
Linguistic <strong>Gender</strong> is a way of categorizing nouns. It always involves universal<br />
features of sex, humanness, and animacy, and is a window into social life and<br />
cognitive patterns. But Linguistic <strong>Gender</strong> classification goes beyond male and female.<br />
As we will see throughout Chapter 2 (particularly in §2.4), Linguistic <strong>Gender</strong>s are<br />
also used to categorize inanimate entities. This is where we might find that a gender<br />
labelled ‘feminine’ would include more than just females, and <strong>the</strong> one labelled<br />
‘masculine’ more than just males. The term ‘neuter’ tends to refer to a gender<br />
which includes inanimate (or irrational) beings, or a residue gender whose semantic<br />
basis is difficult to capture. The choice of Linguistic <strong>Gender</strong> can be based on clues<br />
o<strong>the</strong>r than just <strong>the</strong> meaning of a noun. Feminine and masculine genders often<br />
include inanimate nouns with no connection to female or male sex, e.g. French<br />
maison ‘house’ (feminine), château ‘castle’ (masculine).
6 1 The multifaceted <strong>Gender</strong><br />
In Antoine Meillet’s (1964: 164) words, gender provides an example of ‘a grammatical<br />
category that plays (in a good many of <strong>the</strong> modern Indo-European languages)<br />
a considerable role in morphology without answering, most of <strong>the</strong> time, to a<br />
definite meaning’—especially where inanimate nouns are concerned.<br />
The partial semantic opacity of Linguistic <strong>Gender</strong> has earned it a bad reputation,<br />
among some linguists and lay people who complain that Linguistic <strong>Gender</strong> is hard<br />
to learn. Jespersen (1972: 108) praised Modern English for losing complex agreement<br />
in Linguistic <strong>Gender</strong>—as found in Old English—and making things easier and more<br />
straightforward:<br />
In Old English, as in all <strong>the</strong> old cognate languages, each substantive [noun], no matter whe<strong>the</strong>r<br />
it referred to animate beings or thing or abstract notions, belonged to one or o<strong>the</strong>r of <strong>the</strong> three<br />
gender-classes. Thus masculine pronouns and endings were found with names of a great many<br />
things which had nothing to do with male sex (e.g. horn, ende ‘end’, ebba ‘ebb’, dæg ‘day’) and<br />
similarly feminine pronouns and endings with many words without any relation to female sex<br />
(e.g. sorh ‘sorrow’, glof ‘glove’, plume ‘plum’, pipe). Anyone acquainted with <strong>the</strong> intricacies of<br />
<strong>the</strong> same system (or want of system) in German will feel how much English has gained in<br />
clearness and simplicity by giving up <strong>the</strong>se distinctions and applying he only to male, and she<br />
only to female, living beings. The distinction between animate and inanimate now is much<br />
more accentuated than it used to be.<br />
Throughout this book we will see how important Linguistic <strong>Gender</strong> is for many<br />
aspects of human communication. And Linguistic <strong>Gender</strong> in English is not as simple<br />
and clear as it may seem. The ways in which <strong>the</strong> use of pronouns, especially <strong>the</strong><br />
generic he, have changed in recent years reflect social developments and <strong>the</strong> changing<br />
status of Social <strong>Gender</strong>s across <strong>the</strong> English-speaking world.<br />
What is so special about <strong>the</strong> Linguistic <strong>Gender</strong>? <strong>How</strong> does it interact with Social<br />
<strong>Gender</strong> and Natural <strong>Gender</strong>, across languages and cultures? What makes it a useful<br />
linguistic resource ra<strong>the</strong>r than an encumbrance for poor language learners? And how<br />
does <strong>the</strong> integrated complex of resources of Natural <strong>Gender</strong>, Linguistic <strong>Gender</strong>, and<br />
<strong>the</strong> evolving images of Social <strong>Gender</strong> play out in view of societal changes? <strong>How</strong> can<br />
Linguistic <strong>Gender</strong> and <strong>the</strong> semantic composition of categories related to ‘male’ and<br />
‘female’ undergo restructuring in language planning? These are <strong>the</strong> questions we<br />
approach in <strong>the</strong> present study.<br />
Throughout <strong>the</strong> book, I have chosen to capitalize Linguistic <strong>Gender</strong>, Natural<br />
<strong>Gender</strong>, and Social <strong>Gender</strong>—to stress <strong>the</strong> fact that all of <strong>the</strong>se are ultimately just<br />
nicknames which only partly capture <strong>the</strong> concepts and categories <strong>the</strong>y cover.<br />
1.3 <strong>How</strong> this book is organized<br />
We start, in Chapter 2, with Linguistic <strong>Gender</strong> and its expression. Many languages<br />
of <strong>the</strong> world have a gender system in <strong>the</strong>ir grammar. The size of <strong>the</strong> system varies.
1.3 <strong>How</strong> this book is organized 7<br />
There are two genders in French, three in German, four in Dyirbal (from North<br />
Queensland in Australia), more elsewhere. We seldom find an exact correspondence<br />
between masculine/feminine (Linguistic <strong>Gender</strong>) and male/female sex (Natural<br />
<strong>Gender</strong>). In German most nouns referring to females are feminine but Mädchen<br />
‘girl’ is in neuter gender (because it contains <strong>the</strong> diminutive suffix -chen which is<br />
always neuter). Linguistic <strong>Gender</strong> may span grammar and lexicon; vide he-man, tomboy<br />
in English. <strong>Gender</strong> may be distinguished in personal pronouns only, as in<br />
English. This chapter covers <strong>the</strong> formal properties of Linguistic <strong>Gender</strong>s, gender<br />
agreement, and anaphoric gender, and <strong>the</strong> ways in which Linguistic <strong>Gender</strong> interacts<br />
with o<strong>the</strong>r linguistic categories. We <strong>the</strong>n focus on different principles of Linguistic<br />
<strong>Gender</strong> choice—by meaning and also by form—and look at <strong>the</strong> problem of markedness<br />
in Linguistic <strong>Gender</strong>s.<br />
Linguistic <strong>Gender</strong>s always include semantic parameters of animacy, humanness,<br />
and sex, or Natural <strong>Gender</strong>. In a number of languages, <strong>the</strong> choice of Linguistic<br />
<strong>Gender</strong>s—especially for inanimate entities—is based on <strong>the</strong>ir shape and size. The<br />
meanings of Linguistic <strong>Gender</strong>s may involve value and importance—reflecting<br />
associations with, and stereotypes of, Social <strong>Gender</strong>s. This is <strong>the</strong> topic of Chapter 3,<br />
‘Round women and long men: physical properties in Linguistic <strong>Gender</strong>’.<br />
Linguistic <strong>Gender</strong>s have a plethora of functions—<strong>the</strong>y help highlight different<br />
meanings of <strong>the</strong> same noun, track referents in discourse, and are a source of elaborate<br />
metaphors. In a number of languages a noun can be assigned to more than one<br />
Linguistic <strong>Gender</strong> with a change in meaning: <strong>the</strong>se underscore <strong>the</strong> versatility, and <strong>the</strong><br />
utility of Linguistic <strong>Gender</strong>s as a means of classifying entities of <strong>the</strong> world, debunking<br />
<strong>the</strong> myth of gender as an arbitrary and redundant category. We discuss <strong>the</strong>se in<br />
Chapter 4, ‘What are Linguistic <strong>Gender</strong>s good for?’<br />
Meanings associated with Linguistic <strong>Gender</strong>s—animacy, humanness, and sex—<br />
can be expressed through a variety of o<strong>the</strong>r means. These include noun categorization<br />
devices, or classifiers, and many noun categories, including case and number.<br />
So-called ‘gender-less’ languages have ways of expressing gender meanings, through<br />
using different words for males and females, or different affixes to distinguish sexes.<br />
Attitudes to Social <strong>Gender</strong>s—often downplaying <strong>the</strong> status of women—come to light<br />
through <strong>the</strong> use of terms and forms in ‘gender-less’ languages. Chapter 5, ‘<strong>Gender</strong><br />
meanings in grammar and lexicon’, addresses <strong>the</strong>se issues.<br />
In Chapter 6, ‘The rise and fall of Linguistic <strong>Gender</strong>s’, we turn to where Linguistic<br />
<strong>Gender</strong>s come from, how <strong>the</strong>y may get restructured over time, and how <strong>the</strong>y can be<br />
lost altoge<strong>the</strong>r. If languages are in contact, <strong>the</strong>y often come to share <strong>the</strong>ir Linguistic<br />
<strong>Gender</strong>s. Contact is often to blame for <strong>the</strong> demise of <strong>Gender</strong>s. Adaptability of<br />
Linguistic <strong>Gender</strong> in situations of language contact and language obsolescence<br />
fur<strong>the</strong>r attests to its vital importance, and functionality. This chapter also touches<br />
upon <strong>the</strong> acquisition of Linguistic <strong>Gender</strong>s by children, and <strong>the</strong>ir loss in language<br />
dissolution. Linguistic <strong>Gender</strong>s can be reshaped as part of conscious language
8 1 The multifaceted <strong>Gender</strong><br />
engineering: this issue is mentioned in Chapter 6, and <strong>the</strong>n discussed in some more<br />
detail in Chapter 11.<br />
What happens when men are assigned to <strong>the</strong> feminine Linguistic <strong>Gender</strong> and<br />
women to <strong>the</strong> masculine—that is, if Linguistic <strong>Gender</strong>s are reversed? This may be<br />
done for a joke, or <strong>the</strong> effect may imply offence, praise, solidarity, and endearment—<br />
based on subtle overtones of value underlying Social <strong>Gender</strong>s. Reversing Linguistic<br />
<strong>Gender</strong> of humans—speaking of a man as if he were a woman, and of a woman as if she<br />
were a man—highlights <strong>the</strong> stereotypes associated with male and female Social <strong>Gender</strong>.<br />
We also discuss <strong>the</strong> overtones of <strong>the</strong> word meaning ‘man’ used to refer to people in<br />
general, and markedness, status, and power—intrinsically associated with Social <strong>Gender</strong><br />
categories—as reflected in Linguistic <strong>Gender</strong> choice. These are <strong>the</strong> topics discussed<br />
in Chapter 7, ‘Manly women and womanly men: <strong>the</strong> effects of gender reversal’.<br />
Linguistic <strong>Gender</strong>s are a source of metaphors and poetic imagery. They reflect<br />
myths, beliefs, and traditions of <strong>the</strong> speakers. Linguistic <strong>Gender</strong>s may mirror Social<br />
<strong>Gender</strong> patterns, as we saw in Chapter 7. In addition, Linguistic <strong>Gender</strong>s and <strong>the</strong>ir<br />
meanings may affect cognition and <strong>the</strong> ways in which people perceive <strong>the</strong> world<br />
around <strong>the</strong>m. Men and women have different physical characteristics which relate to<br />
<strong>the</strong>ir Social <strong>Gender</strong>s. These are <strong>the</strong> topics of Chapter 8, ‘The images of gender’.<br />
Social <strong>Gender</strong> finds its linguistic expression through <strong>the</strong> ways in which men and<br />
women speak. Differences between male and female ways of speaking exist in any<br />
language. Natural <strong>Gender</strong> properties—such as higher pitch and narrow vocal tract—<br />
account for some of such differences. Male and female speech distinctions can be<br />
paralinguistic (that is, differences may lie in higher pitch for women’s speech, and<br />
o<strong>the</strong>r phonetic features, plus facial and bodily gestures). Or <strong>the</strong>y may be conventionalized<br />
as an integral part of grammar or lexicon: this is <strong>the</strong> case in ‘genderexclusive<br />
languages’ where men and women have different phonemic systems, or<br />
obligatorily use different words, or different sound correspondences. In a number<br />
of languages—from linguistic minorities in North and South America, Siberia, and<br />
India to Japanese and Thai—such differences between women’s and men’s dialects<br />
are obligatory and striking. Conventionalized registers known as ‘women’s speech’<br />
and ‘men’s speech’ where gender indexicals systematically span phonological, morphological,<br />
and lexical domains have been described for numerous languages in<br />
North America, <strong>the</strong> Chukotko-Kamchatkan family in Siberia, and a few in Amazonia.<br />
The choice of code may be determined by a combination of Natural <strong>Gender</strong> and<br />
Social <strong>Gender</strong>: In Koasati, a Muskogean language from North America, using male<br />
code is a mark of authority. According to <strong>the</strong> male Atsinas, from <strong>the</strong> Algonquian<br />
family, ‘male’ status is acquired by birth and social and cultural maturity; and it is<br />
thus natural that a racial and cultural outsider be addressed with a female form. 15<br />
‘Male’ and ‘female’ are an achieved status, and not an innate property.<br />
Patterns of male and female speech, or male and female dialects, can be deployed<br />
in constructing one’s identity. Members of <strong>the</strong> third gender—including hijras in
1.4 The empirical basis, and a note on conventions 9<br />
India, and gays, lesbians, and transvestites in Western societies—deploy male and<br />
female speech differences, Linguistic <strong>Gender</strong>, and many o<strong>the</strong>r linguistic features, to<br />
project an image of a male or a female in a Social <strong>Gender</strong> sense. Chapter 9, ‘When<br />
women and men speak differently’, focuses on distinct speech patterns corresponding<br />
to <strong>the</strong> divisions in Social <strong>Gender</strong>, and gender-determined variation in speech<br />
practices in gender-variable languages.<br />
Men and women may have different speech styles and master different genres. In<br />
Chapter 10, ‘The rituals of gender’, we fur<strong>the</strong>r explore <strong>the</strong> roles of Social <strong>Gender</strong>s in<br />
traditional societies where women and men used to be associated with distinct<br />
domains, and different speech styles. This is where <strong>the</strong> asymmetry, and <strong>the</strong> lack of<br />
equality, between women and men is particularly apparent. Special languages and<br />
language registers can come to be used in male-only rituals. A whole set of terms may<br />
be forbidden to women. Women can be viewed ei<strong>the</strong>r as keepers and promoters of<br />
prestigious linguistic norm, and traditional language, or as dangerous ‘o<strong>the</strong>rs’ which<br />
lead <strong>the</strong> society in <strong>the</strong> wrong direction. Conventionalized speech practices reflect <strong>the</strong><br />
relative standing of, and <strong>the</strong> asymmetry between, Social <strong>Gender</strong>s.<br />
<strong>How</strong> does Linguistic <strong>Gender</strong> reflect social changes—and <strong>the</strong> improvement of <strong>the</strong><br />
woman’s place (using <strong>the</strong> title of Robin Lakoff ’s 1975 classic book)? In Chapter 11,<br />
‘<strong>Gender</strong> in grammar and society’, we discuss <strong>the</strong> impact of social changes in <strong>the</strong> position<br />
of women on <strong>the</strong> use of Linguistic <strong>Gender</strong>s and classifiers, including <strong>the</strong> avoidance of<br />
generic use for male terms and trends for European languages to become more ‘genderequal’.<br />
The feminist movement plays a particular role in shaping language change.<br />
Throughout <strong>the</strong> history of feminism and <strong>the</strong> backlash against it, language has been used<br />
as a rhetorical weapon—reflecting power relationships, investigating social and linguistic<br />
discrimination, and <strong>the</strong> embodiment and sexualization of women, and <strong>the</strong>ir traditional<br />
activities, as a way of putting women down. Feminist <strong>the</strong>ories have played a<br />
substantial role in working out problems with women’s status and obtaining equality—<br />
and even more, moving away from polarization and classification based on ‘sex’. We<br />
focus on how language reflects gender in all its guises, and how <strong>the</strong> linguistic treatment<br />
of ‘women’ can be seen as a barometer of social change.<br />
The last chapter—‘The heart of <strong>the</strong> matter: envoi’—summarizes <strong>the</strong> main points—<br />
<strong>the</strong> meanings and expression of Linguistic <strong>Gender</strong> and <strong>the</strong>ir correlations with gender<br />
as a social construct, toge<strong>the</strong>r with changes in Linguistic <strong>Gender</strong> choice, and form,<br />
depending on changes in world view, cosmology, and social realities in flux.<br />
1.4 The empirical basis, and a note on conventions<br />
This book is focused on <strong>the</strong> three faces of gender across languages and cultures of<br />
<strong>the</strong> world. It has an empirical inductive focus—relying on facts ra<strong>the</strong>r than on ad hoc<br />
<strong>the</strong>ories and hypo<strong>the</strong>ses. 16 This study is based on an investigation of about 700<br />
languages and—where possible—<strong>the</strong>ir social environment. Special attention has been
10 1 The multifaceted <strong>Gender</strong><br />
paid to data from languages on which I have first-hand expertise, and to minority<br />
languages and groups which have not yet been given enough prominence in <strong>the</strong><br />
existing literature on aspects of gender. Giving prominence to minority languages<br />
and cultures helps avoid a bias towards Western stereotypes and westernized perception<br />
of women and men. For instance, a treatment of Linguistic <strong>Gender</strong>s—which<br />
deviates from what one is accustomed to in familiar Indo-European languages—as<br />
‘non-canonical’ reflects <strong>the</strong> weight, and <strong>the</strong> bias, of post-colonial traditionalism. This<br />
is what I have attempted to eschew.<br />
Throughout this book, I have only been able to mention a portion of <strong>the</strong> available<br />
literature, and only a selection of examples. A fair number of works on gender in its<br />
various guises have not been mentioned here—ei<strong>the</strong>r because <strong>the</strong>y provide additional<br />
instances and anthropological ra<strong>the</strong>r than language-oriented discussion of points<br />
exemplified here, or because <strong>the</strong>y are not exactly relevant to <strong>the</strong> ways linguistic<br />
phenomena are used, or because <strong>the</strong>y make claims which are not fully au<strong>the</strong>nticated<br />
or convincing, or contain mistakes and misinterpretations which make <strong>the</strong>m unreliable.<br />
I could not cite all <strong>the</strong> examples of every particular phenomenon—o<strong>the</strong>rwise <strong>the</strong><br />
book would have become immense. I usually provide a particularly illustrative<br />
example, and mention o<strong>the</strong>r similar ones (in a note). If a certain phenomenon is<br />
found in more than half of <strong>the</strong> languages under consideration I call it ‘relatively<br />
frequent’; if it is found in a restricted number of languages (one to ten), I cite all of<br />
<strong>the</strong>m and indicate its rarity. Note, however, that what appears rare to us at <strong>the</strong> present<br />
stage of knowledge may turn out to be more frequent when we start learning more<br />
about hi<strong>the</strong>rto little-known languages and areas. This is <strong>the</strong> reason why I chose at this<br />
stage not to give any statistical counts. Only about one-tenth of all human languages—<br />
and societies in which <strong>the</strong>y are spoken—have been documented so far; it <strong>the</strong>refore<br />
seems most judicious to follow a qualitative approach, postponing quantitative analysis<br />
some time in <strong>the</strong> future, when more data is available and can be assessed.<br />
This book contains many examples from—and many mentions of—languages from<br />
various areas and genetic groupings. When <strong>the</strong> language is introduced for <strong>the</strong> first<br />
time, its affiliation and where it is spoken is given in brackets—for instance, ‘Kwami<br />
(a Chadic language from Nigeria)’. Later mentions of <strong>the</strong> same language do not<br />
include this information—which is summarized in <strong>the</strong> Index of languages at <strong>the</strong> end<br />
of <strong>the</strong> book. At <strong>the</strong> end of each chapter <strong>the</strong> reader will find notes and sources.<br />
This books aims at unravelling how Linguistic <strong>Gender</strong>, Social <strong>Gender</strong>, and Natural<br />
<strong>Gender</strong> interact, viewed through <strong>the</strong> eyes of a linguist. As <strong>the</strong> societies move towards<br />
greater equality of Social <strong>Gender</strong>s in <strong>the</strong>ir attitudes and practices, <strong>the</strong> languages <strong>the</strong>y<br />
speak evolve. I have tried to capture some of <strong>the</strong> dynamics of how <strong>the</strong> expression of<br />
multifaceted <strong>Gender</strong> reflects <strong>the</strong> world of perception, cognition, and social change. As<br />
Bolinger (1991: 319) puts it, ‘no matter how wide <strong>the</strong> net is cast, a fish or two always<br />
escapes’. There will always be room for upcoming enthusiasts to cast <strong>the</strong>ir nets wider.
1.4 The empirical basis, and a note on conventions 11<br />
NOTES AND SOURCES<br />
1. Bullock, Stallybrass, and Trombley (1977: 348).<br />
2. This is <strong>the</strong> definition on <strong>the</strong> website of <strong>the</strong> American Psychological association. Many<br />
sources focus on issues of Social <strong>Gender</strong> and individual aspects of its linguistic expression.<br />
A full bibliography and critique will be a task on its own. A comprehensive bibliography<br />
on Social <strong>Gender</strong> is in Upton (2012); see also definitions and references in Holmes and<br />
Meyerhof (2003), Kramarae and Treichler (1992: 173); Baron (1986), Coates (1993, 2012),<br />
Mills (2003a, 2003b, 2008), Talbot (2010), Coates and Pichler (2011), McConnell-Ginet<br />
(2014), and also Romaine (1999); Aikhenvald (2015b) is an up-to-date bibliography on<br />
Linguistic <strong>Gender</strong>. Silverstein (1985: 220–3) identifies a fur<strong>the</strong>r meaning of ‘gender’—that<br />
of an ideology of <strong>the</strong> ways in terms of which gendered language and <strong>the</strong> patterns of<br />
variation are understood by speakers <strong>the</strong>mselves.<br />
3. Cheshire (2002: 423).<br />
4. TheOxford English Dictionary Online describes ‘gender’ as ‘<strong>the</strong> state of being male or female’,<br />
also stating that ‘gender’ is a linguistic term and refers to <strong>the</strong> ‘grammatical classification of<br />
nouns and related words, which roughly corresponds to <strong>the</strong> two sexes and sexlessness’.<br />
Terms ‘natural gender’,or‘biological gender’ are sometimes used interchangeably.<br />
5. See also Moore (1994) on‘sex’ and ‘social gender’.<br />
6. Zimman and Hall (forthcoming). The term ‘queer’ is an alternative, which covers gay men<br />
and lesbian women, and ‘transgender’ individuals. The term ‘queer linguistics’ is used in<br />
academic writing to refer to linguistic practices by gays, lesbians, and transgender people.<br />
Barrett (2006) offers an overview of <strong>the</strong> history of <strong>the</strong> term, and of ‘queer’ speech; see also<br />
Leap (2012).<br />
7. See Zimman and Hall (forthcoming); Bucholz and Hall (1995b), Hall (2002), Hall and<br />
O’Donovan (1996) and also McConnell-Ginet (2014). We return to this in §9.3.<br />
8. Bateson (1958); see also Silverman (2001), Herdt (1987), Creed (1984).<br />
9. Trechter (1995); see §9.3.<br />
10. See Aikhenvald (2010: 215–16).<br />
11. See Aikhenvald (2000: 347–50), and references <strong>the</strong>re; and Chapters 4 and 11.<br />
12. Linguistic <strong>Gender</strong> has been subject to many misconceptions and ‘linguistic myths’. One<br />
is, in Philips’s(1980: 530) words, that ‘grammatical gender is relatively rare’, in contrast to<br />
natural gender which is ‘a cultural and linguistic universal’. The statement about ‘rarity’ of<br />
grammatical gender is utterly wrong—grammatical gender is one of <strong>the</strong> most widespread<br />
categories in <strong>the</strong> world, covering much of Africa, Europe, North and South America, and<br />
New Guinea; see Corbett (1991) and map 1 in Aikhenvald (2000: 78). The term ‘linguistic<br />
gender’ has been used in a number of contradictory ways. English has no gender agreement<br />
within a noun phrase. This has led some German scholars to state that English has<br />
‘no gender’, forgetting about <strong>the</strong> gendered pronouns ‘she’, ‘he’, and ‘it’. Pitfalls of <strong>the</strong> term<br />
‘natural gender’ and its current usage have been addressed by McConnell-Ginet (2014).<br />
13. See a comprehensive discussion in Braun (1997b; 2000a; 2001).<br />
14. English is ra<strong>the</strong>r unusual in having just one term to cover <strong>the</strong> three faces of ‘<strong>Gender</strong>’. In<br />
French, genre is used for Linguistic <strong>Gender</strong> (and also in a number of o<strong>the</strong>r meanings,<br />
including kind and genre); <strong>the</strong> term sexe covers Social and Natural <strong>Gender</strong>s. In German,<br />
both Genus and Geschlecht are used for Linguistic <strong>Gender</strong>; <strong>the</strong> term Geschlecht is used to<br />
cover Social and Natural <strong>Gender</strong>. In Modern Greek, <strong>the</strong> term for linguistic gender is yénos<br />
(γένος); social gender is referred to as cinonikó fílo (κοινωνικό φύλο), and natural gender<br />
violoyikó fílo (βιολογικό φύλο) (Angeliki Alvanoudi, p.c.). Contemporary Russian is ra<strong>the</strong>r
12 1 The multifaceted <strong>Gender</strong><br />
striking in that each of <strong>the</strong> three meanings have a distinct one-word term: rod for<br />
Linguistic <strong>Gender</strong>, pol for Natural <strong>Gender</strong>, and <strong>the</strong> recent borrowing from English,<br />
ghender, for Social <strong>Gender</strong>. English has some alternatives to gender as a cover term.<br />
Large systems of noun categorization with more than four terms in Bantu, Australian,<br />
and South American languages are sometimes referred to as ‘noun classes’. An alternative<br />
to Natural <strong>Gender</strong> is ‘sex’,or‘demographic gender’ (Silverstein 2015). Social <strong>Gender</strong> is also<br />
referred to as sociocultural gender.<br />
15. See Saville-Troike (1988) on Koasati and Taylor (1982) on Atsina.<br />
16. As Leonard Bloomfield (1933: 20) put it: ‘The only useful generalizations about language<br />
are inductive generalizations. Features which we think ought to be universal may be absent<br />
from <strong>the</strong> very next language that becomes accessible . . . The fact that some features are, at<br />
any rate, widespread, is worthy of notice and calls for an explanation; when we have<br />
adequate data about many languages, we shall have to return to <strong>the</strong> problem of general<br />
grammar and to explain <strong>the</strong>se similarities and divergences, but this study, when it comes,<br />
will not be speculative but inductive.’
2<br />
Linguistic <strong>Gender</strong> and its expression<br />
Linguistic <strong>Gender</strong>s are grammatical classes of nouns, based on core semantic properties.<br />
These include sex (female and male), animacy, humanness, and also shape and<br />
size (especially for non-humans). We start with general features of Linguistic <strong>Gender</strong><br />
expressed through agreement and anaphoric personal pronouns.<br />
2.1 Linguistic <strong>Gender</strong> in a nutshell<br />
Linguistic <strong>Gender</strong> is one of <strong>the</strong> earliest grammatical categories to be recognized. The<br />
Greek philosopher Protagoras (fifth century BC) was <strong>the</strong> first to use <strong>the</strong> term gender.<br />
In his description of Ancient Greek, he divided Greek nouns into three classes:<br />
‘feminine’, ‘masculine’, and ‘inanimate’ (nowadays called neuter). This typical gender<br />
system is found in many Indo-European languages, including Latin. During historical<br />
change Latin neuter nouns were redistributed between <strong>the</strong> o<strong>the</strong>r two genders,<br />
giving <strong>the</strong> modern system of masculine and feminine in French and Italian.<br />
Traditional grammar has continued using <strong>the</strong> term ‘gender’ for Indo-European<br />
and Semitic languages with a typical division of nouns into masculine and feminine,<br />
or masculine, feminine, and neuter. When Europeans came to study African languages,<br />
<strong>the</strong>y discovered larger gender-like systems with eight or more possibilities in<br />
languages like Swahili. These did not necessarily involve just a masculine–feminine<br />
distinction. The term ‘noun class’ came to be used for larger systems of this kind. 1<br />
Relatively simple two-term gender systems include masculine and feminine, as in<br />
French, Portuguese, Hausa, Hebrew, and many o<strong>the</strong>r Afroasiatic languages. Algonquian<br />
languages of North America have one gender for animates, and ano<strong>the</strong>r one<br />
for inanimates. Tamil and a few o<strong>the</strong>r Dravidian languages divide nouns into two<br />
gender classes. One is labelled ‘rational’ and covers humans, gods, and demons; it<br />
fur<strong>the</strong>r divides into masculine and feminine. The o<strong>the</strong>r one is labelled ‘non-rational’:<br />
it includes all animates and inanimates which do not belong to <strong>the</strong> ‘rational’ class. 2<br />
Latin, German, and Russian have three genders—masculine, feminine, and neuter.<br />
Four genders broadly labelled ‘masculine’, ‘feminine’, ‘edible vegetable’, and ‘all else’<br />
(or ‘neuter’) are a feature of a number of Australian languages, including Dyirbal<br />
(to which we return in §2.2.1).<br />
<strong>How</strong> <strong>Gender</strong> <strong>Shapes</strong> <strong>the</strong> <strong>World</strong>. First edition. Alexandra Y. Aikhenvald.<br />
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.
14 2 Linguistic <strong>Gender</strong> and its expression<br />
Box 2.1 sums up <strong>the</strong> general properties of linguistic genders.<br />
BOX 2.1 General properties of linguistic genders<br />
I. There is a limited, countable number of gender classes.<br />
II. Each noun in <strong>the</strong> language belongs to one (or sometimes more than one) class.<br />
III. There is always some semantic basis to <strong>the</strong> grouping of nouns into gender classes.<br />
Languages vary in how much semantic basis <strong>the</strong>re is. This usually includes animacy,<br />
humanness, and sex, and sometimes also shape, size, and extent.<br />
IV. A constituent outside <strong>the</strong> noun itself must agree in gender with a noun. Agreement can<br />
be with o<strong>the</strong>r words in <strong>the</strong> noun phrase (adjectives, numbers, demonstratives, articles,<br />
etc.) and/or with <strong>the</strong> predicate of <strong>the</strong> clause, or an adverb. <strong>Gender</strong> can be marked in<br />
personal pronouns, reflecting ‘anaphoric’ gender agreement.<br />
2.1.1 Marking Linguistic <strong>Gender</strong><br />
Linguistic <strong>Gender</strong> in a language can be marked on agreeing modifiers and verbs, or<br />
personal and o<strong>the</strong>r pronouns. <strong>Gender</strong> can be marked on <strong>the</strong> noun itself: it is <strong>the</strong>n said<br />
to be marked ‘overtly’. Examples 2.1 and 2.2, from Portuguese, illustrate masculine and<br />
feminine genders. <strong>Gender</strong> is marked on <strong>the</strong> noun and on <strong>the</strong> accompanying definite<br />
article and <strong>the</strong> adjective ‘beautiful’, indicating <strong>the</strong>ir agreement in gender with <strong>the</strong> noun:<br />
2.1 o menin-o lind-o<br />
ARTICLE:MASC.SG child-MASC.SG beautiful-MASC.SG<br />
‘<strong>the</strong> beautiful boy’<br />
2.2 a menin-a lind-a<br />
ARTICLE:FEM.SG child-FEM.SG beautiful-FEM.SG<br />
‘<strong>the</strong> beautiful girl’<br />
Portuguese also has two anaphoric genders in personal pronouns: ele ‘he’, ela ‘she’,<br />
eles ‘<strong>the</strong>y:men’ and elas ‘<strong>the</strong>y:women’. Derivational genders are marked by suffixes<br />
on nouns, e.g. senhor ‘mister’, senhora ‘lady’, patrão ‘owner of (masculine)’, patroa<br />
‘owner of (feminine)’.<br />
<strong>Gender</strong> of a noun may be ‘covert’. Then <strong>the</strong> form of <strong>the</strong> noun tells us nothing<br />
about <strong>the</strong> gender it belongs to. The word Mond ‘moon’ in German is masculine, while<br />
Gabel ‘fork’ is feminine. The form of <strong>the</strong> article <strong>the</strong>y require will indicate <strong>the</strong>ir gender:<br />
we say der Mond ‘<strong>the</strong> (masculine singular) moon’ but die Gabel ‘<strong>the</strong> (feminine<br />
singular) fork’. The gender of a noun is recognizable through <strong>the</strong> agreeing forms.<br />
The form of some German nouns gives us a hint of what gender <strong>the</strong>y belong to: we<br />
will see in §2.2.3 that gender in German is chosen based on a mixture of principles,<br />
semantic and formal.
2.1 Linguistic <strong>Gender</strong> in a nutshell 15<br />
There is no agreement gender in English: 3 <strong>the</strong> same forms of <strong>the</strong> article and <strong>the</strong><br />
adjective are used with ‘boy’ and with ‘girl’ in <strong>the</strong> translations of 2.1 and 2.2. Anaphoric<br />
gender in English features in personal pronouns he for male humans, she for female<br />
humans, and it for non-humans. The actual usage is not fully straightforward: a<br />
non-human—for instance, a beloved cat or dog—can be personified and referred to<br />
as ‘he’ or ‘she’ depending on <strong>the</strong>ir sex. ‘Ships’ and ‘countries’ are referred to as ‘she’<br />
notwithstanding <strong>the</strong>ir being non-humans. In some varieties of English, using ‘he’ or<br />
‘she’ with regard to an inanimate expresses <strong>the</strong> speaker’s attitude—endearment, contempt,<br />
or familiarity. He used to be a pronoun with generic reference. <strong>How</strong>ever, it is no<br />
longer <strong>the</strong> preferred generic pronoun (<strong>the</strong>y is commonly used instead, to avoid what is<br />
perceived as ‘male bias’). This change reflects <strong>the</strong> changing status of male and female<br />
Social <strong>Gender</strong>s (see §§7.5–7, and Chapter 11).<br />
2.1.2 <strong>Gender</strong> agreement and anaphoric gender<br />
English is among numerous languages with just <strong>the</strong> anaphoric gender in personal<br />
pronouns. (We will see, in §6.3, how <strong>the</strong> Old English agreement gender was lost.)<br />
Kaingáng is <strong>the</strong> only Jê language (of sou<strong>the</strong>rn Brazil) to have masculine and feminine<br />
distinction in third person pronoun. Many Carib languages in South America<br />
(including Trio, Apalaí, and Hixkaryana) have animate and inanimate forms of<br />
third person singular pronoun. In Finnish, <strong>the</strong> third person pronoun hän ‘he, she’<br />
can only be used for human referents (<strong>the</strong> proximate demonstrative se is used for<br />
non-humans). 4<br />
Anaphoric gender tends to be found in languages with agreement gender, including<br />
German, French, Portuguese, and Russian. But this is not uniformly <strong>the</strong> case.<br />
Maithili, an Indo-Aryan language from Nepal and nor<strong>the</strong>rn India, has four honorific<br />
forms in its personal pronouns, but no gender specification. Masculine and feminine<br />
genders are expressed through agreement on verbs and on definite adjectives. In<br />
Abui, a Papuan language from eastern Indonesia, gender is marked in object prefixes<br />
on verbs but not in personal pronouns. 5<br />
In many languages, animate nouns tend to agree with modifiers based on <strong>the</strong>ir<br />
meaning ra<strong>the</strong>r than on <strong>the</strong>ir formal allegiance to a gender. In Swahili ki-faru (gender<br />
7-rhinoceros) belongs to <strong>the</strong> ‘inanimate’ class 7. But <strong>the</strong> noun can occur with an adjective<br />
marked for gender 1 (which covers animates): ki-faru m-kubwa ‘a big rhinoceros’.This<br />
is how semantic agreement overrides <strong>the</strong> expected agreement based just on form (also<br />
known as ‘syntactic’ (or ‘mechanical’) agreement). 6<br />
Anaphoric agreement in a language with complex rules of gender choice may show<br />
more straightforward correlations with meaning than agreement with an article or an<br />
adjective in a noun phrase. The word for ‘girl’ in German, das Mädchen, belongs to<br />
<strong>the</strong> neuter gender on morphological grounds: as we will see in §2.2.2, <strong>the</strong> diminutive<br />
suffix -chen it contains is always associated with <strong>the</strong> neuter. We can see that in 2.3 ‘<strong>the</strong>
16 2 Linguistic <strong>Gender</strong> and its expression<br />
(neuter) girl’ requires neuter agreement of <strong>the</strong> article and <strong>the</strong> adjective. <strong>How</strong>ever, in<br />
numerous instances a girl will be referred to with <strong>the</strong> feminine pronoun sie ‘she’<br />
ra<strong>the</strong>r than es ‘it’. In this case, anaphoric agreement follows <strong>the</strong> meaning of <strong>the</strong> noun<br />
ra<strong>the</strong>r than its form. This is known as semantic agreement. The following example<br />
comes from Heinrich Mann’s novel Der Untertan (published in 1914):<br />
2.3 Das junge Mädchen (neuter) gab ihm die Hand. Sie (fem) wollte wohl nett sein,<br />
aber was war mit ihr (fem) anzufangen.<br />
‘The young girl gave him her hand. Apparently, she intended to be nice, but<br />
what could be done with her’ (p. 14).<br />
The girl is introduced with a definite article, in <strong>the</strong> neuter form (reflecting <strong>the</strong><br />
neuter agreement gender of <strong>the</strong> noun itself). She is <strong>the</strong>n referred to with <strong>the</strong> feminine<br />
anaphoric pronoun as ‘she’. In a recent study, Braun and Haig (2010) showed that<br />
speakers tend to favour <strong>the</strong> feminine pronoun sie (and not <strong>the</strong> neuter pronoun es) for<br />
older girls (18 years of age), and neuter pronouns for 2-to12-year-olds who have not<br />
yet attained <strong>the</strong> status of ‘women’. 7<br />
2.1.3 Linguistic <strong>Gender</strong> and o<strong>the</strong>r linguistic categories<br />
<strong>Gender</strong>s can be marked in a variety of ways—with suffixes or prefixes, apophony (or<br />
vowel changes), tone patterns, and change of stress. Linguistic <strong>Gender</strong>s are a feature<br />
of languages with complex morphology, and a fusional or agglutinating (but not an<br />
isolating) profile. <strong>Gender</strong> and number can be combined into a single affix. In Bantu<br />
languages, gender agreement markers come in singular/plural pairs. They appear on<br />
modifiers of different kinds and on <strong>the</strong> predicate. In 2.4, from Swahili, <strong>the</strong> noun class<br />
marker, ki-, also appears on <strong>the</strong> noun itself. That is, noun classes are overtly marked<br />
(Welmers 1973: 171). Ki- marks <strong>the</strong> singular term of noun class 7/8 (see Table 2.2)<br />
which covers inanimates. The plural counterpart of ki- will be vi-.<br />
2.4 ki-kapu ki-kubwa ki-moja ki-li-anguka<br />
NCL7:INANIM-basket<br />
‘One large basket fell’<br />
NCL7:INANIM-large NCL7:INANIM-one NCL7:INANIM-PAST-fall<br />
There may be fewer gender distinctions in plural than in singular. In German and<br />
Russian, genders are not distinguished in plural number. 8 Most frequently, anaphoric<br />
genders are distinguished in third person only: this is what we find in familiar Indo-<br />
European languages. Berber, Semitic, Cushitic, and many Chadic languages have<br />
masculine and feminine forms of second person and third person pronouns, both<br />
singular and plural. Manambu and o<strong>the</strong>r languages from <strong>the</strong> Ndu family in <strong>the</strong> Sepik<br />
area of New Guinea have different forms for masculine and feminine genders in<br />
second and third person singular pronouns. Gala, from <strong>the</strong> same family, has two<br />
genders in first, second, and third persons (singular)—as we will see in Table 7.1. 9
2.1 Linguistic <strong>Gender</strong> in a nutshell 17<br />
Linguistic <strong>Gender</strong> interacts with definiteness in Maithili. Definite forms of adjectives<br />
modifying a noun agree with it in gender, e.g. mot̍-ka mərəd ‘<strong>the</strong> fat man’, and<br />
mot̍-ki jənana ‘<strong>the</strong> fat woman’. Indefinite forms do not, e.g. pəigh mərəd ‘a tall man’,<br />
pəigh məugi ‘a tall woman’. 10<br />
2.1.4 <strong>How</strong> many genders?<br />
A language may distinguish different sets of gender forms in different contexts.<br />
Romanian, a Romance language, distinguishes two gender forms in <strong>the</strong> singular,<br />
and two gender forms in <strong>the</strong> plural. We find three combinations:<br />
(I) Nouns which take -Ø in <strong>the</strong> singular and -i in <strong>the</strong> plural (marked on an<br />
agreeing adjective and o<strong>the</strong>r modifiers).<br />
The core of this group of nouns are men. This is why this form is often called<br />
‘masculine gender’. Examples are un prieten bun (INDEFINITE.ARTICLE:masc.sg friend:<br />
masc.sg good:masc.sg) ‘a good male friend’; and prieten-i bun-i (friend-masc.pl goodmasc.pl)<br />
‘good male friends’.<br />
(II) Nouns which take -ă in <strong>the</strong> singular and -e in <strong>the</strong> plural.<br />
The core of this group of nouns are women. This is why this form is described as<br />
‘feminine gender’. Examples are un-ă prieten-ă bun-ă (INDEFINITE.ARTICLE-fem.sg<br />
friend-fem.sg good-fem.sg) ‘a good female friend’; prieten-e bun-e (friend-fem.pl<br />
good-fem.pl) ‘good female friends’.<br />
(III) Nouns which take -Ø in <strong>the</strong> singular and -e in <strong>the</strong> plural on adjectives.<br />
This group of nouns includes many inanimates which is why some scholars call<br />
this ‘neuter’. Examples are un stilou bun (INDEFINITE.ARTICLE.masc.sg pencil good:<br />
masc.sg) ‘a good pencil’, stilouri bun-e (pencil:pl good-fem.pl) ‘good pencils’.<br />
The meanings of <strong>the</strong> three gender combinations are not entirely transparent. Not<br />
all inanimates belong to <strong>the</strong> ‘neuter’ gender (group III). Table 2.1 summarizes <strong>the</strong><br />
marking of singular and plural feminine, masculine, and neuter forms. 11<br />
TABLE 2.1. <strong>Gender</strong>s in Romanian and <strong>the</strong>ir marking<br />
Markers of singular<br />
number<br />
Distribution in gender<br />
groups of nouns<br />
-a feminine<br />
-e<br />
neuter<br />
-Ø<br />
masculine -i<br />
Markers of plural<br />
number
18 2 Linguistic <strong>Gender</strong> and its expression<br />
Alternatively, a language may have two coexisting subsystems of genders used in<br />
different contexts: one for agreeing pronouns (including demonstratives or verbal<br />
cross-referencing) and one for adjectives and o<strong>the</strong>r modifiers. Gaaguju, an Australian<br />
language, has a system of four genders used on adjectives as modifiers—one for<br />
human males and some animates, one for human females and o<strong>the</strong>r animates, one<br />
for plants and <strong>the</strong>ir parts, and for weapons, and a fur<strong>the</strong>r one for abstract entities,<br />
body parts, and fire. Just two classes—animate and inanimate—are marked on verbs.<br />
Palikur, an Arawak language from Brazil and French Guyana, distinguishes three<br />
genders in demonstratives and personal pronouns. Agreement markers on verbs and<br />
adjectives have two gender forms: one for feminine and one for masculine/neuter.<br />
Languages with several gender systems employed in different environments are<br />
known as ‘split-gender’ systems. 12<br />
2.2 <strong>How</strong> to choose a Linguistic <strong>Gender</strong><br />
Linguistic <strong>Gender</strong> always has some semantic basis, although not everything can be<br />
explained in semantic terms. No matter how complicated <strong>the</strong> overall system of Linguistic<br />
<strong>Gender</strong> assignment may be, its semantic core will always include reference to <strong>the</strong><br />
universal parameters of sex, humanness, and animacy. Non-humans and inanimates<br />
can be assigned to gender classes based on fur<strong>the</strong>r features—including size and shape<br />
(see Chapter 3). Morphological and phonological form can also be at play.<br />
2.2.1 <strong>Gender</strong> choice by meaning<br />
Linguistic <strong>Gender</strong> can be assigned to a noun based entirely on its meaning. The<br />
choice can be fairly transparent. Malto, Kolami, Ollari, and Parji, also from <strong>the</strong><br />
Dravidian family, have one Linguistic <strong>Gender</strong> for human males and <strong>the</strong> o<strong>the</strong>r<br />
one for <strong>the</strong> rest. We can recall from §2.1 that all nouns in Tamil divide into two<br />
genders labelled ‘rational’ and ‘non-rational’. Rational nouns include humans,<br />
gods, and demons, with <strong>the</strong> remainder being ‘non-rational’. Diyari, an Australian<br />
language, and Kaingang, a Jê language from sou<strong>the</strong>rn Brazil, divide all nouns into<br />
female humans and <strong>the</strong> rest. Nouns in Siouan languages from North America, Carib<br />
languages of South America, and a few Austronesian languages spoken in eastern<br />
Indonesia divide into animate and inanimate gender classes.<br />
Sex, humanness, animacy, and rationality may combine in one gender system.<br />
Zande and Ma (Ubangi languages from Central Africa) have four genders—feminine,<br />
masculine, non-human animate, and inanimate. Godoberi, a north-east Caucasian<br />
language, has three genders—feminine, masculine, and ‘non-rational’. 13<br />
<strong>Gender</strong> can be chosen based on somewhat different principles for different groups<br />
of nouns. Tunica, an isolate formerly spoken in Louisiana, had two genders—<br />
masculine and feminine. <strong>Gender</strong> was marked in definite articles and in pronouns.
2.2 <strong>How</strong> to choose a Linguistic <strong>Gender</strong> 19<br />
The gender of humans, birds, and most animals was assigned based on <strong>the</strong> sex:<br />
females were feminine, and males masculine. Fish, shellfish, smaller reptiles, and<br />
insects were masculine. Trees, fruits, vegetables, and tools were also masculine.<br />
Inanimate feminines included terms for substances viewed as masses (such as sand,<br />
salt, metal, cotton). Buildings and <strong>the</strong>ir parts were feminine, and so were abstract<br />
nouns and nouns referring to processes, such as dance, war, noise, and sickness. 14<br />
Systems with semantic gender assignment vary in <strong>the</strong>ir transparency. <strong>Gender</strong><br />
choice in Dyirbal 15 is semantically based, but not straightforward. Four genders are<br />
expressed through article-like noun markers (but not on nouns <strong>the</strong>mselves). Three of<br />
<strong>the</strong>m are associated with one or more concepts:<br />
gender 1 (noun marker bayi)—male humans and non-human animates;<br />
gender 2 (noun marker balan)—female humans; fire; drinkable liquids; fighting;<br />
gender 3 (noun marker balam)—non-flesh food;<br />
gender 4 (noun marker bala)—a residue gender covering everything else, including<br />
body and o<strong>the</strong>r parts, place names, and flesh food (meat and fish).<br />
All animates are distributed between gender 1 and gender 2 (except bees which are<br />
in gender 4). Three general principles determine gender membership of a noun:<br />
I. If a noun has a characteristic X (on <strong>the</strong> basis of which its gender will be chosen)<br />
but is associated with characteristic Y through BELIEF OR LEGEND it will be <strong>the</strong>n assigned<br />
to a different gender based on characteristic Y. This is a principle of MYTHOLOGICAL<br />
ASSOCIATION, orMYTH-AND-BELIEF PRINCIPLE.<br />
It is believed that birds are, as a class, <strong>the</strong> spirits of dead women. Birds are classed<br />
as members of gender 2 (‘feminine’, balan) ra<strong>the</strong>r than bayi on <strong>the</strong> basis of <strong>the</strong>ir nonhuman<br />
animate status. There are exceptions to this, also based on beliefs. Willy<br />
wagtail belongs to gender I (masculine), bayi jigirrjigirr, since he is believed to be <strong>the</strong><br />
metamorphosis of a legendary man (and <strong>the</strong> way <strong>the</strong> bird wiggles its tail is reminiscent<br />
of how men dance in traditional dances, corroborees). Non-edible snakes are<br />
members of gender I (bayi). An exception is balan bima ‘death adder’ who is also a<br />
legendary woman, and thus belongs to gender 2. The sun was believed to be a<br />
woman, and so belongs to gender 2 (balan garri). The moon is her husband, and<br />
so <strong>the</strong> noun belongs to gender 1 (bayi gagara). Thunder, lightning, and rain are<br />
believed to be legendary males, and are included in gender 1.<br />
II. If <strong>the</strong> referent of a noun with a characteristic X is perceived to have a PHYSICAL<br />
ASSOCIATION with <strong>the</strong> characteristic Y, <strong>the</strong>n this may be reflected in <strong>the</strong> gender choice<br />
for this noun. The association can be based on physical association. Terms for fruit<br />
and vegetable food belong to gender 3; <strong>the</strong>y are produced by plants which belong to<br />
gender 4. Honey is balam girñjal (gender 3), and is produced by bees. Because of this<br />
analogy, bees belong to gender 4—<strong>the</strong>y are <strong>the</strong> only animates not in genders 1 or 2.<br />
Physical similarity may be ano<strong>the</strong>r reason for gender choice. Most insects belong to
20 2 Linguistic <strong>Gender</strong> and its expression<br />
gender 1 (bayi). But <strong>the</strong> firefly, balan yugiyam, belongs to gender 2 because <strong>the</strong><br />
flashes of light it emits are similar to sparks from a fire.<br />
Or <strong>the</strong> referent with a characteristic X can be used to make or catch Y. For<br />
instance, matches and a match box used to hold <strong>the</strong>m belong to gender 2 (balan)<br />
since <strong>the</strong>y produce fire which belongs to <strong>the</strong> same gender.<br />
III. If a set of nouns belongs to a certain gender, and members of its subset share a<br />
particular IMPORTANT PROPERTY, <strong>the</strong>n this subset may be assigned to a different<br />
gender. Most trees without edible parts belong to gender 4, but stinging trees are<br />
placed in gender 2, due to <strong>the</strong>ir harmful nature. Most fishes belong to gender 1, as<br />
non-human animates. Three species of fish belong to gender 2, since <strong>the</strong>y are harmful<br />
to humans. The jellyfish, balan jawayi, has tentacles that can inflict a painful sting.<br />
The freshwater stonefish, balan jaŋgan, has venomous spines which may give painful<br />
wounds if stepped on. And eating balan juruŋun, <strong>the</strong> toad fish, can be fatal. Pythons,<br />
<strong>the</strong> only snakes that are eaten, are balan.<br />
These principles of gender choice highlight what <strong>the</strong>re is in <strong>the</strong> legends and what is<br />
important for survival. They are living proof of functionality of genders.<br />
Mythological association, beliefs, and legends play a role in gender in quite a few<br />
languages. In <strong>the</strong> Western Torres Strait language, all nouns denoting males are<br />
masculine, and all <strong>the</strong> rest are feminine. The moon is masculine, due to mythological<br />
association with masculinity. In Tunica, thunder and lightning were personified as<br />
mythical men, and were masculine for this reason. 16 We return to this in §8.1.<br />
The meaning of a small gender system may be transparent to a limited extent.<br />
Nouns in Fox and o<strong>the</strong>r Algonquian languages divide into animate and inanimate<br />
genders. 17 <strong>Gender</strong> is marked on <strong>the</strong> noun, and on demonstratives and verbs. Those<br />
nouns which denote humans and biologically animate beings belong to <strong>the</strong> animate<br />
gender, e.g. neniwa ‘man’, ihkwewa ‘woman’, mahkwa ‘bear’, and maneto . wa ‘spirit,<br />
snake’. Nouns denoting entities that move—including spirits and heavenly bodies—<br />
are also grammatically animate. The noun ni . c᷈a . pa ‘doll’ is animate, since a doll<br />
represents a human.<br />
A number of nouns which refer to inanimates belong to <strong>the</strong> animate gender. Some<br />
names of plants are inanimate, e.g. ahte . himini ‘strawberry’, and o<strong>the</strong>rs are animate,<br />
e.g. wi . tawi . ha ‘raspberry’. The ‘Implacable Raspberry’ in <strong>the</strong> title of a paper by Straus<br />
and Brightman (1982) sums up <strong>the</strong> idea of unpredictability of Algonquian gender<br />
when it comes to entities o<strong>the</strong>r than human beings or animals.<br />
A few animate–inanimate pairs show regular differences in <strong>the</strong>ir meanings. In Fox,<br />
‘grandfa<strong>the</strong>r’ is animate; but a ceremonial pole referred to as ‘our grandfa<strong>the</strong>r’ is<br />
inanimate. A noun with general or collective reference may be inanimate, and individual<br />
items animate. In Cree, ano<strong>the</strong>r Algonquian language, mistik ‘tree’ is animate,<br />
and mistik ‘stick’ is inanimate. In Fox s᷈o˞niya . hi ‘silver, money’ is inanimate, and<br />
s᷈o˞niya . ha ‘a coin, a bill’ is animate; owiya . he . hi ‘animals, small game (collective)’ is<br />
inanimate; and owiye . he . ha ‘a (small) animal’ is animate.
2.2 <strong>How</strong> to choose a Linguistic <strong>Gender</strong> 21<br />
The animate gender is associated with spiritual power and agency. The noun<br />
c᷈i . paya ‘corpse, ghost’ is animate: it is conceived as a former human being, and in<br />
its meaning ‘ghost’ it denotes a spiritually powerful being. The principle of ‘power’<br />
in <strong>the</strong> assignment of nouns to <strong>the</strong> animate class can be illustrated with gender shifts<br />
in traditional narratives. A stone (aseny-) is generally inanimate. But in a traditional<br />
story a man addresses a prayer to a stone which takes an animate agreement marker<br />
on <strong>the</strong> verb. 18 Fur<strong>the</strong>r semantic functions of a difficult-to-capture gender choice are<br />
revealed through variable gender in discourse—we return to this in §4.2.<br />
Noun classes in Bantu languages are a well-known example of a semantically<br />
opaque system. Table 2.2 summarizes a basic semantic grid common to Bantu noun<br />
class systems based on <strong>the</strong> interaction of shape, size, and humanness. These parameters<br />
provide only a partial semantic motivation for <strong>the</strong> noun classes in individual<br />
Bantu languages. 19<br />
<strong>Gender</strong> assignment in modern Bantu languages is to some extent semantically<br />
opaque, though its semantic nucleus is still discernible. Thus, in Babungo, class 1/2 is<br />
basically human; however, it is a much bigger class than it was in Proto-Bantu, and<br />
also contains many animals, some birds and insects, body parts, plants, and household<br />
and o<strong>the</strong>r objects, e.g. necklace, pot, book, rainbow. Shape and size also appear<br />
as semantic parameters: in ChiBemba, class 7/8 is associated with large size and<br />
carries pejorative overtones, while class 12/13 includes small objects and has overtones<br />
of endearment. 20<br />
In Jingulu, an Australian language, nouns divide into four genders, which are only<br />
partially semantically transparent. The ‘vegetable’ class mostly includes objects which<br />
are long, thin, or pointed; this class happens to include most vegetables, as well as<br />
body parts such as <strong>the</strong> colon, penis, and neck, instruments such as spears, fire-drills,<br />
and barbed wire, natural phenomena such as lightning and rainbows, and also roads<br />
and trenches. The ‘feminine’ class includes female humans and higher animates, and<br />
TABLE 2.2. <strong>Gender</strong>–number pairs in Bantu<br />
GENDER CLASS<br />
SEMANTICS<br />
1/2 humans, a few o<strong>the</strong>r animates<br />
3/4 plants, plant parts, foods, non-paired body parts, miscellaneous<br />
5/6 fruits, paired body parts, miscellaneous inanimates<br />
7/8 miscellaneous inanimates<br />
9/10 animals, miscellaneous inanimates, a few humans<br />
11/10 long objects, abstract entities, miscellaneous inanimates<br />
12/13 small objects, birds<br />
6 masses<br />
14 abstract qualities, states, masses, collectives<br />
15 infinitives
22 2 Linguistic <strong>Gender</strong> and its expression<br />
also words for axes, <strong>the</strong> sun, and most smaller songbirds. The semantic content of <strong>the</strong><br />
remaining two classes, ‘masculine’ and ‘neuter’, is harder to define: masculine is<br />
mostly used for <strong>the</strong> rest of animates and neuter for <strong>the</strong> rest of inanimates, except that<br />
flat and/or rounded inanimates (most trees and eggs, and body parts such as <strong>the</strong> liver<br />
and eyebrow) are masculine. 21<br />
A language can be thought of as a long-term repository of speakers’ culture and<br />
lore, and it is unlikely that every detail could be explained. There is no synchronic<br />
reason why ‘raspberry’ should be animate throughout Algonquian languages—any<br />
mythological association which may have existed in <strong>the</strong> past has since been lost.<br />
The great majority of gender choices in Dyirbal can be explained in terms of general<br />
principles and rules of gender transfer and association. But, as Dixon (2015: 42) puts<br />
it, ‘not all. I know of no reason why balan (<strong>Gender</strong> II) is used for prawns and crabs,<br />
for <strong>the</strong> two egg-laying mammals (platypus and echidna), and for <strong>the</strong> dingo or native<br />
dog. (There are a number of o<strong>the</strong>r individual eccentricities.)’<br />
Most languages have such ‘eccentricities’—an unpredictable residue of semantically<br />
unexplainable gender choice. In Dyirbal, ‘<strong>the</strong>re may have been fur<strong>the</strong>r legends<br />
and beliefs, which I [Dixon] did not become familiar with, that would help explain<br />
more about gender assignment. And <strong>the</strong>re could be legends which had been forgotten<br />
but left in <strong>the</strong>ir wake some special gender specification. As in any judicious enquiry<br />
into <strong>the</strong> semantic basis of a grammatical system of genders, a lot can be explained—<br />
but not everything’ (Dixon 2015: 42–3).<br />
2.2.2 <strong>Gender</strong> choice by form<br />
The form of a noun may play a role in Linguistic <strong>Gender</strong> choice. Numerous<br />
derivational affixes in German are each associated with one gender. Nouns containing<br />
<strong>the</strong> suffix -ung ‘action noun’ are feminine, e.g. Bedeutung ‘meaning’, Bewegung<br />
‘movement’. Nouns which contain a diminutive suffix -lein or -chen are always<br />
neuter. The noun Mäd-chen ‘girl’ belongs to <strong>the</strong> neuter gender because it contains<br />
<strong>the</strong> neuter suffix. Similarly, korítsi ‘girl’ in Modern Greek is neuter thanks to <strong>the</strong><br />
neuter suffix -(í)tsi it contains. The augmentative form korítsaros ‘pretty or buxom<br />
girl’ belongs to masculine gender (despite its female referent) because nouns with <strong>the</strong><br />
augmentative suffix -aros are always masculine.<br />
Phonological form of a noun may correlate with Linguistic <strong>Gender</strong>. In Qafar, an<br />
East Cushitic language, nouns with inanimate referent whose citation form ends in a<br />
vowel are feminine, all <strong>the</strong> rest are masculine. In Hausa, all non-sex differentiable<br />
nouns which end in ‐aa are feminine. In Katcha (Kordofanian), any noun beginning<br />
with m- belongs to <strong>the</strong> feminine gender, unless it is ‘notionally masculine’. In<br />
Punjabi, an Indo-Aryan language, ‘<strong>the</strong> rule of thumb is that inflected nouns ending<br />
in -aa are usually assigned masculine gender whereas <strong>the</strong> nouns ending in -ii are<br />
feminine’. 22
2.2 <strong>How</strong> to choose a Linguistic <strong>Gender</strong> 23<br />
Phonological principles of gender choice help determine gender assignment of<br />
loans from a French-based creole, French, and Portuguese in Palikur. The gender of<br />
<strong>the</strong> source language is usually overridden by <strong>the</strong>se phonological principles. Masculine<br />
human nouns usually end in ‐e or -i. Feminine nouns end in -u or -o. Neuter nouns<br />
can end in any vowel or consonant. Loan nouns which end in -o or -u are usually<br />
feminine in gender, e.g. marto ‘hammer’ (from French Creole marto), sitru ‘lemon’<br />
(from French Creole citrõ), tattu ‘armadillo’ (variant of tat; from Portuguese masculine<br />
tatu). 23 Nouns which end in a consonant are usually given neuter gender, e.g.<br />
simis ‘shirt’ (from French chemise).<br />
Morphological and phonological features may help determine <strong>the</strong> Linguistic<br />
<strong>Gender</strong>, but are never <strong>the</strong> only principles for its choice. We now turn to <strong>the</strong> ways<br />
in which form and meaning work toge<strong>the</strong>r in Linguistic <strong>Gender</strong> assignment.<br />
2.2.3 Meaning meets form: mixed principles of gender assignment<br />
Not infrequently, Linguistic <strong>Gender</strong> is chosen based on a combination of meaningbased<br />
and form-based principles. The gender of humans is based on semantics.<br />
O<strong>the</strong>r, formal principles, are at work for <strong>the</strong> rest. Most Albanian varieties have two<br />
genders, masculine and feminine. Nouns denoting females are feminine, and those<br />
denoting males are masculine. Inanimate nouns ending in a consonant (e.g. gjak<br />
‘blood’ or dhëmb ‘tooth’) or a stressed vowel are assigned to <strong>the</strong> masculine Linguistic<br />
<strong>Gender</strong>. Nouns ending in an unstressed vowel tend to be feminine. The semantic<br />
principle overrides <strong>the</strong> formal one: rojë ‘guard’ is masculine since it denotes a male<br />
(although it ends in an unstressed ë typical of feminines). 24<br />
As we saw in §2.2.2, in Greek, ‘to a certain extent, grammatical gender can be<br />
considered to be a property of stem-formatives involved in <strong>the</strong> formation of particular<br />
nouns. Thus, diminutives in -(í)tsi are always neuter, augmentatives in -aros are<br />
always masculine, abstract nouns in -ja are always feminine, and so on’ (Joseph and<br />
Philippaki-Warburton 1987: 152–3).<br />
Semantic and phonological principles underlie <strong>the</strong> assignment of Linguistic <strong>Gender</strong>s<br />
in a number of Papuan languages of New Guinea. Yimas, from <strong>the</strong> Lower Sepik<br />
family in Papua New Guinea, has eleven Linguistic <strong>Gender</strong>s. Four are chosen by <strong>the</strong><br />
meaning of <strong>the</strong> nouns; <strong>the</strong>se cover (i) human males; (ii) human females; (iii) animals;<br />
and (iv) culturally important items. The rest are phonologically motivated: <strong>the</strong> agreeing<br />
constituent repeats <strong>the</strong> last consonant of <strong>the</strong> noun root. Arapesh languages of <strong>the</strong><br />
East Sepik Province in Papua New Guinea work on a similar principle: <strong>the</strong>y have a<br />
noun class for male humans, and one for female humans. Nouns of o<strong>the</strong>r semantic<br />
groups trigger similar ‘alliterative’ or phonologically based, agreement. This is always<br />
overridden by semantics; so, loanwords with human referents in Arapesh languages<br />
are assigned to classes according to whe<strong>the</strong>r <strong>the</strong>y refer to a man or to a woman. Most<br />
nouns which belong to <strong>the</strong> masculine gender contain a final or an initial segment n,
24 2 Linguistic <strong>Gender</strong> and its expression<br />
e.g. aleman ‘man’ or Nekitel ‘personal name’ (Nekitel 1986, and p.c.). The English<br />
loanword nes ‘nurse’ is treated as a member of <strong>the</strong> feminine class (whose members<br />
are marked with ʔ or kw-). Words such as ankelo ‘angel’ and Spiritu Santu ‘Holy<br />
Spirit’ are conceived of as males, and so treated as masculine. That is, meaning-based<br />
choice may override <strong>the</strong> choice based on formal grounds. 25<br />
It has been often stated that <strong>the</strong>re is no real semantic basis for gender assignment<br />
in some well-known Indo-European languages. In a seminal study, Zubin and<br />
Köpcke (1986) provided a semantic rationale for gender assignment of nouns of<br />
different semantic groups in German. In agreement with <strong>the</strong> natural sex principle,<br />
masculine and feminine genders mark <strong>the</strong> terms for male and female adults of each<br />
species of domestic and game animals, and neuter is assigned to non-sex-specific<br />
generic and juvenile terms. Superordinate terms are often neuter. Masculine gender<br />
is used for types of cloth, types of precipitation and wind, and types of minerals.<br />
Types of knowledge (e.g. Kenntnis ‘knowledge’) and disciplines (e.g. Linguistik<br />
‘linguistics’) have feminine gender, and games and types of metal have neuter<br />
gender. Almost all nouns with generic reference, such as das Ding ‘thing’, das<br />
Gerät ‘implement, apparatus’, das Gut ‘goods’, have neuter gender (with one<br />
exception: die Sache ‘thing’ is feminine). Superordinate terms usually belong to<br />
neuter gender, and items of a more basic level are feminine or masculine, more<br />
rarely neuter. Table 2.3 illustrates <strong>the</strong> neuter gender of superordinate nouns and <strong>the</strong><br />
non-neuter (masculine and feminine) of nouns referring to ‘basic level’ objects (see<br />
Zubin and Köpcke 1986: 147ff.).<br />
A combination of meaning-based and form-based principles accounts for gender<br />
choice in <strong>the</strong> majority of cases. But <strong>the</strong>re is always a small residue of exceptions which<br />
cannot be explained. For example, two superordinate terms are feminine: Pflanze<br />
‘plant, herb’ and Farbe ‘colour’, going against <strong>the</strong> principle in Table 2.3.<br />
Morphological and semantic principles interact in German gender assignment for<br />
some groups of nouns. Bird names have masculine gender unless <strong>the</strong>y end in a<br />
derivational suffix which is feminine. Sailing vessels are assigned to <strong>the</strong> feminine<br />
gender unless <strong>the</strong>y end in a masculine derivational suffix (Zubin and Köpcke 1986:<br />
TABLE 2.3. Semantic basis of gender choice in German: an illustration<br />
SUPERORDINATE<br />
Instrument (n) ‘musical instrument’<br />
Obst (n) ‘fruit’<br />
Spielzeug (n) ‘toy’<br />
Land (n) ‘land’<br />
BASIC LEVEL<br />
Guitarre (f) ‘guitar’, Trompete (f) ‘trumpet’<br />
Apfel (m) ‘apple’, Pflaume (f) ‘plum’<br />
Bauklotz (m) ‘block’, Puppe (f) ‘doll’<br />
Wald (m) ‘woods’, Sumpf (m) ‘swamp’,<br />
Wiese (f) ‘meadow’
2.3 Markedness and Linguistic <strong>Gender</strong>s 25<br />
175). The loanword der Drink ‘alcoholic drink’ is masculine, and its gender assignment<br />
is based on a complex of semantic and morphological and phonological<br />
features (Zubin and Köpcke 1984: 44). These are:<br />
(i) Phonological: Consonant clusters in initial and in final position; initial cluster<br />
dr-; final nasal (typical for masculine nouns)<br />
(ii) Morphological: -s ‘plural’ (characteristic of masculine or of neuter gender)<br />
(iii) Semantic: class of drinkable liquids.<br />
Semantic and formal principles can compete with each o<strong>the</strong>r. <strong>Gender</strong> of compounds<br />
is normally determined by <strong>the</strong> last component. Streich-holz (strike-wood)<br />
‘match’ is neuter because <strong>the</strong> last part, Holz, is neuter. But in north Germany<br />
Streichholz is masculine because of its association with ‘stick-like objects’ which are<br />
typically masculine (cf. Stock (masculine) ‘stick’). As expected, most compounds<br />
containing <strong>the</strong> masculine noun Mut ‘courage, valour, spirit’ are masculine. These<br />
include Freimut ‘frankness’, Heldenmut ‘heroism’, Hochmut ‘arrogance’, Lebensmut<br />
‘exhilaration’, Unmut ‘bad temper’, and Übermut ‘bravado’. But some are feminine;<br />
for instance Anmut ‘gracefulness’, Demut ‘humility’, Langmut ‘patience’, and Schwermut<br />
‘melancholy’. An experimental study of -mut compounds (by Zubin and Köpcke<br />
1984) showed that <strong>the</strong> -mut compounds assigned to <strong>the</strong> feminine gender were<br />
associated with introverted feelings, while masculine gender was associated with<br />
extroversion. The same principle appears to hold for nouns ending in <strong>the</strong> suffix<br />
-sal. These are normally neuter, such as Schicksal ‘fate, destiny’. Those that are<br />
feminine describe introverted emotional states, such as Trübsal ‘misery’.<br />
Mixed principles of Linguistic <strong>Gender</strong> choice account for synchronically opaque<br />
gender assignment which may seem close to arbitrary. The complexity, and semantic<br />
opacity, in gender choice creates an impediment for second-language learners of<br />
many European languages. Language learners—linguists included—complain that<br />
Linguistic <strong>Gender</strong>s are arbitrary and redundant and are nothing o<strong>the</strong>r than unnecessary<br />
obstacles for <strong>the</strong> uninitiated. As we will see in Chapter 4, this is far from being<br />
<strong>the</strong> case: Linguistic <strong>Gender</strong>s are a useful and versatile device which helps organize<br />
discourse, express subtle meanings, and reflect <strong>the</strong> status of Social <strong>Gender</strong>s.<br />
2.3 Markedness and Linguistic <strong>Gender</strong>s<br />
Linguistic <strong>Gender</strong>s within a language may not have <strong>the</strong> same status. There is a<br />
fundamental distinction between two kinds of markedness in language—formal<br />
and functional. A formally unmarked term will be <strong>the</strong> only one in its system to<br />
have zero realization (or a zero variant). The Spanish masculine singular pronoun el<br />
is formally unmarked for gender. Its feminine counterpart, ella, is formally marked.<br />
In Portuguese, ele ‘he’ and ela ‘she’ are equally formally marked.
26 2 Linguistic <strong>Gender</strong> and its expression<br />
Functional markedness is different in nature. It relates to how <strong>the</strong> forms are used<br />
(and not what <strong>the</strong>y look like). The marked term(s) may be used each in a restricted,<br />
specifiable situation, with <strong>the</strong> unmarked term being used in all o<strong>the</strong>r circumstances.<br />
The unmarked category appears in neutralized contexts, or when one does not wish<br />
to be specific. In Portuguese, <strong>the</strong> choice of a masculine or a feminine pronoun is<br />
made when <strong>the</strong> sex of a human is known and unambiguous. But when <strong>the</strong>re is a mix<br />
of people of different sexes <strong>the</strong> masculine form is used. One distinguishes o filho<br />
‘son’ and a filha ‘daughter’ in <strong>the</strong> singular. The plural ‘children’ will be referred to<br />
as os filhos (lit. <strong>the</strong> sons). The corresponding feminine form, as filhas, willunambiguously<br />
cover just ‘daughters’. Similarly, a irmã is ‘sister’ and oirmão‘bro<strong>the</strong>r’. If<br />
I ask someone how many siblings <strong>the</strong>y have, I will refer to ‘bro<strong>the</strong>rs and sisters’ as<br />
irmãos, literally, bro<strong>the</strong>rs. That is, <strong>the</strong> masculine—not <strong>the</strong> feminine—form can be<br />
considered a functionally unmarked choice.<br />
In many familiar Indo-European languages <strong>the</strong> masculine form is used to refer to a<br />
group of mixed sexes, to generic referents, or to those whose sex is unknown. In<br />
Spanish, ellos ‘<strong>the</strong>y (masculine)’ refers to a group consisting only of males and also a<br />
group of mixed sex. The form ellas ‘<strong>the</strong>y (feminine)’ refers exclusively to a group<br />
consisting of females. Similarly, <strong>the</strong> masculine plural form les américains in French<br />
can refer to a group of males or to a group of mixed sexes; its feminine counterpart,<br />
les américaines, refers only to a group of female Americans. In Russian, question<br />
words and indefinite pronouns require masculine singular agreement. I will say Kto<br />
pris᷈el (who come+PAST.masc.sg) ‘Who came?’ even when asking about a woman. In<br />
this case, <strong>the</strong> masculine gender is functionally unmarked. Until recently, <strong>the</strong> masculine<br />
he was favoured as a generic pronoun in English, and thus as a functionally<br />
unmarked choice; we return to this in §11.5.1. 26<br />
If nouns belonging to different genders are coordinated, <strong>the</strong> gender form chosen<br />
for agreement will be <strong>the</strong> unmarked one. In Portuguese, Hebrew, and French <strong>the</strong><br />
masculine plural form will be used. One would say, in Portuguese, um menino e uma<br />
menina pequenos (indef.masc.sg boy and indef.fem.sg girl small.masc.pl) ‘a small boy<br />
and girl’, using <strong>the</strong> masculine plural form of <strong>the</strong> adjective ‘small’. Hebrew distinguishes<br />
feminine and masculine forms in singular and in plural. The masculine plural<br />
form is used to refer to a group of mixed linguistic genders, for animates, and for<br />
inanimates. For instance, sus means ‘(male) horse, stallion; horse in general’ and<br />
sus-a (horse-fem.sg) refers to a mare. Sus-im (horse-masc.pl) could refer to ei<strong>the</strong>r<br />
horses in general or only to stallions. Sus-ot (horse-fem.pl) can refer only to mares. If<br />
two humans—a male (of masculine Linguistic <strong>Gender</strong>) and a female (of feminine<br />
Linguistic <strong>Gender</strong>)—are coordinated, <strong>the</strong> masculine plural forms are used. Example<br />
2.5 illustrates <strong>the</strong> masculine plural pronoun hem ‘<strong>the</strong>y’ (in <strong>the</strong> function of a ‘present<br />
tense copula’), <strong>the</strong> masculine plural noun yeladim ‘boys, children’, and a masculine<br />
plural adjective tovim ‘good’, to refer to a girl and a boy. 27
2.3 Markedness and Linguistic <strong>Gender</strong>s 27<br />
2.5 yael ve-xaim hem yelad-im tov-im<br />
Yael(fem.sg) and-Haim(masc.sg) <strong>the</strong>y.masc.pl child-masc.spl good-masc.pl<br />
‘Yael (a girl) and Haim (a boy) are good children’<br />
This criterion—known as gender resolution—will not apply if no genders are<br />
distinguished in <strong>the</strong> plural (as in German, Russian, or Manambu).<br />
In all-female groups, ei<strong>the</strong>r feminine or masculine plural forms can be used. Yishai<br />
Tobin—who used to teach a university course for speech pathology students—<br />
comments that in classes consisting of just women both masculine and feminine<br />
plural forms were used. This resonates with a preference for using masculine forms<br />
(including <strong>the</strong> second person singular masculine pronoun ata) as generics by female<br />
speakers of Hebrew and of Palestinian Arabic (we return to this in §11.5.2).<br />
The generic use of masculine gender forms—especially anaphoric ones—is linked<br />
to <strong>the</strong> issue of an ‘unmarked’‘generic’ man, and <strong>the</strong> presumed supremacy of <strong>the</strong> male<br />
Social <strong>Gender</strong>. Linguistic <strong>Gender</strong> is seen as reflecting social attitudes: we return to<br />
this in §7.6.<br />
Feminine—ra<strong>the</strong>r than masculine—is <strong>the</strong> functionally unmarked choice in a few<br />
languages, scattered around <strong>the</strong> world—in South America, North America, New<br />
Guinea, and Australia. Jarawara, from <strong>the</strong> small Arawá family in Brazil, has two<br />
genders: feminine and masculine. The gender of all humans is determined by <strong>the</strong>ir<br />
sex. For o<strong>the</strong>r nouns, <strong>the</strong> principles of gender choice are complex, and only partly<br />
predictable. For instance, all liquids are feminine (and so are <strong>the</strong> recently introduced<br />
liquids such as gasoline or beer). Bahi ‘sun, thunder’, abariko ‘moon’, and amowa<br />
‘star’ are masculine by mythological association: <strong>the</strong>y are regarded as mythical men.<br />
About 53 per cent of plant names are feminine. Most garden plants with sweet taste<br />
are feminine (sami ‘pineapple’, jifari ‘banana’), and starch foods are masculine.<br />
<strong>Gender</strong> is marked in agreement on adjectives, demonstratives, and verbs, and also<br />
in possessive constructions. There is strong evidence for feminine being <strong>the</strong> functionally<br />
unmarked gender choice.<br />
First, all pronouns are marked on <strong>the</strong> verb with <strong>the</strong> feminine affix. R. M. W. Dixon<br />
reports that having to refer to himself by a pronoun with feminine agreement was<br />
one of <strong>the</strong> difficulties he found in speaking Jarawara. Secondly, <strong>the</strong> interrogative<br />
himata ‘what?’ always takes feminine gender agreement (if <strong>the</strong> speaker does not<br />
know <strong>the</strong> gender of <strong>the</strong> item <strong>the</strong>y are asking about), as in<br />
2.6 himata ama-ri?<br />
what<br />
‘What is it?’<br />
be-CONTENT.INTERROGATIVE.FEMININE<br />
Thirdly, if a general statement is made, a noun with no specific male or female<br />
reference will take feminine gender agreement. For instance, <strong>the</strong> noun wahati ‘person’
28 2 Linguistic <strong>Gender</strong> and its expression<br />
can refer to a man or woman of <strong>the</strong> Jamamadi tribe. A specific Jamamadi person will<br />
be referred to with feminine or with masculine agreement, depending on <strong>the</strong>ir sex.<br />
A generic Jamamadi person will require feminine agreement.<br />
That is, <strong>the</strong> masculine gender in Jarawara is <strong>the</strong> marked choice, only used to refer<br />
to a human male or something classified as belonging to <strong>the</strong> masculine gender. The<br />
feminine—<strong>the</strong> functionally unmarked choice—is used to refer to human females, or<br />
things classified as belonging to <strong>the</strong> feminine gender, and also where no gender is<br />
specified. The feminine gender is functionally unmarked—yet it would be premature<br />
to conclude that feminine Linguistic <strong>Gender</strong> as a default choice automatically accords<br />
women a special Social <strong>Gender</strong> status. We will see, in §7.6, that a particularly important<br />
woman can be referred to with <strong>the</strong> marked masculine gender, as if she were promoted<br />
to masculinity. 28<br />
Markedness relations may operate differently depending on a noun’s meaning. In<br />
Manambu, <strong>the</strong> feminine gender is functionally unmarked with regard to non-human<br />
nouns. A question ‘what is it?’ involves feminine agreement:<br />
2.7 agwa ja:p-al?<br />
what thing-3fem.sg.NOMINAL.CROSS.REFERENCING<br />
‘What is it?’<br />
Feminine forms are used in general statements. A common expression, al-al (that.<br />
fem.sg-3fem.sg.NOMINAL.CROSS.REFERENCING), roughly translatable as ‘that’s it, that’s<br />
how things are’, involves only feminine forms. Nominalized verbs and generic<br />
statements always require feminine forms. So does <strong>the</strong> loanword tenkyu ‘thank<br />
you’. Feminine forms of adjectives can be used as adverbs, e.g. kwasa ‘small (feminine<br />
singular), a little bit’.<br />
With human referents <strong>the</strong> situation is not <strong>the</strong> same. Plural is a functionally<br />
unmarked choice for humans in content questions about <strong>the</strong> identity of a person<br />
of unknown sex. In 2.8, a speaker asks who is going to come. Just one guest was<br />
expected, and <strong>the</strong> answer was expected to be in <strong>the</strong> singular. Plural is a neutral, or<br />
functionally unmarked choice.<br />
2.8 sə ya-kəna-di?<br />
who<br />
‘Who will come?’<br />
come-FUTURE-3pl.SUBJECT<br />
Third person singular masculine or feminine pronoun is never used to refer to an<br />
unknown human. But <strong>the</strong> masculine noun du ‘man’ has generic overtones of ‘a<br />
human’ in general. ‘Humanness’ of non-prototypical humans is defined as being du<br />
‘man’. Manambu women tell stories about underground villages where dead people<br />
live in a manner similar to Europeans. In one such description, <strong>the</strong> speaker stressed<br />
that those who live in <strong>the</strong>se villages are humans, and said: du-adi (man-3pl) ‘<strong>the</strong>y are<br />
human’, using <strong>the</strong> word du in a generic sense ‘human’. When I asked my classificatory
2.4 Linguistic <strong>Gender</strong>s and <strong>the</strong>ir labels: envoi 29<br />
sister about mischievous spirits who live in fig trees and what <strong>the</strong>y look like, I received<br />
an answer: du-adi, ñan-a-pək (man-3pl, we-LINKER-LIKE) ‘<strong>the</strong>y are human (literally<br />
man), like us’.<br />
As we will see in Chapter 3, masculine gender in Manambu is associated with<br />
cultural significance. And <strong>the</strong> term for ‘man’, du, has a broader range of meanings<br />
than ta:kw ‘woman’. Feminine gender is functionally unmarked for all effects and<br />
purposes—except in reference to humans. It is tempting to correlate <strong>the</strong> generic use<br />
of ‘man’ in Manambu with <strong>the</strong> male dominance across <strong>the</strong> board—we return to this<br />
in §7.6 and §11.5.<br />
Not every diagnostic context for ‘markedness’ can be applied to every language.<br />
Dyirbal has a semantically based system of four genders, as we saw in §2.2.1. Nei<strong>the</strong>r<br />
of <strong>the</strong>m can be considered functionally unmarked. Dixon (2015: 42) elaborates on<br />
this:<br />
I tried to investigate every possibility. For a group of people of mixed sex it seems that ei<strong>the</strong>r<br />
bayi [Class I which includes male humans] or balan [Class II which includes female humans]<br />
may be used. There may be a tendency to use balan if most of <strong>the</strong> group are female, or if a<br />
senior member is a woman, and similarly for bayi and men. But <strong>the</strong>se are only tendencies.<br />
What about a baby in <strong>the</strong> womb whose sex is not known? In English, <strong>the</strong> pronoun it is often<br />
used here…Bala (Class IV, residue or neuter) could not be employed, since in Dyirbal<br />
animates are confined to bayi and balan.…I enquired about this and was told that as soon<br />
as a baby was conceived <strong>the</strong> parents would think of it as male or female, thus referring to it by<br />
bayi or balan.<br />
In many languages, masculine Linguistic <strong>Gender</strong> is <strong>the</strong> functionally unmarked<br />
choice—in §7.6 and §11.5.1 we return to <strong>the</strong> generic uses of masculine pronouns in<br />
English, and masculine forms in o<strong>the</strong>r languages. <strong>How</strong> can <strong>the</strong> markedness of a<br />
Linguistic <strong>Gender</strong> correlate with <strong>the</strong> status of a Social <strong>Gender</strong>? Markedness, status,<br />
and power in Linguistic <strong>Gender</strong> choice are <strong>the</strong> topic of §7.7.<br />
2.4 Linguistic <strong>Gender</strong>s and <strong>the</strong>ir labels: envoi<br />
Linguistic <strong>Gender</strong>s are one of <strong>the</strong> means of categorizing nouns through language.<br />
The choice of a Linguistic <strong>Gender</strong> may be semantically transparent only to a partial<br />
extent, and correlate only in part with <strong>the</strong> Natural <strong>Gender</strong>, or sex. The assignment of<br />
Linguistic <strong>Gender</strong> to a noun may be based on its form (ra<strong>the</strong>r than its meaning). This<br />
creates opacity in gender choice, especially for non-humans where Natural <strong>Gender</strong> is<br />
irrelevant. In languages where gender choice is opaque it is often hard—if not<br />
impossible—to offer an exhaustive explanation for gender choice for every noun.<br />
The terms used to ‘label’ genders are thus misleading—a gender labelled ‘feminine’<br />
would include more than just females, and <strong>the</strong> one labelled ‘masculine’ more than<br />
just males. The term ‘neuter’ tends to refer to a gender which includes inanimate (or
30 2 Linguistic <strong>Gender</strong> and its expression<br />
irrational) beings, or a residue gender whose semantic basis is difficult to capture. In<br />
each case, a linguist needs to carefully investigate <strong>the</strong> language-specific principles<br />
obscured by <strong>the</strong> nickname-like terms.<br />
In a number of languages from New Guinea and Africa, where gender is assigned<br />
to humans based on <strong>the</strong>ir sex, physical properties play a role in choosing <strong>the</strong><br />
Linguistic <strong>Gender</strong> of non-humans. This is what Chapter 3 is about.<br />
NOTES AND SOURCES<br />
1. Noun class and gender are often used interchangeably. To avoid confusion, I will use <strong>the</strong><br />
term Linguistic <strong>Gender</strong> throughout. See Kilarski (2014: 59–272), on <strong>the</strong> history of study of<br />
Linguistic <strong>Gender</strong> from antiquity onwards. Typological features of genders and noun<br />
classes and <strong>the</strong> history of terms can be found in Aikhenvald (2012b, 2004, 2006, and<br />
especially 2000: 18–80); and also Corbett (1991, 2014a, 2014b).<br />
The term gender has been used in a different way. For instance, in <strong>the</strong> Athabaskan<br />
linguistic tradition <strong>the</strong> term gender is used to refer to verbal classifiers which mark<br />
agreement with intransitive subject or transitive object, and characterize <strong>the</strong> referent<br />
noun in terms of shape and form (Aikhenvald 2000: 15). The term noun class may refer<br />
to different ways of marking plural on nouns (something which involves classifying nouns<br />
on formal ra<strong>the</strong>r than semantic grounds: an example is in Dimmendaal 2000).<br />
2. See Lehmann (1993: 20); Asher (1985: 136–7).<br />
3. There appears to be a misleading trend, especially among German speaking scholars,<br />
to claim that English has no gender (e.g. Motschenbacher 2010). This shows disregard<br />
of anaphoric gender in English and obscures <strong>the</strong> difference between English and those<br />
languages where gender is restricted to derivation and lexicon, such as Hungarian, Estonian,<br />
and Indonesian.<br />
4. See Aikhenvald (2012a: chapter 10) for South American languages; Aikhenvald (2000: 369,<br />
439) for Finnish and o<strong>the</strong>r languages.<br />
5. See Schapper (2010a: 182–4, 2010b: 420–1).<br />
6. See Heine (1982a: 194); fur<strong>the</strong>r examples in Aikhenvald (2000: 38–9).<br />
7. That anaphoric gender tends to be chosen on semantic ra<strong>the</strong>r than formal grounds, and that<br />
<strong>the</strong> choice of agreement gender may follow formal principles (based on non-semantic<br />
choice of gender) is captured by <strong>the</strong> Agreement Hierarchy (see Corbett 1991: 225–41).<br />
This Hierarchy offers principles behind <strong>the</strong> choice for semantic ra<strong>the</strong>r than form-based<br />
gender agreement. It identifies four types of agreement position:<br />
attributive < predicate < relative pronoun < personal pronoun,<br />
stating that ‘as we move rightwards along <strong>the</strong> hierarchy, <strong>the</strong> likelihood of semantic<br />
agreement will increase monotonically’. This statement assumes that anaphoric gender is<br />
a type of agreement gender (an issue which is debatable).<br />
8. In a few languages, more genders are distinguished in non-singular numbers: see Steinhauer<br />
(1986) on Biak; a discussion in Aikhenvald (2000: 243–5) and Aikhenvald and Dixon<br />
(2011b: 181); Plank and Schellinger (1997). In Mangarrayi and Jawoyn, Australian languages’<br />
gender markers are fused with case (Dixon 2002: 508–9).<br />
9. This agrees with Greenberg’s universal 44, that ‘if a language has gender distinctions in <strong>the</strong><br />
first person, it always has gender distinctions in <strong>the</strong> second or third person’ (Greenberg
2.4 Linguistic <strong>Gender</strong>s and <strong>the</strong>ir labels: envoi 31<br />
1963: 95). Tocharian A, an extinct Indo-European language, is an exception: it is <strong>the</strong> only<br />
language in <strong>the</strong> family to have feminine and masculine forms of first person full pronouns:<br />
näş ‘I (male speaker)’ and ñuk ‘I (female speaker)’. There are no genders in second person<br />
pronouns; two genders are distinguished in demonstratives used in lieu of third person<br />
pronouns: see Matasović (2004: 63) and Jasanoff (1989).<br />
10. See Yadav (1996: 63–5, 124–7).<br />
11. See Matasović (2004: 51–2) on Romanian; Mallinson (1986: 244–6). Surface realizations<br />
(or forms) of genders have been referred to as ‘target’ genders; and agreement genders as<br />
‘controller genders’: see Aikhenvald (2000: 45–7), Corbett (1991: 151). A similar example<br />
of two gender forms with three agreement possibilities comes from Telugu, a Dravidian<br />
language (Krishnamurti and Gwynn: 1985: 56–8). In a number of Cushitic languages,<br />
plural is recognized as ‘<strong>the</strong> third gender’: see Fraizyngier (2012: 523), Mous (2012: 364–9),<br />
and Tsegaye, Mous, and Schiller (2014).<br />
12. See Heine (1982a: 195) on different gender systems for pronouns and for o<strong>the</strong>r agreeing<br />
modifiers in African languages, including Swahili; Harvey (1992, 2002) on Gaagudju,<br />
Aikhenvald (2007) on Baniwa, and Aikhenvald and Green (2011: 405–7) on Palikur.<br />
Fur<strong>the</strong>r instances are discussed in Aikhenvald (2000: 68–75).<br />
13. Aikhenvald (2000: 22–4, 2012a: 281–2) and references <strong>the</strong>re; Kibrik (1996).<br />
14. See Haas (1941: 65).<br />
15. See Dixon (1972: 308–12; 2015: 27–43). Dyirbal has been <strong>the</strong> subject of numerous reanalyses<br />
in secondary sources. That by Lakoff (1987) involves a severe misrepresentation.<br />
16. Bani (1987), Haas (1941: 65); Mithun (1999: 101).<br />
17. Algonquian languages are one of <strong>the</strong> large language families in <strong>the</strong> USA stretching from<br />
Alberta and Montana in <strong>the</strong> west of <strong>the</strong> United States to <strong>the</strong> Atlantic, and from <strong>the</strong> nor<strong>the</strong>ast<br />
(Labrador) to South Carolina; see Mithun (1999: 98–9), Kilarski (2007), Craik (1982),<br />
Dahlstrom (1995), Darnell and Vanek (1976), Goddard (2002), Quinn (2001), Straus and<br />
Brightman (1982), and Wolfart (1973).<br />
18. See Dahlstrom (1995), on Fox; see also Wolfart (1973: 21–3) on gender in Plains Cree.<br />
19. See Spitulnik (1989: 207).<br />
20. See Schaub (1985: 175), Denny (1976), Aikhenvald (2000: 281–3).<br />
21. See Pensalfini (2003: 159–68).<br />
22. See Aikhenvald (2000: 25); Heine (1982a: 200); Bhatia (1993: 216–17).<br />
23. There may be a historical explanation for this phonological principle of gender assignment.<br />
Proto-Arawak had a feminine (non-masculine) affix *-u/o which is still preserved in<br />
<strong>the</strong> form of Palikur gender-sensitive suffixes, and in some lexical items, e.g. tino ‘woman’,<br />
cf. Proto-Arawak *c᷈ina-ru ‘woman’.<br />
24. See Matasović (2004: 70); see also Heine (1982a: 200) on semantic principles of Linguistic<br />
<strong>Gender</strong> assignment in Irakw, a South Cushitic language, and formal ones for <strong>the</strong> rest.<br />
25. See Foley (1991: 119–64) on Yimas, Nekitel (1986) on Abu' Arapesh, and an overview<br />
of Arapesh languages in Dobrin (2012: 84–109). Alliterative agreement has also been<br />
described as one of <strong>the</strong> agreement techniques for a number of varieties of Baïnouk, a West<br />
Atlantic language (Cobbinah 2010).<br />
26. See <strong>the</strong> discussion of markedness in Schane (1970), Corbett (1991: 291), Aikhenvald (2000:<br />
54–6), and Aikhenvald and Dixon (2011b: 177); Bulygina and Shmelev (1996: 103),<br />
Rothstein (1973), Jakobson (1984) on Russian; see Wales (1996), Curzan (2003), Newman<br />
(1997).<br />
27. See Tobin (2001: 183–5) on <strong>the</strong> functionally unmarked masculine gender in Hebrew; Sa'ar<br />
(2007) on masculine forms used by women speakers of Hebrew and Palestinian Arabic.
32 2 Linguistic <strong>Gender</strong> and its expression<br />
See Sadiqi (2003) on masculine plural forms with reference to mixed-sex groups in<br />
Moroccan Arabic.<br />
28. See Dixon (2004: 285–7) on Jarawara (Arawá family), a few o<strong>the</strong>r South American<br />
languages (Aikhenvald 2012a, 2012b), a few Australian languages (e.g. Wangkumara, <strong>the</strong><br />
western dialect of Wagaya, Kala Lagaw Ya, and Murrinhpatha: Alpher 1987: 174, Breen<br />
1976a: 336, 1976b: 340, 590, Walsh 1976: 150–6), and a number of Nor<strong>the</strong>rn Iroquoian<br />
languages (Chafe 2004).
3<br />
Round women and long men:<br />
physical properties in Linguistic<br />
<strong>Gender</strong><br />
Meanings of Linguistic <strong>Gender</strong>s typically involve humanness, animacy, and sex—<br />
female versus male. In a number of languages across <strong>the</strong> world, animates and inanimates<br />
are assigned genders depending on <strong>the</strong>ir shape and size. We have identified two<br />
scenarios.<br />
SCENARIO 1. Masculine Linguistic <strong>Gender</strong> is associated with larger size and long slender<br />
shape and feminine gender is associated with smaller size and roundish shape: this is<br />
<strong>the</strong> topic of §3.1.<br />
SCENARIO 2. Feminine Linguistic <strong>Gender</strong> is associated with larger size and masculine<br />
gender is associated with smaller size; this is <strong>the</strong> topic of §3.2.<br />
In §3.3 we reconcile <strong>the</strong>se seemingly opposing scenarios. Linguistic <strong>Gender</strong> choice<br />
can also correlate with value and importance of an object, partly reflecting Social<br />
<strong>Gender</strong> stereotypes: this is <strong>the</strong> topic of §3.4.<br />
3.1 Small round women and long slender men<br />
Round shape and small size are associated with <strong>the</strong> feminine Linguistic <strong>Gender</strong> in a<br />
dozen languages along <strong>the</strong> Sepik River in New Guinea. The ‘opposite’—long and<br />
large—objects are considered ‘masculine’. Speakers of Manambu, a Ndu language<br />
from this area, are ‘gender-proud’. When asked what is special about <strong>the</strong>ir language,<br />
speakers reply: ‘in our language, everything in <strong>the</strong> world is ei<strong>the</strong>r woman-like or<br />
man-like’. 1 The language has two genders, masculine and feminine. <strong>Gender</strong> is<br />
covert: it cannot be inferred from <strong>the</strong> form of <strong>the</strong> noun. One knows what gender a<br />
noun belongs to by <strong>the</strong> form of an agreeing demonstrative, adjective, possessive<br />
pronoun, or verb. Choice of gender for humans is always based on <strong>the</strong>ir Natural<br />
<strong>Gender</strong>, or sex: asa:y ‘fa<strong>the</strong>r’ and du ‘man’ are masculine and amaey ‘mo<strong>the</strong>r’ and<br />
ta:kw ‘woman’ are feminine (Linguistic <strong>Gender</strong>s for humans can be reversed under<br />
<strong>How</strong> <strong>Gender</strong> <strong>Shapes</strong> <strong>the</strong> <strong>World</strong>. First edition. Alexandra Y. Aikhenvald.<br />
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.
34 3 Round women and long men<br />
exceptional circumstances—see §§7.1–3). If <strong>the</strong>re is one word for both sexes,<br />
gender on accompanying words helps tell <strong>the</strong>m apart: ma:m is <strong>the</strong> word for<br />
‘older sibling’. Amasculineforma-də ma:m (that-masc.sg elder.sibling) means<br />
‘that older bro<strong>the</strong>r’, and a feminine form a ma:m (that+fem.sg elder sibling)<br />
means ‘that older sister’.<br />
Outside <strong>the</strong> realm of humans, gender choice is based on <strong>the</strong> object’s physique.<br />
A big dog or a big pig will have masculine gender, no matter whe<strong>the</strong>r <strong>the</strong>y are male or<br />
female: a-də bal (that-masc.sg pig) means ‘that (big) pig’, and a bal (that+fem.sg pig)<br />
‘that (small) pig’. One can always specify <strong>the</strong> sex by adding a noun, amaey (‘mo<strong>the</strong>r’)<br />
for females, and asa:y ‘fa<strong>the</strong>r’ for males: that is, amaey bal (mo<strong>the</strong>r pig) is a female<br />
pig, and asa:y bal (fa<strong>the</strong>r pig) is a male pig.<br />
Some animals, birds, and heavenly bodies have special roles in myths. Məd<br />
‘cassowary (a non-flying bird)’ is always feminine gender, because cassowaries are<br />
conceived of as mythical women (who turned into cassowaries later). This association<br />
is widespread in New Guinea. Similarly, mæ:n ‘bird of paradise’ is a woman in<br />
disguise, and cannot be made masculine. In §8.1, we turn to mythological associations<br />
in <strong>the</strong> choice of Linguistic <strong>Gender</strong>s.<br />
In choosing gender for inanimates, shape comes into play. Long or large objects<br />
are treated as masculine, and small or round ones as feminine. Val ‘canoe’ is<br />
masculine if big, and feminine if small. Væy ‘spear’ is masculine due to its inherent<br />
long shape and size; it is feminine when <strong>the</strong> word is used to refer to a small spear or a<br />
shotgun. A house of usual size is referred to as feminine; an unusually big house is<br />
masculine. The Manambu people who have visited England all agree that Buckingham<br />
Palace is a ‘masculine type’ house because it is really big. The House of<br />
Parliament in Canberra was also judged big enough to be considered ‘masculine’.<br />
Skyscrapers are ‘masculine’, and so are traditional big men’s houses which used to be<br />
large and impressive (kara:b). A long road (ya:b) will be referred to with masculine<br />
gender, and a short one with feminine gender.<br />
Some inanimates have a ‘typical’ shape and size. For instance, ar ‘lake’, kabak ‘stone’,<br />
ya:l ‘belly’, and ab ‘head’ are typically round, and so assigned to <strong>the</strong> feminine gender.<br />
Changing gender of any of <strong>the</strong>se implies <strong>the</strong>ir unusual, and even grotesque, size.<br />
A woman has an affair with a snake, and becomes pregnant. Her belly gets bigger<br />
and bigger and <strong>the</strong>n becomes grotesquely huge. It is <strong>the</strong>n referred to as ‘masculine’. The<br />
change of gender is shown in example 3.1. <strong>Gender</strong> markers are in bold face.<br />
3.1 lə-kə ya:l ata numa-ø məy ta:l<br />
she-POSS+fem.sg belly <strong>the</strong>n big-fem.sg very become+3fem.sgSUBJECT.PAST<br />
a numa-də ya:l ta:d lə-kə-də ya:l<br />
<strong>the</strong>n big-masc.sg belly become+3masc.sgSUBJECT.PAST she-POSS-masc.sg belly<br />
‘(Her) belly <strong>the</strong>n became very big (feminine), here is an unusually big (masculine)<br />
belly, her (masculine) belly’
3.1 Small round women and long slender men 35<br />
In an English commentary this was described by a speaker as a ‘big man-type belly’.<br />
Changing genders can describe unusual situations. ‘Head’ is feminine because of<br />
its round shape. When a person has a headache, <strong>the</strong> head <strong>the</strong>n feels heavy and<br />
unusually big, and can be referred to with masculine gender. In 3.2, a speaker<br />
complained of a headache, talking about her head with masculine Linguistic<br />
<strong>Gender</strong>:<br />
3.2 wuna-də ab kagəl yi-na-d<br />
I-masc.sg head pain have-PRESENT-3masc.sg.SUBJECT<br />
‘My head (masculine because of its unusual weight) is aching’<br />
Linguistic <strong>Gender</strong> of nouns denoting natural phenomena and physical states is<br />
based on <strong>the</strong>ir extent. A darkish night (ga:n) will be feminine, and a night which<br />
is completely dark is masculine. Hunger (ka:m) is feminine; a very strong hunger is<br />
masculine. Quantity determines <strong>the</strong> Linguistic <strong>Gender</strong> choice of mass nouns. A little<br />
money or a little blood will be referred to with feminine gender; if <strong>the</strong>re is a lot of<br />
money, or blood, <strong>the</strong> Linguistic <strong>Gender</strong> will be masculine. The term gwalugw<br />
‘patrilineal clan’ is masculine. A numerically small clan can be assigned feminine<br />
gender. Figure 3.1 summarizes <strong>the</strong> principles of gender choice in Manambu.<br />
Some entities have a typical shape—and so a typical Linguistic <strong>Gender</strong>. All trees<br />
are masculine due to <strong>the</strong>ir height and vertical stance. Their fruit is feminine independent<br />
of <strong>the</strong> shape, e.g. mi: ‘tree (generic)’ and ma:s ‘betel nut tree’ are masculine;<br />
təkəmi ‘fruit (generic)’ and ma:s ‘betel nut fruit’ are feminine. An unusually big fruit<br />
can be masculine.<br />
If a noun refers to a concept central to <strong>the</strong> Manambu mythology, Linguistic<br />
<strong>Gender</strong> is chosen based on ‘mythological association’ and important property<br />
principles—something we saw for Dyirbal in §2.2.1. Moon (bap) is feminine, because<br />
it is conceived of as a mythological woman; an alternative, endearing, name for it is<br />
bap-a-ta:kw (moon-LK-woman) ‘lady moon’. Similarly, stars (kugar) are conceptualized<br />
as women belonging to <strong>the</strong> clan group which is associated with sun, moon, and<br />
light. No matter how big <strong>the</strong> moon or a star, <strong>the</strong>y will never be referred to with<br />
masculine gender.<br />
Similar examples come from a number of neighbouring languages. Alamblak,<br />
from <strong>the</strong> Sepik Hill family, is spoken to <strong>the</strong> south-east of Manambu. 2 Unlike<br />
Manambu and its relatives, gender is marked on <strong>the</strong> noun itself. All nouns which<br />
denote females and short, squat, or wide objects are feminine and have a form<br />
marked with a feminine suffix -t. Tall, long, and slender objects are masculine gender,<br />
and take <strong>the</strong> masculine suffix -r. Terms for inanimate objects which host <strong>the</strong> feminine<br />
suffix include terms for house, fighting shield, and trees ‘which are typically relatively<br />
shorter and more squat than o<strong>the</strong>r trees’. Those which occur with <strong>the</strong> masculine<br />
suffix include ‘terms for arrows, signal trumpet, typically tall slender-growing trees,<br />
large string bag varieties, etc.’
36 3 Round women and long men<br />
human<br />
adults<br />
children and babies<br />
Natural <strong>Gender</strong><br />
Natural <strong>Gender</strong><br />
age, size<br />
males, and males by association:<br />
masculine<br />
females: feminine<br />
male: masculine<br />
female: feminine<br />
relatively big and/or old: masculine<br />
relatively small and/or young: feminine<br />
non-human<br />
animals<br />
inanimates<br />
lower animates<br />
mass nouns<br />
size<br />
size and shape<br />
quantity<br />
large: masculine<br />
small: feminine<br />
long and large: masculine<br />
round and small: feminine<br />
big: masculine<br />
small: feminine<br />
natural phenomena<br />
extent/intensity<br />
FIGURE 3.1 <strong>Gender</strong> assignment in Manambu<br />
complete/intense: masculine<br />
non-complete/non-intense: feminine<br />
A noun can occur with an ‘atypical’ gender suffix. This indicates that <strong>the</strong> object is<br />
unusual as to its size. The word for ‘house’ usually occurs with feminine suffix. It can<br />
occur with <strong>the</strong> masculine suffix(kuñ-r house-masc): <strong>the</strong>n <strong>the</strong> house is perceived as an<br />
unusually long one. If <strong>the</strong> term for a slit-gong drum is marked with <strong>the</strong> masculine<br />
suffix(nërwi-r), this indicates that <strong>the</strong> drum is ‘unusually slender, which implies that<br />
it was made incorrectly and does not sound good’. For humans and animals, gender<br />
switch indicates change in Natural <strong>Gender</strong>, e.g. yima ‘person’, yima-r (person-masc)<br />
‘man’, yima-t (person-fem) ‘woman’.<br />
Sare (or Kapriman), ano<strong>the</strong>r language from <strong>the</strong> Sepik Hill family, also has two<br />
Linguistic <strong>Gender</strong>s, masculine and feminine. 3 Linguistic <strong>Gender</strong> assignment to<br />
nouns referring to lower animates (including rats, birds, and insects) depends on<br />
shape and size: small, short, or rounded referents, such as mice, will be assigned to<br />
<strong>the</strong> feminine Linguistic <strong>Gender</strong>, and those animates which are big, tall, or slender—
3.1 Small round women and long slender men 37<br />
for instance, dogs—will belong to <strong>the</strong> masculine Linguistic <strong>Gender</strong> (unless <strong>the</strong> sex of<br />
<strong>the</strong> animate needs to be differentiated).<br />
The choice of <strong>the</strong> Linguistic <strong>Gender</strong> of an inanimate is determined by shape and<br />
size: objects which are big, tall, long, or slender are treated as ‘masculine’, and those<br />
which are small, short, or squat are feminine. Some examples are in 3.3.<br />
3.3 MASCULINE: BIG, TALL, LONG, SLENDER FEMININE: SMALL, SHORT, SQUAT<br />
wuni-r ‘big house’<br />
xudari-r ‘big axe’<br />
seboxu-r ‘high table’<br />
wuni-s ‘small house’<br />
xudari-s ‘small axe’<br />
seboxu-s ‘squat table’<br />
As Sumbuk (1999: 116) putsit,‘howbig,tallorhighanobjectmustbetobe<br />
accorded one gender or <strong>the</strong> o<strong>the</strong>r is determined by <strong>the</strong> traditional expectation of<br />
<strong>the</strong> speaker with regard to a particular object. With regard to <strong>the</strong> size of things<br />
like houses, canoes and containers, it is <strong>the</strong> capacity of <strong>the</strong> referents of <strong>the</strong>se nouns<br />
to hold persons or things that determines what gender <strong>the</strong> noun takes. If a house,<br />
for instance, has space enough only for a single family (parents, children and<br />
grandparents), <strong>the</strong>n it would normally be described as a small house, thus taking<br />
<strong>the</strong> feminine gender. If on <strong>the</strong> o<strong>the</strong>r hand, <strong>the</strong> house has a capacity for several<br />
families, <strong>the</strong>n it would normally be described as a big house and be accorded <strong>the</strong><br />
masculine gender.’<br />
Somewhat different principles of Linguistic <strong>Gender</strong> choice apply for different<br />
semantic groups of nouns—similarly to Manambu, in Figure 3.1. <strong>Gender</strong> choice for<br />
plants depends on whe<strong>the</strong>r <strong>the</strong> plant is young or mature. A term for a young plant<br />
will belong to <strong>the</strong> masculine gender (despite its being typically small and short).<br />
A term for a mature plant (normally big and tall) will belong to <strong>the</strong> feminine gender<br />
(e.g. popo-r (pawpaw-masc) ‘young pawpaw tree’, popo-s (pawpaw-fem) ‘mature<br />
pawpaw tree’): Sumbuk (1999: 117) suggests that mature plants ‘are regarded as<br />
<strong>the</strong> female of <strong>the</strong> plants since <strong>the</strong>y bear fruit’; young plants, ‘despite being short and<br />
small, which are features that we would ordinarily associate with feminine gender,<br />
are regarded as masculine by <strong>the</strong> Sare because of <strong>the</strong>ir inability to bear fruit’.<br />
A similar principle applies to fruit and seeds. Mature fruit are feminine, and small<br />
immature ones are masculine, e.g. gonxa-r (banana-masc) ‘immature banana’,<br />
gonxa-s (banana-fem) ‘mature banana’, popoyioka-r (pawpaw-masc) ‘immature pawpaw’,<br />
popoyioka-s (pawpaw-fem) ‘mature pawpaw’. The ‘important property’ of<br />
maturity overrides <strong>the</strong> shape- and size-based choice of Linguistic <strong>Gender</strong>.<br />
The choice of Linguistic <strong>Gender</strong> for inanimates depends on shape and size in a few<br />
more languages across <strong>the</strong> island of New Guinea and surrounds. In Wära, an isolate<br />
from <strong>the</strong> Fly River region in New Guinea, long objects are assigned to <strong>the</strong> masculine<br />
gender and round objects, and objects consisting of multiple parts, are feminine. In<br />
Olo, from <strong>the</strong> Torricelli grouping in New Guinea, feminine gender is associated with<br />
<strong>the</strong> ‘small nature’ of <strong>the</strong> entity. The choice of Linguistic <strong>Gender</strong> for inanimates in
38 3 Round women and long men<br />
Savosavo, a Papuan language from <strong>the</strong> Solomons, depends on <strong>the</strong> size of <strong>the</strong> object:<br />
small objects are feminine, and large ones are masculine. 4<br />
Correlations between <strong>the</strong> size and <strong>the</strong> Linguistic <strong>Gender</strong> go beyond New Guinea.<br />
The majority of Afroasiatic languages of North Africa have two Linguistic <strong>Gender</strong>s—<br />
masculine for males and feminine for females. <strong>Gender</strong> is marked on <strong>the</strong> noun itself, in<br />
agreeing modifiers, on verbs, and in personal pronouns. The marker of <strong>the</strong> feminine<br />
Linguistic <strong>Gender</strong> is homophonous with diminutive marking across <strong>the</strong> family. And<br />
<strong>the</strong> feminine Linguistic <strong>Gender</strong> tends to correlate with small size of an object. Ayt<br />
Seghrushen (Moroccan Berber) men belong to <strong>the</strong> masculine Linguistic <strong>Gender</strong> and<br />
women to <strong>the</strong> feminine gender (e.g. arba ‘male child’, t-arba-tt ‘female child’). For<br />
inanimates and lower animates (whose sex is irrelevant and difficult to determine<br />
anyway), a masculine form denotes ‘something bigger than <strong>the</strong> feminine’:<br />
3.4 t-fus-tt (feminine) ‘little baby hand’ fus (masculine) ‘hand’<br />
t-ams̍at̍t̍(feminine) ‘thigh’ amsad̍(masculine) ‘very big thigh’<br />
t-axxuy-t (feminine) ‘louse’ axxuy (masculine) ‘very big louse’<br />
Changing Linguistic <strong>Gender</strong> for humans and sex-differentiable animals implies<br />
difference in size (and not in Natural <strong>Gender</strong>, or sex):<br />
3.5 t-amət̍t̍ut̍t̍(feminine) ‘woman’ amət̍t̍ud̍(masculine) ‘very big woman’<br />
t-yis-tt (feminine) ‘little stallion’ yis (masculine) ‘stallion, horse’<br />
t-aymar-t (feminine) ‘mare’ aymar (masculine) ‘very big mare’<br />
t-ɣat̍t̍(feminine) ‘goat (female)’ ɣad̍(masculine) ‘big female goat, difficult to<br />
handle’<br />
Masculine Linguistic <strong>Gender</strong> can have overtones of something big, or an ‘augmentative’.<br />
Some nouns have three forms: augmentative (masculine), a neutral<br />
(masculine or feminine), and diminutive (feminine). For instance, an Arabic loanword<br />
lkursi ‘chair’ is masculine (and neutral in size). A feminine form t-akursi-tt<br />
means ‘little chair’, and a masculine form a-kursi means ‘very big chair’ (<strong>the</strong> Arabic<br />
article l- is replaced with Berber a-, an erstwhile article-like morpheme). 5<br />
The association between <strong>the</strong> feminine Linguistic <strong>Gender</strong> and small size on <strong>the</strong><br />
one hand, and masculine Linguistic <strong>Gender</strong> and larger size, on <strong>the</strong> o<strong>the</strong>r hand, is<br />
a pervasive feature of ano<strong>the</strong>r subgroup within Afroasiatic—<strong>the</strong> Omotic languages<br />
of Ethiopia (including Wolaitta, Maale, Benchnon, Sheko, Dime, and Dizin). In<br />
3.6a, from Wolaitta, <strong>the</strong> masculine Linguistic <strong>Gender</strong> of a noun with a human<br />
referent (‘child’) indicates male sex. The feminine Linguistic <strong>Gender</strong> in 3.6b indicates<br />
female sex:<br />
3.6a naʔá-y y-iísi<br />
child-MASC:NOMINATIVE come-3masc.sg:PERFECTIVE<br />
‘The boy came’
3.1 Small round women and long slender men 39<br />
3.6b naʔ-íya y-aásu<br />
child-FEM:NOMINATIVE come-3fem.sg:PERFECTIVE<br />
‘The girl came’<br />
An inanimate ‘tree’ in 3.7 is large in size—this is signalled through its masculine<br />
Linguistic <strong>Gender</strong>:<br />
3.7 mítta-y kúnd-iísi<br />
tree-MASC:NOMINATIVE fall-3masc.sg:PERFECTIVE<br />
‘The (big) tree fell’<br />
The tree in 3.8 is small—and this is reflected in its feminine Linguistic <strong>Gender</strong>:<br />
3.8 mítt-íya kúnd-aásu<br />
tree-FEM:NOMINATIVE fall-3fem.sg:PERFECTIVE<br />
‘The little (feminine) tree fell’<br />
Feminine Linguistic <strong>Gender</strong> with nouns with inanimate referents in Benchnon, also<br />
Omotic, indicates that <strong>the</strong> object is small, as in 3.9:<br />
3.9 két-àɕ-ī versus két-àn-ā<br />
house-THIS.MASC-NOMINATIVE.MASC<br />
‘this (big) house’<br />
house-THIS.FEM-NOMINATIVE.FEM<br />
‘this (small) house’<br />
The feminine Linguistic <strong>Gender</strong> may be used to vent a speaker’s ‘derogatory attitude’,<br />
as in 3.10.<br />
3.10 tá tʃˈāmá-ī versus tá tʃˈāmá-ā<br />
I shoe-NOMINATIVE.MASC I shoe-NOMINATIVE.FEM<br />
‘my shoe’<br />
‘my shoe I don’t care about’<br />
A similar principle applies to o<strong>the</strong>r subgroups of Afroasiatic. 6 In <strong>the</strong> Harar dialect<br />
of Oromo, a Cushitic language, if <strong>the</strong> speaker wishes to ‘evaluate <strong>the</strong> referent as<br />
major’, <strong>the</strong> noun acquires masculine gender marking. If a noun is assigned to<br />
masculine gender but refers to something evaluated as minor, it may be treated as<br />
feminine, via agreement. In eastern and sou<strong>the</strong>rn varieties of Oromo, ‘if a speaker<br />
wishes to indicate that <strong>the</strong> referent is marked as augmented with respect to a<br />
particular aspect in <strong>the</strong> domain of discourse…<strong>the</strong> noun may be treated as masculine’<br />
(Clamons 1993: 275–6).<br />
The noun ablee ‘knife’ triggers feminine agreement on <strong>the</strong> modifier (tun ‘this:<br />
feminine’) and <strong>the</strong> predicate (doom-tuu ‘dull-feminine’):<br />
3.11 ablee-n tun doom-tuu<br />
knife-SUBJECT.TOPIC this:fem dull-fem<br />
‘This knife is dull’
40 3 Round women and long men<br />
In 3.12, <strong>the</strong> same noun, ‘knife’, is masculine. This is reflected in <strong>the</strong> agreement<br />
forms of <strong>the</strong> modifier and <strong>the</strong> predicate. The Linguistic <strong>Gender</strong> shift indicates that<br />
<strong>the</strong> knife is considered big by <strong>the</strong> speaker:<br />
3.12 ablee-n xun doom-aa<br />
knife-SUBJECT.TOPIC this:masc dull-masc<br />
‘This (big) knife is dull’ 7<br />
Similar principles have been described for o<strong>the</strong>r languages in East Africa. In<br />
Turkana, an East Nilotic language from Kenya, changing <strong>the</strong> gender of a noun<br />
correlates with <strong>the</strong> size of a referent. 8 In 3.13, <strong>the</strong> same root can occur with <strong>the</strong><br />
masculine, <strong>the</strong> feminine, and <strong>the</strong> neuter prefix. The referent of <strong>the</strong> masculine noun is<br />
typically big, <strong>the</strong> referent of <strong>the</strong> feminine noun is smaller in size, and <strong>the</strong> referent of<br />
<strong>the</strong> neuter noun is very small:<br />
3.13 e-mor-u` ‘rocky mountain, big stone’ —masculine<br />
a-mor-u` ‘hill, stone’ —feminine<br />
i-mor-u` ‘pebble’ —neuter<br />
In Camus, from <strong>the</strong> Maa subgroup of East Nilotic, small objects are classed as<br />
feminine, and big items are assigned to <strong>the</strong> masculine gender. In Khwe, a Central<br />
Khoisan language spoken in Namibia (and also in Angola, Botswana, and Zambia),<br />
inanimate referents are assigned to <strong>the</strong> masculine Linguistic <strong>Gender</strong> if <strong>the</strong>y are<br />
long, high, narrow, pointed, big, or strong, and to feminine Linguistic <strong>Gender</strong> if<br />
<strong>the</strong>y are short, small, round, thick, broad, or weak. In Katcha, a Kadugli-Krongo<br />
language spoken in <strong>the</strong> Nuba mountains in <strong>the</strong> Sudan, masculine Linguistic <strong>Gender</strong><br />
includes long, thick, and solid objects; feminine Linguistic <strong>Gender</strong> subsumes<br />
hollow, deep, flat, and thin objects. 9 In Piapoco, a North Arawak language from<br />
Colombia, inanimate objects which are roundish, such as càyuwa ‘hat’, belong to<br />
<strong>the</strong> feminine gender. Objects which are slender and long, such as énu ‘shotgun’, are<br />
masculine. 10<br />
Shape and size can be a gendering dimension in a language without Linguistic<br />
<strong>Gender</strong>. Nungon is a Papuan language spoken in Morobe province. Natural <strong>Gender</strong><br />
is expressed in lexical items referring to man, woman, boy, girl, and a plethora of kin<br />
terms. Hannah Sarvasy (2016) reports that ‘<strong>the</strong> only domain in which inanimate<br />
objects are classed as male or female based on size is that of yams’. Referring to yams<br />
as ‘man’ or as ‘woman’ is ‘largely based on <strong>the</strong> size and shape of <strong>the</strong> tubers—smaller<br />
and rounder are female, while longer and larger are male’.<br />
So far we have seen how <strong>the</strong> choice of feminine Linguistic <strong>Gender</strong> may correlate<br />
with small size, and also round (and sometimes thin and narrow) shape. This is<br />
especially so for inanimates and lower animates (whose Natural <strong>Gender</strong> is not known<br />
or not important). Masculine Linguistic <strong>Gender</strong> is assigned to objects which are large,<br />
long, and strong. We now turn to a different scenario.
3.2 When ‘women’ are larger than ‘men’ 41<br />
3.2 When ‘women’ are larger than ‘men’<br />
In a number of languages, large size (and also squat extended shape) is a correlate of<br />
feminine gender. Small size and slender shape are a correlate of masculine gender.<br />
The choice of Linguistic <strong>Gender</strong> in Tiwi, an Australian language, is based on <strong>the</strong><br />
Natural <strong>Gender</strong> of humans and animals: all males are assigned to <strong>the</strong> masculine<br />
gender and all female human beings and animals are assigned to <strong>the</strong> feminine gender.<br />
Masculine and feminine genders are marked on <strong>the</strong> noun, and also in agreement and<br />
anaphora.<br />
For inanimate entities, <strong>the</strong> choice of Linguistic <strong>Gender</strong> is determined by physical<br />
size and shape. The semantic features correlating with <strong>the</strong> masculine gender are<br />
‘small’, ‘straight’, and ‘thin’. The features correlating with <strong>the</strong> feminine gender<br />
are ‘large’, ‘round’, and ‘ample’. These features are also used to assign genders to<br />
those animals who play no role in <strong>the</strong> traditional mythology, and whose sex is hard to<br />
determine. Similar kinds of objects or animals are distinguished through Linguistic<br />
<strong>Gender</strong>: a smaller or thinner one is classed as masculine, and a larger one as feminine.<br />
Some examples are in 3.14:<br />
3.14 MASCULINE: SMALL, THIN FEMININE: LARGE, ROUND, AMPLE<br />
waliwalini ‘small ant’<br />
miaɹti ‘small pandanus’<br />
muŋkwani ‘small stone axe’<br />
man̪t̪ani ‘small stick’<br />
t̪uŋkwaliti ‘narrow-headed spear’<br />
waliwaliŋa ‘large ant’<br />
miaɹiŋa ‘large pandanus’<br />
muŋkwaŋa ‘large stone axe’<br />
man̪t̪aŋa ‘large stick’<br />
aɹawuniŋkiɹi ‘broad-headed spear’<br />
In Tiwi, ‘all trees are feminine…because <strong>the</strong>y are large, round and ample compared<br />
with mere sticks which are all masculine’. 11<br />
In Mali, a Baining language from East New Britain (Papua New Guinea), lower<br />
animates and inanimates are assigned to <strong>the</strong> feminine Linguistic <strong>Gender</strong> if <strong>the</strong>y are<br />
large in size. If <strong>the</strong>y are of average size, <strong>the</strong>y are assigned to <strong>the</strong> masculine Linguistic<br />
<strong>Gender</strong>. This is shown in 3.15.<br />
3.15 MASCULINE: AVERAGE SIZE FEMININE: UNUSUALLY LARGE SIZE<br />
thinem-ka ‘a fish’<br />
bang-ka ‘a house’<br />
lulen-ga ‘a dugout canoe’<br />
thinem-ki ‘a big fish’<br />
bang-ki ‘large building’<br />
lulen-gi ‘a long dugout’<br />
A three-way contrast between large (feminine), regular (masculine), and small<br />
(diminutive) nouns can be seen in some borrowings from Tok Pisin, an Englishbased<br />
Creole. A large serving spoon is a feminine sipun-ki, a dessert spoon (normal<br />
size) is a masculine sipun-ka, and a tiny teaspoon is a diminutive sipun-ini. Feminine<br />
plastik-ki refers to a largish plastic bag, or a large cocoa, flour, or rice bale. Masculine<br />
plastik-ka is a plastic container of smaller capacity, or a plastic shopping bag. 12
42 3 Round women and long men<br />
Hadza is a language isolate from central Tanzania. Masculine and feminine<br />
genders are marked on nouns, and also in agreement and anaphora. Humans are<br />
assigned to genders according to <strong>the</strong>ir sex. The choice of genders for o<strong>the</strong>r nouns<br />
follows a number of principles; among <strong>the</strong>m shape and size. In many instances, a<br />
masculine noun refers to a smaller object, and a feminine noun (marked with <strong>the</strong><br />
feminine suffix -ko) to a larger one, as shown in 3.16.<br />
3.16 MASCULINE NOUNS: SMALLER SIZE FEMININE NOUNS: LARGER SIZE<br />
ʔato ‘small axe’<br />
waɬi ‘testicle’<br />
muʔa ‘twig, thin stick’<br />
ʔato-ko ‘large axe’<br />
waɬi-ko ‘swollen testicle’<br />
muʔa-ko ‘stick’<br />
Size is not <strong>the</strong> only physical property relevant for choosing a Linguistic <strong>Gender</strong> in<br />
Hadza. Masculine nouns can refer to a longer and thinner object; <strong>the</strong>ir feminine<br />
counterparts are short and stout, as shown in 3.17.<br />
3.17 MASCULINE NOUNS: THIN, LONG, DEEP FEMININE NOUNS: SHORT, STOUT, SHALLOW<br />
ts'ahu ‘long, thin tail’<br />
ts'aho-ko ‘short, stout tail’<br />
ʔutume ‘long spear’ ʔutume-ko ‘short spear’ 13<br />
Just like o<strong>the</strong>r Omotic languages, Hamar has a feminine and a masculine gender<br />
which are chosen on <strong>the</strong> base of sex (or Natural <strong>Gender</strong>) for humans and sexdifferentiable<br />
animals, e.g. k'úl-ta ‘he goat’ and k'úllo ‘she-goat’. Hamar appears to<br />
be unique in its family in one way. The feminine Linguistic <strong>Gender</strong> is associated with<br />
large size and importance. Smaller items belong to <strong>the</strong> masculine Linguistic<br />
<strong>Gender</strong>—see 3.18.<br />
3.18 MASCULINE NOUNS: SMALL FEMININE NOUNS: LARGE<br />
amā ‘small field’<br />
nuta ‘individual fire’<br />
ammo ‘large field’<br />
nuno ‘large fire’<br />
Inanimates assigned to <strong>the</strong> masculine Linguistic <strong>Gender</strong> are less significant than<br />
feminine ones: for instance, goitā ‘path: masculine’ is a little used path, and goinno<br />
‘path:feminine’ is <strong>the</strong> main path. 14 We return to Lydall’s explanation for this feature<br />
of Hamar in §3.4.<br />
In Cantabrian Spanish 15 Linguistic <strong>Gender</strong> assignment to higher animates, and<br />
humans, generally follows <strong>the</strong>ir sex, or Natural <strong>Gender</strong>. We find pairs such as bellu<br />
‘steer’, bella ‘heifer’; lobu ‘male wolf ’, loba ‘female wolf ’; oveju ‘ram’,andoveja ‘ewe’.<br />
Inanimates and lower animates which are ei<strong>the</strong>r tall or of smaller size, or of narrow<br />
shape, or of vertical orientation are assigned to masculine Linguistic <strong>Gender</strong>. Objects<br />
which are unusually large, wide, horizontal, or small and squat are feminine. Some<br />
examples are in Table 3.1.<br />
The choice of Linguistic <strong>Gender</strong> for inanimates correlates with size in various<br />
forms of Standard Spanish. In many cases, <strong>the</strong> feminine noun refers to a larger entity,
3.3 Physical properties in Linguistic <strong>Gender</strong> choice 43<br />
TABLE 3.1. Semantic features in gender choice for nouns in Cantabrian Spanish<br />
MASCULINE<br />
MALE, e.g. hiju ‘son’<br />
SMALLER SIZE, e.g.<br />
anguilu ‘very small eel-like fish’<br />
tortu ‘small cake prepared in frying pan’<br />
montón ‘stack (of hay)’<br />
NARROW, e.g.<br />
picu ‘metal spike’<br />
VERTICAL, e.g.<br />
coteru ‘rising mountain meadow’<br />
castru ‘large protruding rock’<br />
TALL, e.g. tree names:<br />
manzanu ‘apple tree’<br />
naranju ‘orange tree’<br />
FEMININE<br />
FEMALE, e.g. hija ‘daughter’<br />
UNUSUALLY LARGE OR AVERAGE, e.g.<br />
anguila ‘eel’<br />
torta ‘larger cake prepared in oven’<br />
montona ‘very large stack of hay’<br />
WIDE, e.g.<br />
pica ‘hammerlike instrument with a wide head’<br />
HORIZONTAL, e.g.<br />
cotera ‘mountain meadow (flat)’<br />
castra ‘large flat rock (horizontal)’<br />
SQUAT, SMALL, e.g. fruit names:<br />
manzana ‘apple’<br />
naranja ‘orange fruit’<br />
e.g. garbanzo ‘chick-pea’ (masculine), garbanza ‘large species of chick-pea’ (feminine),<br />
panero ‘small basket’ (masculine), panera ‘large oval basket’ (feminine), cesta<br />
‘large hand basket’, cesto ‘smaller basket’. In some cases, a masculine ra<strong>the</strong>r than a<br />
feminine noun would refer to a larger entity, e.g. barreno ‘large-size auger (a boring<br />
tool)’, barrena ‘auger’. 16<br />
3.3 Physical properties in Linguistic <strong>Gender</strong> choice: contrasting<br />
<strong>the</strong> two scenarios<br />
The choice of Linguistic <strong>Gender</strong> based on shape, size, and o<strong>the</strong>r physical properties of<br />
an entity involves meanings summarized in Table 3.2.<br />
A complex of physical features is at play in <strong>the</strong> choice of Linguistic <strong>Gender</strong>. They<br />
include shape, dimensionality, and size. Size as a gendering parameter offers contradictory<br />
results. In some languages <strong>the</strong> feminine Linguistic <strong>Gender</strong> is applied to larger<br />
objects, in o<strong>the</strong>rs, to smaller ones. Shape and dimensionality are more consistent.<br />
Feminine Linguistic <strong>Gender</strong> subsumes objects that are round and squat, and also<br />
horizontal. Masculine Linguistic gender includes narrow and elongated, and<br />
also vertical things.<br />
Shape and dimensionality are salient semantic parameters in all noun categorization<br />
devices (or classifiers). 17 Linguistic <strong>Gender</strong> is one of <strong>the</strong>se linguistic categorization<br />
of entities. Orientation (vertical versus horizontal) is a preferred semantic<br />
parameter for verbal classifiers, and especially classificatory verbs. A classificatory
44 3 Round women and long men<br />
TABLE 3.2. Physical properties in Linguistic <strong>Gender</strong> choice<br />
FEMININE GENDER MASCULINE GENDER EXAMPLES DISCUSSED HERE<br />
short, squat, wide<br />
objects<br />
small, flat objects<br />
large, wide, round<br />
objects<br />
big, tall, long, slender<br />
objects<br />
large, tridimensional<br />
objects<br />
small, straight, thin<br />
objects<br />
Languages of New Guinea<br />
(Manambu, Iatmul, Gala, Yalaku,<br />
Kwoma, Wosera, Boiken, Alamblak,<br />
Sare, Olo, Wära, Savosavo);<br />
Afroasiatic languages (including<br />
Cushitic, Omotic, Amharic (Semitic));<br />
East Nilotic;<br />
Khwe (Central Khoisan)<br />
Abau (isolate, New Guinea area)<br />
Tiwi (Australian region)<br />
larger objects smaller objects Mali (Baining); Hadza (isolate);<br />
Hamar (Omotic)<br />
larger, wide, horizontal,<br />
squat objects<br />
smaller, narrow,<br />
vertical, tall objects<br />
Cantabrian Spanish<br />
positional verb ‘sit’ can be used with referents judged to be squat and horizontal,<br />
including women. A classificatory verb ‘stand’ is used of tall, large, and vertical<br />
referents, including men. This type of system has been described for numerous<br />
Papuan languages. 18 In §5.1, we turn to <strong>the</strong> ways in which gender meanings are<br />
expressed through noun categorization devices o<strong>the</strong>r than Linguistic <strong>Gender</strong>.<br />
The shape and dimensionality associated with <strong>the</strong> choice of masculine Linguistic<br />
<strong>Gender</strong> can be conceived as involving ‘phallic’ imagery. The objects associated with<br />
feminine Linguistic <strong>Gender</strong> will <strong>the</strong>n be in some sense ‘opposite’. In this way,<br />
Linguistic <strong>Gender</strong> choice involves projecting physical characteristics typical of female<br />
and male Natural <strong>Gender</strong>s onto <strong>the</strong> inanimate world in general. Such Linguistic<br />
<strong>Gender</strong> assignment is based on ‘important property’ (in line with examples in §2.2.1).<br />
The roots of this ‘important property’ lie in conventionalized images associated with<br />
<strong>the</strong> Natural <strong>Gender</strong>—to which we turn in Chapter 8.<br />
3.4 Beyond mere physique: attitude, value, and importance<br />
in Linguistic <strong>Gender</strong><br />
Changing Linguistic <strong>Gender</strong> of an inanimate may correlate with speakers’ attitudes.<br />
Linguistic <strong>Gender</strong> may be a token of endearment, or disdain—see §3.4.1. Or it may<br />
reflect value and cultural importance—see §3.4.2.
3.4 Beyond mere physique 45<br />
3.4.1 Endearment and disdain through Linguistic <strong>Gender</strong><br />
Small roundish objects can be regarded as particularly cute, or dear to <strong>the</strong> speaker.<br />
Overtones of endearment are common to many diminutive forms in <strong>the</strong> world’s<br />
languages: in Portuguese, cabecinha ‘little head’ is a diminutive of cabeça ‘head’; this<br />
word is often used to mean ‘dear little head’,or‘head of a dear little person’, such as a<br />
child or a baby. 19<br />
We can recall, from §3.1, that <strong>the</strong> feminine Linguistic <strong>Gender</strong> marker doubles as a<br />
diminutive across <strong>the</strong> Afroasiatic language family. The association between <strong>the</strong><br />
feminine Linguistic <strong>Gender</strong> and endearment is pervasive. In Amharic, ‘almost any<br />
noun can be treated as feminine if occasion calls for it…Thus, yih mas'haf “this book<br />
(masculine)” is normal, while yicc mas'haf “this book (feminine)” might refer to a<br />
favourite little booklet’ (Cowley et al. 1976: 84). In Oromo (see §3.1, and Clamons<br />
1993: 276–8) something considered cute, or regarded affectionately, will be assigned<br />
to <strong>the</strong> feminine Linguistic <strong>Gender</strong>. An example is at 3.19.<br />
3.19 waan-ti tun jiidh-tuu<br />
thing-FEM.SUBJECT.TOPIC this.fem wet-fem<br />
‘This (cute little) thing is wet’<br />
The masculine Linguistic <strong>Gender</strong> is <strong>the</strong> opposite of feminine: it will have pejorative<br />
overtones. A ‘nasty thing’ belongs to <strong>the</strong> masculine Linguistic <strong>Gender</strong>:<br />
3.20 waan-i xun jiidh-aa<br />
thing-MASC.SUBJECT.TOPIC this.masc wet-masc<br />
‘This (nasty) thing is wet’<br />
The feminine Linguistic <strong>Gender</strong> in Ayt Seghrushen, a Berber language, has<br />
overtones of affection: <strong>the</strong> masculine afus is ‘hand’, and <strong>the</strong> feminine t-fus-tt is ‘little<br />
baby hand’ (as we can recall from example 3.4). The masculine Linguistic <strong>Gender</strong><br />
may indicate that something is grotesquely big, ugly, or hard to handle. The form<br />
t-ɣat̍t̍(feminine) refers to a female goat. The masculine counterpart is ɣad̍(masculine)<br />
which means ‘big female goat, difficult to handle’. The feminine noun tit̍t̍‘eye’<br />
has a masculine counterpart at̍t̍aw ‘very big eye’ with pejorative overtones. The<br />
feminine tamziyda ‘mosque’ can be made masculine; <strong>the</strong> resulting form amziyda<br />
means ‘a ridiculously big mosque’.<br />
We mentioned in §2.2.2, that gender choice in Palikur is determined by <strong>the</strong> form<br />
of <strong>the</strong> noun, and also its meaning. Feminine Linguistic <strong>Gender</strong> for lower animates<br />
can be associated with small size. Masculine Linguistic <strong>Gender</strong> implies larger size,<br />
and may have overtones of unpleasantness. Rat is a small animal, but is assigned<br />
masculine gender because it is looked upon as dirty and bad. But a cute little baby rat<br />
would be referred to as feminine. Turtles are usually feminine; but a turtle which is a<br />
nuisance and has to be got rid of would be referred to as masculine; all insects are
46 3 Round women and long men<br />
masculine in spite of <strong>the</strong>ir small size, according to an explanation by a native<br />
consultant, ‘because none of <strong>the</strong>m are any good for food and all <strong>the</strong>y do is bo<strong>the</strong>r<br />
people, eat crops and cause sickness’. 20<br />
In Cantabrian Spanish, <strong>the</strong> masculine Linguistic <strong>Gender</strong> has overtones of thinness,<br />
meagre proportions, and generally smaller size than <strong>the</strong> feminine Linguistic <strong>Gender</strong><br />
(Table 3.1). Changing Linguistic <strong>Gender</strong> correlates with <strong>the</strong> speaker’s attitude.<br />
A coarse and bumpy road would be referred to as masculine, and a smooth one as<br />
feminine. The term oveju ‘male sheep’ can be used in a derogatory sense to refer to a<br />
particularly meagre exemplar of <strong>the</strong> species: a cattle raiser was reminiscing about a<br />
particularly bad meal he’d once had, consisting of a sheep’s head, a few green<br />
vegetables, and a few potatoes, and remarked that <strong>the</strong> sight of <strong>the</strong> head of that<br />
oveju (male sheep) ‘peering at him from <strong>the</strong> pot was something he would never<br />
forget’. In Spanish, <strong>the</strong> feminine form oveja is normally used in a generic sense; and<br />
here a ‘shock or humour’ has resulted from <strong>the</strong> selection of <strong>the</strong> male form based not<br />
on <strong>the</strong> sex of <strong>the</strong> unfortunate animal, but as a ‘deprecative expression of <strong>the</strong><br />
revulsion’ felt by <strong>the</strong> speaker (Holmquist 1991: 60). Along similar lines, hiju míu,<br />
literally, ‘my son’, was used to refer to a young girl about 12 years of age, a female not<br />
yet developed, in a deprecatory reference. A little girl was treated on a par with a nonhuman.<br />
The masculine Linguistic <strong>Gender</strong> of non-humans in Lokono, an Arawak language<br />
from Guyana, has overly positive overtones. The feminine Linguistic <strong>Gender</strong> is <strong>the</strong><br />
opposite. Animals and birds which are thought of as having a ‘positive personality’<br />
are masculine—<strong>the</strong>y include turtles and hummingbirds. Domestic animals to which<br />
speakers have a special attachment, for instance, a dog, are masculine; however, one’s<br />
neighbour’s dog (whom one does not particularly like) is more likely to be feminine.<br />
Nice and cute animals are masculine, while bigger animals are feminine (van Baarle<br />
and Sabajo 1997). This is—so far—<strong>the</strong> only example of a language I have been able to<br />
find where masculine Linguistic <strong>Gender</strong> is associated with endearment and special<br />
affection. Linguistic <strong>Gender</strong> reversals for humans have similar overtones—we return<br />
to <strong>the</strong>se in §7.2.<br />
Things can be seen differently. A small object may be treated as substandard and<br />
not good enough. Feminine Linguistic <strong>Gender</strong> for some nouns in Ayt Seghrushen<br />
has pejorative overtones. An Arabic loan (originally from French camion) lkamyun<br />
(masculine) means ‘truck’. The Ayt Seghrushen form akaymun (masculine) means<br />
‘gigantic truck’; <strong>the</strong> feminine form t-kamyun-t means ‘little worthless truck’. The<br />
masculine form of ‘female goat’ has overtones of ‘a big female goat difficult to<br />
handle’. 21 We saw in 3.10 how in Benchnon, an Omotic language, a shoe assigned<br />
to <strong>the</strong> feminine Linguistic <strong>Gender</strong> is <strong>the</strong> one that <strong>the</strong> speaker doesn’t care about.<br />
Table 3.3 summarizes <strong>the</strong> opposite attitudes of speakers, encoded in <strong>the</strong> feminine<br />
and masculine Linguistic <strong>Gender</strong>s of animals, inanimates, and—in Spanish—young<br />
children.
3.4 Beyond mere physique 47<br />
TABLE 3.3. Endearment and disdain through Linguistic <strong>Gender</strong>s<br />
FEMININE GENDER MASCULINE GENDER EXAMPLES<br />
endearment, affection<br />
negative overtones (being<br />
grotesquely big, nasty, or<br />
meagre and substandard)<br />
Oromo, Palikur,<br />
Ayt Seghrushen,<br />
Cantabrian Spanish<br />
substandard, derogatory neutral Ayt Seghrushen, Benchnon<br />
negative personality, dislike positive personality, affection Lokono<br />
In summary: small size and <strong>the</strong> associated Linguistic <strong>Gender</strong> can express cuteness<br />
and endearment. Or it may involve ‘belittling’ something. The opposite Linguistic<br />
<strong>Gender</strong> may <strong>the</strong>n refer to a substandard and annoying entity. The overtones of<br />
endearment or a negative attitude for shape- and size-based Linguistic <strong>Gender</strong> are<br />
associated with an evaluative aspect of size: something small and thin can be nice and<br />
dear, or no good at all. Something too big tends to be ridiculously grotesque. This is<br />
what we have so far seen for non-humans; <strong>the</strong> effects of Linguistic <strong>Gender</strong> reversals<br />
for men and women are <strong>the</strong> topic of §§7.2–3. We now turn to <strong>the</strong> associations<br />
between Linguistic <strong>Gender</strong>, importance, and cultural value, and <strong>the</strong>ir possible links to<br />
Social <strong>Gender</strong> concepts, and societal values.<br />
3.4.2 Value and importance in Linguistic <strong>Gender</strong><br />
Typical features of Social <strong>Gender</strong> can be projected onto entities which have no Natural<br />
<strong>Gender</strong>, in <strong>the</strong> form of Linguistic <strong>Gender</strong> assignment. The choice of Linguistic <strong>Gender</strong><br />
may reflect <strong>the</strong> place of <strong>the</strong> referent in <strong>the</strong> conceptual system of beliefs and rituals. We<br />
start with an example of how <strong>the</strong> masculine Linguistic <strong>Gender</strong>—associated with large<br />
size of an object—may reflect ritual life, and cultural importance.<br />
Manambu society—just like o<strong>the</strong>r groups in <strong>the</strong> Sepik—is male-oriented. Traditional<br />
ownership of land and of valuables is inherited through one’s fa<strong>the</strong>r’s line. So<br />
is membership of a clan. When a woman gets married, she is expected to move in<br />
with her husband’s line. (A husband who moves in with his wife after marriage can<br />
be ridiculed as a ‘female husband’ of sorts: see §7.2.) The centrepiece of Manambu<br />
culture is, traditionally, male cults (including initiation rituals, now obsolete).<br />
A man’s reputation and status used to depend primarily on esoteric knowledge of<br />
totemic names, names belonging to <strong>the</strong> subclan, spells, and myths. Male proficiency<br />
in <strong>the</strong>se matters was crucial for men’s prowess in ritual debates—including debates<br />
on ownership of names and of <strong>the</strong> land—and also magic and sorcery. In <strong>the</strong> past,<br />
men’s central role used to depend on achievements in wars against neighbouring
48 3 Round women and long men<br />
groups and in head-hunting. 22 According to <strong>the</strong> tradition, men have access to<br />
esoteric information concerning secret names and spells, and have <strong>the</strong> upper hand<br />
in political and ritual affairs. Women are not supposed to have access to highly<br />
valued traditional knowledge (though some of <strong>the</strong>m in fact do), and are denied active<br />
participation in traditional activities such as <strong>the</strong> yam ritual and name debates.<br />
This traditional importance of ‘male-hood’, or masculine Social <strong>Gender</strong>, is iconically<br />
reflected in <strong>the</strong> assignment of masculine gender based on ‘important property’.<br />
Names of rituals are masculine. It is not appropriate to refer to Saki ‘name debate’<br />
with feminine gender (notwithstanding <strong>the</strong> fact that name debates may involve<br />
female and male names, and can be relatively short in time). Similarly, Kəkətəp,<br />
<strong>the</strong> ritual which involves mortuary payments, is also masculine (no matter whe<strong>the</strong>r<br />
<strong>the</strong> deceased was a man or a woman, or how many valuables change hands).<br />
Along similar lines, kara:b ‘large ceremonial house’ and sa:y ‘ceremonial house for<br />
uninitiated men’ are both masculine. Ceremonial houses in present-day Avatip (<strong>the</strong><br />
main village of <strong>the</strong> Manambu) are smallish (much smaller than normal dwellings).<br />
But referring to a ceremonial house with feminine gender is considered unacceptable<br />
and even offensive.<br />
<strong>Gender</strong> assignment based on importance is relevant for terms to do with <strong>the</strong> art of<br />
speech. The term ma:j ‘story’ requires masculine agreement, if it is a traditional story,<br />
or if it refers to head-hunting raids performed by men, no matter how long it is. If it<br />
refers to a casual story or a biography of someone it is likely to be feminine. Similarly,<br />
ba:gw ‘performance, dance’ is masculine only when it refers to a traditional activity.<br />
A casual performance—such as a dance party organized by Manambu expatriates in<br />
Canberra, or someone casually playing <strong>the</strong> guitar—is feminine. ‘<strong>Gender</strong>-proud’<br />
speakers of Manambu refer to unimportant stories as ‘woman-type’ stories, and to<br />
important ones as ‘men-type’.<br />
Association with ‘male’ and ‘female’ Social <strong>Gender</strong> attributes determines <strong>the</strong><br />
Linguistic <strong>Gender</strong> of culturally important notions, and objects. Certain cultural<br />
concepts have inherently ‘female’ reference, and are always assigned to <strong>the</strong> feminine<br />
gender. O<strong>the</strong>rs are associated with masculinity, and trigger masculine agreement.<br />
The notion of ja:m ‘a set of hereditary magical and ritual powers’ is personified as a<br />
named female spirit, whose role is to punish incest and violation of <strong>the</strong> marriage<br />
principles. Each subclan has a væy, its ancestor, with a literal meaning ‘spear’. The<br />
connotations of væy are masculine and phallic (which goes toge<strong>the</strong>r with its elongated,<br />
‘masculine’, shape), while ja:m is represented as a womb. Toge<strong>the</strong>r væy and ja:m<br />
‘signify <strong>the</strong> “male” and “female” aspects of a group’s social identity’ (Harrison 1990:<br />
33). This agrees with <strong>the</strong> assignment of masculine and feminine genders to <strong>the</strong>m.<br />
It is traditionally believed that human bones (ap) are formed from fa<strong>the</strong>r’s semen<br />
and transmitted through one’s fa<strong>the</strong>r. In contrast, blood (ñiki) derives from mo<strong>the</strong>r’s<br />
womb blood and is transmitted by matrifiliation (Harrison 1990: 33). Consequently,<br />
ap ‘bone’ is assigned masculine gender, while ñiki ‘blood’ is usually feminine (unless
3.4 Beyond mere physique 49<br />
it comes in large quantities). The noun ap also carries <strong>the</strong> connotations of centrality<br />
and importance, ano<strong>the</strong>r corollary of masculine gender. Thus, speakers derive <strong>the</strong><br />
name of Avatip, considered <strong>the</strong> most important of <strong>the</strong> four Manambu villages, from<br />
ap-a təp (bone-LINKER village) ‘<strong>the</strong> strong, large, central village’ (lit. <strong>the</strong> bone-village).<br />
Cultural roles—and <strong>the</strong> status of <strong>the</strong> masculine Social <strong>Gender</strong>—can be considered<br />
<strong>the</strong> basis of Linguistic <strong>Gender</strong> choice. This is also how speakers explain <strong>the</strong>m: an<br />
important story relating a clan’s origin or wartime exploits will be referred to as<br />
‘man-type’, and a biography or a fairy-tale as a ‘woman-type’. Similarly, in Yangoru<br />
Boiken, a language related to Manambu, ‘size is a gendering dimension for referents’:<br />
large objects are masculine and small ones are feminine. Masculine gender is<br />
associated with strength, importance, and value, and gender is assigned to culturally<br />
significant rituals, stories, and objects in numerous o<strong>the</strong>r languages spoken in<br />
societies focused on male activities and cults. 23<br />
The male ‘supremacy’ in cultural activities, and <strong>the</strong> status of masculine Social<br />
<strong>Gender</strong> is reflected in Linguistic <strong>Gender</strong> even if <strong>the</strong>re are no obvious correlations<br />
between <strong>the</strong> assignment of Linguistic <strong>Gender</strong> and an object’s size. In Oyda, an<br />
Omotic language, Linguistic <strong>Gender</strong> is assigned to ceremonial objects—drums—<br />
based on <strong>the</strong>ir importance. A small roundish drum—described as ‘<strong>the</strong> major one,<br />
<strong>the</strong> leader of all drums’ is assigned masculine Linguistic <strong>Gender</strong>; and a very large<br />
oval-shaped drum which has less importance belongs to <strong>the</strong> feminine Linguistic<br />
<strong>Gender</strong> (Amha 2013).<br />
Hamar, from <strong>the</strong> Omotic subfamily of Afroasiatic, is unusual for its family in that<br />
<strong>the</strong> feminine, ra<strong>the</strong>r than <strong>the</strong> masculine, Linguistic <strong>Gender</strong> correlates with <strong>the</strong><br />
referent’s importance—as we saw in §3.2. In Lydall’s words (1988: 78),<br />
feminine gender is used to indicate large and major things ra<strong>the</strong>r than small and unimportant<br />
ones, while masculine gender is used to indicate small and minor things ra<strong>the</strong>r than big and<br />
important ones…Upon telling this to fellow linguists and anthropologists, I have been met<br />
with expressions of disbelief or astonishment. <strong>How</strong> can <strong>the</strong> Hamar have arrived at such a<br />
formula? <strong>How</strong> can a society in which men generally enjoy a higher status than women speak a<br />
language in which masculine gender is equated with smallness?<br />
To explain this apparent puzzle, Lydall (1988: 89) hypo<strong>the</strong>sizes that in agricultural<br />
and o<strong>the</strong>r tasks in <strong>the</strong> life of <strong>the</strong> Hamar,<br />
any event or enterprise requires, on <strong>the</strong> one hand, male initiative and determination, and, on<br />
<strong>the</strong> o<strong>the</strong>r hand, female performance and substantiation. Male initiative and determination is<br />
conceived of as small in scale, and occasional, while female performance and substantiation<br />
is large-scale and frequent. These ideas explain why <strong>the</strong> masculine form of nouns is used to<br />
indicate small, infrequently used or seldom found items, while <strong>the</strong> feminine form is used<br />
to indicate large, frequently used or commonly found items.<br />
An association of a Linguistic <strong>Gender</strong> with cultural importance and value appears<br />
to mirror stereotypes of Social <strong>Gender</strong>. But a word of warning is in order: as Mithun
50 3 Round women and long men<br />
(2014: 159) put it, ‘though grammatical gender may indeed reflect culture, <strong>the</strong><br />
relationship is not necessarily immediate and direct’. In Manambu, nouns væy<br />
‘spear’ and ap ‘bone’ are assigned masculine gender—and this also can be thought<br />
of as based on <strong>the</strong>ir longish size and shape (and not necessarily associations with<br />
‘masculinity’). Similarly, ya:l ‘womb, belly’—which is typically feminine—is associated<br />
with women. Since it is also typically ‘round’ in shape, its gender assignment can<br />
be attributed to a pervasive correlation between shape and feminine Linguistic<br />
<strong>Gender</strong>. Moon and stars are mythological women, but <strong>the</strong>y are also roundish in<br />
<strong>the</strong>ir shape.<br />
Which one comes first—gender choice by shape, or by mythology? Which one is<br />
contingent upon <strong>the</strong> o<strong>the</strong>r? We can recall that ab ‘head’ is feminine, due to its round<br />
shape, and despite its perceived importance for mental processes (which correlates<br />
with ‘masculinity’ as exponent of ‘importance’). This may imply that shape is primary<br />
in <strong>the</strong> gender assignment of ‘spear’, ‘bone’, ‘womb; belly’, ‘moon’, and‘star’, and that<br />
<strong>the</strong> cultural extension to masculinity or femininity is, at least historically, a corollary of<br />
<strong>the</strong> erstwhile shape-based association. But synchronically, we are faced with Linguistic<br />
<strong>Gender</strong> choice which reflects a male orientation of values.<br />
Correlations between lack of cultural importance and small size in Manambu do<br />
not imply a subservient and secondary position of women in <strong>the</strong> society. The spheres<br />
of men’s and women’s knowledge and experience are complementary and separate.<br />
We return to this in §10.1 and §10.5.1.<br />
NOTES AND SOURCES<br />
1. Fur<strong>the</strong>r details on gender assignment in Manambu are in Aikhenvald (2008; 2012a).<br />
2. Bruce (1984: 97).<br />
3. Sumbuk (1999: 109–23).<br />
4. Riisto Sarsa, p.c. on Wära, McGregor and McGregor (1982: 55) on Olo, Wegener (2012:<br />
63–5) on Savosavo. Similar systems have been described for Abau, a linguistic isolate<br />
spoken to <strong>the</strong> west of <strong>the</strong> Sepik River in <strong>the</strong> adjacent Sandaun Province (Lock 2011:<br />
47–52), Kwoma, from <strong>the</strong> small Kwoma-Nukuma family (Bowden 1997, Kooyers 1974),<br />
and Iatmul, Gala, Wosera, and Boiken, from <strong>the</strong> Ndu family (see references in Aikhenvald<br />
2012a). In Yalaku, closely related to Manambu, a small object, a bird, or an animal will be<br />
referred to with feminine gender, and a larger one with masculine gender. A very small male<br />
bat was described as feminine, and referred to as tu-il (man-3fem.sg), literally ‘she is a man’.<br />
5. See Kossmann (2014), on Ayt Seghrushen, and fur<strong>the</strong>r intricacies of <strong>the</strong> system; see<br />
Diakonoff (1990), for an overview of Afroasiatic languages.<br />
6. See Seyoum (2008: 43), Hellenthal (2010: 153) on Sheko, Amha (2001: 71–2) on Maale,<br />
Amha (2012: 444) on Wolaitta; Rapold (2006: 182) on Benchnon. See Sasse (1974: 419) on<br />
Galab (or Dasenech: Cushitic); Castellino (1975) on links between small size and feminine<br />
gender across Cushitic languages, and Leger (1998: 207) on Kwami and o<strong>the</strong>r Chadic<br />
languages.
3.4 Beyond mere physique 51<br />
7. Hoben (1976: 287). Similarly, in Amharic, an Ethio-Semitic language from Ethiopia, a<br />
noun can be ‘masculine or feminine according to whe<strong>the</strong>r its referent is considered to<br />
be large and distant (masculine) or small and intimate (feminine)’: see Cowley et al. (1976:<br />
84), and Pankhurst (1992: 168).<br />
8. See Dimmendaal (1983: 219–21) for Turkana, and similar examples from <strong>the</strong> related<br />
Maasai.<br />
9. See Heine (1982a: 210–11), Kilian-Hatz (2008: 41–2), Köhler (1962) on Khwe, Heine<br />
(1982a: 205) on Camus and Katcha.<br />
10. See Klumpp (1990: 67).<br />
11. See Osborne (1974: 51).<br />
12. See Stebbins (2005: 102).<br />
13. See Edenmyr (2004: 14–17).<br />
14. See Lydall (1988: 78). Note that in this source she refers to <strong>the</strong> masculine as ‘minor’ and to<br />
<strong>the</strong> feminine as ‘major’ gender.<br />
15. Also known as Montañés. See Holmquist (1991); Pountain (2005); Bergen (1980: 52–4);<br />
also see <strong>the</strong> historical explanation <strong>the</strong>re and in Priestly (1983); see Butt and Benjamin<br />
(2004: 1–15) for Linguistic <strong>Gender</strong> in Standard Spanish. Similar examples from Italian are<br />
in Ferrari (2005: 39–40), Bonfante (1946: 847) (disputed by Ervin 1962).<br />
16. That nouns belonging to <strong>the</strong> feminine Linguistic <strong>Gender</strong> (and ending in -a) denote objects<br />
of larger size may be due to <strong>the</strong>ir origin: many such nouns may have originally been neuter<br />
plural or collectives (which also ended in -a in Classical Latin) and later undergone<br />
reinterpretation (Pountain 2005: 334–5); see also Kahane and Kahane (1948–9) on <strong>the</strong><br />
productivity of <strong>the</strong> ‘feminine augmentative’ in Romance languages.<br />
17. An outline of preferred semantic parameters of various noun categorization devices is in<br />
Aikhenvald (2000: 306).<br />
18. See Lang (1975) on Enga, Brown (1981) on Waris; fur<strong>the</strong>r references in Aikhenvald (2000:<br />
158–60, 166–8).<br />
19. See Mendes (2014a) on how diminutives in Brazilian Portuguese may convey <strong>the</strong> idea of<br />
something nice and cosy; see also Grandi (2002) on meaning overtones of diminutives and<br />
augmentatives across European languages.<br />
20. See Aikhenvald and Green (2011).<br />
21. See Kossmann (2014), on Ayt Seghrushen.<br />
22. Aikhenvald (2009).<br />
23. Nouns belonging to <strong>the</strong> feminine Linguistic <strong>Gender</strong> in Khwe, a Central Khosian language<br />
(Kilian-Hatz 2008: 42), Galab (or Dasenech), a Cushitic language (Sasse 1974), and Angave,<br />
an Angan language from Papua New Guinea (Speece n.d.) are characterized as ‘small, less<br />
important’. Nouns of masculine Linguistic <strong>Gender</strong> are big and important.
4<br />
What are Linguistic <strong>Gender</strong>s<br />
good for?<br />
Linguistic <strong>Gender</strong>s, as a way of grammatical categorization of nouns, impose a way of<br />
speaking on men and women. The Natural <strong>Gender</strong> of a speaker of Modern Hebrew,<br />
Russian, Modern Greek, and many o<strong>the</strong>r languages with pervasive gender agreement<br />
will be obvious from <strong>the</strong> gendered forms <strong>the</strong>y will use. A statement ‘I love you’ in<br />
Modern Hebrew has to be specified for gender of ‘I’ (on <strong>the</strong> verb) and ‘you’ (in <strong>the</strong><br />
pronoun). A woman will say to a man, Ani ohevet otxa ‘I love(feminine) you(masculine)’,<br />
and a man will say to a woman, Ani ohev otax ‘I love(masculine) you(feminine)’.<br />
A failure to use <strong>the</strong> appropriate Linguistic <strong>Gender</strong> in a language can lead to<br />
spectacular results. In a short story ‘The entertaining episode of <strong>the</strong> article in<br />
question’ (1925), detective Lord Peter Wimsey solves a burglary mystery by use of<br />
gender. At a London railway station, Lord Peter overhears what appears to be a<br />
young French girl saying to her partner Me prends-tu pour un imbécile? (‘Do you take<br />
me for an imbecile?). Lord Peter notices <strong>the</strong> masculine form of <strong>the</strong> indefinite article<br />
‘un’ (ra<strong>the</strong>r than <strong>the</strong> feminine form ‘une’), and this serves to reveal that <strong>the</strong> supposed<br />
girl is in reality a renowned male jewel thief. As Lord Peter explains, ‘in France, every<br />
male child is brought up to use masculine adjectives about himself ’. For <strong>the</strong> thief, it<br />
wasn’t possible to kick this habit. <strong>Gender</strong> gave <strong>the</strong> game away. 1<br />
Linguistic <strong>Gender</strong> serves to differentiate <strong>the</strong> meanings of a noun (and help<br />
distinguish males from females), as we will see in §4.1. It helps understand who<br />
does what to whom—that is, track referents in discourse—this is what we will see in<br />
§4.2. <strong>Gender</strong> individuates some participants and backgrounds o<strong>the</strong>rs, and highlights<br />
<strong>the</strong>ir properties. Linguistic <strong>Gender</strong> helps enrich <strong>the</strong> lexicon (for one thing, creating<br />
male and female pairs). All <strong>the</strong>se functions are shared with o<strong>the</strong>r noun categorization<br />
devices—especially numeral classifiers, noun classifiers, and classifiers on verbs. 2<br />
Linguistic <strong>Gender</strong>s are a means of categorizing entities. In <strong>the</strong> first place, <strong>the</strong>y<br />
differentiate between men and women. In Portuguese, <strong>the</strong> forms menino ‘boy’ and<br />
menina ‘girl’ differ in <strong>the</strong>ir linguistic gender marking and in agreement. Meanings<br />
of genders in German reflect what Zubin and Köpcke (1986) called ‘hierarchical<br />
categorization’ of nouns. The neuter gender tends to express general concepts, so<br />
<strong>How</strong> <strong>Gender</strong> <strong>Shapes</strong> <strong>the</strong> <strong>World</strong>. First edition. Alexandra Y. Aikhenvald.<br />
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.
4.1 Variable choice of Linguistic <strong>Gender</strong> 53<br />
that superordinate nouns tend to be neuter. More specific and less general terms will<br />
be assigned to non-neuter gender (see Table 2.3 in Chapter 2). So, <strong>the</strong> superordinate<br />
das Tier ‘animal (fauna)’ is neuter. More basic level terms der Fisch ‘fish’ and der Vogel<br />
‘bird’ are masculine, and <strong>the</strong> names of species, including die Eule ‘owl’ and der Karpfen<br />
‘carp’, are feminine and masculine respectively. 3<br />
In many familiar languages, each noun belongs to just one gender. <strong>Gender</strong> membership<br />
is fixed—like a non-removable birthmark. In German, die Sonne ‘<strong>the</strong> sun’ is<br />
feminine. It is nonsensical to try and refer to it as *der Sonne, with <strong>the</strong> masculine<br />
gender. Yet in many languages gender of a noun can vary—and this highlights <strong>the</strong><br />
versatility of what <strong>the</strong> genders express. 4<br />
4.1 Variable choice of Linguistic <strong>Gender</strong><br />
Some nouns have one form for females and for males. These are sometimes called<br />
epicenes. 5 Different forms of Linguistic <strong>Gender</strong>—in agreement or in anaphora—will<br />
disambiguate <strong>the</strong>ir sex. A baby can be referred to as ‘she’ or as ‘he’ depending on<br />
whe<strong>the</strong>r it is a girl or a boy. In many languages, epicene nouns occur with forms of<br />
different genders, as a means of <strong>the</strong>ir disambiguation. In Dyirbal, jaja ‘baby’ and<br />
ñalŋga ‘young child’ can be specified as ei<strong>the</strong>r masculine or feminine, depending on<br />
<strong>the</strong>ir sex. Similarly, inamatewe ‘child’ in Jarawara, an Amazonian language with two<br />
genders, can be cross-referenced as ei<strong>the</strong>r masculine or feminine as appropriate.<br />
Archi, a north-east Caucasian language, has four Linguistic <strong>Gender</strong>s. The noun lo<br />
‘youngster’ can be assigned to three of <strong>the</strong>se. When assigned to <strong>the</strong> masculine class 1<br />
it means ‘boy’. When assigned to <strong>the</strong> feminine class 2, it means ‘girl’, and when<br />
assigned to class 4 (which includes animals), it means ‘young animal’.<br />
In Dyirbal bimu is both ‘fa<strong>the</strong>r’s elder sister’ (taking feminine gender marker,<br />
balan) and ‘fa<strong>the</strong>r’s elder bro<strong>the</strong>r’ (used with <strong>the</strong> masculine marker bayi). The terms<br />
for ‘child’ in Dyirbal distinguish <strong>the</strong> sex of <strong>the</strong> parent: daman is a child (a son or a<br />
daughter) of mo<strong>the</strong>r, or mo<strong>the</strong>r’s sibling, and galbin is child of fa<strong>the</strong>r or fa<strong>the</strong>r’s<br />
sibling. <strong>Gender</strong> markers help distinguish between sons and daughters:<br />
4.1 balan daman ‘mo<strong>the</strong>r’s daughter’ balan galbin ‘fa<strong>the</strong>r’s daughter’<br />
bayi daman ‘mo<strong>the</strong>r’s son’ bayi galbin ‘fa<strong>the</strong>r’s son’<br />
Many kinship terms in Manambu—including ma:m ‘elder sibling’, ñamus<br />
‘younger sibling’, babay ‘maternal grandparent’, yanan ‘child of one’s daughter’<br />
and gwa:l ‘child of one’s son’—can be masculine or feminine. Agreement gender<br />
will disambiguate <strong>the</strong>m: one will say wuna ma:m (my+feminine.sg elder sibling) ‘my<br />
elder sister’ and wuna-də ma:m (my-masculine.sg elder sibling) ‘my elder bro<strong>the</strong>r’.<br />
(A sister will be referred to as lə ‘she’, and a bro<strong>the</strong>r as də ‘he’.)<br />
Similarly, nouns denoting professions may show variable agreement depending on<br />
whe<strong>the</strong>r <strong>the</strong>y refer to men or to women. Doctor in English can be referred to as he or
54 4 What are Linguistic <strong>Gender</strong>s good for?<br />
she depending on <strong>the</strong> sex of <strong>the</strong> person. In Russian, vrac᷈‘doctor’ or sudja ‘judge’ will<br />
require feminine or masculine agreement depending on <strong>the</strong> person’s sex.<br />
The degree of variability in Linguistic <strong>Gender</strong> depends on <strong>the</strong> language. In Dyirbal,<br />
each name of an animal has a fixed class membership; however, exceptionally, noun<br />
class assignment can be changed to stress <strong>the</strong> sex of a particular animal, e.g. ‘to point out<br />
that a certain dog is male bayi guda can be used’ (Dixon 1982: 182). Usually, guda ‘dog’<br />
belongs to Class II (Dixon 1982: 180), and so <strong>the</strong> ‘unmarked’ usage would be balan guda.<br />
Very occasionally, changing noun class can create a pragmatic effect. In Dyirbal, yara<br />
‘man’ belongs to Class I; and so would be referred to as bayi yara.<strong>How</strong>ever,Dixon(1982:<br />
166) reports that a hermaphrodite was once jokingly referred to as balan yara, witha<br />
feminine Class II marker, pointing out his female characteristics. In this case, <strong>the</strong><br />
manipulation of Linguistic <strong>Gender</strong> has a pragmatic, as well as semantic effect.<br />
Variable Linguistic <strong>Gender</strong> may be important for ascertaining <strong>the</strong> meaning of a<br />
noun. Kwami is a West Chadic language with two genders, feminine and masculine.<br />
<strong>Gender</strong> is covert—that is, not marked on <strong>the</strong> noun itself. It is realized through<br />
agreement on demonstratives (-mè ‘masculine’, -jè ‘feminine’). <strong>Gender</strong> choice for<br />
human nouns is based on sex. The gender choice of inanimates is more complex.<br />
Nouns which refer to <strong>the</strong> domestic sphere are typically feminine, and nouns referring<br />
to objects outside <strong>the</strong> household are usually masculine. Some fruits and plants belong<br />
to masculine gender, but only as long as <strong>the</strong>y grow. As soon as <strong>the</strong>y are harvested and<br />
brought home for manufacturing, <strong>the</strong>ir gender changes to feminine, as in <strong>the</strong><br />
following examples: 6<br />
4.2 GROWING ON A FARM HARVESTED OR AT HOME (FOR PREPARATION) MEANING<br />
lòolów-mè (masculine) lòolów-jè (feminine) cotton<br />
gùyà-mè (masculine) gùyà-jè (feminine) pumpkin<br />
sháagúm-mè (masculine) sháagúm-jè (feminine) millet<br />
Variable gender marking in Maung (Capell and Hinch 1970: 47–52) serves to<br />
differentiate related meanings of polysemous nouns. Maung has five genders whose<br />
choice is largely semantic. <strong>Gender</strong> I consists of names of male beings; <strong>Gender</strong> II of<br />
names of female beings; <strong>Gender</strong> III includes objects associated with <strong>the</strong> ground<br />
(except plants); <strong>Gender</strong> IV subsumes trees and <strong>the</strong>ir parts; and <strong>Gender</strong> V includes<br />
vegetable foods, plants, and items associated with housing. Noun classes are overtly<br />
marked on nouns, and realized through subject and object agreement on verbs,<br />
adjectives, and possessive expressions. This is how variable gender works.<br />
The noun -nimi expresses <strong>the</strong> idea of a ‘long bone’. Used with a ‘human male’<br />
<strong>Gender</strong> I prefix, (j)i-nimi means ‘his backbone’; with a <strong>Gender</strong> II, female human<br />
class, ninj-imi means ‘her backbone’. With a <strong>Gender</strong> IV prefix, ma-nimi means ‘trunk<br />
of a tree’, and with gender V prefix ad-imi means ‘main radicle’ of a potato plant or<br />
yam vine (Capell and Hinch 1970: 47). Table 4.1 illustrates how noun classes are used<br />
with ano<strong>the</strong>r noun root, mawur ‘arm, long part of ’.
4.1 Variable choice of Linguistic <strong>Gender</strong> 55<br />
TABLE 4.1. Variable <strong>Gender</strong> assignment in Maung<br />
<strong>Gender</strong> I i-mawur ‘man’s arm’<br />
<strong>Gender</strong> II ninj-mawur ‘woman’s arm’<br />
<strong>Gender</strong> III u-mawur ‘tributary of a river’<br />
<strong>Gender</strong> IV ma-mawur ‘branch of a tree’<br />
<strong>Gender</strong> V a-bawur ‘tendril of a vine’<br />
In a follow-up study, Singer (2010: 397) gave fur<strong>the</strong>r illustrations to <strong>the</strong> semantic<br />
effect of changing genders. The noun jampakang ‘corrugated iron’ is usually<br />
assigned to <strong>the</strong> masculine gender (<strong>Gender</strong> I) (like most o<strong>the</strong>r European-origin<br />
artefacts). When used as building material, it occurs with <strong>Gender</strong> V, by analogy with<br />
o<strong>the</strong>r items associated with housing which also belong to this gender. Linguistic<br />
<strong>Gender</strong>s allow <strong>the</strong> creation of new lexical items, and manipulate <strong>the</strong> meanings of <strong>the</strong><br />
existing ones.<br />
Changing Linguistic <strong>Gender</strong>s may have a special discourse effect. Algonquian<br />
languages have two Linguistic <strong>Gender</strong>s—animate and inanimate—whose meaning<br />
is only partly predictable (see §2.2.1). Quite a few inanimate nouns belong to <strong>the</strong><br />
animate gender. In Goddard’s(2002: 225) words, ‘In discussing <strong>the</strong> phenomenon, <strong>the</strong><br />
usual and convenient explanation given by speakers is something like, “you talk about<br />
it like it’s something living”, or“it sounds like it’s alive”’. Changing a noun’s gender<br />
from inanimate to animate highlights its agency in a story. The comb (normally<br />
inanimate) shifts to animate as it starts talking to its owner, in a Menominee story.<br />
An inanimate skull becomes animate when it takes on <strong>the</strong> role of a Rolling Skull, a<br />
mythical ogre. Some names for trees are usually inanimate, but may be referred to with<br />
animate gender if used ‘in <strong>the</strong> winter story’ or ‘addressed religiously’. 7<br />
In Ojibwe, a grammatically inanimate noun is often treated as ‘animate’ in a story,<br />
‘due to its assuming animate properties, such as ability to speak to exert spiritual<br />
power’, as Valentine (2001: 118) puts it. In one traditional story, a young man<br />
receives a blessing from a mirror, which later on provides him with protection in<br />
battle. When <strong>the</strong> mirror first visits <strong>the</strong> man, it is referred to with inanimate gender.<br />
Later on, when it confers its blessing upon <strong>the</strong> man, it becomes animate. As a speaker<br />
of Ojibwe put it, ‘in Odawa [a dialect of Ojibwe] <strong>the</strong> concept of animateness is<br />
limitless. It can be altered by <strong>the</strong> mood of <strong>the</strong> moment, <strong>the</strong> mood of <strong>the</strong> speaker, <strong>the</strong><br />
context, <strong>the</strong> use, <strong>the</strong> circumstances, <strong>the</strong> very cosmos or our totality.’ 8<br />
Bunaq, an Austronesian language spoken in Timor, has two genders—animate<br />
and inanimate marked in object prefixes on verb and in demonstratives. The term for<br />
‘clan’ and terms for vehicles are inanimate. But <strong>the</strong>y can be reclassified as animate, if<br />
<strong>the</strong> fact that <strong>the</strong>y are carrying humans is emphasized. Nouns referring to natural<br />
elements are inanimate. If human agency is involved in <strong>the</strong>ir creation or behaviour,
56 4 What are Linguistic <strong>Gender</strong>s good for?<br />
<strong>the</strong>y can be reclassified as animate. So, il ‘water’ is inanimate if it refers to water<br />
drawn from a river. If <strong>the</strong> same noun refers to water artificially flooded into a rice<br />
paddy from a river, it is treated as animate. 9<br />
Variable gender with humans may have overtones of value and of attitude.<br />
Women in Oneida, a Nor<strong>the</strong>rn Iroquoian language, may be referred to with feminine<br />
gender if <strong>the</strong>y are small and graceful, and with neuter, if <strong>the</strong>y are big and aggressive. 10<br />
In Swahili, a Bantu language, most nouns occur with prefixes of just one class, with a<br />
few notable exceptions. M-zee means ‘old person’ (and contains <strong>the</strong> human class<br />
prefix m-). This prefix can be replaced with ki- (inanimate class); <strong>the</strong> resulting form<br />
ki-zee means ‘scruffy old person’. Switching genders can thus be a mark of intimacy<br />
or derision. <strong>Gender</strong> choice may correlate with politeness. Addressing an older<br />
woman with feminine gender in Lak, a north-east Caucasian language from Daghestan,<br />
is not as polite as addressing her with a more general gender which subsumes a<br />
larger class of nouns, including inanimates (see Khaidakov 1963: 49–50). This is<br />
similar to what we find in Mohawk, an Iroquoian language (see §7.3, and Mithun<br />
2014: 138).<br />
We return to <strong>the</strong> effects of gender reversals for women and men in §§7.2–3.<br />
Variable Linguistic <strong>Gender</strong>s can help differentiate objects in terms of size and<br />
shape. We saw in Chapter 3 how Linguistic <strong>Gender</strong> choice in Cantabrian Spanish<br />
accords with <strong>the</strong> Natural <strong>Gender</strong> of humans and animates. Inanimates and lower<br />
animates which are ei<strong>the</strong>r tall or of smaller size, or of narrow shape, or of vertical<br />
orientation are assigned to masculine gender. Objects which are unusually large,<br />
wide, horizontal, or small and squat are feminine. Variable gender may reflect <strong>the</strong><br />
shape and <strong>the</strong> state of <strong>the</strong> object. In Ket, a language with masculine and feminine<br />
gender, a growing tree is masculine, and a cut-down tree, or a log, is feminine; an<br />
upright tree is masculine, and a tree with a curved trunk is feminine. In Khwe, a<br />
Central Khoisan language, an inanimate noun can be allocated to masculine or<br />
feminine gender depending on its shape: masculine is associated with big, long,<br />
rectangular, and feminine with small, round, broad. 11<br />
Savosavo, a Papuan language spoken in <strong>the</strong> Solomon Islands, has two genders<br />
marked on verbs, pronouns, and demonstratives. Feminine gender includes humans,<br />
birds, and animals. Masculine gender covers males and inanimates. An inanimate<br />
noun can be assigned feminine gender, if it is very small. Molo ‘knife’ is usually<br />
masculine. A small knife can be referred to as feminine. <strong>Gender</strong> variation may have a<br />
fur<strong>the</strong>r, pragmatic function: if a noun is particularly salient in a story, it <strong>the</strong>n acquires<br />
feminine agreement. 12 The meanings of shape and size in gender choice correlate<br />
with stereotypical features associated with Natural <strong>Gender</strong>. They may also reflect <strong>the</strong><br />
Social <strong>Gender</strong> stereotypes—that is, <strong>the</strong> relative position of men and women within a<br />
society, and <strong>the</strong>ir importance: this is what we saw in §3.4.2.<br />
Variable gender can be of a different nature. A recent loan may not as yet have<br />
settled into a gender. The word kivi ‘kiwi fruit’ in Russian is sometimes treated as
4.2 Linguistic <strong>Gender</strong> in discourse 57<br />
neuter (based on its form and morphological features: it is an indeclinable noun<br />
ending in a vowel), and sometimes as masculine (based on <strong>the</strong> analogy with frukt<br />
‘fruit’ which is masculine).<br />
Nouns with similar meanings may differ in gender only. In German, <strong>the</strong> masculine<br />
der See means ‘<strong>the</strong> lake’, and <strong>the</strong> feminine die See means ‘<strong>the</strong> sea’. The masculine der<br />
Band means ‘<strong>the</strong> volume (of a book)’, <strong>the</strong> neuter das Band means ‘<strong>the</strong> ribbon, tape,<br />
band, wavelength (of a radio)’, and <strong>the</strong> feminine die Band means ‘<strong>the</strong> (rock)band’;<br />
<strong>the</strong> masculine der Leiter means ‘<strong>the</strong> director, head, leader, manager’ and <strong>the</strong> feminine<br />
die Leiter means ‘<strong>the</strong> ladder, stepladder’. 13 Nouns which differ only in <strong>the</strong>ir gender<br />
can be semantically linked. The word gabal in Dyirbal is feminine (balan gabal) in<br />
its meaning ‘egret, heron’. The same form gabal is neuter (bala gabal) with <strong>the</strong><br />
meaning ‘sand, sandbank’. Herons live and walk on sand—this could be <strong>the</strong> connection<br />
between <strong>the</strong> two words which differ in gender only. The blossom on a tree is<br />
maŋga—this is treated as belonging to <strong>the</strong> edible gender if it will develop into an<br />
edible fruit (balam maŋga); any o<strong>the</strong>r blossom is neuter—bala maŋga.<br />
Dyirbal had a special Mo<strong>the</strong>r-in-law, or Jalnguy, style used to communicate with<br />
in-laws. This register was much less specific than <strong>the</strong> everyday register. In many<br />
cases, <strong>the</strong> Mo<strong>the</strong>r-in-law register has just one term where <strong>the</strong> everyday language had<br />
several. This is where gender helps disambiguate <strong>the</strong> homonyms in Jalnguy.<br />
The everyday language style has three different words for ‘rainbow’, ‘wompoo<br />
pigeon’, and ‘flame kurrajong’ tree. One term covers all <strong>the</strong> three in <strong>the</strong> Mo<strong>the</strong>r-inlaw<br />
style. The wompoo pigeon was given <strong>the</strong> same name as <strong>the</strong> rainbow because of its<br />
colour being like <strong>the</strong> green of <strong>the</strong> rainbow. The flame kurrajong tree was held to be<br />
sacred to <strong>the</strong> rainbow snake—this explains it having <strong>the</strong> same name. Three gender<br />
markers distinguish <strong>the</strong> three terms: <strong>the</strong> rainbow is masculine, <strong>the</strong> pigeon is feminine,<br />
and <strong>the</strong> tree is neuter:<br />
4.3 Everyday style Mo<strong>the</strong>r-in-law or Jalnguy style<br />
rainbow bayi yamani bayi gagilbarra<br />
wompoo pigeon balan bagamu balan gagilbarra<br />
flame kurrajong tree bala dila bala gagilbarra<br />
Linguistic <strong>Gender</strong> may correlate with how discourse is organized, and it may be<br />
instrumental in identifying who is doing what.<br />
4.2 Linguistic <strong>Gender</strong> in discourse<br />
The use of Linguistic <strong>Gender</strong> may depend on <strong>the</strong> definiteness and topicality of <strong>the</strong><br />
noun. This is a feature Linguistic <strong>Gender</strong> shares with o<strong>the</strong>r systems of noun categorization.<br />
14 In Motuna (a Papuan language from Bougainville in Papua New Guinea)<br />
<strong>the</strong> verb has to agree in Linguistic <strong>Gender</strong> with its object only if <strong>the</strong> object is <strong>the</strong> topic
58 4 What are Linguistic <strong>Gender</strong>s good for?<br />
of discourse. The object nii ‘I (masculine)’ is topical: ‘I’ is what <strong>the</strong> text was about.<br />
The verb bears <strong>the</strong> masculine agreement marker.<br />
4.4 nii Aanih-ki tangumu-u-ng<br />
I(MASC):ABS Aanih(FEM)-ERG slap+1sgO+3sgA-NEAR.PAST-MASC<br />
‘Aanih (a female name) slapped me (topic).’<br />
In 4.5, <strong>the</strong> subject, Aanih, is topical and its feminine gender appears on <strong>the</strong> verb:<br />
4.5 Aanih nii tangumu-i-na<br />
Aanih(FEM) I(MASC):ABS slap+1sgO+3sgA-NEAR.PAST-FEM<br />
‘Aanih (a female name) (topic) slapped me.’ 15<br />
Overt gender marking on a noun can signal its definiteness or specificity. In Gola,<br />
a West Atlantic language of Liberia, gender-marking prefixes and suffixes act similarly<br />
to definite articles. A definite noun is marked for gender, an indefinite noun is not,<br />
e.g. kul ‘atree’, ke-kul ‘<strong>the</strong> tree’; gbalia ‘a dwarf antelope’, o-gbalia-a ‘<strong>the</strong> dwarf antelope’<br />
(note that here noun class is marked by a combination of a prefix and a suffix).<br />
The presence of an overt gender marker can correlate with a specific individuated<br />
referent, and its absence with a more generic, or indefinite, referent. In Turkana, an<br />
East Nilotic language from Kenya, genders are marked on nouns <strong>the</strong>mselves (we can<br />
recall from §3.1 how changing gender of a noun may correlate with <strong>the</strong> size and <strong>the</strong><br />
state of an entity). The overt gender prefix on nouns ‘is frequently omitted with<br />
names of animals in folk tales when <strong>the</strong> names are used in <strong>the</strong>ir generic sense’. The<br />
word for ‘lion’ is ɛ-ŋatùɲ, and for ostrich ɛ-kalèes, if used in isolation (Dimmendaal<br />
1983: 221). At <strong>the</strong> start of a story about two friends, a lion and an ostrich, <strong>the</strong> two<br />
words were used without <strong>the</strong>ir prefixes, as ŋatùɲ kà kalèes ‘lion and ostrich’ (generic<br />
or indefinite). 16<br />
Overt gender marking can depend on <strong>the</strong> function of a noun in discourse. In <strong>the</strong><br />
Australian languages Nunggubuyu and Warray <strong>the</strong> presence of a gender prefix ona<br />
noun is correlated with definiteness or givenness, and its absence indicates focus and<br />
foregrounding. In Wardaman noun class prefixes are used when introducing a new<br />
participant and for maintaining reference to <strong>the</strong> noun throughout <strong>the</strong> story. 17<br />
Linguistic <strong>Gender</strong> helps disambiguate what is being talked about. In German, Krug<br />
‘jug’ is masculine, and Schale ‘vase’ is feminine. The jug is referred to as ‘he’ and <strong>the</strong><br />
vase as ‘she’. As a consequence, <strong>the</strong>re is no ambiguity in German concerning what got<br />
broken. 18<br />
4.6 Der Krug (masculine) fiel in die Schale (feminine),<br />
ART.MASC.SG jug fell into ART.FEM.SG bowl<br />
aber er (masc) zerbrach nicht<br />
but he broke NEG<br />
‘The jug fell into <strong>the</strong> bowl but it (<strong>the</strong> jug, lit. he) didn’t break.’
4.3 The utility of Linguistic <strong>Gender</strong> 59<br />
In English, ‘it’ would refer to both ‘jug’ and ‘bowl’, and <strong>the</strong> disambiguation will be<br />
less straightforward. Different forms for different genders allow us to communicate in a<br />
succinct fashion. Failing to understand gender cues may produce an embarrassment.<br />
Dixon (2016: 83–4) tells an amusing anecdote about <strong>the</strong> discourse potential of<br />
genders. He went to a shop in Porto Velho, a regional capital in Brazil, to buy one<br />
mirror (um espelho, masculine) which one of his Indian friends had asked for. The<br />
shopkeeper produced a whole box of mirrors (uma caixa, feminine), took one mirror<br />
from it, and quoted <strong>the</strong> price, saying Dez reais por uma (ten real for one: feminine)<br />
‘ten reais for one’. Dixon protested that ten reais (about five dollars) was too much for<br />
one small mirror. The shopkeeper continued bargaining <strong>the</strong> price down, and <strong>the</strong>n, in<br />
Dixon’s words,<br />
Then it hit me. My gender-starved brain has failed to pick up that he said por uma. The<br />
feminine form of <strong>the</strong> indefinite article must be referring to <strong>the</strong> box, uma caixa. The price being<br />
quoted was for a box of ten mirrors. If he had been giving <strong>the</strong> price for a mirror, um espelho,he<br />
would have said por um. I purchased one mirror, for one Real, and left <strong>the</strong> shopkeeper shaking<br />
his head at foreigners who can’t understand simple language. If <strong>the</strong> transaction had been<br />
conducted in English, a price would have had to be quoted for a box or for a mirror. It can be<br />
seen that marking gender on articles, in Portuguese, makes for a more succinct discourse.<br />
Linguistic <strong>Gender</strong>s help animate <strong>the</strong> world. A noun whose Linguistic <strong>Gender</strong> is<br />
feminine can be depicted, and conceptualized, as a ‘woman’. In Portuguese and<br />
Russian <strong>the</strong> word for ‘death’ is feminine. And so, in folk tales and pictures, Death<br />
is represented as a woman. In Portuguese, it is politely referred as Dona Morte, Ms<br />
Death. In German, <strong>the</strong> word for death is masculine—and it is depicted as an old man.<br />
In §§8.1–2 we turn to <strong>the</strong> imagery associated with Linguistic <strong>Gender</strong>.<br />
Not only does Linguistic <strong>Gender</strong> help in categorizing entities. Linguistic <strong>Gender</strong> is<br />
a repository of beliefs and social practices. We can recall, from §2.2.1, that in Dyirbal<br />
birds are classed as members of gender 2 (‘feminine’, balan) ra<strong>the</strong>r than bayi on <strong>the</strong><br />
basis of <strong>the</strong>ir non-human animate status. This is so because birds are believed to be<br />
<strong>the</strong> spirits of dead women. In numerous languages of <strong>the</strong> Sepik area of New Guinea,<br />
cassowary and bird of paradise are mythological women; and <strong>the</strong>y are invariably<br />
assigned to <strong>the</strong> feminine gender. More on this in §8.1.<br />
4.3 The utility of Linguistic <strong>Gender</strong><br />
As Jakobson (1959: 116) put it, ‘languages differ essentially in what <strong>the</strong>y must convey,<br />
and not in what <strong>the</strong>y may convey’. Like any grammatical category, Linguistic <strong>Gender</strong><br />
imposes partly semantic, and partly formal restrictions, on what one is to say in a<br />
particular language. Some linguists, feminists, and language learners bemoan <strong>the</strong> fact<br />
that Linguistic <strong>Gender</strong>s are imposing and restrictive—especially so in languages where
60 4 What are Linguistic <strong>Gender</strong>s good for?<br />
<strong>the</strong>ir basis is not fully semantically transparent. We can recall from Chapter 2 that<br />
Mädchen ‘girl’ in German belongs to neuter gender (based on its morphological make<br />
up: a diminutive suffix -chen which automatically requires neuter gender choice). The<br />
word for ‘turnip’, Rübe, belongs to <strong>the</strong> feminine gender (partly based on its phonological<br />
form, as most nouns ending in -e are feminine). In his scurrilous attack on ‘The<br />
Awful German Language’, Mark Twain complained that ‘In German, a young lady has<br />
no sex, while a turnip has’. Too much is being made of a purely linguistic fact. And we<br />
can recall, from §2.1.2, that a girl in German—especially if she is grown up—will be<br />
referred to with <strong>the</strong> pronoun ‘she’ reflecting her Natural <strong>Gender</strong> and a mature female<br />
Social <strong>Gender</strong> position (see <strong>the</strong> details in Braun and Haig 2010). That Linguistic<br />
<strong>Gender</strong>s have no function (o<strong>the</strong>r than hindering second-language learners) is simply<br />
wrong. Their versatility, and <strong>the</strong> ways in which <strong>the</strong>y may mirror myths and beliefs—<br />
and can be a source of poetic metaphors—makes languages with Linguistic <strong>Gender</strong> far<br />
more fun to work with than <strong>the</strong> ones which lack <strong>the</strong>m. If languages are in contact, <strong>the</strong>y<br />
will be likely to come to share important features—Linguistic <strong>Gender</strong>s among <strong>the</strong>m, as<br />
we will show in Chapter 6. We saw in Chapter 3 how positive features associated with<br />
masculine gender can be projected onto culturally important entities. The meanings<br />
associated with typical properties of males and females and <strong>the</strong>ir attributes get transferred<br />
onto essentially gender-less, inanimate objects and phenomena. Expressive<br />
power in poetry and metaphor is gained through deploying Linguistic <strong>Gender</strong>s—as<br />
we will see in Chapter 8. Those who accuse languages with Linguistic <strong>Gender</strong>s of being<br />
unnecessarily ‘irrational’ or ‘ornate’ simply haven’t looked far enough—or perhaps are<br />
trying to be different for <strong>the</strong> sake of it. 19<br />
Some feminists—speakers of languages with anaphoric gender—complain that <strong>the</strong><br />
obligatory choice between ‘he’ and ‘she’ in singular pronouns makes it difficult to<br />
promote gender equality: speakers are forced to make a gender choice even if <strong>the</strong>y<br />
would ra<strong>the</strong>r avoid it. But <strong>the</strong> way <strong>the</strong> choice is made has ano<strong>the</strong>r side to it. This<br />
choice may give us an insight into <strong>the</strong> ways Social <strong>Gender</strong> stereotypes work, and how<br />
speakers perceive <strong>the</strong>m. The ways in which <strong>the</strong> choices change reflect changes in <strong>the</strong><br />
position of women within <strong>the</strong> society. These visible changes through language can be<br />
monitored and traced more easily than if <strong>the</strong>y are covert. Linguistic <strong>Gender</strong> is among<br />
<strong>the</strong> lucky few grammatical categories which reflect social and cultural meanings, and<br />
is amenable to linguistic engineering and conscious change, as a social environment<br />
transforms. As we will see in subsequent chapters, Linguistic <strong>Gender</strong> offers a window<br />
into <strong>the</strong> dynamics of social developments, and perhaps <strong>the</strong> human mind. Speakers<br />
are aware of Linguistic <strong>Gender</strong> meanings and implications—this is something <strong>the</strong>y<br />
eagerly discuss, and may wish to reform—we return to this in §11.5. It is not <strong>the</strong> case<br />
that languages with Linguistic <strong>Gender</strong> are more ‘sexist’ than those without it; we will<br />
see in §5.3 and <strong>the</strong>n in Chapter 11 how linguistic sexism can find its expression in<br />
languages with scarcely any Linguistic <strong>Gender</strong>.
4.3 The utility of Linguistic <strong>Gender</strong> 61<br />
Linguistic <strong>Gender</strong> is one of <strong>the</strong> means of categorizing nouns, and it shares its<br />
many functions with o<strong>the</strong>r noun categorization devices of o<strong>the</strong>r kinds, especially<br />
numeral, noun, and verbal classifiers. When applied to humans, Linguistic <strong>Gender</strong><br />
correlates with Natural and Social <strong>Gender</strong>. And <strong>the</strong>se can be, and usually are,<br />
expressed through o<strong>the</strong>r means. ‘<strong>Gender</strong>-less’ languages—for some, <strong>the</strong> lucky<br />
escapees from <strong>the</strong> ‘prison-house’ of Linguistic <strong>Gender</strong> and its limitations—hardly<br />
ever achieve <strong>the</strong> ‘ideal’ of gender neutrality. Igbo has no Linguistic <strong>Gender</strong> and, in<br />
Amadiume’s (2015: 89) words,‘has not built up rigid associations’ between various<br />
attributes, objects, and genders. ‘The genderless word mmadu, humankind, applies<br />
to both sexes. There is no usage, as <strong>the</strong>re is in English, of <strong>the</strong> word “man” to represent<br />
both sexes, nei<strong>the</strong>r is <strong>the</strong>re <strong>the</strong> cumbersome option of saying “he or she”, “his or<br />
her”, “him or her”.’ <strong>How</strong>ever, this ‘linguistic gender’ neutrality did not stop men in<br />
<strong>the</strong> traditional Igbo society from having more powers and more privileges than<br />
women. With <strong>the</strong> advent of colonial powers, <strong>the</strong> inequality of Social <strong>Gender</strong>s became<br />
more drastic. 20<br />
We now turn to how meanings of sex, animacy, humanness, and also Social <strong>Gender</strong><br />
find <strong>the</strong>ir expression beyond <strong>the</strong> strict division of nouns into Linguistic <strong>Gender</strong> classes.<br />
NOTES AND SOURCES<br />
1. Sayers (1995: 24, 33).<br />
2. Functions of genders as part of a continuum of noun categorization devices (including<br />
numeral classifiers, noun classifiers, verbal classifiers, possessive and relational classifiers,<br />
and additional types) are discussed in Aikhenvald (2000: chapter 12), and summarized in<br />
Aikhenvald (2006). The same functions are discussed (sometimes with different examples)<br />
in Contini-Morava and Kilarski (2013). Additional discussion is in Weber (2000).<br />
3. See Holmquist (1991) on Cantabrian Spanish, Ferrari (2005) on Italian, Pountain (2005)<br />
on Spanish; Audring (2013) offers examples of <strong>the</strong> individuating functions of gender<br />
choice in Dutch.<br />
4. Typologists who choose to focus on <strong>the</strong> ‘canons’ of a few European languages relegate<br />
systems with variable gender assignment to a questionable status of ‘non-canonical<br />
systems’ (e.g. Corbett 2014b: 122). But if one ventures beyond one’s preconceptions,<br />
languages with variable assignment of Linguistic <strong>Gender</strong> (which are not just <strong>the</strong> ‘exotic<br />
few’) offer a fertile field for investigating <strong>the</strong> meanings of genders.<br />
5. Aikhenvald (2000: 41); Mat<strong>the</strong>ws (2010: 124). And see §6.6 on one of <strong>the</strong> meanings of<br />
‘common gender’.<br />
6. See Leger (1998: 206) and Dinslage, Leger, and Storch (2000: 125).<br />
7. Goddard (2002: 221).<br />
8. Similar examples, from Fox, are Goddard (2002) and Dahlstrom (1995).<br />
9. Schapper (2010a: 182–4).<br />
10. See Abbott (1984), Mithun (1999: 100–1; 2014), Bonvillain (1973: 86–7), and §7.3.<br />
11. Dul'son (1968: 62ff.) on Ket; Kilian-Hatz (2008: 41–2) for Khwe; also see Hagman (1977:<br />
23) for similar phenomena in Nama.<br />
12. Wegener (2012: 63–5).
62 4 What are Linguistic <strong>Gender</strong>s good for?<br />
13. See Zgusta (1971: 82–3) for more examples, and how to deal with <strong>the</strong>m in dictionarymaking.<br />
14. See a comprehensive discussion in chapter 12 of Aikhenvald (2000).<br />
15. Masa Onishi (1994), p.c.<br />
16. Westermann (1947: 17); Heine (1982a: 193) on Gola, Dimmendaal (1983: 219–21) on<br />
Turkana.<br />
17. See Heath (1984: 169–70) on Nunggubuyu, Harvey (1987: 53) on Warray, and Merlan<br />
(1994) on Wardaman and Merlan, Roberts, and Rumsey (1997: 85) on o<strong>the</strong>r nor<strong>the</strong>rn<br />
Australian languages with prefixes marking genders; see Bruce (1984: 97) on Alamblak.<br />
18. Example from Zubin and Köpcke (1986: 174).<br />
19. See Deutscher’s(2010: 203–4) witty criticism of <strong>the</strong> misleading idea that Linguistic <strong>Gender</strong><br />
is ‘irrational’.<br />
20. Amadiume (2015) offers a comprehensive analysis of <strong>the</strong> traditional Igbo society with its<br />
division of labour and cultural and social roles for men and women, and <strong>the</strong> ways in which<br />
women lost <strong>the</strong>ir place and <strong>the</strong>ir powers as a consequence of colonialization; we return to<br />
this in Chapter 11.
5<br />
<strong>Gender</strong> meanings in grammar<br />
and lexicon<br />
Linguistic <strong>Gender</strong> is one of many means of categorizing a noun, in terms of its<br />
animacy, humanness, or sex, across <strong>the</strong> world’s languages. Classifiers of various<br />
types—including numeral classifiers, noun classifiers, and classifiers on verbs—<br />
encode <strong>the</strong> same set of meanings on a par with inherent properties of objects: <strong>the</strong>ir<br />
shape, size, consistency, and dimensions. We briefly delve into <strong>the</strong>se in §5.1. O<strong>the</strong>r<br />
categories of a noun—number, case, and declension type—can reflect meanings<br />
associated with Linguistic <strong>Gender</strong>s—see §5.2. Then, in §5.3, we turn to how gender<br />
meanings can be expressed in ‘gender-less’ languages.<br />
5.1 Sex, humanness, and animacy in classifier systems<br />
Grammatical means for <strong>the</strong> linguistic categorization of nouns can be found in every<br />
language. Noun categorization devices o<strong>the</strong>r than Linguistic <strong>Gender</strong> include a variety<br />
of types. NUMERAL CLASSIFIERS appear next to a numeral or a quantifier, categorizing a<br />
noun in terms of its meaning. These are typical of languages of East and South-East<br />
Asia, North and South America, and New Guinea. NOUN CLASSIFIERS categorize <strong>the</strong><br />
noun on its own, independently of any o<strong>the</strong>r element in a phrase or a clause. They are<br />
a feature of many Australian, Mayan, and Austronesian languages. VERBAL CLASSIFIERS<br />
occur on a verb and categorize its object or intransitive subject. There can also be<br />
classifiers in possessive constructions, and classifiers with locative and deictic expressions.<br />
We will see, in Chapter 6, how Linguistic <strong>Gender</strong> can arise through reinterpretation<br />
of noun categorization devices of o<strong>the</strong>r kinds (thus confirming <strong>the</strong>ir<br />
ultimate unity). 1<br />
One language can combine several types of noun categorization devices. In<br />
‘multiple classifier’ languages, <strong>the</strong> same morphemes occur in several grammatical<br />
contexts—with numerals, adjectives, demonstratives, and also verbs. Classifiers can<br />
encode meanings typically associated with gender systems—sex, animacy, and<br />
humanness. We start with numeral classifiers, as <strong>the</strong> most frequent type across <strong>the</strong><br />
world’s languages.<br />
<strong>How</strong> <strong>Gender</strong> <strong>Shapes</strong> <strong>the</strong> <strong>World</strong>. First edition. Alexandra Y. Aikhenvald.<br />
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.
64 5 <strong>Gender</strong> meanings in grammar and lexicon<br />
5.1.1 Numeral classifiers<br />
Most languages with a numeral classifier system have special terms for humans and<br />
for animates. Jacaltec, a Mayan language, has three numeral classifiers: -wanʺ for<br />
humans, -c'onʺ for animals, and -(e)b' for inanimates. The three most frequent<br />
classifiers in Indonesian are orang for humans, as in se-orang guru (one-CL:HUMAN<br />
teacher) ‘one teacher’, ekor for non-human animates, as in dua ekor ular (two CL:<br />
NONHUMAN.ANIMATE snake) ‘two snakes’, and buah for inanimates, as in tiga buah<br />
meja (three CL:INANIMATE table) ‘three tables’. There are a few more classifiers for<br />
inanimates whose choice is based on shape and function, e.g. batang for ‘cylindrical<br />
objects’ (Sneddon 1996: 135–6). Halkomelem, a Salish language of <strong>the</strong> Pacific Northwest<br />
of North America, has a special classifier for counting grown up people, and<br />
ano<strong>the</strong>r one for counting children, e.g. lx w =elə ‘three people’, lx w =eyl ‘three children’.<br />
Categorization of humans through numeral classifiers may involve more than just<br />
sex. Korean has at least 154 obligatory numeral classifiers. Nouns with non-human<br />
reference are classified based on <strong>the</strong>ir nature, shape, or arrangement. Humans are<br />
categorized depending on <strong>the</strong>ir status and sex. The numeral classifier pwun refers to<br />
respected persons (and also <strong>the</strong> Christian God, Buddha, and o<strong>the</strong>r gods); <strong>the</strong> classifier<br />
myeng is a neutral way of referring to a person. The classifier nom refers to a<br />
despicable or insignificant male, and nyen to a despicable or insignificant female,<br />
e.g. kkangphay twu nom (gangster two NUM.CL:DESPICABLE.MALE) ‘two male gangsters’,<br />
kkangphay twu nyen (gangster two NUM.CL:DESPICABLE.FEMALE) ‘two female gangsters’. 2<br />
Burmese is a Tibeto-Burman language with several score numeral classifiers.<br />
Table 5.1 features classifiers used to count humans and animates. 3<br />
Large systems of numeral classifiers in languages of East and South-East Asia tend<br />
to reflect societal organization in <strong>the</strong>ir classifiers for humans. Thai, a Tai-Kadai<br />
language with several dozen numeral classifiers, is spoken in a monarchy. There is<br />
a special classifier phráʔoŋ for ‘emperor, king’, ʔoŋ for ‘God, Buddha, Buddhist<br />
monk, emperor, king’, rûup for Buddhist monk, thân for ‘prime minister and high<br />
rank officer’, naay ‘policeman’, in addition to a general human classifier khon. 4<br />
There is no monarchy in <strong>the</strong> Republic of Lao; and <strong>the</strong> system of human classifiers<br />
lacks a term for ‘king’. A numeral classifier khon2 ‘person’ covers people o<strong>the</strong>r than<br />
TABLE 5.1. Animate classifiers in Burmese<br />
hsu<br />
Buddha and his attributes: relics, images, <strong>the</strong> Law<br />
pa<br />
deities, saints, monks, royalty<br />
u: people of status, teachers, scholars<br />
jau' ordinary humans<br />
kaun animals, ghosts, dead bodies, depraved people, children
5.1 Sex, humanness, and animacy in classifier systems 65<br />
monks; classifiers gong3 and huup4 are used for monks. O<strong>the</strong>r nor<strong>the</strong>rn Tai languages,<br />
including Zhuang and Maonan, have no classifiers whose choice would be<br />
determined by <strong>the</strong> social rank, or religious position. Lu (2012: 112–13), himself a native<br />
speaker of Maonan and of Zhuang, notes that <strong>the</strong> fact that Zhuang lacks classifiers for<br />
various social and clerical ranks (such as ‘monk’,or‘high official’) is a direct reflection<br />
of <strong>the</strong> social organization of <strong>the</strong> Zhuang people: <strong>the</strong>re is no social stratum of clergy, and<br />
<strong>the</strong> society is much more socially equal than in Thailand or in Laos.<br />
What about special classifiers for women? Numeral classifiers in Zhuang categorize<br />
humans in terms of sex and age. The classifier pou 4 simply states that <strong>the</strong> noun refers to<br />
a ‘human being’, asinsa:m1 pou 4 lɯk 8 (three CL:HUMAN child) ‘three children’. The<br />
classifier ta 4 is used for young women, as in sam 1 ta 4 lɯk 8 (three CL:YOUNG.FEMALE child)<br />
‘three daughters’. The classifier tak 8 covers young males, as in sam 1 tak 8 lɯk 8 (three CL:<br />
YOUNG.MALE child) ‘three sons’. Two fur<strong>the</strong>r classifiers cover adults of each sex: me 6<br />
‘adult female’, and kɔŋ 1 ‘adult male’, e.g. sa:m 1 me 6 a:ŋ 6 jɯ 1 (three CL:ADULT.FEMALE<br />
doctor) ‘three female doctors’, and sa:m 1 kɔŋ 10 a:ŋ 6 jɯ 1 (three CL:ADULT.MALE doctor)<br />
‘three male doctors’. The human classifier, ra<strong>the</strong>r than female-specific forms, in<br />
Zhuang is used to refer to professional women (instead of female-specific forms<br />
which have somewhat demeaning overtones).<br />
The use of a human classifier mirrors woman’s place in society, and speakers’<br />
attitudes, in Maonan. A human is typically referred to as ʔai 1 (CL:HUMAN) and a nonhuman<br />
animal as tɔ 2 (CL:ANIMAL). Infants, children, and women are classified with tɔ 2 ,<br />
<strong>the</strong> non-human animal category. As Tian-Qiao Lu, a native speaker of Maonan, put<br />
it, ‘judging from <strong>the</strong> application of this nonhuman classifier to a child or a woman,<br />
we can say that a child or a woman is deemed as undersocialized in <strong>the</strong> Maonan<br />
society’. As more women become integrated into professional society, <strong>the</strong>y start<br />
being referred to with <strong>the</strong> human classifier ʔai 1 (Lu 2012: 83–4, 115). The use of<br />
classifiers for women reflects <strong>the</strong>ir status in <strong>the</strong> society, and <strong>the</strong>ir exclusion—just as<br />
Linguistic <strong>Gender</strong> often does. Classifiers reflect social changes—just as Linguistic<br />
<strong>Gender</strong>s may do. We return to this in §11.4.<br />
<strong>Gender</strong> systems always have animacy, humanness, or sex as <strong>the</strong>ir core semantic<br />
feature. Not so for numeral classifiers. An example of a language with numeral<br />
classifiers with no animacy or humanness comes from Kana, a Cross-River language<br />
from Nigeria. The language has nineteen numeral classifiers whose choice is based on<br />
<strong>the</strong> shape of person or an object. So, a dog or a human will be referred to with a general<br />
classifier. Kana does not have genders. Human nouns can be differentiated from <strong>the</strong><br />
rest by o<strong>the</strong>r means: for instance, emphatic pronouns usually refer just to humans. 5<br />
5.1.2 Noun classifiers<br />
Noun classifiers are independent words with generic meaning which categorize <strong>the</strong><br />
noun with which <strong>the</strong>y co-occur. Three of more than twenty noun classifiers in Yidiñ,
66 5 <strong>Gender</strong> meanings in grammar and lexicon<br />
an Australian language, are used with human nouns—<strong>the</strong>se are waguja ‘man, human<br />
male, masculine’, buña ‘woman, human female, feminine’, and bama ‘person’. One<br />
would not generally say just ‘<strong>the</strong> girl dug up <strong>the</strong> yam’: it is more felicitous to say ‘<strong>the</strong><br />
person girl dug up <strong>the</strong> vegetable yam’: 6<br />
5.1 [mayi jimirr] [bama-al yaburu-ŋgu] jula-al<br />
vegetable+ABS yam+ABS CL:PERSON-ERG girl-ERG dig-PAST<br />
‘The person girl dug up <strong>the</strong> vegetable yam’<br />
Noun classifiers for humans in Australian languages can reflect socially significant<br />
categories. Yankunytjatjara has a special noun classifier for ‘initiated men’. Murrinhpatha<br />
has a classifier for Aboriginal people (which also covers human spirits) and<br />
ano<strong>the</strong>r one for non-Aboriginal people, which also covers all o<strong>the</strong>r animates and<br />
<strong>the</strong>ir products. 7<br />
Noun classifiers in Mayan languages typically refer to sex, age, and social status of<br />
humans. In Jacaltec, noun classifiers contain information about kinship relations,<br />
respect, and age—see Table 5.2. There are special classifiers for male and female<br />
deities, and one for infants. Mam has classifiers for men and women; for young and<br />
old men and women; for old men and women to whom respect is due, and for<br />
someone of <strong>the</strong> same status as <strong>the</strong> speaker. 8<br />
Noun classifiers typically have one or more forms for humans and animates. In<br />
contrast, classifiers on verbs and possessive classifiers do not have to have animacy,<br />
humanness, or sex as one of <strong>the</strong>ir meanings. We will see, in §6.1.2, how noun<br />
classifiers can give rise to a system of Linguistic <strong>Gender</strong>s.<br />
TABLE 5.2. Noun classifiers for humans and deities in Jacaltec (a selection)<br />
cumam<br />
cumi7<br />
ya7<br />
naj<br />
ix<br />
naj ni7an<br />
ix ni7an<br />
ho7<br />
xo7<br />
xo7 ni7an<br />
unin<br />
male deity<br />
female deity<br />
respected human<br />
male non-kin<br />
female non-kin<br />
young male non-kin<br />
young female non-kin<br />
male kin<br />
young male kin<br />
young female kin<br />
infant
5.1 Sex, humanness, and animacy in classifier systems 67<br />
5.1.3 Verbal classifiers and o<strong>the</strong>r classifier types<br />
Verbal classifiers (also called verb-incorporated classifiers) occur on <strong>the</strong> verb, categorizing<br />
a noun—typically in S (intransitive subject) or O (direct object) function—<br />
in terms of its shape, size, structure, or position, and sometimes also animacy.<br />
A subtype of verbal classifiers are special classificatory verbs. In Athabaskan languages,<br />
<strong>the</strong>se are regularly chosen depending on <strong>the</strong> inherent properties of <strong>the</strong> object,<br />
subject, and even <strong>the</strong> oblique. In some languages, <strong>the</strong>re is a special form for an<br />
‘animate’ object. Mescalero Apache, an Athabaskan language, has ten classificatory<br />
verbs. The choice involves animacy, shape, and form. A selection is in Table 5.3. The<br />
stem given here means ‘be located’. 9<br />
O<strong>the</strong>r Athabaskan languages do not include animacy in <strong>the</strong>ir system of classificatory<br />
verbs. This is <strong>the</strong> case in Carrier (where animacy is expressed in numerals and in<br />
demonstratives, while classificatory verbs are chosen based on <strong>the</strong> shape of <strong>the</strong> noun).<br />
Innu, an Algonquian language, has eight verbal classifiers, referring to shape, nature,<br />
and consistency of S and O, but not animacy or sex. Palikur, an Arawak language<br />
from Brazil and French Guyana, with three genders and several score numeral<br />
classifiers, has a special set of verb classifiers with no term for animates or humans<br />
(a classifier -pit ‘irregular shape’ is used on verbs to refer to a human, or an animate,<br />
or something of irregular shape). 10<br />
The choice of an existential, locative, and positional verb may depend on what<br />
<strong>the</strong>ir subject is like—whe<strong>the</strong>r it is animate, human, round, and so on. Japanese<br />
distinguishes animacy in its existential verbs iru ‘be:animate’ and aru ‘be:inanimate’.<br />
Hua, a Papuan language from <strong>the</strong> Highlands of New Guinea, has a copula bai with<br />
animate subjects and mo with inanimates. Small systems of existential verbs ‘be’ in<br />
Tibeto-Burman languages tend to include a special verb ‘be’ for animates. Ersu has<br />
five verbs ‘be’, one for animates, one for living plants, two for inanimates, and one for<br />
abstract concepts. 11<br />
In a number of Papuan languages of New Guinea, classificatory verbs categorize<br />
<strong>the</strong> intransitive subject argument in terms of its orientation or stance in space, and<br />
also its inherent properties. In Enga, a verb meaning ‘stand’ is used with referents<br />
TABLE 5.3. Mescalero Apache classificatory verb categories<br />
stem form ‘be located’<br />
semantics of <strong>the</strong> intransitive subject<br />
1. -'ą ‘single, solid, round inanimate object’<br />
2. -tiį ‘single animate object’<br />
3. -la ‘dual objects of any kind; a rope-like object’<br />
4. -tą ‘elongated, rigid object; a stick-like object’<br />
5. -ɬ-tsuu ‘flexible object; a cloth-like object’
68 5 <strong>Gender</strong> meanings in grammar and lexicon<br />
judged to be tall, large, strong, powerful, standing, or supporting, e.g. men, houses,<br />
trees; ‘sit’ is used with referents judged to be small, squat, horizontal, and weak, e.g.<br />
women, possums, and ponds. This is reminiscent of shape and dimensionality<br />
associated with masculine and feminine Linguistic <strong>Gender</strong>s discussed in §3.3. Most<br />
systems of classificatory verbs in Papuan languages have no special form for animates,<br />
or for men and women: all of <strong>the</strong>se are classified by what is conventionalized<br />
as <strong>the</strong>ir typical position. 12<br />
A number of languages have possessive classifiers—special morphemes which<br />
characterize a possessed noun within possessive constructions. Palikur has five possessive<br />
classifiers—one for domesticated animals, one for food, one for plants, one for<br />
animals caught to eat, and one for child. So, a classifier construction nu-kamkayh tino<br />
(1sg-CL:CHILD woman) is <strong>the</strong> only way of saying ‘my daughter’. A number of Uto-<br />
Aztecan languages (including Cora and Papago) have animate and inanimate classifiers<br />
in possessive constructions. Some Oceanic languages in New Guinea have special<br />
possessive classifiers for domestic animals. Hmong, a Hmong-Mien language of Laos,<br />
has a special classifier used for ‘possession’ of living beings illustrated in 5.2.<br />
5.2 nws tus txiv ntxawn tus ntxhais<br />
he CL:LIVING.BEING uncle CL:LIVING.BEING daughter<br />
‘<strong>the</strong> daughter of his uncle’<br />
Classifiers in Hmong are also used with numerals, demonstratives, and on nouns<br />
<strong>the</strong>mselves. It is a ‘multiple classifier’ language where classifiers are used in many<br />
environments. 13 Kilivila, an Oceanic language with many classifiers used on<br />
numerals, demonstratives, and adjectives, has classifiers te and to for humans and<br />
males, and na for females (see §6.1.1 on <strong>the</strong>ir origin). 14 Classifiers of different types in<br />
one language can classify humans and animates along different lines. In Jacaltec noun<br />
classifiers for humans offer a detailed categorization based on sex, age, kinship<br />
relations, and respect (Table 5.2).<br />
Numeral classifiers may categorize nouns into humans, animals, and inanimates<br />
only. Carrier, an Athabaskan language from British Columbia in Canada, has twelve<br />
distinct kinds of noun categorization through classificatory verb stems, demonstratives,<br />
numerals, and prefixes on verbs. Special forms for humans and non-humans<br />
are distinguished for numerals, demonstratives, relativizers, and <strong>the</strong> quantifier ‘how<br />
many’. 15<br />
5.1.4 Linguistic <strong>Gender</strong>s and classifiers as noun categorization devices:<br />
commonalities and differences<br />
Linguistic <strong>Gender</strong> and classifiers as noun categorization devices share <strong>the</strong>ir functions,<br />
and can be related historically. They can combine in one language, each with<br />
somewhat different meanings. Malto, a South Dravidian language, has two genders:
5.2 Sex, humanness, and animacy in noun categories 69<br />
male human versus <strong>the</strong> rest in <strong>the</strong> singular number, and human versus non-human<br />
in plural. Thirty numeral classifiers include a special form for humans (jen) and for<br />
non-human animals (maq) which include ghosts (Mahapatra 1979: 120).<br />
Dyirbal, an Australian language, has a system of four genders whose choice is<br />
meaning-based. Yidiñ, a neighbouring language, has about twenty noun classifiers.<br />
The two encode similar meaning, but in a different way. Dixon (2015: 58–9) offers a<br />
comparison of <strong>the</strong> semantic fields covered by <strong>the</strong> two techniques in <strong>the</strong> two languages.<br />
The semantic field of ‘people’ is covered by Yidiñ classifiers waguja ‘man’ and<br />
buña ‘woman’, and by Dyirbal noun gender markers bayi ‘masculine’ and balan<br />
‘feminine’. Yidiñ forms are restricted to living beings, while Dyirbal forms are also used<br />
for male and female spirits and legendary people. We have also seen that balan has<br />
o<strong>the</strong>r concepts associated with it: fire, drinkable liquids, and fighting. Two of <strong>the</strong>se are<br />
associated with different noun classifiers in Yidiñ: buri for fire, bana for drinkable<br />
liquid. There is no special classifier in Yidiñ for fighting or fighting implement.<br />
Every noun has to be assigned to a Linguistic <strong>Gender</strong> if a language has one.<br />
Classifiers are less restrictive—in many languages, not every noun has to be accompanied<br />
by a classifier. The choice of Linguistic <strong>Gender</strong> choice may be semantically<br />
opaque. The choice of classifier tends to be more straightforward. For instance,<br />
Palikur has three genders whose choice is mostly semantic, but with a degree of<br />
opacity, especially for non-humans. There are a dozen numeral classifiers, which<br />
include one classifier for humans with a masculine and a feminine form, chosen<br />
exclusively by a person’s sex. Classifiers can never be chosen by a noun’s morphological<br />
features or <strong>the</strong>ir form, unlike Linguistic <strong>Gender</strong>s. We will see, in Chapter 6,<br />
that classifiers can give rise to Linguistic <strong>Gender</strong>s.<br />
Just like Linguistic <strong>Gender</strong>s, classifiers may reflect social values and stereotypes,<br />
especially where women are concerned. This is what we saw in Maonan in §5.1.1—a<br />
topic to which we return in §11.1.<br />
5.2 Sex, humanness, and animacy in noun categories<br />
The majority of languages have some grammatical distinctions between humans and<br />
non-humans, or animates and inanimates. Uralic languages have no linguistic<br />
genders or classifiers. All <strong>the</strong> languages of <strong>the</strong> family differentiate between question<br />
words ‘who’ (Proto-Uralic *ki) and ‘what’ (Proto-Uralic *mi). Having different<br />
forms for ‘who’ and ‘what’ is not universal: <strong>the</strong> same form covers both in Australian<br />
languages Marra, Ngandi, and Kayardild and in Ewe, a Kwa language from Ghana. 16<br />
Universal semantic parameters of sex, animacy, and humanness can find <strong>the</strong>ir<br />
reflection in o<strong>the</strong>r noun categories—number, case, and declension classes.<br />
There are typically more number distinctions made for human nouns and for<br />
nouns referring to animates than for inanimates. In Koasati, a Muskogean language
70 5 <strong>Gender</strong> meanings in grammar and lexicon<br />
from <strong>the</strong> USA, only human nouns can be optionally marked for plural; o<strong>the</strong>r nouns<br />
do not take a plural marker at all. 17<br />
A human or an animate noun is most likely to be in <strong>the</strong> function of a transitive<br />
subject. If it appears in <strong>the</strong> object position it may acquire special marking. This is<br />
known as Differential Object Marking (DOM). 18 A typical example comes from<br />
Spanish. A direct object (O) occurs with a marker (<strong>the</strong> preposition a) if it is specific<br />
and animate, as in 5.3.<br />
5.3 Juan vio [a mi papá]<br />
Juan saw DIRECT.OBJECT my Dad<br />
‘Juan saw my Dad’<br />
If it is inanimate it will be unmarked:<br />
5.4 Juan vio [el libro]<br />
Juan saw <strong>the</strong> book<br />
‘Juan saw <strong>the</strong> book’<br />
A division of nouns into declension classes may correlate with animacy and<br />
humanness. Lezgian, a north-east Caucasian language, lost <strong>the</strong> genders reconstructed<br />
for <strong>the</strong> proto-language. The choice of its oblique stem suffixes correlates with <strong>the</strong><br />
meaning of nouns: <strong>the</strong> suffix -Adi is used with nouns referring to a non-discrete mass,<br />
and -rA is used with most nouns which denote animals and people. Armenian, an<br />
Indo-European language without Linguistic <strong>Gender</strong> in anaphora or in agreement,<br />
has <strong>the</strong> distinction between person and non-person in its inflectional paradigm. 19<br />
Most Slavonic languages distinguish three genders—masculine, feminine, and<br />
neuter, chosen on <strong>the</strong> basis of semantics and partly form (especially for inanimates).<br />
In addition, many have developed a special ‘animate’ subgender. Animate nouns take<br />
different agreement form from inanimates in <strong>the</strong> accusative case, e.g. Serbo-Croat<br />
òvog sȉna (this.ACC.ANIM son.ACC.ANIM) ‘this son’ (object) and òvāj prózor (this.ACC.<br />
INAN window.ACC.INAN) ‘this window’ (object). The inanimate object form is <strong>the</strong> same<br />
as <strong>the</strong> nominative, and <strong>the</strong> animate object form is <strong>the</strong> same as genitive. The animacy<br />
distinction is semantically transparent.<br />
In addition to this, some West Slavic languages have developed special agreement<br />
for nouns denoting males (known as ‘virile’ subgender). Three agreement genders<br />
(masculine, feminine, and neuter) are distinguished in singular and not in plural in<br />
Polish. Two different plural past tense agreement forms distinguish males from<br />
everyone else, in byli ‘were’ (males) versus byly ‘were’ (o<strong>the</strong>r subjects, non-males).<br />
Macedonian, a South Slavic language, has three genders (masculine, feminine, and<br />
neuter) in agreement and anaphora. A special ‘virile’ form of cardinal numbers<br />
covers men and mixed groups, e.g. trojca ‘three (male humans and mixed sex<br />
groups)’ and tri ‘three (any o<strong>the</strong>r genders; non-virile)’. 20
5.3 <strong>Gender</strong> in gender-less languages 71<br />
Animacy, humanness, and sex expressed through categories o<strong>the</strong>r than Linguistic<br />
<strong>Gender</strong> are typically binary, and semantically transparent. Once developed, <strong>the</strong>y can<br />
evolve into Linguistic <strong>Gender</strong>s: we turn to this in §6.1.<br />
We now turn to ways of expressing gender-related meanings in languages which<br />
have no Linguistic <strong>Gender</strong> classification in agreement, or in anaphora.<br />
5.3 <strong>Gender</strong> in gender-less languages<br />
In every language one can distinguish a bull from a cow, and a man from a woman,<br />
using different words. Words for ‘man’ and ‘woman’ can be added to nouns, to<br />
differentiate between sexes. Turkish and Indonesian have no anaphoric Linguistic<br />
<strong>Gender</strong> and no gender agreement. The lexeme hanɪm ‘woman, lady’ in Turkish is<br />
used in combination with terms referring to professions, e.g. hoca hanɪm ‘Lady<br />
Teacher’. In Indonesian, <strong>the</strong> masculine form lelaki or laki-laki can be added to<br />
a noun to refer to a male, e.g. orang laki-laki ‘man’, and perempuan to refer to a<br />
woman, e.g. organg perempuan ‘woman’. The feminine betina can be added to a term<br />
for an animal to denote <strong>the</strong> female, as in singa betina ‘lioness’, and djantan to denote<br />
a male, e.g. singa djantan ‘lion’. 21<br />
Languages without Linguistic <strong>Gender</strong> include <strong>the</strong> majority of Austronesian languages,<br />
Turkic, Tungusic, Uralic, and most Tupí and Carib languages in South<br />
America. Creole and Pidgin languages almost universally lack Linguistic <strong>Gender</strong>. 22<br />
Standard Hausa, a major language of Nor<strong>the</strong>rn Nigeria, has two genders, masculine<br />
and feminine. These are expressed in personal pronouns and a variety of agreement<br />
contexts, including possessive constructions. Pidginized Hausa (used as a means of<br />
inter-ethnic communication by non-native speakers) uses just one, erstwhile ‘masculine’,<br />
form in all contexts, without distinguishing masculine and feminine: 23<br />
5.5 Standard Hausa<br />
masc<br />
fem<br />
yā<br />
tā<br />
zō<br />
zō<br />
‘he has come’<br />
‘she has come’<br />
Pidgin Hausa<br />
yā zō ‘she/he has come’<br />
We can recall, from §2.1.2, that Swahili, just like many o<strong>the</strong>r Bantu languages, has<br />
adjectival agreement markers which reflect gender and o<strong>the</strong>r properties of referents.<br />
In <strong>the</strong> Kenya Pidgin Swahili, <strong>the</strong> zero gender marker has replaced all <strong>the</strong> agreement<br />
markers. Fanagalo, a Bantu-based creole spoken in Botswana, has lost genders typical<br />
for a Bantu language. Chinook Jargon was a pidgin based on a number of indigenous<br />
languages of <strong>the</strong> American Northwest (especially Lower Chinook, Nootka, and<br />
Chehalis), with elements of English, French, Hawaiian, and Chinese. Chinookan<br />
languages have three genders in <strong>the</strong> singular (masculine, feminine, and neuter); this<br />
feature is absent from <strong>the</strong> Chinook Jargon.
72 5 <strong>Gender</strong> meanings in grammar and lexicon<br />
Most European-based Creole languages use one form for ‘she’ and ‘he’ (e.g. Tok<br />
Pisin em, Kristang el, Papiamentu e, Nubi uwo ‘s/he’). A rare exception appears to be<br />
<strong>the</strong> Negerhollands Creole Dutch, spoken until recently in <strong>the</strong> United States Virgin<br />
Islands, which distinguishes between animate (h)am ‘he, she’ (cf. Dutch hem ‘him’)<br />
and inanimate di ‘it’ (cf. Dutch dit ‘this: neuter gender’). Adjectives in Europeanbased<br />
Creoles are typically based on a feminine form, but <strong>the</strong> same form is used for<br />
men and for women. A feminine form of an adjective ‘fat’ has been preserved in<br />
Guyanais Creole French with <strong>the</strong> meaning of ‘pregnant’ (grós, cf. French grosse); <strong>the</strong><br />
erstwhile masculine form, gro (French gros) has <strong>the</strong> meaning of ‘fat’.<br />
Some languages without Linguistic <strong>Gender</strong> have derivational affixes producing<br />
forms with male and female reference. Indonesian has a few masculine–feminine<br />
pairs, like dew-a ‘god’, dew-i ‘goddess’, borrowed from Sanskrit. Estonian, a language<br />
without Linguistic <strong>Gender</strong>, has many lexical pairs, including naine ‘woman’ and mees<br />
‘man’, tütar ‘daughter’ and poeg ‘son’, and so on. Similar to o<strong>the</strong>r Uralic languages,<br />
Estonian has no agreement gender or anaphoric gender. The same third person<br />
pronoun tema is used for any sex or animacy. A few suffixes on nouns help distinguish<br />
a man from a woman, e.g. kangelane ‘hero’ and kangelanna ‘heroine’, kuningas ‘king’<br />
and kuninganna ‘queen’. We will see in §6.4.2 that <strong>the</strong> feminine derivational suffix -tar<br />
in Estonian (and its Finnish counterpart -ttare) was borrowed from a Germanic source.<br />
This suffix is now used with numerous terms (native and borrowed) referring to<br />
women, e.g. runo-tar ‘poetess’, from runo ‘poem’ (native word), kuninga-tar ‘queen’.<br />
The word sõber ‘friend’ can refer to a man or a woman, and can occur with two<br />
feminine suffixes, with a slight change in meaning. The feminine sõbranna ‘female<br />
friend’ is nowadays used among women talking about female friends, and <strong>the</strong> less<br />
frequent sõbratar tends to refer to a man’s girlfriend. Ross (2012: 167) remarks that<br />
sõbranna has overtones of someone ‘indisputably more lightweight and unreliable<br />
than simply sõber’ and tends to occur with negative attributes such as ‘jealous’ or<br />
‘grumpy’. This is reminiscent of negative overtones associated with many female<br />
derivations in English—to which we return in §7.5.<br />
The Eastern Maroon Creole of Suriname (also known as Ndyuka) is Englishbased.<br />
Personal nouns can be marked as masculine (with man ‘man’) or as feminine<br />
(uman ‘woman’). Nouns marked with uman may simply denote <strong>the</strong> functions<br />
performed by women, and by men in a particular setting. Olo-man (grave-man) is<br />
a grave digger; olo-uman is <strong>the</strong> one who prepares, brings, and distributes food to <strong>the</strong><br />
olo-man (Migge 2001: 99). Or a male counterpart may have a generic meaning, as in<br />
waka-man ‘traveller’, and <strong>the</strong> female counterpart will have a more negative and oversexualized<br />
connotation: waka-uman is not just a ‘travelling woman’: this has strong<br />
connotations of a woman of easy virtue. Once again, this is reminiscent of negative<br />
overtones of terms referring to women—see §7.5 and §11.2. 24<br />
It is hard to translate expressions like ‘his and hers’ into a language where ‘he’ and<br />
‘she’ share <strong>the</strong> same form. In order to remedy <strong>the</strong> gap, a new partly jocular form
5.3 <strong>Gender</strong> in gender-less languages 73<br />
tema-ke has been recently introduced into Modern Estonian. Tema-ke consists of <strong>the</strong><br />
pronoun tema ‘he, she’ and <strong>the</strong> diminutive suffix. The form is far from being in active<br />
use: it is found in jocular columns in newspapers and in <strong>the</strong> internet, and in<br />
translations from English: <strong>the</strong> title of <strong>the</strong> film Her, by Spike Jonze (made in 2013),<br />
is Temake. Web-pages advertise getaways for him and her, tema and temake. That<br />
<strong>the</strong> diminutive suffix has been used to create a feminine form of a pronoun goes<br />
toge<strong>the</strong>r with small size associated with ‘femaleness’—something we mentioned in<br />
<strong>the</strong> preceding chapter. 25<br />
Languages without Linguistic <strong>Gender</strong> may have additional distinctions which<br />
most languages do not make. The absence of Linguistic <strong>Gender</strong> is a feature of<br />
Hungarian. A special suffix -nő marks a female referent, in pairs such as tanitó<br />
‘teacher’ and tanitó-nő‘female teacher’. A fur<strong>the</strong>r suffix -né refers to a ‘matrimonial<br />
feminine’, someone married to who <strong>the</strong> suffix is attached to. Tanító is ‘teacher’, and<br />
tanító-né is ‘teacher’s wife’. One can thus distinguish members of <strong>the</strong> royal family:<br />
ra<strong>the</strong>r than using a general term ‘queen’, one will refer as király-né to king’s wife, and<br />
as király-nőto a queen as <strong>the</strong> female head of state in her own right. 26<br />
As Amadiume (2015: 89) points out for her native Igbo, no distinction in gender is<br />
made in pronouns, ‘thus <strong>the</strong>re is no reminder in speech to distinguish between <strong>the</strong><br />
sexes’. In Turkish, Estonian, and Hungarian nouns denoting persons may refer to<br />
men and women. In Estonian, politseinik may refer to a policeman or a policewoman,<br />
arst to a female or to a male doctor, and õpetaja to a male or a female teacher.<br />
Speakers of languages without genders require extralinguistic—usage- and culturebased—clues<br />
in distinguishing men from women, in terms of sex and social standing.<br />
But <strong>the</strong> ways in which reference to men and to women is made remain asymmetrical<br />
and essentially male-biased.<br />
The word for ‘man’ is used as a generic noun meaning ‘person’ in many languages<br />
without Linguistic <strong>Gender</strong>. In Indonesian, mahasiswa ‘student’ and karyawan ‘clerk’<br />
often occur as generic terms, and <strong>the</strong>ir derivational female counterparts are used only<br />
occasionally. Kuntjara (2001) describes this as a way of ‘excluding’ women. Braun’s<br />
(2001) investigation of Turkish showed that forms which are gender-neutral can be<br />
biased: an experiment with 130 university students (78 females and 52 males) showed<br />
that nouns such as köylü ‘villager’, kis̹i ‘voyager’,andyoncu ‘passenger’, and <strong>the</strong> pronoun<br />
birisi ‘someone’, were spontaneously conceived of as males in <strong>the</strong> first place.<br />
This inherent gender bias reflects male dominance across Turkish society: ‘men<br />
are simultaneously <strong>the</strong> privileged group and <strong>the</strong> leading figures’ in <strong>the</strong> Turkish<br />
economy, labour market, politics, law, and religion. If one needs to specify that <strong>the</strong><br />
person is a woman, a word ‘woman’ or ‘lady’ will be added. Generally speaking,<br />
Turkish terms for person reference contain information about <strong>the</strong> division of labour<br />
in Turkish society. Futbolcu ‘football player’ will be immediately understood as<br />
relating to a man, and c̹ocuk bakɪcɪsɪ ‘nursery school teacher’ as relating to a<br />
woman. The use of gendered and generic nouns in Turkish discourse and in proverbs
74 5 <strong>Gender</strong> meanings in grammar and lexicon<br />
provides clues about which terms are more ‘appropriate’ for a woman and which<br />
ones for a man. Generic terms such as esi-mees (first-man) for ‘chairman’ used to<br />
refer to a man and to a woman reflect <strong>the</strong> erstwhile dominance of ‘men’ in Estonian<br />
(though <strong>the</strong> term esi-naine ‘chairwoman’ is now also used). Male—ra<strong>the</strong>r than<br />
female—reference can still be considered functionally unmarked.<br />
In Braun’s (2001: 305) words, in Turkish and in o<strong>the</strong>r languages, ‘men are central<br />
to <strong>the</strong> “human” category’, and ‘a peripheral or secondary position is what remains for<br />
females. This “extra-ordinary” status of women is linguistically communicated by a<br />
pronounced tendency towards gender marking in female reference’: a child (çocuğu)<br />
will be likely to be associated primarily with a male child, while female reference has<br />
to be explicitly expressed, as in kɪz çocuğu ‘girl child’.<br />
The stereotypes associated with Social <strong>Gender</strong>, and <strong>the</strong> male bias, will be reflected<br />
in any language—with or without Linguistic <strong>Gender</strong>—as long as <strong>the</strong>y are relevant for<br />
<strong>the</strong> community which speaks it. Unravelling <strong>the</strong>m will be a task as demanding as that<br />
involved in understanding <strong>the</strong> social and cultural underpinnings in noun categorization<br />
through dedicated means—including Linguistic <strong>Gender</strong>.<br />
NOTES AND SOURCES<br />
1. Advocated for in Aikhenvald (2000, 2006, and forthcoming).<br />
2. See Lee (2014: 42–3) on Korean. See Adams (1989: 56–7) on examples of special classifiers<br />
for men and women in Austroasiatic languages; see Gerdts and Hinkson (2004: 254) on<br />
Halkomelem, Craig (1986a: 265) on Jacaltec, Sneddon (1996) on Indonesian, Thompson<br />
(1987: 194) and Daley (1998: 96–7) on Vietnamese; Enfield (2007: 125, 142) on Lao;<br />
Lu (2012: 83) on Maonan and Zhuang. A full list of human classifiers in Zhuang is in Lu<br />
(2012: 114).<br />
3. From Becker (1975: 116). See also Aikhenvald (2000: 315) on metaphorical extensions of<br />
human classifiers.<br />
4. See Lu (2012: 110), Diller (1985).<br />
5. See Ikoro (1996: 90–1) on Kana; o<strong>the</strong>r systems of numeral classifiers are discussed in<br />
Aikhenvald (2000: 287–8).<br />
6. Dixon (1982: 185; 2015: 44).<br />
7. See Goddard (1985: 94) on Yankunytjatjara, and Walsh (1997: 256) on Murinhpatha.<br />
8. See Craig (1986a: 266–7), on Jacaltec; England (1983) on Mam. Number 7 indicates a<br />
glottal stop.<br />
9. See Rushforth (1991: 253). See Aikhenvald (2000: chapter 5), (2006) on how classificatory<br />
verbs differ from lexical pairs like English eat (solid food) and drink (liquid food), and<br />
fur<strong>the</strong>r features of verbal classifiers.<br />
10. See Poser (2005) on Carrier, Drapeau and Lambert-Brétière (2011: 302–4) on Innu;<br />
Aikhenvald and Green (2011) on Palikur.<br />
11. Haiman (1980) on Hua, Martin (1975) on Japanese, Zhang (2014: 445) on Ersu.<br />
12. Similar systems have been described for Ku Waru and Kewa, also from <strong>the</strong> Papuan region.<br />
See Lang (1975) on Enga, Franklin (1981) on Enga and Kewa, Merlan, Roberts, and<br />
Rumsey (1997) on Ku Waru.
5.3 <strong>Gender</strong> in gender-less languages 75<br />
13. Jarkey (2015), Bisang (1993: 29–30).<br />
14. See Senft (1986: 76) for Kilivila.<br />
15. See Poser (2005: 162–3).<br />
16. See Collinder (1965: 138) for Uralic, Dixon (2002: 329) and references <strong>the</strong>re for Australian<br />
languages, Ameka (2012: 53–4) for Ewe.<br />
17. This tendency was summarized as <strong>the</strong> principle of Nominal Hierarchy by Smith-Stark<br />
(1974).<br />
18. See Bossong (1985, 1991).<br />
19. Haspelmath (1993: 76–7), Alexeyev (1985: 27ff.), Matasović (2004).<br />
20. On Macedonian, see Friedman (1993: 267); for <strong>the</strong> development and characteristics of<br />
animate subgenders in Slavonic languages see Comrie and Corbett (1993: 16); Schenker<br />
(1993: 108).<br />
21. See Kwee (1965: 55) on Indonesian, Braun (2000a: 53) for Turkish, Collinder (1965: 57)<br />
for Finnish, Hasselblatt (2015) and Wiedemann (1875: 196–7) for Estonian.<br />
22. Pidgins and Creoles arise as a makeshift means for filling communicative needs, ra<strong>the</strong>r<br />
than evolving in <strong>the</strong> way natural, non-contact languages would. See Holm (2000: 216–17,<br />
222) on general features of Creoles and <strong>the</strong>ir lack of genders; Silverstein (1972a: 397–8)on<br />
traditional Chinook, and Silverstein (1972b: 618) on <strong>the</strong> Chinook Jargon, Owens (2001:<br />
349, 355; 2014: 283) on Arabic-based Creoles, Anderson and Janson (1997: 183–5) on<br />
Fanagalo. As pointed out in Aikhenvald (2000: 389), Creole and Pidgin languages uniformly<br />
lack classifiers or o<strong>the</strong>r noun categorization devices.<br />
23. See Heine and Reh (1984: 42–3); Hodge (1958).<br />
24. Braun (2001) refers to <strong>the</strong> expression of a combination of social gender and natural gender<br />
in languages without linguistic gender, and <strong>the</strong> social dichotomy of male and female roles<br />
in language expressed through lexicon, as ‘covert’ gender. This is misleading, as such usage<br />
can lead to confusion with covert linguistic gender (which is not marked on nouns<br />
<strong>the</strong>mselves).<br />
25. See Ross (1996, 2012), Hasselblatt (2015) on Estonian.<br />
26. See Vasvári (2015: 210–11), on <strong>the</strong> use and origin of <strong>the</strong>se forms.
6<br />
The rise and fall of Linguistic <strong>Gender</strong>s<br />
Linguistic <strong>Gender</strong>s have <strong>the</strong>ir histories. They can develop out of lexical items or noun<br />
categories with meanings of animacy or humanness. The meanings of Linguistic<br />
<strong>Gender</strong>s can change over time. Linguistic <strong>Gender</strong>s can be gained, or lost, if languages<br />
are in contact with each o<strong>the</strong>r. Social changes may affect <strong>the</strong> use of Linguistic<br />
<strong>Gender</strong>s applied to human beings. But losing—or gaining—a Linguistic <strong>Gender</strong> is<br />
independent of increase or decrease in equality of Social <strong>Gender</strong>s. We now turn to<br />
<strong>the</strong> mechanisms of developing Linguistic <strong>Gender</strong>.<br />
6.1 Developing Linguistic <strong>Gender</strong><br />
Nouns referring to humans—man, woman, person—can develop into a closed<br />
system of anaphoric gender. Or <strong>the</strong>y can give rise to derivational gender markers.<br />
This process is known as grammaticalization. 1 Once a language has acquired a<br />
marker of anaphoric or derivational gender, this can be reinterpreted and reanalysed<br />
as agreement gender.<br />
Pathways in <strong>the</strong> development of Linguistic <strong>Gender</strong> are summarized in Box 6.1.<br />
BOX 6.1 <strong>How</strong> to gain a linguistic gender: pathways of development<br />
PATHWAY I. Generic nouns > ANAPHORIC GENDER > AGREEMENT GENDER—§6.1.1 and<br />
PATHWAY IV<br />
PATHWAY II. Generic nouns > DERIVATIONAL GENDER—§6.1.1; <strong>the</strong>n see also PATHWAY V<br />
PATHWAY III. Generic nouns > closed system of NOUN CLASSIFIERS > DERIVATIONAL GENDER ><br />
AGREEMENT GENDER—§6.1.2 and PATHWAY IV<br />
PATHWAY IV. ANAPHORIC GENDER > AGREEMENT GENDER—§6.1.3<br />
PATHWAY V. DERIVATIONAL GENDER > AGREEMENT GENDER—§6.1.4<br />
Pathways I–III involve grammaticalization of nouns into exponents of Linguistic<br />
<strong>Gender</strong> (and <strong>the</strong>ir subsequent reanalysis and reinterpretation). Pathways IV–V<br />
<strong>How</strong> <strong>Gender</strong> <strong>Shapes</strong> <strong>the</strong> <strong>World</strong>. First edition. Alexandra Y. Aikhenvald.<br />
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.
6.1 Developing Linguistic <strong>Gender</strong> 77<br />
involve reanalysis and reinterpretation of markers of Linguistic <strong>Gender</strong>, and a change<br />
from one kind of gender to ano<strong>the</strong>r.<br />
As we saw in §2.1.3, Linguistic <strong>Gender</strong> interrelates with o<strong>the</strong>r noun categories. Core<br />
meanings of Linguistic <strong>Gender</strong>s—humanness, sex, and animacy—can be expressed<br />
through number, case, and classifiers of various types (§§5.1–2). Reinterpretation of<br />
<strong>the</strong>se categories may account for <strong>the</strong> emergence of Linguistic <strong>Gender</strong> classification.<br />
6.1.1 From generic nouns to Linguistic <strong>Gender</strong>s<br />
Nouns referring to humans (‘man’, ‘person’, and ‘woman’) can give rise to a closed<br />
grammatical system of anaphoric gender in personal pronouns. Kokota, an Oceanic<br />
language spoken in <strong>the</strong> Solomon Islands, distinguishes masculine and feminine<br />
genders in its third person pronouns. The masculine pronoun manei comes from<br />
mane ‘man’. 2 In Bari and Toposa, East Nilotic languages, masculine personal pronoun<br />
lo ‘he’ and its feminine counterpart na come from nouns, *lV ‘member/person<br />
of ’ and *nyaa- ‘girl, daughter’. 3<br />
Newly grammaticalized personal pronouns—that is, exponents of anaphoric<br />
gender—can be reanalysed as agreement gender markers. Zande, an Ubangian<br />
language spoken in <strong>the</strong> Democratic Republic of Congo, distinguishes four genders<br />
in <strong>the</strong> singular, and three in plural personal pronouns. Feminine and masculine<br />
genders in <strong>the</strong> singular collapse into one ‘human’ gender in <strong>the</strong> plural. In agreement<br />
with a well-established typological tendency, <strong>the</strong>re are fewer genders in plural than<br />
in <strong>the</strong> singular (see §2.1.3). <strong>Gender</strong> and number distinctions in Zande personal<br />
pronouns are shown in Scheme 6.1.<br />
SINGULAR PLURAL<br />
Masculine<br />
Human<br />
Feminine<br />
Animal Animal<br />
Inanimate Inanimate<br />
SCHEME 6.1 <strong>Gender</strong> and number in Zande personal pronouns<br />
A number of terms in <strong>the</strong> system come from independent nouns—as shown in 6.1:<br />
6.1 Lexical origins of Zande pronouns<br />
NOUN<br />
GENDERED PRONOUN<br />
man, male > masculine<br />
person > human<br />
animal, meat (pl) > animal (pl)<br />
thing > inanimate
78 6 The rise and fall of Linguistic <strong>Gender</strong>s<br />
Some of <strong>the</strong> original nouns which gave rise to pronouns are still used as nouns in<br />
related languages, e.g. Zande kɔ́‘he (masculine)’, Banda kɔ́, Ngala kwã ‘man’.<br />
Personal pronouns (lexical in origin) have given rise to agreement markers used on<br />
adjectives (as copula complements). In 6.2, <strong>the</strong> personal pronoun kɔ́appears as a<br />
suffix and as a prefix marking agreement on <strong>the</strong> verb and <strong>the</strong>n on <strong>the</strong> adjective ‘big’:<br />
6.2 kɔ́-ni bakέrέ-kɔ́<br />
he-be big-he<br />
‘He is big’<br />
Sometermswithinlargergendersystemsmayhavedevelopedoutofnouns.The<br />
diminutive gender marker -pi in Bantu languages of Cameroon comes from Niger-Congo<br />
*bi ‘child’, and <strong>the</strong> locative gender marker -ku comes from <strong>the</strong> noun *ku ‘outside’. 4<br />
Nouns meaning ‘man’ or ‘person’ can give rise to numeral classifiers referring to<br />
humans. For instance, <strong>the</strong> Korean numeral classifier salam ‘CLASSIFIER:PERSON/ADULT’<br />
is related to <strong>the</strong> noun sālam. 5 Feminine and masculine terms within a system of<br />
classifiers in multiple environments can come from grammaticalized forms for ‘man’<br />
and ‘woman’. Kilivila, an Oceanic language spoken on <strong>the</strong> Trobriand Islands in<br />
Papua New Guinea, has over a hundred classifiers which characterize a noun in<br />
terms of its sex, animacy, shape, and o<strong>the</strong>r intrinsic features. Classifiers are prefixes<br />
to numerals and adjectives and infixed to demonstratives. 6 One of <strong>the</strong> sex-based<br />
classifiers, to/te- ‘man, human being in general’, comes from <strong>the</strong> noun tau ‘man’. The<br />
o<strong>the</strong>r sex-based classifier na- ‘women, females’ comes from vivina ‘woman’.<br />
Noun classifiers typically stem from grammaticalized nouns with generic meanings.<br />
In Akatek, a Mayan language, naj, a noun classifier meaning ‘man’ developed<br />
from <strong>the</strong> noun winaj ‘man’. The noun classifier ʔis᷈̍ ‘woman’ developed from ʔis᷈̍<br />
‘woman’. In Mam, also Mayan, sex- and social gender-sensitive classifiers developed<br />
from nouns referring to males and females—see Table 6.1.<br />
TABLE 6.1. Noun classifiers for humans in Mam and nouns <strong>the</strong>y originated from<br />
CLASSIFIER<br />
q'a ‘classifier: young man’<br />
txin ‘classifier: young woman’<br />
ma ‘classifier: man’<br />
xu7j ‘classifier: woman’<br />
swe7j ‘classifier: old man’<br />
xhyaa7 ‘classifier: old woman’<br />
xnuq ‘classifier: old man, respectfully’<br />
xuj ‘classifier: old woman, respectfully’<br />
COMMON NOUN<br />
q'aa ‘young man’<br />
txiin ‘young woman’<br />
matiij ‘big’<br />
xu7j ‘woman’<br />
swe7j ‘old man’<br />
yaab'aj ‘grandmo<strong>the</strong>r’<br />
xiinaq ‘old man, respectfully’<br />
xu7j ‘old woman, respectfully’
6.1 Developing Linguistic <strong>Gender</strong> 79<br />
Noun classifiers in Coatzoquitengo Mixtec, a Mixtecan language from Mexico,<br />
identify a referent as male, female, deity, animal, longish object, or liquid. Of <strong>the</strong>se,<br />
classifiers for ‘male’ and ‘female’ come from nouns meaning ‘man’ and ‘woman’:<br />
Proto-Mixtec *tee > tà ‘noun classifier: male’, Proto-Mixtec *ñaha > ñà ‘noun<br />
classifier: female’. Classifiers are used similarly to definite articles, e.g. tà loho (CL:MALE<br />
young.person) ‘<strong>the</strong> boy’, ñà loho (CL:FEMALE young.person) ‘<strong>the</strong> girl’. 7<br />
Historical development from ‘man’ to ‘person, human in general’ can be compared<br />
to <strong>the</strong> use of <strong>the</strong> masculine form, or a masculine pronoun, to refer to humans as a<br />
class. No language so far has a similar development, for a noun meaning ‘woman’ to<br />
‘person, human in general’.<br />
‘Inherent nature’ noun classifiers in Australian languages include terms for ‘man’,<br />
‘woman’, and ‘person’. They transparently come from independent nouns with<br />
generic reference. Yidiñ forms ‘person’, ‘woman’, ‘man’ can be used as full nouns<br />
and as noun classifiers. In some languages <strong>the</strong>se noun classifiers gave rise to derivational<br />
gender markers and to agreement genders, as we will see in §6.1.2.<br />
A noun meaning ‘person’, ‘man’, ‘woman’,or‘daughter’ can grammaticalize into a<br />
derivational gender affix referring to a male or to a female. In Mupun, a West Chadic<br />
language, masculine and gender can be marked by prefixes on proper names and<br />
some common nouns, and on a pronominal element meaning ‘such and such’. The<br />
feminine prefix nà- has developed via grammaticalization of naa ‘woman’. The<br />
masculine prefix dà- comes from daa ‘man’, e.g. nà-kómtàk ‘such and such a female’,<br />
dà-kómtàk ‘such and such a male’.<br />
The suffix -(a)nɪm in Turkish appears in a few words with female reference, e.g.<br />
hocanɪm ‘female teacher’. It clearly comes from grammaticalization of <strong>the</strong> lexeme<br />
hanɪm ‘woman, lady’ in combination with terms referring to professions, especially in<br />
address, e.g. hoca hanɪm ‘Lady Teacher’. 8<br />
No examples of generic nouns developing into gender-sensitive derivational affixes<br />
and <strong>the</strong>n into agreement genders have been reported so far. We will see in §6.1.4 how<br />
derivational gender markers can develop into agreement genders.<br />
6.1.2 From generic nouns to noun classifiers and <strong>the</strong>n to Linguistic <strong>Gender</strong>s<br />
Once generic nouns have developed into noun classifiers, <strong>the</strong>y can give rise to<br />
Linguistic <strong>Gender</strong>s. This pathway starts with grammaticalization of nouns into a<br />
closed system of noun classifiers. The next step involves reinterpreting noun classifiers<br />
as affixes to a noun <strong>the</strong>y occur with, and as agreement markers on accompanying<br />
modifiers. The development of agreement gender does not have to pass through<br />
<strong>the</strong> stage of derivational gender.<br />
Ngan.gityemerri, a language from nor<strong>the</strong>rn Australia, has fifteen noun classifiers. 9<br />
Six of <strong>the</strong>se are free words (which have grammaticalized out of generic nouns), five<br />
are prefixes (or proclitics), and four are both free words and prefixes (or proclitics).
80 6 The rise and fall of Linguistic <strong>Gender</strong>s<br />
Ngan.gityemerri also has five gender classes whose origin lies in noun classifiers. The<br />
development from noun classifiers to agreement genders involves several steps.<br />
STEP 1: involves a generic-specific ‘pairing’ of nouns: a generic noun classifier precedes<br />
<strong>the</strong> specific noun it refers to.<br />
6.3 [gagu wamanggal] NOUN.CL-NOUN kerre ngeben-da<br />
NCL:ANIMAL wallaby big 1sgS+AUX-shoot<br />
‘I shot a big wallaby’<br />
STEP 2: since noun classifiers are independent words, often favoured over specific<br />
nouns especially to maintain reference within a story, a noun phrase can consist of a<br />
generic noun plus a modifier, as in 6.4.<br />
6.4 gagu kerre ngeben-da<br />
NCL:ANIMAL big 1sgS+AUX-shoot<br />
‘I shot a big animal [wallaby]’<br />
STEP 3: if specific nouns are included, both <strong>the</strong> specific noun and modifiers tend to<br />
‘attract’ generics. This ‘repetition’ of noun classifiers is <strong>the</strong> predecessor of agreement:<br />
6.5 [gagu wamanggal] NOUN.CL-NOUN [gagu kerre] NOUN.CL-ADJ ngeben-da<br />
NCL:ANIMAL wallaby animal big 1sgS+AUX-shoot<br />
‘I shot a big wallaby’ (lit. I shot an animal-wallaby animal-big)<br />
STEP 4: Repeated noun classifiers cliticize to <strong>the</strong> specific nouns, and are reduced to<br />
proclitics which effectively mark agreement, as in 6.6.<br />
6.6 wa=ngurmumba wa=ngayi darany-fipal-nyine<br />
male=youth male=big 3sgS+AUX-return-FOCUS<br />
‘My initiand son has just returned’<br />
STEP 5: Noun class marking proclitics become obligatory prefixes on <strong>the</strong> nouns and<br />
on agreeing modifiers. At this stage, <strong>the</strong>y become fully grammaticalized:<br />
6.7 wú-pidìrri wu=mákarri<br />
NCL:CANINE-dingo<br />
‘a bad dog’<br />
NCL:CANINE=bad<br />
STEP 6: Thisstageinvolves‘prefix absorption’: that is, a gender class prefix onanoun<br />
becomes lexicalized as part of it. Some prefixed roots may be interpreted as stems which<br />
can take fur<strong>the</strong>r gender marking. This may lead to double class marking, e.g. wa-mumu<br />
(male-police) ‘policeman’; wur-wa-mumu (female-male-police) ‘policewoman’.<br />
Similar scenarios applied in o<strong>the</strong>r Australian languages. In Dyirbal, an Australian<br />
language with no prefixes, <strong>the</strong> ‘edible vegetable’ gender developed out of a generic<br />
noun *mayi ‘non-flesh food’. The generic noun reinterpreted as a prefix to
6.1 Developing Linguistic <strong>Gender</strong> 81<br />
deictic determiners, and underwent morphological reduction to -m. 10 But unlike<br />
Ngan.gityemerri, Dyirbal developed gender agreement markers without ever going<br />
through <strong>the</strong> stage of a derivational gender.<br />
6.1.3 From anaphoric gender to agreement gender<br />
Agreement genders can arise via reinterpretation of anaphoric gender expressed<br />
in demonstratives or third person pronouns. This scenario has been described for a<br />
number of Australian languages. Warndarran has five Linguistic <strong>Gender</strong>s. The<br />
marker for <strong>the</strong> class of human males rna- transparently relates to <strong>the</strong> recurrent free<br />
form pronoun masculine na. The marker for <strong>the</strong> class of human females (which also<br />
includes some animals) ŋa- relates to <strong>the</strong> free pronoun ŋa(l). Along similar lines,<br />
Greenberg (1978: 61ff.) proposed a chain of historical development, from demonstrative<br />
to definite article to specific article to marker of a noun, or an overt gender<br />
marker, in Voltaic languages. 11<br />
Different terms in one gender system in one language may come from a variety of<br />
sources. In Warndarran <strong>the</strong> prefix for <strong>the</strong> class of plants with edible parts ma- is<br />
linked to <strong>the</strong> grammaticalization of noun classifier mayi ‘edible plant’.<br />
6.1.4 From derivational gender to agreement gender<br />
An already existing derivational gender marker can evolve into an exponent of gender<br />
agreement. <strong>Gender</strong> systems are a salient feature of most Indo-European languages. The<br />
Anatolian languages are thought to be <strong>the</strong> most archaic. Hittite, <strong>the</strong> first Anatolian<br />
language to be deciphered and still <strong>the</strong> best understood, had just two genders—neuter<br />
(or ‘inanimate’) and‘common’ (or ‘animate’). It is now <strong>the</strong> consensus in <strong>the</strong> Indo-<br />
European scholarship that <strong>the</strong> feminine gender in <strong>the</strong> branches of Indo-European<br />
(o<strong>the</strong>r than Anatolian) evolved after <strong>the</strong> Anatolian languages split off.<br />
A later origin for feminine gender is supported by a number of arguments. For<br />
instance, <strong>the</strong> Proto-Indo-European question word *k w is ‘who?’ referred to both<br />
feminine and masculine, and a separate form *k w id ‘what?’ was used for <strong>the</strong> neuter.<br />
Some compound adjectives in Greek have one form for masculine and feminine, e.g.<br />
rhodo-dáktulos ‘rosy-fingered (he or she)’, and a separate one for neuter, e.g. rhododáktulon<br />
‘rosy-fingered (it)’. 12 Exactly how Indo-European languages acquired <strong>the</strong>ir<br />
feminine gender is a matter of some debate. Most scholars believe that agreement<br />
gender came about through reinterpretation of derivational affix *-h 2 used to form<br />
abstract and action nouns. This may also have involved analogy with <strong>the</strong> term for<br />
‘woman’, *g w enh 2 (<strong>the</strong> noun which gave rise to words for ‘woman’ in many languages,<br />
including Greek gine ‘woman’, and incidentally queen in English). Due to its<br />
ending in <strong>the</strong> laryngeal *h 2 this noun may have become associated with <strong>the</strong> abstract<br />
and collective nouns with <strong>the</strong> same ending, which was reanalysed as a derivational<br />
marker for <strong>the</strong> new feminine gender. Once reanalysed as a gender marker, it<br />
extended to mark agreement. 13
82 6 The rise and fall of Linguistic <strong>Gender</strong>s<br />
6.1.5 Linguistic <strong>Gender</strong> from o<strong>the</strong>r nominal categories<br />
Linguistic <strong>Gender</strong> distinctions can originate in reinterpretation of some o<strong>the</strong>r nominal<br />
category—typically number and case.<br />
Languages tend to have more number distinctions for human and animate nouns<br />
than for inanimates (see §2.1.3). As a consequence, categorization based on animacy<br />
may be encoded through number marking. A combination of number and animacy<br />
forms <strong>the</strong> basis of gender-like agreement classes in Kiowa (a Kiowa-Tanoan language<br />
of Oklahoma). Nouns of Class 1 are animate (and some inanimates which have<br />
animacy in legends, e.g. ‘sun’). Nouns of Class 2 are mostly inanimates, while those of<br />
Class 3 are inherently paired (and thus dual) (Class 4 can be considered a residue)<br />
(Watkins 1984: 81–91).<br />
The marking of <strong>the</strong> agreement classes is on nouns <strong>the</strong>mselves and in pronominal<br />
prefixes on verbs. Balochi, a Western Iranian language, has no gender distinctions—a<br />
feature shared with its relatives Persian, Kurdish, and Zazaki. The category ‘human’<br />
is marked just on one plural form: a special marker of group plurality kadag<br />
(originally meaning ‘settlement, village’) occurs only with humans. 14<br />
A development of a special ‘masculine’ (or ‘virile’) form out of reinterpreted case–<br />
number combination has been reported for Polish. The dual genitive-accusative<br />
came to be used with nouns of masculine gender exclusively, giving rise to ‘virile’<br />
forms initially used just with number word ‘two’ and <strong>the</strong> modifier ‘both’. Over <strong>the</strong><br />
course of <strong>the</strong> fourteenth–seventeenth centuries, dual forms were gradually replaced<br />
with plural forms and extended to constructions with o<strong>the</strong>r numbers. The virile<br />
genitive construction in Polish appears in number words in 6.8, for ‘men’, and 6.9, for<br />
‘women’:<br />
6.8 pie̹ciu/dzie̹sieciu/wielu panów<br />
five:VIRILE/ten:VIRILE/many:VIRILE<br />
‘five/ten/many men’<br />
6.9 pie̹ć̩/dzie̹siḛć/wiele kobiet<br />
five/ten/many women:GEN.PL<br />
‘five/ten/many women’ 15<br />
men:GEN.PL<br />
New gender distinctions can be developed through reanalysis of existing case<br />
forms. 16 We can recall, from §5.2, how Spanish has come to distinguish animate<br />
and inanimate nouns through marking objects (examples 5.3–5.4).<br />
During <strong>the</strong> Common Slavic period, <strong>the</strong> nominative and <strong>the</strong> accusative singular<br />
forms of masculine nouns fell toge<strong>the</strong>r, due to phonological shortening and erosion<br />
of final segments. The masculine gender thus became similar to neuter (where<br />
nominative and accusative forms were <strong>the</strong> same). The homonymy of two different<br />
cases (or case ‘syncretism’) was solved by substituting <strong>the</strong> genitive for <strong>the</strong> accusative<br />
for masculine animate nouns. Originally, this may have started just from nouns
6.2 Reshaping Linguistic <strong>Gender</strong>s 83<br />
referring to singular males. In thirteenth- and fourteenth-century Russian, only free<br />
adult males were covered by <strong>the</strong> ‘animate’ declension—thus adding a sociolinguistic<br />
dimension to <strong>the</strong> classification of nouns. The pattern of different declension for<br />
animate and inanimate masculine nouns in Slavic languages can be accounted for by<br />
<strong>the</strong> principles behind Differential Object Marking (mentioned in §5.2): animate, and<br />
especially human nouns tend to acquire special marking when used as objects. 17<br />
Proto-Indo-European gender has been lost in a few languages of <strong>the</strong> family. In<br />
some, this loss was accompanied by emergence of a new, gender-like, distinction.<br />
Armenian lost gender distinctions before its first attestation. There is now a distinction<br />
between animate versus inanimate, or person–non-person, in case paradigms. 18<br />
Bengali, an Indo-Aryan language, does not have any gender distinctions in agreement.<br />
A new, human versus non-human, distinction appears in personal pronouns.<br />
Third person pronouns typically refer to humans. To talk about a non-human, one<br />
uses a demonstrative. 19<br />
6.2 Reshaping Linguistic <strong>Gender</strong>s<br />
Linguistic <strong>Gender</strong>s change, expand, and contract in a variety of ways. Derivational<br />
gender markers on nouns can start being used with personal pronouns, and <strong>the</strong>n<br />
develop into an anaphoric gender. South Dravidian languages (which originally had a<br />
masculine/non-masculine gender distinction) innovated a special feminine gender<br />
distinction in personal pronouns by extending <strong>the</strong> feminine derivation suffix *-al̍to<br />
<strong>the</strong> pronominal root *aw. The feminine personal pronoun awal̍was created following<br />
<strong>the</strong> analogy of pairs such as *mak-antu ‘son’, *mak-al̍<br />
‘daughter’. 20 This new<br />
anaphoric gender has its roots in <strong>the</strong> erstwhile derivational marker.<br />
The principles of Linguistic <strong>Gender</strong> choice may change over time. In Hausa,<br />
feminine gender is now chosen on <strong>the</strong> basis of <strong>the</strong> noun’s phonological form:<br />
nouns in -aa are feminine. This ending goes back to a fossilized derivational suffix<br />
used to form feminine nouns. What used to be a morphological principle of assignment<br />
is now phonological. 21<br />
Linguistic <strong>Gender</strong>s may develop new meanings. Konkani, an Indo-Aryan language<br />
spoken on <strong>the</strong> west coast of India, has three Linguistic <strong>Gender</strong>s—masculine, feminine,<br />
and neuter. There is anaphoric gender in pronouns, and gender agreement on<br />
some adjectives and verbs. In some dialects, a young woman can be referred to with<br />
<strong>the</strong> neuter pronoun tε̃. The erstwhile feminine pronoun ti refers to an older female. If<br />
a romantic novel or a movie with a young heroine has <strong>the</strong> title She its appropriate<br />
translation would involve <strong>the</strong> erstwhile neuter pronoun tε̃‘(younger) she, formerly<br />
it’, and not ti ‘she’. 22<br />
Markers of Linguistic <strong>Gender</strong>s may be lost altoge<strong>the</strong>r. Then nouns regroup, based<br />
on <strong>the</strong>ir meanings and on <strong>the</strong>ir form. Examples come from <strong>the</strong> history of Romance<br />
languages.
84 6 The rise and fall of Linguistic <strong>Gender</strong>s<br />
Latin had three genders—masculine, feminine, and neuter. These were marked on<br />
<strong>the</strong> noun itself, and also by agreement on adjectives and a variety of pronouns. Most<br />
Romance languages lost <strong>the</strong> neuter gender, and now have just masculine and<br />
feminine. 23 The erstwhile Latin neuter nouns were redistributed between masculine<br />
and feminine. So, neuter nouns of <strong>the</strong> second declension (which contained numerous<br />
masculine nouns) became masculine, e.g. Latin ferrum ‘iron’, Italian and Portuguese<br />
ferro, Spanish hierro, French fer, Catalan ferre, Sardinian ferru. Some neuter plurals<br />
ending in -a were drawn into <strong>the</strong> feminine gender, as -a was reinterpreted as a feminine<br />
gender marker, e.g. arma ‘arms’, Italian, Sardinian, Catalan, Spanish, Portuguese arma,<br />
French arme.<br />
Some neuter nouns had variable gender in Latin; <strong>the</strong>y became masculine in some<br />
languages and feminine in o<strong>the</strong>rs. The descendant of <strong>the</strong> Latin neuter noun mare ‘sea’<br />
is feminine in French (la mer), but masculine in Portuguese (o mar) and in Spanish<br />
(el mar; but old Spanish la mar). Analogy played a role. According to Pope (1934:<br />
304), mer ‘sea’ in French was co-opted into <strong>the</strong> feminine gender by analogy with its<br />
counterpart, terre ‘earth, ground’ (Latin terra, also feminine). Similarly, Old French<br />
vallis ‘valley’ changed its gender from masculine to feminine under <strong>the</strong> influence of<br />
mons ‘mountain’. Latin aestas ‘summer’ (French été) was originally feminine, but<br />
became masculine under analogical pressure of o<strong>the</strong>r terms for seasons. The Old<br />
French dent ‘tooth’ changed its gender from masculine to feminine, under <strong>the</strong><br />
influence of bouche ‘mouth’. The history of ‘redistribution’ of neuter nouns boils<br />
down to a history of individual words.<br />
The loss of neuter occurred in o<strong>the</strong>r Indo-European languages. In Celtic, <strong>the</strong><br />
neuter disappeared fairly early. Old Irish had three genders. The neuter was lost in<br />
<strong>the</strong> Middle Irish period, c. tenth–eleventh centuries AD. Hardly any traces of neuter<br />
gender have been recorded for Welsh and Breton. Latvian, a Baltic language, has only<br />
masculine and feminine genders. Neuter has been lost in a few Slavic varieties, among<br />
<strong>the</strong>m <strong>the</strong> Sele Fara dialect of Slovene spoken in Austria.<br />
In a number of languages, <strong>the</strong> neuter gender was marked by an unstressed vowel at<br />
<strong>the</strong> end of a noun and of agreeing forms. The reduction of <strong>the</strong> final unstressed o to -ə<br />
in neuter nouns made <strong>the</strong>m undistinguishable from masculine nouns. The ‘death’ of<br />
<strong>the</strong> neuter gender (as Priestly 1983 puts it) in Sele Fara took a few decades. During<br />
his fieldwork in 1935, Isačenko (1939) documented a tendency towards a ‘masculinization’<br />
of neuter nouns. The neuter gender was completely gone by <strong>the</strong> time Priestly<br />
did his own fieldwork in 1979 and 1982. Most neuter nouns became masculine (both<br />
in form and in <strong>the</strong> agreement <strong>the</strong>y take), and just a few shifted to <strong>the</strong> feminine<br />
gender. 24<br />
Phonological erosion of nouns’ endings—where genders are marked—may put <strong>the</strong><br />
whole category in peril. East Indo-Aryan languages Bengali, Assamese, and Oriya lost<br />
all <strong>the</strong> traces of masculine, feminine, and neuter genders so prominent in Sanskrit,<br />
<strong>the</strong>ir oldest Indo-Aryan relative. In Masica’s (1991: 221) words, ‘far from being
6.3 Partial loss and reinterpretation of gender 85<br />
aberrant, this Eastern Old Indo-Aryan loss of gender can be seen as <strong>the</strong> most natural<br />
outcome of <strong>the</strong> phonological erosion’. The Old Iranian languages Avestan and Old<br />
Persian had three genders, just as did Sanskrit. The final syllables of nouns and<br />
agreeing modifiers were weakened and lost in Middle West Iranian languages—<br />
Parthian and Pahlavi—and in numerous modern Iranian languages, including Modern<br />
Persian, Baluchi, Tat, and Talysh; and <strong>the</strong> genders were lost. 25<br />
We now turn to <strong>the</strong> partial loss of Linguistic <strong>Gender</strong> in English, and its transformations<br />
throughout <strong>the</strong> history of <strong>the</strong> language.<br />
6.3 Partial loss and reinterpretation of gender: <strong>the</strong> story of English<br />
<strong>Gender</strong> in English—if compared to its direct ancestor, Old English (and also to<br />
Proto-Germanic)—is distinctly impoverished. Old English had three genders—<br />
masculine, feminine, and neuter. <strong>Gender</strong>s were marked on nouns <strong>the</strong>mselves, and<br />
on agreeing modifiers. Different agreement forms of <strong>the</strong> demonstrative ‘this’ feature<br />
in 6.10. They have all been replaced with one form for all in <strong>the</strong> modern language, as<br />
can be seen from <strong>the</strong> translation.<br />
6.10 se cyning<br />
this.MASC king.MASC<br />
‘this king’<br />
sēō cwēn<br />
this.FEM queen.FEM<br />
‘this queen’<br />
Þæt sćip<br />
this.NEUTER<br />
‘this ship’<br />
ship.NEUTER<br />
Most nouns denoting men in Old English were grammatically masculine (e.g.<br />
man, secg, wer ‘man’) and nouns denoting women were feminine (e.g. mægð ‘maid’,<br />
cwen ‘queen’). There were a number of notable exceptions for nouns referring to<br />
females—<strong>the</strong> masculine noun wifmann ‘woman’; and neuter nouns wif ‘woman’,<br />
mægden ‘maiden’. Anaphoric gender—marked in personal pronouns—was mostly<br />
based on <strong>the</strong> grammatical gender of a noun. So, wif ‘woman’, grammatically neuter,<br />
would be referred to with a neuter pronoun. Studies of Old English show that <strong>the</strong><br />
seeds of semantic agreement (similar to what we saw in §2.1.2 and 2.3 for German)<br />
were already visible as early as <strong>the</strong> eleventh century. As Curzan (2003: 45) puts it, in<br />
Old English documents:<br />
Within <strong>the</strong> noun phrase, grammatical gender agreement in <strong>the</strong> inflectional endings of <strong>the</strong> noun<br />
and its modifiers is obligatory (with occasional exceptions for unnaturally gendered animate
86 6 The rise and fall of Linguistic <strong>Gender</strong>s<br />
nouns such as masculine wifmann ‘woman’). Outside <strong>the</strong> noun phrase, however, grammatical<br />
gender agreement between anaphoric pronouns and <strong>the</strong>ir antecedents is often variable. In <strong>the</strong><br />
few cases in Old English where <strong>the</strong> grammatical gender of a noun referring to a human being<br />
conflicts with its gender (e.g. wif ‘woman’ [neuter]), <strong>the</strong> anaphoric pronoun almost always<br />
agrees with <strong>the</strong> referent’s sex. In o<strong>the</strong>r words, nouns such as wif and wifmann are referred back<br />
to with forms of she. 26<br />
In later documents, inanimate objects started being referred to with ‘it’ independently<br />
of <strong>the</strong>ir grammatical genders. By <strong>the</strong> end of <strong>the</strong> fourteenth century agreement<br />
gender and gendered forms of nominal declensions were all but lost, as part of<br />
general decay of case and agreement endings.<br />
The three anaphoric genders survived in personal pronouns (and also in reflexives,<br />
such as herself and himself ). Their choice shifted to being almost exclusively based<br />
on sex and animacy: nouns denoting male humans were referred to by he, nouns<br />
denoting female humans referred to by she, and nouns denoting non-humans<br />
referred to by it (irrespective of <strong>the</strong>ir sex). This is of course a rough approximation:<br />
additional uses of gendered pronouns (such as ‘she’ for ships, cities, and countries)<br />
make <strong>the</strong> system less straightforward.<br />
The history of gender loss and restructuring in English involved two processes:<br />
A. Loss of agreement gender and gendered declensions was due to <strong>the</strong> decay of<br />
case and agreement ending (<strong>the</strong> result of phonological attrition of word-final<br />
syllables). This process was complete by mid- to late fourteenth century.<br />
B. The anaphoric gender has never been lost. <strong>Gender</strong> agreement became almost<br />
straightforwardly semantic—from that based on <strong>the</strong> grammatical gender of a<br />
noun (independently of <strong>the</strong> referent’s sex or animacy) to that based on animacy<br />
and sex. The process started in Old English (before declensional distinctions<br />
and agreement were lost), and must have become complete by <strong>the</strong> late Middle<br />
English period. 27<br />
In all likelihood, <strong>the</strong> two processes took place simultaneously. The exact reason for<br />
<strong>the</strong> demise of agreement gender and gendered declensions in English (point A)<br />
remains a matter of some debate. As Curzan (2003: 48) puts it, ‘<strong>the</strong> facts of <strong>the</strong><br />
gender shift in English potentially argue for at least some external causes’, as stress<br />
shift ‘alone cannot account for why English lost its complex inflectional system and<br />
grammatical gender system’. 28<br />
The changing status of Social <strong>Gender</strong>s has affected <strong>the</strong> use of gendered pronouns<br />
in English, and especially of <strong>the</strong> ‘generic’ he; this is <strong>the</strong> topic of §11.5.1.<br />
A number of o<strong>the</strong>r languages followed <strong>the</strong> same path as English—losing <strong>the</strong>ir<br />
agreement gender, but maintaining Linguistic <strong>Gender</strong> in anaphoric pronouns. This is<br />
what happened in Cross-River languages, from Nigeria. In Yazgulami and Farizandi,<br />
two Iranian languages, masculine and feminine genders remain in personal pronouns<br />
only (Edelman 1966). A number of languages went <strong>the</strong> o<strong>the</strong>r way. In <strong>the</strong> Dardic
6.4 Linguistic <strong>Gender</strong> in language contact 87<br />
languages Garvi and Gavar, and in Punjabi, an Indo-Aryan language, <strong>the</strong> masculine<br />
and feminine survive in nouns, some adjectives and verbal forms, but not in personal<br />
pronouns. 29<br />
So far we have focused on language-internal reasons for changes in Linguistic<br />
<strong>Gender</strong>s. We now turn to <strong>the</strong> role of language contact in how genders are shaped.<br />
6.4 Linguistic <strong>Gender</strong> in language contact<br />
As speakers of different languages come in contact with each o<strong>the</strong>r, <strong>the</strong>ir languages<br />
become similar in various ways. Linguistic <strong>Gender</strong> is among <strong>the</strong> categories which<br />
tend to spread from one language to <strong>the</strong> next in a contact situation. If a language with<br />
Linguistic <strong>Gender</strong>s comes in contact with a gender-less language, chances are that<br />
Linguistic <strong>Gender</strong> is likely to fall into disuse. 30<br />
6.4.1 To lose a Linguistic <strong>Gender</strong><br />
Intensive language contact has resulted in <strong>the</strong> loss of Linguistic <strong>Gender</strong>s in a few<br />
well-documented instances. Literary Latvian (from <strong>the</strong> Baltic subgroup of Indo-<br />
European) has two genders, masculine and feminine. The Tamian dialect of Latvian<br />
used to be in close contact with Livonian, a Balto-Finnic language with no genders.<br />
The loss of grammatical gender in Tamian Latvian is <strong>the</strong> outcome of Livonian<br />
influence.<br />
Brahui is unusual for a Dravidian language in that it has no gender distinctions. In<br />
<strong>the</strong> last centuries, <strong>the</strong> Brahui people appear to have been in close contact with <strong>the</strong><br />
Balochi, speakers of a West Iranian language with no gender. Earlier on, <strong>the</strong> Brahui<br />
may have been in contact with Persians (whose language also has no gender). Once<br />
again, <strong>the</strong> complete loss of gender in Brahui results from language contact.<br />
Along similar lines, <strong>the</strong> Albanian dialect of Mandres in nor<strong>the</strong>rn Greece completely<br />
lost <strong>the</strong> system of two genders (feminine and masculine) found in all o<strong>the</strong>r<br />
Albanian dialects. The entire village moved to its present location from <strong>the</strong> adjacent<br />
areas of Turkey. In all likelihood <strong>the</strong> loss of gender is due to Turkish influence.<br />
Similarly, Tangale, a Chadic language, lost gender distinctions as a result of contact<br />
with unrelated Adamawa, with no Linguistic <strong>Gender</strong>. The loss of gender in Arawak<br />
languages Mawayana, Amuesha, and Chamicuro was due to <strong>the</strong> impact from <strong>the</strong><br />
surrounding and unrelated languages with no genders. Two genders—masculine and<br />
feminine—are a typical feature of Omotic languages of Ethiopia; Nor<strong>the</strong>rn Mao is <strong>the</strong><br />
only exception. Its gender was lost due to <strong>the</strong> influence from <strong>the</strong>ir immediate<br />
neighbours—speakers of unrelated Surmic languages. 31<br />
When languages are in contact, with people speaking several languages, <strong>the</strong>ir<br />
linguistic features adjust to each o<strong>the</strong>r. This is a way of reducing processing and<br />
cognitive overload. <strong>Gender</strong> plays a role in cognition and conceptualization of <strong>the</strong>
88 6 The rise and fall of Linguistic <strong>Gender</strong>s<br />
outside world. It is also a source of metaphors, and <strong>the</strong> basis of cultural stereotypes<br />
which determine human behaviour and relationship. What makes Linguistic <strong>Gender</strong><br />
so susceptible to contact is its role in cognition and communication.<br />
The loss of Linguistic <strong>Gender</strong>s may be conditioned by a number of factors—<br />
linguistic contact being just one of <strong>the</strong>m. Greek varieties, spoken in Turkey, were very<br />
much unlike Modern Greek in <strong>the</strong>ir grammar. As a result of contact with Turkish,<br />
<strong>the</strong>y lost many of <strong>the</strong> typical Greek features, and developed Turkish-like ways of<br />
saying things. Agreement gender and anaphoric gender (also absent from Turkish)<br />
were lost. But was Turkish influence <strong>the</strong> only culprit? In all likelihood, Turkish<br />
influence was an accelerating factor in <strong>the</strong> process of loss of a three-gender system,<br />
already started before contact with Turkish. 32<br />
6.4.2 Evolving a Linguistic <strong>Gender</strong><br />
A reverse situation may occur. Suppose that <strong>the</strong> majority of languages in a family do<br />
not have any grammatical gender distinction, and just one language does. If this<br />
language is in contact with, or located close to, an unrelated language with grammatical<br />
gender, it is fair to suggest that gender has probably been developed independently, as a<br />
result of contact-induced change. Tualatin was <strong>the</strong> only member of <strong>the</strong> Kalapuyan<br />
language family (formerly spoken in Western Oregon) to have two genders marked in<br />
demonstratives. One gender subsumed human females, <strong>the</strong> o<strong>the</strong>r everything else.<br />
Tualatin bordered Clackamas Chinook (a language with three genders), and in all<br />
likelihood <strong>the</strong> emergence of gender in Tualatin was due to language contact. 33<br />
In languages with no o<strong>the</strong>r linguistic means of expressing gender, derivational<br />
gender may come about through borrowed forms. A borrowed noun can be used in<br />
compounds, and <strong>the</strong>n reinterpreted as derivational gender.<br />
Finnish has very little in terms of linguistic gender. Its major marker of derivational<br />
gender is of a foreign origin. One of <strong>the</strong> oldest Germanic loanwords into<br />
Finnish is tytär ‘daughter’, from Germanic *tuttare ‘daughter’. The borrowed form<br />
appeared in many compounds and was later grammaticalized as a suffix -ttare. This<br />
suffix is now used with numerous terms (native and borrowed) referring to women,<br />
e.g. runo-tar ‘poetess’, from runo ‘poem’ (native word), kuninga-tar ‘queen’, from<br />
kuningas ‘king’ (a Germanic loan). This form was borrowed from Finnish into closely<br />
related Estonian by Johannes Aavik, <strong>the</strong> major proponent of <strong>the</strong> Estonian language<br />
reform in <strong>the</strong> early twentieth century (it is most used in occupational terms, e.g.<br />
laulja ‘singer’, laulja-tar ‘female singer’). 34<br />
Developing a gender marker does not have to involve grammaticalization. Suppose<br />
a language without Linguistic <strong>Gender</strong> comes in contact with ano<strong>the</strong>r language with<br />
productive masculine and feminine forms. It may <strong>the</strong>n borrow gendered words.<br />
<strong>Gender</strong> markers may get reanalysed as separate morphemes and extended to native<br />
words, echoing <strong>the</strong> reinterpretation in <strong>the</strong> history of Agreement <strong>Gender</strong> (see §6.1).
6.4 Linguistic <strong>Gender</strong> in language contact 89<br />
This is what happened in Indonesian, a language with no Linguistic <strong>Gender</strong>. Some<br />
loanwords with human reference from Sanskrit, an Indo-Aryan language with<br />
genders, come in pairs: forms denoting males end in -a and forms denoting females<br />
end in -i—for instance, putra ‘son’ and putri ‘daughter’, saudara ‘male sibling or<br />
relative’ and saudari ‘female sibling or relative’, dewa ‘god’ and dewi ‘goddess’. This<br />
principle has been extended to a few native words. Pemuda originally meant ‘young<br />
person’, but nowadays this refers to ‘young man’, and a new word, permud-i ‘young<br />
woman’, was created as its female counterpart.<br />
Masculine and feminine in Turkish can be expressed through a handful of<br />
derivational markers. Feminine gender markers -ic̹e and -c̹a are borrowings, from<br />
<strong>the</strong> Serbo-Croat feminine derivational gender marker -ica. They occur in a few Slavic<br />
borrowings into Turkish, all referring to women, e.g. kralic̹e ‘queen’ (masculine kral<br />
‘king’), c̹aric̹e ‘tsarina’ (masculine c̹ar ‘tsar’). The neologism tanric̹a ‘goddess’ was<br />
formed by analogy from <strong>the</strong> name of <strong>the</strong> old Turkish sky-god Tanrɪ. Just like in<br />
Indonesian, borrowed nouns have been reanalysed, and <strong>the</strong> loan suffix expanded to<br />
fur<strong>the</strong>r nouns.<br />
The derivational gender marker -nna in Estonian is currently used with loanwords<br />
and with native words. It penetrated <strong>the</strong> language from German loans, such as<br />
kuninganna ‘queen’ (from German Königin) with its masculine counterpart kuningas.<br />
As a result of reanalysis, it was expanded to native words, such as sõbranna<br />
‘female friend’ (sõber ‘male friend, friend in general’). 35<br />
Foreign markers of derivational gender are common in languages which have<br />
linguistic genders of o<strong>the</strong>r kinds. The feminine suffix -ess in English goes back to<br />
Greek -issa from where it was borrowed into Latin as -issa and French as -isse. Pairs<br />
such as host and host-ess and patron and patron-ess were borrowed into Middle<br />
English. From about <strong>the</strong> late fourteenth century onwards, <strong>the</strong> suffix expanded to<br />
native English nouns, creating forms like goddess, danceress, and dwelleress. The<br />
suffix is quite productive—we often hear hostess, waitress, governess, and stewardess.<br />
But some -ess nouns shifted <strong>the</strong>ir meanings: mistress is not just a female counterpart<br />
of a master. Many a female writer of poetry would object to being called a poetess,asa<br />
generic poet is devoid of ‘woman-like’ overtones and is thus more desirable in<br />
status—similarly to <strong>the</strong> treatment of ‘gender-specific’ terms for professions with<br />
<strong>the</strong> same suffix in o<strong>the</strong>r languages—§7.5.<br />
In Old English, a derivational gender marker could be added to nouns referring to<br />
a person who made (or was concerned with) <strong>the</strong> item referred to by <strong>the</strong> noun:<br />
6.11 BASE NOUN MALE PERFORMER FEMALE PERFORMER<br />
sang ‘song’ sang-ere ‘male singer’ sang-estre ‘female singer’<br />
sēam ‘seam’ sēam-ere ‘male tailor’ sēam-estre ‘seamstress, female dressmaker’<br />
The original masculine suffix -ere has developed into modern -er (now used with any<br />
gender reference). Suffix -estre gave rise to <strong>the</strong> modern -ster. This suffix lost its
90 6 The rise and fall of Linguistic <strong>Gender</strong>s<br />
exclusively feminine reference beginning from late Old English times. Once this<br />
happened, <strong>the</strong> suffix -ess—already established as a marker of derivational gender—<br />
could follow it. Seamster originally referred to a woman, <strong>the</strong>n to someone of ei<strong>the</strong>r<br />
sex, and <strong>the</strong>n <strong>the</strong> suffix -ess was added, to refer to a female maker of garments,<br />
creating seamstress. The form seamster dropped out of use, but <strong>the</strong> pair songster and<br />
songstress remains.<br />
Ano<strong>the</strong>r foreign feminine suffix borrowed from <strong>the</strong> Old French is -ette, first as a<br />
diminutive as in kitchenette, and <strong>the</strong>n, from <strong>the</strong> early twentieth century, as a<br />
feminine: usher-ette in a cinema, or undergradu-ette at a university. Many feminines<br />
in -ette have a negative feel about <strong>the</strong>m; some were created as pejoratives in <strong>the</strong> first<br />
place. The term suffragette was coined in 1906 by a British journalist writing for The<br />
Daily Mail who was sceptical of <strong>the</strong> activities of a suffragist—a man or a woman<br />
advocating equal rights to vote for both sexes. 36 We return to <strong>the</strong> overtones of<br />
feminine forms in §7.5, and §11.3.<br />
Borrowed gendered forms are not limited to derivation. Romanian, a Romance<br />
language, was influenced by Slavic languages. The vocative singular feminine ending<br />
-o was borrowed from Bulgarian into Romanian. Examples include soră ‘sister’,<br />
vocative soro!, and mamă ‘mum’, vocative mamo (Puşcariu 1943: 435).<br />
6.4.3 Readjusting Linguistic <strong>Gender</strong><br />
If languages are in contact, <strong>the</strong>ir genders may become similar. A prime example<br />
comes from <strong>the</strong> language contact situation described by Gumperz and Wilson (1971:<br />
156) in <strong>the</strong> Indian village of Kupwar. The languages spoken in Kupwar are Kannada,<br />
a Dravidian language, and Marathi and Hindi-Urdu, two Indo-Aryan languages. The<br />
Kupwar Kannada gender system is identical to that of standard Kannada. There are<br />
three genders whose assignment is semantically based: nouns referring to males<br />
belong to <strong>the</strong> masculine gender, nouns referring to females belong to <strong>the</strong> feminine<br />
gender, and all o<strong>the</strong>r animate and inanimate nouns are neuter. The standard Marathi<br />
has three genders (masculine, feminine, and neuter). Inanimates can belong to any<br />
gender. Not so in <strong>the</strong> Kupwar Marathi: here non-human nouns which correspond to<br />
Marathi masculines or feminines are neuter, replicating <strong>the</strong> Kannada pattern. Only<br />
human males are masculine, and human females are feminine. The gender choice has<br />
been restructured and adjusted to <strong>the</strong> dominant language—<strong>the</strong> Kupwar Kannada.<br />
The noun nadī ‘river’ belongs to <strong>the</strong> feminine gender in standard Marathi; in Kupwar<br />
Marathi it is treated as neuter. 37<br />
Adjacent languages in <strong>the</strong> Australian area typically have similar gender systems. In<br />
<strong>the</strong> Mayali dialect chain, <strong>the</strong> western-most dialects are becoming more like <strong>the</strong>ir<br />
neighbour Jawoyn, with only three, instead of four agreement classes, and <strong>the</strong><br />
eastern-most dialects are tending to become more like <strong>the</strong>ir neighbours Rembarrnga<br />
and Dangbon, both of which lack noun class agreement (Dixon 2002: 500). The
6.4 Linguistic <strong>Gender</strong> in language contact 91<br />
Australian language Yanyuwa developed a five-term gender system marked on nouns<br />
and on verbs (human masculine, non-human masculine, feminine, vegetable, and<br />
trees). The terms in <strong>the</strong> system come from different sources. The system itself is<br />
remarkably similar to Yanyuwa’s westerly neighbours, which had such a system<br />
(Dixon 2002: 500–1).<br />
<strong>Gender</strong> can be particularly stable in <strong>the</strong> history of a language if <strong>the</strong> speech<br />
community is aware of it as a particularly salient feature. <strong>Gender</strong>—viewed as dividing<br />
all <strong>the</strong> referents into ‘masculine-like’ and ‘feminine-like’—is what speakers of two<br />
languages from <strong>the</strong> Sepik area in New Guinea—Manambu (Ndu family) and<br />
Kwoma—pinpoint as a striking feature of <strong>the</strong>ir language. <strong>Gender</strong> is an emblematic<br />
feature in Nilotic languages Luwo and Labwor—which has helped maintain it in<br />
language contact. 38<br />
6.4.4 Linguistic <strong>Gender</strong> in language obsolescence<br />
As a minority language becomes obsolescent and loses ground to a dominant<br />
language, its Linguistic <strong>Gender</strong> is likely to be affected. Dyirbal, a vibrant language<br />
in <strong>the</strong> 1960s, is now declining. Traditional Dyirbal had four agreement genders based<br />
on a complex network of semantic features, as we can recall from §2.2.1. Young<br />
People’s Dyirbal has gradually adjusted its gender system towards that of English and<br />
made <strong>the</strong> gender choice more like that for English anaphoric gender. Dyirbal’s<br />
‘typical’ Australian gender covering edible non-flesh food simply disappeared: <strong>the</strong><br />
speakers of Young People’s Dyirbal simply stopped using it. The scope of ‘feminine’<br />
gender 2 became reduced and became reserved only for females (it used to include<br />
drinkable liquids, fire, and things associated with fighting). <strong>Gender</strong> assignment by<br />
mythic association was lost; exceptions became regularized; and <strong>the</strong> use of <strong>the</strong> residue<br />
‘neuter’ gender expanded—subsuming all <strong>the</strong> inanimates, covered by ‘it’ in English. 39<br />
Scottish Gaelic is ano<strong>the</strong>r example of a language in decline. Gaelic has masculine<br />
and feminine agreement genders and also anaphoric genders. Agreement genders are<br />
marked in very specific ways—by initial mutation of <strong>the</strong> noun after <strong>the</strong> definite<br />
article, by agreement of an adjective (again using initial mutation), by <strong>the</strong> form of<br />
<strong>the</strong> definite article. There is also a gender-marked diminutive suffix. Younger and<br />
not-so-fluent speakers tend to overuse <strong>the</strong> masculine pronoun a for all inanimate<br />
nouns. Adjective agreement via consonantal lenition shows variation. Dorian (1981:<br />
148) predicts that if Eastern Su<strong>the</strong>rland Gaelic had ‘future generations of fluent<br />
speakers’, gender marking in pronouns and agreement on adjectives ‘would surely<br />
disappear’, but head-marking gender on nouns would survive, ‘producing a kind of<br />
lexicalisation of gender specification linked to specific nouns’. 40<br />
If speakers are shifting to a language with no linguistic gender, <strong>the</strong>y may well start<br />
losing <strong>the</strong>ir own. The last speakers of Baré, a North Arawak language from Venezuela<br />
and adjacent regions of Brazil, were fluent speakers of Língua Geral, a Tupí-Guaraní
92 6 The rise and fall of Linguistic <strong>Gender</strong>s<br />
lingua franca. Língua Geral has no gender distinctions. As a consequence, masculine<br />
and feminine agreement on adjectives, and even anaphoric gender, in Baré became<br />
optional. The last speaker of Baré would often use ‘he’ to refer to a woman.<br />
The demise of Linguistic <strong>Gender</strong> in language attrition is not universal. In Arvanitika,<br />
an endangered Albanian variety spoken by Albanians who emigrated to Greece<br />
in <strong>the</strong> eleventh and fifteenth centuries, <strong>the</strong> three genders remain distinct. This may be<br />
due to <strong>the</strong> fact that <strong>the</strong> three-way gender distinctions in Arvanitika and <strong>the</strong> dominant<br />
Greek are structurally similar. 41<br />
6.5 Linguistic <strong>Gender</strong> in language acquisition and language dissolution<br />
When children learn <strong>the</strong>ir first language, Linguistic <strong>Gender</strong> is among <strong>the</strong> first features<br />
<strong>the</strong>y master. By <strong>the</strong> age of 3, most Hebrew-speaking children have gained full<br />
command of agreement gender. Similar results have been obtained for speakers of<br />
German, Polish, and Sesotho, a Bantu language. 42<br />
Complex marking of Linguistic <strong>Gender</strong> may slightly delay its full acquisition.<br />
Polish children acquire three genders earlier than Russian children because Linguistic<br />
<strong>Gender</strong> in Russian may be confusing: in <strong>the</strong> spoken language, <strong>the</strong> typical feminine<br />
ending -a sounds very similar to <strong>the</strong> typical neuter ending -o since both are reduced<br />
if <strong>the</strong>y are not stressed. This is not <strong>the</strong> case in Polish. Russian also has a number of<br />
diminutive and endearing forms referring to boys and men, but ending in -a, e.g.<br />
Kolja ‘diminutive for Nicholas’, orVitja ‘diminutive for Victor’. And quite a few<br />
common nouns refer to men, but end in -a (and decline as feminines), including<br />
djadja ‘uncle’. Young children would say djadja side-l-a (uncle sit-PAST-feminine.<br />
singular) instead of djadja side-l (uncle sit-PAST+masculine.singular) ‘uncle was<br />
sitting’. These problems disappear by <strong>the</strong> age of about 3.<br />
In addition to this, diminutives of female names can take masculine endings<br />
‐ok/-ik, asinSveta ‘diminutive of Svetlana’, Svetik ‘diminutive of Sveta’. And <strong>the</strong>re<br />
are nouns which can refer to men and to women, with a masculine form, such as vrac᷈<br />
‘doctor’. A recent study by Rodina (2014) showed that even when caregivers preferred<br />
feminine forms, 2- and 3-year-olds tend to use masculine forms for <strong>the</strong>se<br />
cases—being led by <strong>the</strong> form of <strong>the</strong> noun ra<strong>the</strong>r than by <strong>the</strong> meaning. This overgeneralization<br />
appears to have continued until about <strong>the</strong> age of 7. That is, <strong>the</strong><br />
acquisition of complex agreement rules can be delayed by <strong>the</strong>ir complexity and a<br />
less than straightforward input—but only slightly. 43<br />
<strong>Gender</strong> in German is marked in many different contexts—personal pronouns,<br />
adjectives, articles, and so on. This extensive system of expressing grammatical<br />
gender gives <strong>the</strong> German child ample opportunities to practise it, seek regularities<br />
in <strong>the</strong> system, and produce <strong>the</strong> forms accurately at an early age. German children<br />
have an advantage over English children who are exposed just to <strong>the</strong> anaphoric<br />
gender in personal pronouns. And indeed, a study conducted by Mills (1986: 103–6)
6.6 Linguistic <strong>Gender</strong> and language reforms 93<br />
showed that German children ‘are clearly in advance in <strong>the</strong>ir production of <strong>the</strong><br />
masculine and feminine pronouns’, compared with <strong>the</strong> English children within <strong>the</strong><br />
same age bracket. The German children make hardly any mistakes by <strong>the</strong> age of 3,<br />
while English children do until <strong>the</strong> age of 4.<br />
Learners of a foreign language project <strong>the</strong> Linguistic <strong>Gender</strong> of <strong>the</strong> language <strong>the</strong>y<br />
know best onto <strong>the</strong> system <strong>the</strong>y are trying to master. Speakers of Russian learning<br />
Hebrew tend to transfer <strong>the</strong> gender of <strong>the</strong>ir native language to Hebrew. Speakers of<br />
English (which only has anaphoric gender) reply on <strong>the</strong> meaning of nouns, and<br />
largely disregard <strong>the</strong> form of nouns (Armon-Lotem and Amiram 2014).<br />
The human brain can be damaged—as a result of a stroke, or an accident.<br />
Language skills may <strong>the</strong>n be impaired or lost. Those affected by it can sometimes<br />
produce non-grammatical sentences, or have difficulties understanding <strong>the</strong>ir first<br />
language. Studies which looked at <strong>the</strong> behaviour of gender agreement forms in<br />
German, Italian, and Polish found that sufferers of ‘agrammatical’ aphasia often<br />
use <strong>the</strong> wrong gender form—typically, masculine for feminine. But no matter how<br />
serious <strong>the</strong> condition, Linguistic <strong>Gender</strong>s persist in some form even in those patients<br />
who are most severely affected. 44<br />
6.6 Linguistic <strong>Gender</strong> and language reforms<br />
Speakers of languages with Linguistic <strong>Gender</strong> are typically aware of it as a salient<br />
feature, and are often prepared to discuss it (see §4.3 and §7.7). Speakers of Manambu,<br />
a Papuan language of New Guinea, are ‘gender-proud’: <strong>the</strong> two genders<br />
(masculine and feminine) are pinpointed as a salient feature of <strong>the</strong> language. We<br />
recall from §6.4.2 that Johannes Aavik borrowed a derivational marker for feminine<br />
gender into Estonian, to fill what he felt to be a gap.<br />
Linguistic <strong>Gender</strong> has been <strong>the</strong> target of language reform in <strong>the</strong> history of a number<br />
of European languages. Norwegian is a case in point. Until <strong>the</strong> early twentieth<br />
century, literary Norwegian had a system of two Linguistic <strong>Gender</strong>s: a ‘common’<br />
gender and a neuter gender, as in Danish and in Swedish, ra<strong>the</strong>r than <strong>the</strong> original Old<br />
Norse three genders (masculine, feminine, and neuter). The feminine gender was<br />
restored with <strong>the</strong> development of Nynorsk, a language largely created by Ivar<br />
Aasen, in <strong>the</strong> context of <strong>the</strong> rise of Norwegian nationalism. The reform drew on<br />
south-west dialects of Norwegian, where <strong>the</strong> feminine gender remained in use. The<br />
1938 language reform made <strong>the</strong> feminine form obligatory for about 1,000 nouns. As<br />
Gregersen (1979: 11) remarks, ‘nationalism ra<strong>the</strong>r than feminism was <strong>the</strong> motivation<br />
for <strong>the</strong> change: <strong>the</strong> feminine gender was felt to be a particularly Norwegian trait’. At<br />
present Nynorsk is one of <strong>the</strong> two official languages of Norway. 45<br />
The forms and <strong>the</strong> uses of Linguistic <strong>Gender</strong>s for humans may change as <strong>the</strong> status<br />
of Social <strong>Gender</strong>s evolves. The growing equality of women in many Western societies<br />
has brought about an increase in derivational forms with female reference. Language
94 6 The rise and fall of Linguistic <strong>Gender</strong>s<br />
reforms have targeted ‘sexist’ language—ensuring <strong>the</strong> visibility of women, and<br />
avoiding <strong>the</strong> generic use of masculine pronouns and terms for ‘man’. Social motivations<br />
for change in Linguistic <strong>Gender</strong>s highlight its functionality, and overall<br />
importance. We turn to this in §§11.4–5.<br />
6.7 Linguistic <strong>Gender</strong>s, <strong>the</strong>ir development, demise, and transformations<br />
Linguistic genders can come from a number of sources. A language can gain an<br />
anaphoric gender by making a gender affix out of a noun meaning ‘person’ or ‘man’.<br />
This can <strong>the</strong>n be expanded into derivational gender, or into agreement gender.<br />
A gender-sensitive pronoun can become a marker of gender agreement. Noun classifiers<br />
can give rise to genders. Nominal categories of number and case are intrinsically<br />
prone to reflecting meanings to do with animacy and humanness.<br />
The existing genders can change <strong>the</strong>ir meanings. <strong>Gender</strong>s can be lost, and nouns<br />
regrouped based on <strong>the</strong>ir phonological shape or meaning.<br />
Agreement gender and gender-based declensions were rife in Old English. By <strong>the</strong><br />
end of <strong>the</strong> Middle English period, <strong>the</strong>y were as good as gone. The anaphoric gender<br />
survived in personal pronouns. But its choice gradually came to be based on <strong>the</strong><br />
animacy and sex of <strong>the</strong> noun—ra<strong>the</strong>r than its agreement gender, now lost. The seeds<br />
of this change were present in Old English: a neuter noun wif, <strong>the</strong> precursor of<br />
modern-day wife, could be referred to as ‘it’ or as ‘she’. The process of making gender<br />
in English semantically transparent appears now to be complete. But its simplicity is<br />
illusory: Linguistic <strong>Gender</strong>—especially applied to human beings—is a target for language<br />
reform, and a barometer of social status, and equality, for women (we return to<br />
<strong>the</strong>se issues in §§11.5.1–2).<br />
If languages are in contact with each o<strong>the</strong>r, <strong>the</strong>ir genders are likely to be affected.<br />
They can be gained, or lost. Or <strong>the</strong>ir meanings can readjust to each o<strong>the</strong>r. Languages<br />
with no grammatical genders often borrow gendered forms. These are reanalysed,<br />
and a derivational gender emerges. This is a means which many languages exploit,<br />
expanding derivational genders <strong>the</strong>y may already have.<br />
If a language becomes obsolescent and gradually falls into disuse, it will adjust to<br />
<strong>the</strong> language which is taking control. If <strong>the</strong> dominant language has no grammatical<br />
gender, <strong>the</strong> obsolescent language will lose it. Or its genders can get restructured, to fit<br />
<strong>the</strong> mould of <strong>the</strong> dominant system. <strong>Gender</strong> is acquired early and stays on in many<br />
forms of language dissolution in aphasia.<br />
Linguistic <strong>Gender</strong> can be a target of language reform, especially inasmuch as it<br />
concerns <strong>the</strong> perceived asymmetries of Social <strong>Gender</strong>s reflected through linguistic<br />
expression of genders. Social motivations for changes in Linguistic <strong>Gender</strong>s point<br />
towards speakers’ awareness of <strong>the</strong>m, and <strong>the</strong>ir functionality and importance in <strong>the</strong><br />
social life of many peoples.
6.7 Linguistic <strong>Gender</strong>s, <strong>the</strong>ir development, demise 95<br />
<strong>Gender</strong>—once it exists in a language—is an inextricable part of <strong>the</strong> linguistic<br />
machinery, <strong>the</strong> way <strong>the</strong> language works. The mastery of genders offers speakers an<br />
array of expressive devices, making <strong>the</strong> language more colourful and versatile. This is<br />
what we turn to next.<br />
NOTES AND SOURCES<br />
1. See Heine and Kuteva (2002), Heine and Reh (1984), and Trask (2000) on grammaticalization.<br />
Grammaticalization is often accompanied by a reanalysis of forms and by <strong>the</strong>ir<br />
reinterpretation (when <strong>the</strong>y acquire new and different meanings).<br />
2. See Palmer (2009: 69), and van den Berg (2015), for fur<strong>the</strong>r examples.<br />
3. In a few o<strong>the</strong>r Eastern Nilotic languages, <strong>the</strong>se markers of anaphoric gender develop into<br />
agreement markers (step 2 in Pathway I); see Heine and Vossen (1983); cf. Heine and Reh<br />
(1984: 219–20, 230–2), on Nilotic languages.<br />
4. Heine (1982b: 214), Claudi (1985).<br />
5. See Martin (1992: 180, 768), Lee (2014: 42) on Korean; see also Aikhenvald (2000: 442–6).<br />
6. The demonstratives used as anaphoric gender pronouns are mtona ‘he’ and minana ‘she’,<br />
and contain <strong>the</strong> classifier forms (see Senft 1986, 1996: 18–22, 353 and van den Berg 2015).<br />
7. See Zavala (1992: 152–3, 2000: 134–5) on Akatek; England (1983: 158) on Mam; de León<br />
(1987: 148, 157–8), and also Craig (1986a: 266–7) on classifiers in Jacaltec. Number 7<br />
stands for a glottal stop.<br />
8. The derivational markers are not used in agreement. See Frajzyngier (1993: 49) on Mupun;<br />
similar examples have been described for !Xoo, <strong>the</strong> nor<strong>the</strong>rn dialect of !Xun, a Khoisan<br />
language, and Hunde, a Bantu language: Heine and Kuteva (2002: 133, 209, 213, 315). See<br />
Braun (2000a: 53) on Turkish.<br />
9. See Reid (1997: 215–18); a somewhat similar example is in Green (1997). A general<br />
historical analysis is in Dixon (2002: 467ff.).<br />
10. A similar scenario could apply to Bandjalang (Dixon 2002: 464–6). Details on <strong>the</strong>se<br />
scenarios can be found in Dixon (1982: 171; 2002: 464–6, 497–506), Reid (1997); a brief<br />
summary is in Aikhenvald (2000: 394–400).<br />
11. See Dixon (2002: 499), and Heine and Reh (1984: 232–4), on <strong>the</strong> mechanisms of development<br />
of gender agreement based on anaphoric gender.<br />
12. See, for instance, Clackson (2007: 91–113), Luraghi (2011). Note that three genders in<br />
Latin quis ‘who (masculine)?’, quae ‘who (feminine)?’, and quid ‘who (neuter)?’ introduced<br />
by analogy with o<strong>the</strong>r forms: Clackson (2007: 105).<br />
13. A similar scenario is sketched out by Meillet (1931: 17–20), a major classic in Indo-<br />
European scholarship. He suggested that <strong>the</strong> new gender-marking suffix-h 2 was extended<br />
to <strong>the</strong> stem of <strong>the</strong> animate demonstratives, creating a subcategorization between masculine<br />
and feminine nouns. Fur<strong>the</strong>r discussion and suggestions are in Luraghi (2009, 2011).<br />
A somewhat different scenario and explanation are in Brosman (1979) and Melchert<br />
(1994); see also Fodor (1959: 213).<br />
14. Axenov (2006: 62, 69). Sex differences of animates can be expressed with nouns nar ‘male’<br />
and maādag ‘female’.<br />
15. The scenario is outlined by Janda (1999: 80–2); see also Klenin (1983).<br />
16. The development of genders in Iroquoian languages has been linked to prefixes marking<br />
‘agent’ and ‘patient’. The non-specific agent marker in Proto-Nor<strong>the</strong>rn-Iroquioan came to
96 6 The rise and fall of Linguistic <strong>Gender</strong>s<br />
be used just for humans, possibly, based on <strong>the</strong> frequency of its use. Its specific counterpart<br />
developed into a non-human marker. A fur<strong>the</strong>r non-human agent prefix has its origins in<br />
reinterpretation of a number marker (*wa): see Chafe (1977: 515–16), and Mithun (2014).<br />
17. Huntley (1980) contains a snap-shot of <strong>the</strong> history of <strong>the</strong> ‘animate’ and ‘inanimate’<br />
declensions in Slavic languages; see also Luraghi (2014: 223–4), and Priestly (1983:<br />
357–8). Some semantically inanimate nouns which follow <strong>the</strong> animate conjugation in<br />
Polish, and a fur<strong>the</strong>r division between masculine ‘person’ and masculine ‘animal’ are<br />
discussed by Wertz (1977: 50, 53, 57–9). More examples of gender variation in Contemporary<br />
Polish are in Kryk-Kastovsky (2000). Tocharian A, an extinct Indo-European<br />
language, is <strong>the</strong> only language in <strong>the</strong> family with feminine and masculine forms of first<br />
person full pronouns: näs̍‘I (male speaker)’ and ñuk ‘I (female speaker)’. Jasanoff (1989)<br />
demonstrated that <strong>the</strong> masculine–feminine gender distinctions developed through reinterpretation<br />
of <strong>the</strong> old accusative and nominative case forms.<br />
18. Garibian (1976), Priestly (1983: 357); Clackson (2007).<br />
19. See Onishi (1996); and Priestly (1983: 357); animate and inanimate pronouns are also<br />
distinguished in Assamese in Oriya (fur<strong>the</strong>r Indo-Aryan languages which lost Indo-<br />
European genders).<br />
20. See Krishnamurti (2003: 211).<br />
21. Newman (2000: 208).<br />
22. In a formal context ti can also refer to a young female. See Miranda (1975: 209–11). In all<br />
likelihood, <strong>the</strong> change in <strong>the</strong> use of gendered forms was triggered by change in meaning of<br />
a neuter noun c᷈ed̩ũ from ‘child’ to ‘girl’. See also Masica (1991: 220–1) on how <strong>the</strong> threegender<br />
system in Sinhalese was restructured under Tamil influence, as basically animate<br />
(with a subdivision into masculine and feminine) and inanimate.<br />
23. See also Posner (1966: 132–6), Pope (1934: 304), Matasović (2004: 49–52) on gender in<br />
Romance languages. See discussion of <strong>the</strong> neuter gender in Romanian in Corbett (1991:<br />
150–2), and Loporcaro, Faraoni, and Gardani (2014) for <strong>the</strong> discussion of neuter in Italian<br />
dialects.<br />
24. Masculine and feminine genders have fallen toge<strong>the</strong>r into a ‘common’ gender in a number<br />
of Germanic languages—including Dutch, Frisian, and standard varieties of Swedish,<br />
Norwegian, and Danish. The common gender now contrasts with neuter gender, for<br />
reasons which are largely open to discussion. Harbert (2007: 93–103) offers a comprehensive<br />
discussion of <strong>the</strong> history of gender in Germanic; a summary is in Matasović (2004:<br />
59–61).<br />
25. See Matasović (2004: 41–6), on <strong>the</strong> redistribution and loss of gender in individual Indo-<br />
Aryan and Iranian languages. A phonological explanation for gender loss and restructuring<br />
is echoed by <strong>the</strong> discussion by Ibrahim (1973: 86–7).<br />
26. As Quirk and Wrenn (1957: 75) put it, ‘strict concord in grammatical gender is <strong>the</strong> rule in<br />
O(ld) E(nglish). It is particularly regular with demonstratives and modifiers…But pronouns,<br />
particularly when relating to human beings, are sometimes used with natural<br />
gender in spite of <strong>the</strong> normal requirements of grammatical concord: Þæd maēden ˜ hēō<br />
wearð “<strong>the</strong> maiden ˜ (n) she was”.’<br />
27. Many scholars refer to A as ‘<strong>the</strong> loss of grammatical gender’, and to B as ‘<strong>the</strong> spread of<br />
natural gender’ (e.g. Baron 1971). This terminology is confusing: it implies that anaphoric<br />
gender is not grammatical, and does not account for ‘unnatural’, that is, non-sex-based,<br />
uses of personal pronouns in English.<br />
28. A comprehensive, source-based investigation of <strong>the</strong> use of anaphoric gendered pronouns<br />
and <strong>the</strong> gradual shift from grammatical to semantic basis in anaphoric gender is in Curzan
6.7 Linguistic <strong>Gender</strong>s, <strong>the</strong>ir development, demise 97<br />
(2003: 58–132); see also Baron (1971) and Karatsareas (2009: 205–6, 2013), and a summary<br />
in Corbett (1991: 101–2). The demise of grammatical agreement gender and <strong>the</strong> loss of<br />
inflectional endings is discussed in Curzan (2003: 29–57); see also Kastovsky (2000) and<br />
Harbert (2007: 93–101), Matasović (2004: 58–9). Some scholars suggested influence from<br />
Old Norse, or Norman French as a likely trigger; Curzan (2003: 47–54) addresses <strong>the</strong><br />
possible external causes in <strong>the</strong> demise of gender agreement in English (a fur<strong>the</strong>r <strong>the</strong>ory of<br />
‘creolization’ in <strong>the</strong> Middle English period is dismissed by her as ‘glib’).<br />
29. See Demuth, Faraclas, and Marchese (1986: 459) and Heine (1982b: 196) on Cross-River<br />
and Kru languages, Edelman (1966) on Yazgulami and Farizandi; fur<strong>the</strong>r examples are<br />
in Priestly (1983: 346). See Bhatia (1993: 216–17, 222–9) on Punjabi, Edelman (1999) on<br />
Garvi, Grünberg (1999) on Gawar. Fur<strong>the</strong>r generalizations, and examples, on loss of<br />
genders and gender agreement are in Aikhenvald (2000: 398–9). Dardic and Punjabi<br />
provide exceptions to Greenberg’s universal (1963: universal 43), stating that ‘if a language<br />
has gender categories in <strong>the</strong> noun, it has gender categories in <strong>the</strong> pronoun’.<br />
30. See Aikhenvald (2006) and Storch (2006) on spread of Linguistic <strong>Gender</strong> in situations of<br />
language contact.<br />
31. Mat<strong>the</strong>ws (1956: 316). See Emeneau (1980: 157–9) on Brahui; Hamp (1965: 146) on <strong>the</strong><br />
Albanian of Mantres; Jungraithmayr (1995: 200–1) on Tangale, Adelaar (2006) and<br />
Aikhenvald (2012a) on Mawayana, Amuesha, and Chamicuro, and Ahland (2012: 41)<br />
and Amha (forthcoming) on Nor<strong>the</strong>rn Mao. Haugen (1976: §11.4.6) describes dialects of<br />
Swedish which apparently lost genders under <strong>the</strong> influence of Finnish.<br />
32. Karatsareas (2009: 207); see Janse (2002: 366), Winford (2005: 405), and Dawkins (1916).<br />
33. See Mithun (1999: 432). The emergence of Linguistic <strong>Gender</strong>s in Austronesian languages<br />
of Alor-Pantar area in Indonesia and in a couple of Oceanic languages in West New<br />
Britain are attributed to language contact (Schapper 2010a, 2010b, van den Berg 2015: 39).<br />
34. See Hasselblatt (2015: 131–3) for more examples and <strong>the</strong> etymology.<br />
35. See Abondolo (1998: 178–9) on Finnish, Tadmor (2007: 311–13), Sneddon (1996) on<br />
Indonesian, Braun (1997a, 1997b), Lewis (2000: 23) on Turkish. Wiedemann (1875: 196)<br />
offers fur<strong>the</strong>r examples and an explanation of <strong>the</strong> historical path of <strong>the</strong> development of this<br />
borrowed feminine affix in Estonian.<br />
36. See Dixon (2015: 320–4), for a detailed analysis, and also Baron (1986: 120–1). The dates<br />
for <strong>the</strong> words quoted here are based on <strong>the</strong> Oxford English Dictionary Online.<br />
37. See Gumperz and Wilson (1971) for fur<strong>the</strong>r details; Pandharipande (1997: 368–9) on<br />
standard Marathi.<br />
38. See Storch (2006) on Nilotic languages.<br />
39. Details on Young People’s Dyirbal are in Schmidt (1985); see also Dixon (2015).<br />
40. See Dorian (1981: 124–9, 147–8).<br />
41. See Trudgill (1977: 35); Sasse (1985). As we recall from §3.2, gender choice in Traditional<br />
Tiwi was partly determined by <strong>the</strong> size of objects. Young people and children—who speak<br />
a simplified and partly Anglicized ‘Modern Tiwi’—tend to assign gender on <strong>the</strong> basis of<br />
sex, and are inconsistent in gender assignment to nouns o<strong>the</strong>r than those denoting<br />
humans and dogs (Lee 1987: 84).<br />
42. See Berman (1981, 1985: 299–301), Levy (1983a, 1983b), Demuth (1988); similar results<br />
were discussed by Popova (1973) for Russian, French, and German (Karmiloff-Smith<br />
1979).<br />
43. See Popova (1973), Rodina (2014) on <strong>the</strong> child language acquisition of Russian. Pérez-<br />
Pereira (1991: 585–8) notes delayed acquisition of genders whose marking is ‘ambiguous,<br />
barely transparent and scarcely predictable’, as is <strong>the</strong> case with three genders in German,
98 6 The rise and fall of Linguistic <strong>Gender</strong>s<br />
Czech, Serbo-Croatian, and Russian. This is in contrast to languages like Hebrew, Egyptian<br />
Arabic, Spanish, or French, with clearly differentiated and regular marking of just two<br />
genders; see also Aikhenvald (2000: 414–16).<br />
44. See Menn, O’Connor, and Holland (1995: 87, 113); Perlak and Jarema (2003) on Polish;<br />
Herbert (1991) on <strong>the</strong> maintenance of gender differentiation in Bantu-speaking aphasic<br />
patients.<br />
45. See §6.6 on <strong>the</strong> label ‘common gender’. See Gregersen (1979: 11) for Norwegian and<br />
similar changes in Danish and Swedish, and <strong>the</strong> introduction of feminine pronouns in<br />
written Chinese. See Hagège (2004: 109–10) on Norwegian and o<strong>the</strong>r Germanic languages.
7<br />
Manly women and womanly men:<br />
<strong>the</strong> effects of gender reversal<br />
Using <strong>the</strong> opposite Linguistic <strong>Gender</strong>—feminine for a man, and masculine for a<br />
woman—can be downright wrong. The main reason for my visit to <strong>the</strong> remote<br />
Swakap in <strong>the</strong> Sepik area of New Guinea, <strong>the</strong> only village where <strong>the</strong> Gala language<br />
is spoken, was to find out whe<strong>the</strong>r <strong>the</strong>y really do have different forms for ‘me:<br />
woman’ and ‘me:man’. I travelled with a group of Manambu speakers, and we<br />
spoke Manambu. The two languages—Manambu and Gala—are related (as close as<br />
English and German). The first thing I had to face was <strong>the</strong> pronouns. Manambu<br />
uses <strong>the</strong> form wun for ‘me’, no matter what <strong>the</strong> sex of <strong>the</strong> speaker. ‘You’ distinguishes<br />
gender of <strong>the</strong> addressee—mən ‘you man’ and ñən ‘you woman’. Personal<br />
pronouns in <strong>the</strong> two languages are similar in form, but not in meaning—Table 7.1<br />
summarizes this.<br />
The form meaning ‘you feminine’ in Manambu is used as ‘me feminine’ in Gala.<br />
My attempts at speaking Gala were met with some derision. As a language learner<br />
with some Manambu language background, I found it difficult to use <strong>the</strong> Gala form<br />
ñən to refer to myself (I, a woman, speaking): it sounded to me as if I was using<br />
‘you feminine’ for myself. The Gala speakers laughed at me when I referred to<br />
myself as wun—which means ‘me’ in Manambu, but just ‘me masculine’ in Gala—<br />
TABLE 7.1. Personal pronouns in Gala 1 compared with Manambu<br />
Gala Manambu<br />
1 person singular masculine<br />
wun<br />
1 person singular feminine ñәn<br />
wun<br />
2 person singular masculine mәn mәn<br />
2 person singular feminine yin ñәn<br />
3 person singular masculine<br />
kәl, kәr dә<br />
3 person singular feminine ki<br />
lә<br />
<strong>How</strong> <strong>Gender</strong> <strong>Shapes</strong> <strong>the</strong> <strong>World</strong>. First edition. Alexandra Y. Aikhenvald.<br />
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.
100 7 Manly women and womanly men<br />
saying that I was trying to pretend I was a man. ‘<strong>Gender</strong> reversal’ was out of <strong>the</strong><br />
question.<br />
Not so for female speakers of Palestinian Arabic in Tire (Israel): a woman<br />
may refer to herself with a masculine form if she is talking about being tired, sad,<br />
unhappy, or nervous, as in ʔana mažru:ħ min illi Sa:r (I hurt.masc.sg from that.<br />
which happened) ‘I am hurt by what happened’. A normal feminine form—expected<br />
to be used by a female speaker—is mažru:ħa (hurt.fem.sg). Switching Linguistic<br />
<strong>Gender</strong>s in self-reference is a mark of something unpleasant. No changes <strong>the</strong> o<strong>the</strong>r<br />
way round have been observed: men always refer to <strong>the</strong>mselves with masculine<br />
Linguistic <strong>Gender</strong>. 2<br />
<strong>Gender</strong> reversal can serve a different purpose—that of protection, and deception.<br />
Pankhurst (1992: 169) describes how a family of Amharic speakers use pronouns<br />
to address <strong>the</strong>ir little daughter—sometimes with <strong>the</strong> masculine pronoun, and<br />
sometimes <strong>the</strong> feminine. In Pankhurst’s words,<br />
when applied to children, this blurring of gender identity is used as a strategy to defeat <strong>the</strong> evil<br />
eye, and o<strong>the</strong>r malignant forces conspiring against <strong>the</strong> life of a child. By referring to <strong>the</strong> boy as<br />
‘she’ or <strong>the</strong> girl as ‘he’, people believe <strong>the</strong>y can trick <strong>the</strong> ‘death-wishers’ and increase <strong>the</strong> likelihood<br />
of survival. 3<br />
We can recall, from §2.3, that in a system of Linguistic <strong>Gender</strong> one term may be<br />
functionally unmarked. The masculine Linguistic <strong>Gender</strong> is <strong>the</strong> unmarked choice in<br />
Hebrew (see example 2.5 and discussion <strong>the</strong>re). Female speakers of Hebrew tend to<br />
use masculine forms to refer to <strong>the</strong>mselves and to o<strong>the</strong>r women extending this<br />
unmarked usage. As 'Anat, a Hebrew-speaking hairdresser, put it, ‘talking in <strong>the</strong><br />
masculine frames <strong>the</strong> conversation as more general and less personal’. 4 Markedness<br />
in Linguistic <strong>Gender</strong> systems is <strong>the</strong> topic of §7.7. We return to an association between<br />
<strong>the</strong> ‘unmarked masculine’ and sexist language in §11.5.2.<br />
Changes in Linguistic <strong>Gender</strong> assignment to humans may reflect role reversals in<br />
traditional jocular relationships—see §7.1. <strong>Gender</strong> reversals may be offensive, or<br />
carry overtones of praise—see §7.2. Or <strong>the</strong>y can imply endearment and solidarity—<br />
see §7.3.<br />
7.1 Reversing Linguistic <strong>Gender</strong>s with jocular effects<br />
Among <strong>the</strong> traditional Manambu, classificatory sisters-in-law—female ego’s bro<strong>the</strong>rs’<br />
wives (kajal)—would taunt and tease each o<strong>the</strong>r. For a woman, her bro<strong>the</strong>rs’ wives are<br />
intruders and rivals of sorts. They would come from a different clan group from <strong>the</strong>ir<br />
husband and his sisters, and are like a foreign element in a family. A conventional way<br />
of jokingly greeting a bro<strong>the</strong>r’s wife is by saying:
7.1 Reversing Linguistic <strong>Gender</strong>s with jocular effects 101<br />
7.1 ñən wun-a ta:kw-añən<br />
you.fem.sg I-LK+fem.sg wife-2fem.sg.NOMINAL.PREDICATE<br />
wun ñən-a la:n-adəwun<br />
I you.fem.sg-LK+fem.sg husband-1masc.sg.NOMINAL.PREDICATE<br />
‘You (feminine) are my wife, I (masculine) am your husband (feminine)’ (said<br />
by a woman to a woman)<br />
The first clause in this sentence is fully grammatical—<strong>the</strong> feminine agreement with<br />
<strong>the</strong> second person feminine pronoun is correct. The second clause is nonsensical,<br />
from <strong>the</strong> point of view of Manambu grammar. The Linguistic <strong>Gender</strong> mismatch here<br />
is as follows. The masculine gender is marked on <strong>the</strong> predicate of <strong>the</strong> second clause<br />
(‘I am your husband’), but <strong>the</strong> agreement on <strong>the</strong> possessive ‘your’ is feminine. The<br />
‘correct’ way to say ‘I (man) am your husband’ is ñən-a-də la:n-adəwun (you.fem.sg-<br />
LK-masc.sg husband-1masc.sg.NOMINAL.PREDICATE). The feminine cross-referencing<br />
here is part of <strong>the</strong> joke—a woman saying 7.1 presents herself as a kind of ‘female<br />
husband’. She is virtually assuming <strong>the</strong> Social <strong>Gender</strong> identity of a patrilineal<br />
representative of her bro<strong>the</strong>r’s—and her own—clan. By jokingly assuming <strong>the</strong><br />
masculine Linguistic <strong>Gender</strong> and thus flaunting her ‘quasi-male’ Social <strong>Gender</strong><br />
identity, she takes <strong>the</strong> upper hand, and demonstrates her mock-superiority. This is<br />
achieved through an anomalous gender mismatch, unacceptable in any o<strong>the</strong>r circumstances.<br />
Such pronouncements are typically accompanied by roars of laughter<br />
from <strong>the</strong> speaker and <strong>the</strong> audience (and often followed by fur<strong>the</strong>r nasty comments,<br />
e.g. ‘I (masculine) will kick your fat stomach’, ‘I (masculine) will come at night and<br />
strangle you’, and so on).<br />
Switching genders in this context is not insulting. It is, ra<strong>the</strong>r, a way of manipulating<br />
one’s Social <strong>Gender</strong> identity reflected in anomalous gender use. Traditional<br />
women in <strong>the</strong> villages pride <strong>the</strong>mselves on being able to produce elaborate ‘bro<strong>the</strong>r’s<br />
sister to bro<strong>the</strong>r’s wife’ jokes: this is gradually becoming a mark of proficiency in<br />
cultural knowledge (which is nowadays on <strong>the</strong> wane). 5 Culturally permitted ‘gender<br />
switch’ is reminiscent of <strong>the</strong> Naven ritual described by Bateson (1958) for <strong>the</strong> closely<br />
related Iatmul. The ritual involved cross-dressing and mock Social <strong>Gender</strong> switch<br />
(we have no record of how this used to correlate with <strong>the</strong> use of Linguistic <strong>Gender</strong>;<br />
<strong>the</strong> ritual is hardly performed at present). 6<br />
The female version of Naven was described by Hauser-Schäublin (1995: 49, 1977:<br />
83–95). Naven used to be performed whenever a male or a female child would<br />
undertake a gender-specific task for <strong>the</strong> first time. During <strong>the</strong> all-women Naven<br />
ceremony, women would dress as men, and imitate aggressive and proud male<br />
behaviour. As Hauser-Schäublin (1995: 49) puts it, ‘Naven rites celebrate . . . sociocultural<br />
achievements’,and‘<strong>the</strong>ir transvestite aspects involve women structurally identifying<br />
<strong>the</strong>mselves with <strong>the</strong>ir husbands’. We know nothing about <strong>the</strong> use of Linguistic<br />
<strong>Gender</strong> in such rituals. Their existence confirms <strong>the</strong> cultural appropriateness of mock
102 7 Manly women and womanly men<br />
Social <strong>Gender</strong> shifts in order to emphasize <strong>the</strong> person’s Social <strong>Gender</strong> identity, and<br />
patrilineal allegiance. Manambu joking relationships point towards a similar principle.<br />
In Machiguenga, a Peruvian Arawak language, <strong>the</strong> non-masculine Linguistic<br />
<strong>Gender</strong> covers females and inanimates. It can be used by men in jocular reference<br />
to a ‘third man’, asa‘part of typical male joking behaviour in which men feminise<br />
one ano<strong>the</strong>r with sexual, and especially homosexual, comments and jokes’ (Shepard<br />
1997: 53). In 7.2 <strong>the</strong> verb is marked for feminine gender, and <strong>the</strong>re is a feminine<br />
pronoun. This can be part of a funny story told by men about ano<strong>the</strong>r man who is<br />
ei<strong>the</strong>r not present, or who is within earshot but not participating actively in <strong>the</strong><br />
conversation. A personal pronoun iro can be added.<br />
7.2 tata-ka iro o-ogo-ge-t-ak-a<br />
what-INDEF she 3fem-know/thinking-DISTRIBUTIVE-EPENTHETIC-PERFECTIVE-REALIS<br />
‘What in <strong>the</strong> world is she thinking?’ (referring to something funny <strong>the</strong> man did)<br />
Glenn Shepard (p.c.) comments that ‘<strong>the</strong> function of <strong>the</strong> feminine seems to be to<br />
actively exclude <strong>the</strong>m from <strong>the</strong> conversation and make <strong>the</strong>m <strong>the</strong> “object” of <strong>the</strong> joke<br />
or anecdote. They emphasize <strong>the</strong> “O” on each feminine pronominal marker, in fact<br />
emphasis on <strong>the</strong> first syllable of every word in <strong>the</strong> phrase with a sort of downward<br />
tone with vocal cry, <strong>the</strong> way women talk to young children to scold <strong>the</strong>m.’ Jocular<br />
switches of Linguistic <strong>Gender</strong> make people laugh, because <strong>the</strong>y draw attention to<br />
something out of <strong>the</strong> ordinary and essentially grotesque.<br />
7.2 Offence and praise in Linguistic <strong>Gender</strong> reversals<br />
In <strong>the</strong> day-to-day Manambu, changing <strong>the</strong> Linguistic <strong>Gender</strong> of humans means<br />
trouble. The Linguistic <strong>Gender</strong> of inanimates or non-humans depends on <strong>the</strong>ir size<br />
and shape (as we saw in §3.1). If a woman is referred to with masculine Linguistic<br />
<strong>Gender</strong>, and a man with feminine Linguistic <strong>Gender</strong>, <strong>the</strong>y are downgraded to <strong>the</strong><br />
level of inanimate things. In casual conversations (but hardly ever at open meetings),<br />
a smallish fat woman-like man can be referred to with <strong>the</strong> feminine Linguistic<br />
<strong>Gender</strong>. But kə-ø numa-ø du (this-fem.sg big-fem.sg man) ‘this fat round man’<br />
(smallish) can only be said behind <strong>the</strong> man’s back: this is very rude and offensive. 7<br />
A man who displays wrong social behaviour can also be referred to as ‘woman’—but<br />
again, behind his back, not to cause trouble. Someone who stayed in his wife’s village<br />
ra<strong>the</strong>r than taking her away, and thus breached <strong>the</strong> normal patrilocal practices, was<br />
described to me as:<br />
7.3 kə-ø ta:kw la:n-ad<br />
this-fem.sg woman husband-3masc.SG.NOMINAL.PREDICATE<br />
‘This (feminine) woman is (masculine) a husband’<br />
Here, <strong>the</strong> masculine Linguistic <strong>Gender</strong> is used to mark agreement with a feminine<br />
referent (instead of feminine Linguistic <strong>Gender</strong>). This mismatch makes speakers
7.2 Offence and praise in Linguistic <strong>Gender</strong> reversals 103<br />
laugh—it emphasizes <strong>the</strong> grotesqueness of <strong>the</strong> culturally inappropriate situation of<br />
Social <strong>Gender</strong> switch: a man shows patterns of social behaviour associated with a<br />
woman. The man is a kind of ‘female husband’—in fact, someone who has failed to<br />
live up to his Social <strong>Gender</strong> status. We can recall, from §3.4.2, that a man’s house, no<br />
matter how small in size, cannot be referred to with feminine Linguistic <strong>Gender</strong>—this<br />
is potentially demeaning.<br />
A woman who is too boisterous, large in size, and also imposing and ‘too big for<br />
her boots’ can be called kə-də numa-də ta:kw (this-masc.sg big-masc.sg woman) ‘this<br />
(unusually) big, boisterous, or bossy woman’. This is derogatory, and is not something<br />
one would say to that woman’s face. A woman is classified by her size, as if she<br />
were downgraded to <strong>the</strong> status of an inanimate referent, reflecting behaviour<br />
inappropriate for her Social <strong>Gender</strong>.<br />
On ano<strong>the</strong>r occasion, kə-də numa-də ta:kw (this-masc.sg big-masc.sg woman)<br />
‘this (unusually) big, boisterous or bossy woman’ was said of a woman who sported<br />
an unusually extensive knowledge of totemic names which is traditionally <strong>the</strong><br />
domain of a male. In terms of her Social <strong>Gender</strong> status, she behaved inappropriately;<br />
hence negative connotations of applying <strong>the</strong> masculine Linguistic <strong>Gender</strong> to her.<br />
Once again, classifying a human being by <strong>the</strong>ir size is demeaning: this implies<br />
bringing a human down to <strong>the</strong> level of a mere object.<br />
In a number of Berber languages from <strong>the</strong> Middle Atlas area of Morocco, referring<br />
to a woman with a masculine Linguistic <strong>Gender</strong> form is equally offensive. In Aït<br />
Mguild and Aït Wirra dialects, <strong>the</strong> form marked for feminine derivational Linguistic<br />
<strong>Gender</strong>, tamt̍t̍ut̍t̍, means ‘woman’. The masculine form, amet̍t̍u, denotes a woman<br />
with man-like manners and has derogatory overtones of French femme hommasse<br />
‘overly masculine woman’. 8<br />
In many proverbs from across <strong>the</strong> world assembled by Schipper (2003) a woman<br />
who is taller than her partner, or who displays male-type behaviour, is looked down<br />
upon. In Schipper’s (2003: 20) words,<br />
an Ashanti proverb presents <strong>the</strong> problem as follows, ‘When a tall woman carries palm nuts,<br />
birds eat <strong>the</strong>m off her head’. A tall woman who proudly carries nuts in a bowl on her head is<br />
presented as one who is showing off, and she is warned that she will be punished for it. The<br />
explanation given is that such ‘male’ behaviour is condemned in a woman. The proverb is<br />
warningly quoted to girls who display what society considers to be male traits, and <strong>the</strong> palm<br />
nuts are a metaphor for ‘<strong>the</strong> male world’: as long as men are alive and around, a woman is not<br />
supposed to crack nuts. The proverb reminds women to refrain from getting involved in<br />
designated male roles.<br />
The title of Schipper’s book, Never marry a woman with big feet,reflects that same<br />
attitude: women with big man-like feet are physically and socially inappropriate.<br />
An example of how treating a woman as if she were a man may have a demeaning<br />
effect comes from one incident of <strong>the</strong> male initiation applied to a woman, in <strong>the</strong> Iatmul
104 7 Manly women and womanly men<br />
village of Palimbei, from <strong>the</strong> same Sepik region as Manambu (Hauser-Schäublin 1995:<br />
50–2). A girl who had inadvertently seen men blowing long flutes in a fenced-off<br />
enclosure (something a woman is not allowed to see) was gang-raped, and <strong>the</strong>n<br />
subjected to scarification and a shortened version of male initiation. The initiation<br />
was meant as a severe punishment for a ‘crime’ and stigmatization: <strong>the</strong> girl felt<br />
degraded and shamed, despite <strong>the</strong> fact that she had gained what was considered<br />
important ritual knowledge. It was as if she’d lost her appropriate Social <strong>Gender</strong> status.<br />
The girl never recovered from that experience and soon died.<br />
Changing Linguistic <strong>Gender</strong> for humans can have overtones of praise, and solidarity.<br />
The masculine Linguistic <strong>Gender</strong> in Lokono Arawak has connotations of<br />
something good for animates, inanimates, and also humans (as we can recall from<br />
§3.4.1). A man from <strong>the</strong> speaker’s tribe would normally be referred to with masculine<br />
Linguistic <strong>Gender</strong>. But if he is despised, <strong>the</strong> gender will switch to feminine. Males of a<br />
different tribe are referred to with <strong>the</strong> feminine Linguistic <strong>Gender</strong>. If one is friends<br />
with <strong>the</strong>m, <strong>the</strong>y become ‘masculine’ in <strong>the</strong>ir Linguistic <strong>Gender</strong>. 9<br />
An Amharic-speaking man can address ano<strong>the</strong>r man with a feminine pronoun ‘as<br />
a term of insult, to belittle’ him (Wołk 2009: 131). Such ‘derogatory’ connotations<br />
come about when an elderly person can address a male youngster as ‘feminine’. One<br />
can refer to a male enemy with a feminine pronoun, to show one’s superiority. Hoben<br />
(1976: 287) mentions how <strong>the</strong> second person singular masculine pronoun antɛ can<br />
be replaced with <strong>the</strong> corresponding feminine pronoun anci, ‘if <strong>the</strong> speaker wishes to<br />
insult a male’. This form was used ‘to refer to <strong>the</strong> rebel groups as a put down’, and<br />
also to express social distance—‘in <strong>the</strong> context of an older man using <strong>the</strong> feminine to<br />
address a younger boy’ (Pankhurst 1992: 169). The form anci can also be used ‘in a<br />
humorously belittling sense for <strong>the</strong> smallest in a group of friends or for <strong>the</strong> clown of a<br />
group’ (Hoben 1976: 286).<br />
Addressing women as if <strong>the</strong>y were men has an opposite effect. To use a masculine<br />
pronoun addressing a woman in Amharic means praising her—<strong>the</strong> implication being<br />
that a woman is acting ‘like a man’ and is thus ‘promoted’ to a manly status<br />
(Pankhurst 1992: 169). In Figuig, a North Berber language from Morocco, tamət̍t̍ut<br />
‘woman’ can be made masculine (by removing feminine derivational gender marking).<br />
The resulting masculine form, amət̍t̍u, means ‘a courageous or a corpulent<br />
woman’. This implies approbation (in contrast to <strong>the</strong> examples from o<strong>the</strong>r Berber<br />
languages where <strong>the</strong> overtones of a ‘masculine woman’ are distinctly pejorative). 10<br />
In a few South American languages, reversing Linguistic <strong>Gender</strong> is a mark of<br />
higher Social <strong>Gender</strong> status. Jarawara is an Arawá language from sou<strong>the</strong>rn Amazonia<br />
(with two genders, masculine and feminine). A woman can be referred to with<br />
masculine gender if she is particularly important in <strong>the</strong> society, or close to <strong>the</strong><br />
speaker. A narrator referred to his wife as ‘he’, as‘a mark of affinity and of respect’<br />
towards her (Dixon 2004: 287).
7.3 Linguistic <strong>Gender</strong> reversal: endearment and solidarity 105<br />
Among <strong>the</strong> Tariana and <strong>the</strong>ir Tucanoan neighbours from <strong>the</strong> Vaupés River<br />
Basin in Brazilian Amazonia, women were <strong>the</strong> erstwhile owners of <strong>the</strong> magic<br />
Yurupary flutes but lost <strong>the</strong>m and cannot even look at <strong>the</strong>m any more. Nowadays,<br />
women are denied access to <strong>the</strong> magic powers associated with <strong>the</strong> flute; any word<br />
that sounds similar to <strong>the</strong> flute’s nameusedtobea‘secret’ from women (we return<br />
to <strong>the</strong> anti-women taboo register, and <strong>the</strong> status of women among <strong>the</strong> Tariana, in<br />
Chapter 10). But a particularly respected and important woman is promoted to an<br />
honorary ‘manhood’. A woman in Tariana is normally referred to with feminine<br />
gender, on <strong>the</strong> verb, or <strong>the</strong> pronoun duha ‘she’ or a classifier for females (‐ma). The<br />
Woman-Creator or a particularly powerful woman is consistently talked about as diha<br />
‘he’ and requires a human classifier (-ite), just as any man would. So was a particularly<br />
knowledgeable woman (<strong>the</strong> mo<strong>the</strong>r of my major consultants). A common English<br />
expression man up! meaning ‘get your act toge<strong>the</strong>r in such a way as a real man would’—<br />
especially if applied to a woman—would be an uncanny analogy here. ‘Man’ in <strong>the</strong><br />
Tariana lore epitomizes strength and prowess, while women are to blame for many<br />
mishaps, past and present. We turn to this again in §11.3. This is very different from<br />
Linguistic <strong>Gender</strong> switches in Manambu which are derogatory in essence: <strong>the</strong>y reflect<br />
behaviour inappropriate for one’sSocial<strong>Gender</strong>.<br />
7.3 Linguistic <strong>Gender</strong> reversal: endearment and solidarity<br />
Overtones of Linguistic <strong>Gender</strong> reversal may differ depending on <strong>the</strong> context. Using<br />
a feminine pronoun to a man in colloquial Amharic does not always imply an insult.<br />
Second person feminine pronouns are widely employed by men to address o<strong>the</strong>r<br />
men, as a term of rapprochement, or endearment (Wołk 2009: 131–2, Pankhurst<br />
1992). According to Hoben (1976: 287), in Addis Ababa, but not in <strong>the</strong> countryside,<br />
male friends may address each o<strong>the</strong>r as anci ‘you singular feminine’, to express<br />
affection. 11 This reminds us of how feminine Linguistic <strong>Gender</strong> assignment to<br />
lower animates and inanimates may have overtones of endearment and diminution<br />
(as we saw for Oromo, in §3.4.1). The masculine pronoun can also be used between<br />
women as a mark of ‘attachment and closeness’ (Pankhurst 1992: 170).<br />
In Arabic, a feminine noun, pronoun, adjective, or verb form can be used to a boy,<br />
and a masculine one to a girl. So, one can say, in Arabic, wēn ruħti yā,ʼ binti ‘Where<br />
did you go, little girl?’ to a boy, and inta žuʔān ‘Are you (masculine) hungry<br />
(masculine)?’ to a girl. 12<br />
In Russian, a woman can be affectionately addressed with a form marked for<br />
masculine gender: <strong>the</strong> author, a woman, has often been addressed as moj horoshij ‘my<br />
good one-masculine-singular’ or moj malenjkij ‘my little one-masculine singular’ by<br />
older native speakers of <strong>the</strong> language (all of <strong>the</strong>m women). A younger woman can<br />
address a group of women as Rebjata ‘Guys!’ (using a masculine Linguistic <strong>Gender</strong>
106 7 Manly women and womanly men<br />
form). This is a mark of camaraderie, similar to girl-only groups addressing each<br />
o<strong>the</strong>r as Hey, guys! in <strong>the</strong> USA. 13<br />
In Modern Hebrew, a man can address close female friends, relatives, associates,<br />
and partners with masculine pronouns, and verbal and adjectival forms, as a sign of<br />
affection, intimacy, and solidarity. Close female friends and relatives can do <strong>the</strong> same<br />
to each o<strong>the</strong>r, with similar effect. Tobin (2001: 187–91) describes Linguistic <strong>Gender</strong><br />
reversal in a family of native Hebrew speakers, with two non-identical twin daughters<br />
8½ years of age: one twin is ‘bigger’ and <strong>the</strong> o<strong>the</strong>r one is ‘smaller’. The smaller twin is<br />
<strong>the</strong> one who most frequently uses <strong>the</strong> masculine Linguistic <strong>Gender</strong> forms to refer to<br />
herself; o<strong>the</strong>rs also use <strong>the</strong> masculine Linguistic <strong>Gender</strong> to talk to her. In one<br />
example, <strong>the</strong> smaller twin talks about herself as if she were a boy, asking her<br />
grandmo<strong>the</strong>r to take care of her ra<strong>the</strong>r than of her bigger sister:<br />
7.4 ki hi gdolah ve-ani katan<br />
because she big.fem.sg and-I small.masc.sg<br />
‘Because she is big (feminine) and I am small (masculine)’<br />
The twins were once sent to <strong>the</strong>ir room as a punishment. They both reversed<br />
genders while talking to each o<strong>the</strong>r as a sign of solidarity, as <strong>the</strong>y negotiated <strong>the</strong>ir<br />
uncomfortable situation:<br />
7.5a rotseh lesaxek?<br />
want.masc.sg to.play<br />
‘Ya wanna play?’<br />
7.5b lo, aval tixtov mixtav le-aba ve-tivakesh slixa<br />
no but write.IMPV.masc.sg letter to-daddy and-ask.IMPV.masc.sg sorry<br />
‘No, but write a letter to Daddy and apologize’<br />
The fa<strong>the</strong>r was telling <strong>the</strong> smaller twin what she was like as a baby using feminine<br />
forms. He switched to <strong>the</strong> masculine form when he tells her that she had to stay in <strong>the</strong><br />
hospital alone to become stronger before <strong>the</strong>y could take her home:<br />
7.6 ki hayita tsarix lehithazek ktsat<br />
because you.were.masc.sg necessary.masc.sg to.become.strong little<br />
‘Because you had to become a little stronger’<br />
Masculine Linguistic <strong>Gender</strong> marks solidarity, intimacy, and protection. Feminine<br />
Linguistic <strong>Gender</strong> is not used this way.<br />
The neuter Linguistic <strong>Gender</strong> in Bulgarian includes inanimates. But it may have<br />
overtones of joking endearment if applied to humans. In colloquial Bulgarian a neuter<br />
gender form can be used to refer to a man or to a woman, marking condescension<br />
or endearment. A young teacher (daskal, masculine) can be addressed as daskalče<br />
(neuter) by an older and a wealthier man. A woman behaving in a silly way can be<br />
addressed as prosto (simple+neuter gender) ‘dear silly one’. 14
7.3 Linguistic <strong>Gender</strong> reversal: endearment and solidarity 107<br />
A somewhat different kind of Linguistic <strong>Gender</strong> reversal has been described for <strong>the</strong><br />
Iroquoian languages Mohawk, Oneida, and Onondaga. A woman can be referred to<br />
with ei<strong>the</strong>r <strong>the</strong> special feminine, or <strong>the</strong> neuter form. The feminine form is also used in<br />
<strong>the</strong> meaning of ‘people in general, and when sex is not specified. The neuter form<br />
covers animals.’ 15 The differences between <strong>the</strong> two ways of talking about a woman<br />
are subtle. The choice of <strong>the</strong> prefix in Oneida correlates with what <strong>the</strong> woman is like:<br />
a small and graceful woman will be referred to as ‘feminine’, and a large and<br />
aggressive one would be treated as ‘neuter’. According to Bonvillain (1973: 86–7),<br />
<strong>the</strong> feminine pronouns and agreement prefixes (ye-/ya-) in Mohawk are used for<br />
‘female human beings who are regarded in some prestigeful or respectful way by <strong>the</strong><br />
speaker’, or‘for whom <strong>the</strong> speaker has <strong>the</strong> feeling of affection or closeness’. The<br />
alternative, <strong>the</strong> ‘neuter’ (ka-/wa-), is used o<strong>the</strong>rwise. In Bonvillain’s words, ‘it may or<br />
may not reflect a momentary or general negative feeling on <strong>the</strong> part of <strong>the</strong> speaker<br />
toward <strong>the</strong> female who is spoken about. Some speakers feel that . . . a girl or a woman<br />
who is awkward or aggressive may more readily fall into this category, whereas one<br />
who is graceful and quiet may tend to be classed in <strong>the</strong> o<strong>the</strong>r way.’ 16<br />
The two ways of referring to women vary across Mohawk communities. Some<br />
people employ <strong>the</strong> feminine-indefinite form as a mark of endearment, when talking<br />
about ‘gentler’ speakers, and <strong>the</strong> ‘neuter’ form for ‘someone who is really rough’. The<br />
feminine form ie-ksa'tí:io ‘she is pretty’ would be used for someone ‘a bit more<br />
ladylike’, and <strong>the</strong> neuter form ka-ksa'tí:io ‘she is pretty’ would be preferred for someone<br />
attractive <strong>the</strong> man would ‘like to go out with’ (but perhaps not to marry). Some<br />
comment that a neuter form ‘would be used for someone you don’t respect: an animal<br />
or a stranger’. And indeed, <strong>the</strong>se forms are common for non-Mohawk women.<br />
A neuter form may imply that <strong>the</strong> speaker does not like <strong>the</strong> woman. In general,<br />
feminine-indefinite forms may signal endearment toward family members, and neuter<br />
forms may signal familiarity among close friends. The feminine-indefinite forms mark<br />
respect, but <strong>the</strong> neuter forms are <strong>the</strong> ones used to talk about people one does not know<br />
well. The history, and reconstruction, of Nor<strong>the</strong>rn Iroquoian languages and Mohawk<br />
shows that <strong>the</strong> erstwhile indefinite pronoun and prefix on verbs came to be used to<br />
refer to women as a sign of respect—reflected in some of <strong>the</strong> current usage. The<br />
feminine-neuter is <strong>the</strong> result of reinterpretation of <strong>the</strong> third person category once <strong>the</strong><br />
new masculine was introduced. 17<br />
Once again, a word of warning. In Mithun’s words (2014: 140), <strong>the</strong> feminineindefinite<br />
category ‘did not emerge from a view of women as quintessential or<br />
prototypical human beings’, nor did <strong>the</strong> ‘feminine-neuter’ category ‘develop from a<br />
view of women as akin to animals’. Mithun (2014: 137) reports that ‘speakers are<br />
usually surprised when it is brought to <strong>the</strong>ir attention. Some are horrified to realize<br />
that <strong>the</strong>y are using <strong>the</strong> same forms for women and animals. They report that <strong>the</strong>y<br />
never thought about it, and that “No one notices”.’ We return to <strong>the</strong> status of<br />
‘women’ in Iroquoian languages in §7.7. 18
108 7 Manly women and womanly men<br />
Reversal in self-reference may serve as marker of gender identity, for transgender<br />
and transsexual people. We can recall, from §1.1, how <strong>the</strong> title of <strong>the</strong> autobiography<br />
of Brigitte Martel, a transsexual male who became female, captured <strong>the</strong> change of her<br />
Natural and Social <strong>Gender</strong> from male to female: Né homme, comment je suis devenue<br />
femme—‘Born (masculine) a man, how I became (feminine) woman’. We return to<br />
men’s and women’s speech and transgender practices in §9.3.<br />
7.4 Men as women, women as men: a summary<br />
of Linguistic <strong>Gender</strong> reversals<br />
Changing Linguistic <strong>Gender</strong> of inanimates may highlight <strong>the</strong>ir physical features,<br />
mark endearment or value. Changing Linguistic <strong>Gender</strong> of humans reflects changing<br />
Social <strong>Gender</strong> relationships. When Linguistic <strong>Gender</strong>s are reversed in jocular contexts<br />
in Machiguenga and Manambu, <strong>the</strong> grotesque effect of it makes people laugh.<br />
Reversing genders may result in mortal offence, or in ‘promoting’ a woman to a<br />
higher manly status. Addressing a woman as if she were a man may signal affection<br />
or solidarity. Addressing a man as if he were a woman may be a sign of endearment.<br />
Or it may imply dragging him down to a ‘woman’s’ level. Linguistic <strong>Gender</strong> reversal<br />
can have an opposite effect in <strong>the</strong> same language: addressing men as women in<br />
Amharic may sound insulting in one context, and endearing in ano<strong>the</strong>r (depending<br />
on <strong>the</strong> relative age of people and <strong>the</strong>ir relationships). We can recall, from earlier in<br />
this chapter, that women speakers of Palestinian Arabic refer to <strong>the</strong>mselves as ‘men’<br />
if <strong>the</strong>re is something wrong with <strong>the</strong>m—changing Linguistic <strong>Gender</strong>s implies an<br />
unusual and an uncomfortable state of affairs.<br />
Linguistic <strong>Gender</strong> reversals highlight positive and negative associations with men<br />
and women as social constructs. Here, Linguistic <strong>Gender</strong> reflects <strong>the</strong> stereotypes of<br />
Social <strong>Gender</strong> entombed in <strong>the</strong> language. Tables 7.2 and 7.3 summarize examples<br />
discussed so far.<br />
Idiomatic expressions reflect attitudes to Social <strong>Gender</strong>s and <strong>the</strong>ir comparative<br />
value. English drama queen or old wife when mockingly applied to a non-manly man<br />
have a similar pejorative and insulting effect as do gender switches in Manambu,<br />
Amharic, and Lokono in Table 7.2. Saying ‘man up!’ to a woman urges a woman to<br />
act in a manly way and to be as good as a man—similar to Lokono, Amharic, Figuig,<br />
TABLE 7.2. Treating ‘men’ as ‘women’: masculine to feminine Linguistic <strong>Gender</strong><br />
reversal<br />
Masculine to feminine reversal<br />
pejorative and insulting<br />
endearment and solidarity<br />
Language examples discussed<br />
Manambu, Amharic, Lokono<br />
Amharic, Arabic, and Marathi baby talk
7.5 Attitudes to Social <strong>Gender</strong>s through Linguistic <strong>Gender</strong> reversals 109<br />
TABLE 7.3. Treating ‘women’ as ‘men’: feminine to masculine Linguistic <strong>Gender</strong><br />
reversal<br />
Feminine to masculine reversal<br />
pejorative: downgrading a woman as if she<br />
were too bossy and ‘too big for her boots’<br />
adding value: ‘promoting’ a woman to male<br />
status<br />
endearment and solidarity<br />
Language examples discussed<br />
Manambu; Aït Mguild, Aït Wirra (North<br />
Berber)<br />
Lokono, Amharic, Figuig, Jarawara, Tariana<br />
Arabic and Marathi baby talk, Russian,<br />
Modern Hebrew<br />
Jarawara, and Tariana in Table 7.3. We now turn to fur<strong>the</strong>r ways of expressing<br />
attitudes to Social <strong>Gender</strong>s through <strong>the</strong> use of Linguistic <strong>Gender</strong>s.<br />
7.5 Attitudes to Social <strong>Gender</strong>s through Linguistic <strong>Gender</strong> reversals<br />
Social <strong>Gender</strong> stereotypes associated with Natural <strong>Gender</strong> can be seen through <strong>the</strong><br />
use of pronouns marked for Linguistic <strong>Gender</strong> in English. Traditional grammarians<br />
insist that <strong>the</strong> choice of gender-sensitive third person singular pronouns is based<br />
on Natural <strong>Gender</strong>: ‘she’ for women, ‘he’ for men, and ‘it’ for inanimates. <strong>How</strong>ever,<br />
in many varieties of English <strong>the</strong> situation is not that straightforward. ‘She’ can have<br />
overtones of ‘smallness’ (in agreement with <strong>the</strong> principles we saw in Chapter 3): a bee<br />
was referred to as ‘she’ as early as John Lily’s Euphues (1578). In <strong>the</strong> following<br />
modern conversation, an ant was referred to as he—emphasizing <strong>the</strong> speaker’s<br />
empathy with <strong>the</strong> insect who is referred to as if it was a human. 19<br />
7.7 Speaker A: Ah <strong>the</strong>re’s an ant<br />
Speaker B: Well catch it<br />
Speaker A: Well put him outside/ let him go on to—/ Look he’s on <strong>the</strong> toaster /<br />
Now put him outside nicely .../he may be someone else’s ant you<br />
know<br />
In some varieties of Australian English, ‘she’ can be used to refer to unruly objects for<br />
which one has antipathy (Wales 1996: 139):<br />
7.8 She’s an absolute bastard, this truck.<br />
In <strong>the</strong> English poetic tradition, ‘“masculine”-marked words were grouped according<br />
to supposedly “manly” attributes (mostly positive) and “feminine”-marked words<br />
according to “womanly” attributes (some positive, many negative): such as strong,<br />
active, aggressive, powerful, clever, big fierce, giving (“manly”); versus weak, timid,<br />
passive, loving, soft, helpful, beautiful, small, moral, receptive (“womanly”)’ (Wales
110 7 Manly women and womanly men<br />
1996: 148). This constellation of semantic features of Natural and Social <strong>Gender</strong>s as<br />
reflected in Linguistic <strong>Gender</strong> usage can be conceived as a kind of ‘sexist symbolism’. 20<br />
Mathiot (1979b) conducted a comprehensive study of sex roles of men and women<br />
(that is, <strong>the</strong>ir Social <strong>Gender</strong>) as revealed through Linguistic <strong>Gender</strong> reference, to see<br />
how <strong>the</strong> images of males and females can be reflected in <strong>the</strong> use of personal pronouns<br />
in English. The use of <strong>the</strong> pronouns he and she observed with inanimate referents in<br />
American English was found to correlate with a number of stereotyped features—<br />
part of <strong>the</strong> inherent image and role image American men and women have of<br />
<strong>the</strong>mselves, and of each o<strong>the</strong>r.<br />
The semantic opposition BEAUTIFUL versus UGLY manifests men’s conception of<br />
women’s versus men’s appearance; and <strong>the</strong> semantic opposition manifesting men’s<br />
conception of women’s and men’s achievement potential is INCOMPETENT versus<br />
COMPETENT. Thus a beautiful flower is referred to as ‘she’, and an ugly cactus as<br />
‘he’. 21 A car referred to as a prized possession is ‘she’. An unruly storm is also ‘she’;<br />
and so is <strong>the</strong> troublesome refrigerator.<br />
In contrast, <strong>the</strong> inherent image and role image American women have of <strong>the</strong>mselves<br />
and of men can be formulated in one semantic opposition: MATURE versus<br />
INFANTILE (Mathiot 1979b: 25). A chair which has lost its leg is a ‘he’, and so is a<br />
typewriter that would not work. Table 7.4 summarizes features that consistently<br />
came about in men talking about women and about <strong>the</strong>mselves, and women talking<br />
about men and <strong>the</strong>mselves.<br />
English spoken in Tasmania (described by Pawley 2004) has certain rules governing<br />
<strong>the</strong> choice of pronouns he and she with non-human referents. When <strong>the</strong> sex of a<br />
higher animal is not known, <strong>the</strong> animal is referred to as ‘he’. He is used with plants<br />
or parts of living plants, and any item of goods or portable property (o<strong>the</strong>r than<br />
vehicles) that is viewed as trade goods ra<strong>the</strong>r than a personal possession. As Pawley<br />
(2004: 116; 127) puts it, ‘using “she” for nouns referring to portable goods (not<br />
vehicles) is a marker of “attachment”, or something one manipulates’. For instance, a<br />
soccer ball has been kicked into a tree. Player, shaking <strong>the</strong> branch, says: ‘She’s stuck.<br />
Come down, you bitch!’.<br />
TABLE 7.4. Meanings of ‘she’ vs ‘he’ in American English (Mathiot 1979b) 22<br />
Men’s conception of men and women<br />
men<br />
women<br />
ugly<br />
beautiful<br />
competent incompetent<br />
brave<br />
a challenge to, or a reward for, men<br />
good-natured men’s prized possessions<br />
Women’s conception of<br />
men and women<br />
men women<br />
infantile mature
7.5 Attitudes to Social <strong>Gender</strong>s through Linguistic <strong>Gender</strong> reversals 111<br />
The negative connotations of feminine Linguistic <strong>Gender</strong> surface in many derivational<br />
forms belonging to feminine Linguistic <strong>Gender</strong> with feminine Natural <strong>Gender</strong><br />
reference. As Baron (1986: 114–15) puts it,<br />
feminine English nouns tend to acquire negative connotations at a much faster rate than<br />
masculine or neuter ones, creating semantic imbalances in originally parallel masculine/<br />
feminine pairs like fox—vixen and governor—governess. Efforts on <strong>the</strong> part of feminists and<br />
usage critics to eliminate feminine nouns like authoress in favour of unmarked equivalents on<br />
<strong>the</strong> grounds that <strong>the</strong> marked terms are demeaning have been only partially successful.<br />
Many feminine derivations in ‐ette in English have a negative feel about <strong>the</strong>m; some,<br />
like suffragette, were created as somewhat derogatory in <strong>the</strong> first place (as we recall<br />
from §6.4.2). The word spin-ster originally meant ‘a woman (or, rarely, a man) who<br />
spins’ (Oxford English Dictionary). In <strong>the</strong> seventeenth century, it came to be used as<br />
‘<strong>the</strong> proper legal title of one still unmarried’ (presumably by association of women with<br />
<strong>the</strong> needle-trade). In <strong>the</strong> eighteenth century, spinster acquired <strong>the</strong> pejorative meaning<br />
it has today, of ‘a woman still unmarried; especially one beyond <strong>the</strong> usual age for<br />
marriage, an old maid’. In <strong>the</strong>ir feminist dictionary, Kramarae and Treichler (1992:<br />
429–30) fur<strong>the</strong>r add to <strong>the</strong> description of spinster: ‘Like most terms connected with<br />
women, it became a euphemism for mistress, or prostitute’,a‘figure of fun and ridicule<br />
to those men who see <strong>the</strong>mselves as essential to a woman’s existence’. 23<br />
Deprecatory overtones of ‘femininity’ appear in fur<strong>the</strong>r English expressions such<br />
as female logic, old wives’ tale, old woman. In his letter to his friend James Hogg on 24<br />
March 1814, Lord Byron referred depreciatingly to his fellow poets Coleridge and<br />
Sou<strong>the</strong>y as ‘mere old wives’. 24 One can say He is a drama queen about a particularly<br />
fussy man. He is an old woman was once used to refer to a now deceased linguist who<br />
was incapable of fulfilling his editorial responsibilities and was worrying about every<br />
minute detail of a task he was not up to. This is reminiscent of <strong>the</strong> deprecatory use of<br />
feminine Linguistic <strong>Gender</strong> in Manambu (in §7.2), when talking about a man who is<br />
not up to <strong>the</strong> standards of <strong>the</strong> Social <strong>Gender</strong> expectations in <strong>the</strong> society. Unlike<br />
Manambu, English does not have a corresponding masculine equivalent: saying *She<br />
is a drama king or *She is an old man is not idiomatic.<br />
English is not unique in having negative overtones ascribed to derivational forms of<br />
feminine Linguistic <strong>Gender</strong>. The general term for ‘dean’ in Russian is dekan (masculine<br />
Linguistic <strong>Gender</strong>). Feminine derivations—dekanša, or dekanixa—have distinctly<br />
pejorative overtones. The general term for postgraduate student is masculine aspirant.<br />
Zemskaya, Kitajgorodskaja, and Rozanova (1993: 126) cite Saratov as a city where <strong>the</strong><br />
corresponding feminine form aspirantka is perceived as slighting (this overtone is<br />
absent from <strong>the</strong> standard language). 25 The female version of poet ‘poet’, poètessa, did<br />
not have slighting overtones in <strong>the</strong> nineteenth century. <strong>How</strong>ever, in <strong>the</strong> 1960s Anna<br />
Akhmatova, one of <strong>the</strong> greatest Russian poets, told Yuri Aikhenvald that she should be<br />
referred to as a ‘poet’, and not downgraded to a mere ‘poètessa’. 26
112 7 Manly women and womanly men<br />
Many professions in Brazilian Portuguese have a generic masculine term and a<br />
feminine counterpart, with no negative feelings about <strong>the</strong>m. A male teacher (from<br />
kindergarten to university) is professor, and <strong>the</strong> female one is professor-a, a male<br />
judge is juiz, and a female judge is juiz-a; a male lawyer is advogado, and a female is<br />
advogad-a. A male president is presidente. On <strong>the</strong> surface of it, <strong>the</strong> female form<br />
president-a refers ei<strong>the</strong>r to a president’s wife, or to a woman who ‘presides’. <strong>How</strong>ever,<br />
in common usage presidenta ‘female president’ used to have strong overtones of a<br />
bossy and merciless lady—as big for her boots as a masculine woman among <strong>the</strong><br />
Manambu. Things changed drastically when Dilma Rousseff was elected <strong>the</strong> first<br />
female president of Brazil. A special law of <strong>the</strong> Brazilian government passed on<br />
3 April 2012 ‘determined an obligatory usage of gender inflexion for professions<br />
and degrees’, leaving no option but to use <strong>the</strong> term presidenta for <strong>the</strong> female<br />
president. 27 This is one way in which Linguistic <strong>Gender</strong> reflects social changes and<br />
can be sensitive to language planning: we return to changes in anaphoric and<br />
derivational gender as targets in <strong>the</strong> struggle against ‘sexist’ language in §11.5.<br />
Words of <strong>the</strong> masculine Linguistic <strong>Gender</strong> and masculine pronouns can have<br />
generic senses and be used as a functionally unmarked option. This is what we turn<br />
to now.<br />
7.6 ‘Women’ as a subtype of ‘men’? The overtones<br />
of masculine generics<br />
Traditional practice in English was to employ <strong>the</strong> masculine Linguistic <strong>Gender</strong><br />
pronoun he or man when no Linguistic <strong>Gender</strong> specification was intended, e.g.<br />
Man is mortal. The use of <strong>the</strong> masculine pronoun he in English for generic referents<br />
and those whose Natural <strong>Gender</strong> is not known has been a controversial issue for<br />
some time. Until recently, he in English was generally used as a term for human<br />
reference and also as one subordinate term, for male reference (see <strong>the</strong> discussion in<br />
Alpher 1987). The meanings of quite a few nouns in English show a bias: high-status<br />
occupational terms such as lawyer, physician, orscientist are traditionally referred to<br />
by <strong>the</strong> masculine pronoun he even in <strong>the</strong> contexts where Natural <strong>Gender</strong> (or sex) is<br />
irrelevant or not known. In contrast, a nurse, a primary schoolteacher, or a secretary<br />
would be more likely to be a woman, and referred to as she. The Social <strong>Gender</strong><br />
stereotypes are encroaching upon <strong>the</strong> use of Linguistic <strong>Gender</strong>.<br />
During recent years, it has become <strong>the</strong> custom that <strong>the</strong> generic unmarked pronoun<br />
‘<strong>the</strong>y’ should be generally used, to avoid what was perceived as linguistic sexism. (We<br />
return to <strong>the</strong> story of ‘he’ and ‘man’ in <strong>the</strong> context of language planning and social<br />
change in Chapter 11.) A similar concern applies to <strong>the</strong> use of ‘man’ as a generic term<br />
in modern English: Baron (1986: 137) offers a history of attempts to get rid of ‘man’<br />
as a generic term seen as a reflection of Social <strong>Gender</strong>-based male dominance:
7.6 ‘Women’ as a subtype of ‘men’? 113<br />
Perhaps most troublesome to linguists, feminists and usage critics alike has been <strong>the</strong> use of <strong>the</strong><br />
word man itself, in phrases such as <strong>the</strong> man in <strong>the</strong> street and compounds like mankind and<br />
chairman. Some authorities argue that man, atfirst a word in which both genders were<br />
combined, now refers primarily to males, while o<strong>the</strong>rs claim that <strong>the</strong> neutral sense of man is<br />
not dead and that <strong>the</strong> word still retains <strong>the</strong> primary meaning ‘human being’. In any case, one<br />
must admit that in actual use it is often unclear whe<strong>the</strong>r man refers to people in general or to<br />
men only.<br />
Masculine forms used as generics may be seen as inherently sexist—for some, <strong>the</strong>y<br />
automatically reflect <strong>the</strong> assumption that <strong>the</strong> male of <strong>the</strong> species represents <strong>the</strong><br />
species itself. The male interpretation of referents is hard to avoid if we use <strong>the</strong><br />
generic masculine ‘he’ (as McConnell-Ginet 1979: 77 puts it). The generic ‘he’ and<br />
‘man’ reduce <strong>the</strong> ‘visibility’ of women. Susan Sontag (1973: 186) offers a stronger<br />
statement: ‘grammar, <strong>the</strong> ultimate arena of sexist brainwashing, conceals <strong>the</strong> very<br />
existence of women—except in special situations.’ Many sources on English indirectly<br />
confirm <strong>the</strong> view. Samuel Johnson, in his 1755 classic dictionary of English,<br />
defines ‘man’ as a ‘human being, not a woman, not a boy’; a woman is ‘<strong>the</strong> female of<br />
<strong>the</strong> human race’. The 1828 Webster dictionary defines ‘man’ as ‘mankind, <strong>the</strong> human<br />
race, <strong>the</strong> whole species of human beings’, while woman is ‘<strong>the</strong> female of <strong>the</strong> human<br />
race, grown to adult years’. 28<br />
It may appear that <strong>the</strong> issue of <strong>the</strong> generic masculine pronoun does not arise in<br />
languages with pervasive agreement gender. In Portuguese, Greek, Russian, Spanish,<br />
French, Manambu, and many o<strong>the</strong>r languages a pronoun used to refer to a person, or<br />
an entity, will be strictly chosen by its Linguistic <strong>Gender</strong>. A turnip in German (die<br />
Rübe) is feminine, and will have to be referred to as ‘she’. A man (der Mensch) is<br />
masculine, and a woman (die Frau) is feminine: <strong>the</strong>y will be referred to as ‘he’ and<br />
‘she’ respectively. We saw in §2.1.2 that a girl (das Mädchen) is neuter, and is often<br />
referred to as es ‘it’; however, grown up girls may be talked about as sie ‘she’. But<br />
<strong>the</strong> generic term for humans is ei<strong>the</strong>r <strong>the</strong> same as ‘man’ (German Mensch, French<br />
homme) or is masculine, in its Linguistic <strong>Gender</strong> (as in Russian čelovek). Vinogradov<br />
(1947: 62) stressed <strong>the</strong> idea of a ‘general human being’ expressed through masculine<br />
forms in Russian.<br />
Experimental studies point towards sexist overtones of masculine generic in<br />
German. 29 Masculine terms der Leser ‘reader’, der Student ‘student’, der Bewerber<br />
‘applicant’ tend to be interpreted by speakers as predominantly referring to ‘man’.<br />
This was shown in an experiment conducted by Braun, Sczesny, and Stahlberg<br />
(2005). To avoid a choice between plural forms like Studenten ‘masculine students’<br />
and Studentinnen ‘female students’, ‘feminized’ generics were introduced—similar in<br />
form to feminine plurals but with <strong>the</strong> capital letter I, e.g. StudentInnen, subsuming<br />
male and female students. Feminized generics with capital letter I (visible only in<br />
written language) were perceived by speakers as being more inclusive of women than<br />
male generics. O<strong>the</strong>r studies yielded different results. Nissen (2013) shows that
114 7 Manly women and womanly men<br />
generic masculine forms in Spanish did show a sexist bias in 1995, but not in 2005.<br />
During <strong>the</strong> span of ten years, <strong>the</strong> visibility of women in social life has increased<br />
significantly—and as a consequence, generic terms have grown to be more inclusive<br />
of women. The interpretation of generics can be linked to recent social changes (we<br />
return to this in Chapter 11). The generic masculine seen as an inherently ‘sexist’<br />
reflection of Social <strong>Gender</strong> stereotypes makes Linguistic <strong>Gender</strong> reference a phenomenon<br />
ripe for linguistic reforms. But do languages with a generic feminine ‘she’<br />
and <strong>the</strong> feminine Linguistic <strong>Gender</strong> as an unmarked choice reflect a better world?<br />
7.7 Markedness, status, and power in Linguistic <strong>Gender</strong> choice<br />
Speakers—and linguists—of many European languages are acutely aware of a connection<br />
between <strong>the</strong> usage of form with male Natural <strong>Gender</strong> reference, and male<br />
dominance—viewed as ‘sexist’ language. The linguistic phenomenon—functional<br />
unmarkedness of masculine Linguistic <strong>Gender</strong> in its many forms—is associated<br />
with male dominance in <strong>the</strong> sphere of Social <strong>Gender</strong>: we can recall, from §2.3, that<br />
<strong>the</strong> masculine, or <strong>the</strong> feminine, gender can be an unmarked choice. This is <strong>the</strong> case<br />
in many Indo-European and Semitic languages—so much so that all speakers of<br />
Modern Hebrew have no qualms about using second singular masculine pronoun to<br />
refer to people in general (men and women—as we will see in §11.5.2). At <strong>the</strong> same<br />
time, language reforms aimed at restoring <strong>the</strong> balance between <strong>the</strong> use of masculine<br />
and feminine forms reflect <strong>the</strong> intertwined societal perception of Natural <strong>Gender</strong>,<br />
Linguistic <strong>Gender</strong>, and Social <strong>Gender</strong> (we return to this in §11.5). In o<strong>the</strong>r words,<br />
lack of markedness—or having <strong>the</strong> masculine form as a default choice—is viewed as<br />
<strong>the</strong> epitome of Social <strong>Gender</strong> status, and power.<br />
In a few languages of <strong>the</strong> world, <strong>the</strong> feminine Linguistic <strong>Gender</strong> is functionally<br />
unmarked. That is, if I don’t know whe<strong>the</strong>r a man or a woman is coming, I will use<br />
feminine Linguistic <strong>Gender</strong> to refer to <strong>the</strong> ‘unknown’. Feminine Linguistic <strong>Gender</strong> is<br />
<strong>the</strong> unmarked choice in Manambu anaphoric reference and agreement (as we saw in<br />
§2.3). But in <strong>the</strong> lexicon things are different. The noun du ‘man’ is used as a generic<br />
term for ‘human beings’. Some of <strong>the</strong> spirits of <strong>the</strong> jungle are human-like; and so are<br />
<strong>the</strong> spirits of <strong>the</strong> deceased people. When telling stories about <strong>the</strong>m, speakers stress that<br />
<strong>the</strong>y are humans, by saying du-adi (man-3plural), ‘<strong>the</strong>y are people’. The expression<br />
du kui (man meat) ‘human flesh’—which reflects earlier cannibalistic practices—is<br />
ano<strong>the</strong>r instance of a ‘generic’ man. The noun du ‘man’ can also mean ‘human body’.<br />
The word for woman is not used this way. 30 We can recall that valuable and important<br />
objects belong to <strong>the</strong> masculine Linguistic <strong>Gender</strong>—which is congruent with <strong>the</strong><br />
Manambu focus on male-oriented cults (§3.4.2).<br />
In Jarawara, masculine is <strong>the</strong> functionally marked gender (see Dixon 2004: 186–7,<br />
for discussion). It is used for reference to human males (and o<strong>the</strong>r referents assigned<br />
to masculine Linguistic <strong>Gender</strong>). The feminine Linguistic <strong>Gender</strong> is <strong>the</strong> default
7.7 Markedness, status, and power 115<br />
choice (as we can recall from §2.3). All personal pronouns require feminine agreement<br />
forms—R. M. W. Dixon reports that he found it ra<strong>the</strong>r difficult to refer to<br />
himself ‘as if ’ he were a woman. The functionally unmarked choice—feminine, or<br />
‘non-masculine’—is used for reference to human females (and referents belonging to<br />
<strong>the</strong> feminine Linguistic <strong>Gender</strong>), and also when <strong>the</strong>re is no gender specification. But<br />
this does not imply that women are higher in status than men: as we recall from §7.2,<br />
an important woman can be referred to with masculine gender—as if she were<br />
promoted to <strong>the</strong> status of an honorary ‘male’. There is no reason to believe that<br />
having feminine Linguistic <strong>Gender</strong> as a default choice automatically accords <strong>the</strong>m a<br />
privileged status in <strong>the</strong> society.<br />
In Mohawk and o<strong>the</strong>r North Iroquoian languages, <strong>the</strong> same Linguistic <strong>Gender</strong><br />
category and forms are used for women and generic reference to humans (as we saw<br />
in §7.3). As Chafe (1967: 13–16) puts it, for Seneca:<br />
<strong>the</strong> masculine [third person subject prefix] morpheme . . . denotes a male human being. The<br />
feminine morpheme . . . ei<strong>the</strong>r denotes a female human being or is an indefinite reference to<br />
people in general, translatable as people, <strong>the</strong>y, one. The neuter morpheme . . . denotes ei<strong>the</strong>r an<br />
animal or something inanimate.<br />
<strong>How</strong> does <strong>the</strong> ‘distinctiveness’ of <strong>the</strong> masculine gender forms, as <strong>the</strong> functionally<br />
and formally marked choice in Nor<strong>the</strong>rn Iroquoian languages, correlate with <strong>the</strong><br />
cultural patterns and <strong>the</strong> Social <strong>Gender</strong> status of women? 31 Sources on traditional<br />
practices in Nor<strong>the</strong>rn Iroquoian societies (including <strong>the</strong> Huron, <strong>the</strong> Seneca, and also<br />
<strong>the</strong> Onondaga) suggest that<br />
sex roles were distributed in Iroquois society in such a way that men were conspicuous, often<br />
even flamboyant, and invested with decision-making powers, whereas women stayed in <strong>the</strong><br />
background, a position from which <strong>the</strong>y never<strong>the</strong>less exerted considerable influence on what<br />
men did. Women were nei<strong>the</strong>r unimportant nor undervalued. On <strong>the</strong> contrary, <strong>the</strong>y were<br />
responsible for keeping life going, both from day to day and from generation to generation.<br />
The importance of women in Iroquoian culture has been emphasized by <strong>the</strong> anthropologist<br />
Cara Richards, who went so far as to exclaim, ‘If you must be born a woman, try to be an<br />
Onondaga’.<br />
According to Richards (1974: 401), <strong>the</strong> ‘relatively high status of Iroquois women’<br />
was reflected in <strong>the</strong> matrilineal descent, and also <strong>the</strong> fact that land belonged to women,<br />
and women were <strong>the</strong> ones who appointed <strong>the</strong> chiefs. The men ‘stood out as highly<br />
visible figures against this essentially female background’. This matches <strong>the</strong> principles<br />
of Linguistic <strong>Gender</strong> marking across Nor<strong>the</strong>rn Iroquoian languages—with <strong>the</strong>ir special<br />
marking for masculine gender and ‘its association of women with undifferentiated<br />
people in general’ (Chafe 2004: 106). This is echoed by Mithun (2014: 137):<br />
The use of <strong>the</strong> same forms for female persons, generics, and indefinites suggests a link between<br />
grammar and culture. Iroquoian culture is characterized by longstanding matrilineal and
116 7 Manly women and womanly men<br />
matrilocal traditions. Clan membership is inherited through <strong>the</strong> mo<strong>the</strong>r. Land was traditionally<br />
passed down from mo<strong>the</strong>rs to daughters. Women have always been highly respected, and <strong>the</strong>y<br />
serve important community roles. Chiefs were traditionally men, but <strong>the</strong>y were appointed (and<br />
potentially recalled) by <strong>the</strong> clan mo<strong>the</strong>rs. It is perhaps no surprise that generic persons should be<br />
represented by Feminine gender forms.<br />
But in actual fact, ‘<strong>the</strong> associations between language and culture suggested by use<br />
of one category of women and generics on <strong>the</strong> one hand, and a separate category for<br />
some women and animals on <strong>the</strong> o<strong>the</strong>r, are not as direct as might be assumed’<br />
(Mithun 2014: 140; see also note 16 in Chapter 6 for <strong>the</strong>ir history). That <strong>the</strong> same<br />
gender category happens to be used ‘for women and generic human beings’ is <strong>the</strong><br />
result of historical development, from <strong>the</strong> use ‘of original indefinite category to<br />
certain women as a sign of respect’ (p. 159).<br />
A challenge to a straightforward correlation between Linguistic <strong>Gender</strong> markedness<br />
and women’s societal status comes from fur<strong>the</strong>r quarters. Social <strong>Gender</strong> roles in<br />
many indigenous societies in North America can be described in terms similar to<br />
Nor<strong>the</strong>rn Iroquoian, with men having high ‘visibility’. <strong>How</strong>ever, far from all of <strong>the</strong>m<br />
have developed feminine as <strong>the</strong> unmarked Linguistic <strong>Gender</strong> (see Chafe 2004:<br />
106–7). A culture pattern may at some point in time have motivated <strong>the</strong> linguistic<br />
pattern; however, <strong>the</strong> exact causal relationship is hard to argue.<br />
Correlations between functional markedness of Linguistic <strong>Gender</strong> and Social<br />
<strong>Gender</strong> roles can be interpreted in two, mutually exclusive ways:<br />
1. Having masculine gender forms as a functionally ‘unmarked’ category can be<br />
understood as a token of ‘male dominance’ (and fought against). This is often<br />
assumed to be <strong>the</strong> case in Modern English, and is reflected in <strong>the</strong> contemporary<br />
debate against <strong>the</strong> use of terms with masculine referents as generic ones—see<br />
§11.5.<br />
2. Alternatively, having masculine gender as a ‘special’, marked one may be<br />
understood as a token of special importance and particular ‘visibility’ of<br />
males in cultural practices. This appears to be <strong>the</strong> case in Jarawara, and in <strong>the</strong><br />
Nor<strong>the</strong>rn Iroquoian examples.<br />
As Mithun (2014: 159) puts it, <strong>the</strong>re is no doubt that ‘relations do exist between<br />
grammatical gender categories and culture’. But <strong>the</strong>y are rarely straightforward and<br />
simplistic. Cultural norms and <strong>the</strong> Social <strong>Gender</strong> status cannot be directly correlated<br />
with, or inferred from, <strong>the</strong> linguistic form and markedness relations in Linguistic<br />
<strong>Gender</strong>. <strong>How</strong>ever, <strong>the</strong> choice of Linguistic <strong>Gender</strong> may reflect stereotypes and<br />
expectations associated with Social <strong>Gender</strong> and with Natural <strong>Gender</strong>. This is especially<br />
salient for humans with <strong>the</strong>ir defined social roles, and particularly so in<br />
languages whose speakers are aware of <strong>the</strong> meanings of genders.
7.7 Markedness, status, and power 117<br />
NOTES AND SOURCES<br />
1. Based on my notes 2004. Laycock (1965: 133) wrote down <strong>the</strong> first person forms correctly;<br />
he erroneously interpreted second feminine form as having third feminine reference.<br />
2. Rosenhouse and Dbayyat (2006: 174–5); see <strong>the</strong>re for fur<strong>the</strong>r details and additional<br />
explanation of <strong>the</strong> gender switch.<br />
3. Hoben (1976: 287–8) tells a similar story: ‘Once in a while in <strong>the</strong> countryside, parents call a<br />
son anci (you feminine) or a daughter antɛ (you masculine—A.Y.A.). The usual explanation<br />
is that earlier children of <strong>the</strong> couple died in infancy. The pronoun switch is made in<br />
an effort to conceal <strong>the</strong> true identity of this child from <strong>the</strong> evil forces that attacked <strong>the</strong><br />
o<strong>the</strong>r babies. This is not a case of expressive variation: it is ra<strong>the</strong>r one of deception.’ See<br />
also Repp (1996: 49), for comparable practices in Arabic.<br />
4. See Sa'ar (2007: 424–5); <strong>the</strong> same principle applies in Palestinian Arabic.<br />
5. See Aikhenvald (2008: 618–23), on loss of cultural knowledge among <strong>the</strong> Manambu.<br />
6. Cf. also Houseman and Severi (1998). The Manambu used to have a reduced version of<br />
Naven (Harrison 1990), now all but forgotten.<br />
7. We can recall, from §4.1, thatinDyirbalyara ‘man’ can be used with <strong>the</strong> feminine class<br />
marker, instead of masculine, to point out <strong>the</strong> female characteristics of a hermaphrodite<br />
(see Dixon 1972: 306–12; 1982: 178–83). I am not aware of such examples in Manambu:<br />
<strong>the</strong>re do not appear to be any hermaphrodites, nor homosexuals, in any of <strong>the</strong><br />
settlements.<br />
8. See Taïfi (1991: 53–4), Oussikoum (2013: 820–1). I am grateful to Maarten Kossmann for<br />
<strong>the</strong>se references, and his comments on <strong>the</strong> meanings of derivational gender across Berber<br />
languages.<br />
9. See Pet (1987: 26–7). We do not know enough about <strong>the</strong> traditional Lokono society to<br />
establish any association with <strong>the</strong> Social <strong>Gender</strong> patterns.<br />
10. See Benamara (2013: 411). I am grateful to Maarten Kossmann for making me aware of<br />
this, and o<strong>the</strong>r references on Berber.<br />
11. Using a feminine form in Amharic may also have an opposite function: men may use <strong>the</strong><br />
feminine address form towards men whom <strong>the</strong>y consider high in status (Wołk 2009:<br />
131–2).<br />
12. See Ferguson (1964: 106, 109) on affectionate Linguistic <strong>Gender</strong> reversals in Arabic and<br />
Marathi. Similar instances of gender switch in Egyptian Arabic are described by Woidich<br />
(2005).<br />
13. See Yokoyama (1999: 423, 427). Based on Zemskaya, Kitajgorodskaja, and Rozanova<br />
(1993), Yokoyama (1999: 423) reports that masculine forms used by both sexes to refer<br />
to women carry affectionate connotations. Diminutives of female names may take masculine<br />
endings in both Russian and Polish (Kasia Wojtylak, p.c.), with specially endearing<br />
overtones. So, my Russian friend Masha can be lovingly addressed as Mashik (Masha<br />
+diminutive.masculine.singular); <strong>the</strong> resulting form may trigger masculine ra<strong>the</strong>r than<br />
feminine agreement. Both Milyj Mashik (dear+masc.sg Mashik) and Milaya Mashik (dear<br />
+fem.sg Mashik) are possible (see also Doleschal and Schmid 2001: 265, and references<br />
<strong>the</strong>re). Rothstein (1973: 464) comments on ‘<strong>the</strong> expressive use of nouns of feminine<br />
gender, especially hostile epi<strong>the</strong>ts, to apply to male human beings. Thus, it is more effective<br />
to call a man dura “idiot”, <strong>the</strong> form with feminine gender, than durak (a corresponding<br />
form of masculine gender)’ (a similar example is in Doleschal and Schmid 2001: 265). This
118 7 Manly women and womanly men<br />
example is attributed to Roman Jakobson and is ‘often cited in <strong>the</strong> oral tradition of<br />
Jakobson’s students’ (Yokoyama 1999: 423–7).<br />
14. See Mladenova (2001: 37–9) on what she refers to as ‘value-laden neuters’ in Bulgarian and<br />
also Greek, and <strong>the</strong>ir use in nineteenth-century literature and modern languages.<br />
15. Iroquoianists call it ‘Feminine-zoic’: Mithun (1999: 100; 2014: 137).<br />
16. Chafe (1977: 514) adds to that: ‘<strong>the</strong> situation in Onondaga is similar, though speakers of<br />
<strong>the</strong> language have emphasized to me that ka-/wa- is not derogatory, but only detached and<br />
impersonal’.<br />
17. See Chafe (1977), Mithun (2014: 138, 141–3), and §4.1. A somewhat similar example of<br />
Linguistic <strong>Gender</strong> reversal as a mark of politeness was described by Khaidakov (1963:<br />
49–50) for Lak.<br />
18. Changing genders may correlate with <strong>the</strong> participant’s agency: we can recall, from §4.1,<br />
how an inanimate Comb is treated as ‘animate’ as soon as it starts speaking and acting on<br />
its own, in stories in Fox and o<strong>the</strong>r Algonquian languages (see Goddard 2002). Polish has<br />
three genders (masculine, feminine, and neuter). Masculine nouns fur<strong>the</strong>r divide into<br />
animate/inanimate (in <strong>the</strong> singular) and personal/non-personal (in plural). Non-personal<br />
forms of personal nouns can be used for a pejorative effect, as a way of downgrading a<br />
human to a non-human status; occasionally <strong>the</strong>y may have positive overtones (see<br />
Rothstein 1973: 697).<br />
19. See Wales (1996: 143, 146–52); <strong>the</strong> example has been shortened. Similarly, speakers of<br />
American English may use he to refer to animals <strong>the</strong>y have empathy for (Morris 1999: 188).<br />
20. See Baron (1986: 94); also see Yaguello (1979), for a similar approach to French.<br />
Svartengren (1927 and 1954: 263–83) provides numerous examples, in American English,<br />
of ‘she’ applied to inanimates in English (including tools, hollow objects, bridges, and<br />
man-made tools and machinery, clothing, small objects, furniture, water, seasons, and a<br />
few more).<br />
21. See Mathiot (1979b: 18–19). See also Clamons (1993: 277) and §3.1 on Oromo.<br />
22. See Mathiot (1979b: 12–29).<br />
23. See Baron (1986: 117–20) on <strong>the</strong> history of <strong>the</strong> word; Dixon (2014: 323–4) on <strong>the</strong> history<br />
of <strong>the</strong> suffix -ster and its negative overtones in o<strong>the</strong>r formations, such as gang-ster and<br />
mob-ster.<br />
24. See p. 163 of Byron’s letters and journals, ed. Richard Lansdown (Oxford: Oxford University<br />
Press, 2015).<br />
25. This is contrary to Connors’s(1971: 598) claim that <strong>the</strong> terms for female occupations have<br />
no derogatory or facetious connotations, thanks to <strong>the</strong> egalitarian structure of Soviet<br />
society and equal rights for men and women. See Yokoyama (1999: 420, 427), Zemskaya,<br />
Kitajgorodskaja, and Rozanova (1993).<br />
26. Yuri A. Aikhenvald, p.c.<br />
27. Law N o 12.605, 3 April 2012: . This law reinforces a similar Federal law, of 2 April 1956 (Law 2.749<br />
proposed by <strong>the</strong> Senator Mozart Lago and sanctioned by Juscelino Kubitschek, <strong>the</strong> <strong>the</strong>n<br />
president of Brazil), to <strong>the</strong> same effect. The term presidenta is morphologically unusual in<br />
<strong>the</strong> language. The majority of agentive nouns in -ente, including agente ‘agent’, cliente<br />
‘client’, and paciente ‘patent’, in Portuguese do not have a special feminine form (*clienta,<br />
*agenta, *pacienta are ungrammatical). Just a few do: pairs such as sargente ‘sergeant’ and<br />
sargenta ‘woman sargent; wife of a sergeant’ and parente ‘male relative, relative in general’<br />
and parenta ‘female relative’ are in general use. The ‘legalization’ of <strong>the</strong> term presidenta—<br />
whose referent is for now uniquely identifiable—may result in <strong>the</strong> expansion of productivity
7.7 Markedness, status, and power 119<br />
of <strong>the</strong> feminine -enta (see, for instance, <strong>the</strong> discussion at ).<br />
28. Sexism in English has been discussed at length: a useful summary is in Mills (2008: 145–7);<br />
see also Hellinger and Pauwels (2007), McConnell-Ginet (2003), Eckert and McConnell-<br />
Ginet (2003, 2006), and Coates (1993).<br />
29. See Brauer and Landry (2008), for similar results in French.<br />
30. See Aikhenvald (2015a) on subtle differences in grammatical behaviour between du ‘man,<br />
human being’ and du ‘human body’.<br />
31. See Chafe (2004: 99–104) for fur<strong>the</strong>r discussion.
8<br />
The images of gender<br />
The meanings of Linguistic <strong>Gender</strong>s mirror myths and beliefs. Linguistic <strong>Gender</strong>s are<br />
a rich source of poetic metaphors. What physical and o<strong>the</strong>r properties do Linguistic<br />
<strong>Gender</strong>s correlate with? What role do Linguistic <strong>Gender</strong>s play in human cognition,<br />
and in <strong>the</strong> creation of mental images and representations? Do Linguistic <strong>Gender</strong>s<br />
affect <strong>the</strong> ways in which speakers perceive and construe <strong>the</strong> world around <strong>the</strong>m?<br />
These are <strong>the</strong> questions we address in this chapter.<br />
8.1 Myth-and-belief in <strong>the</strong> choice of Linguistic <strong>Gender</strong><br />
As Dixon (2015: 29) puts it, ‘<strong>the</strong> tribes’ repertoire of legends and beliefs, and <strong>the</strong>ir<br />
general perception of <strong>the</strong> world also help explain <strong>the</strong> ways in which genders are<br />
chosen’. Linguistic <strong>Gender</strong> of an object or an animal can correlate with its role in<br />
myth. We can recall, from §3.1, that ‘moon’ in Manambu is a mythical woman, and is<br />
always assigned to <strong>the</strong> feminine gender (no matter what its size or brightness). In<br />
Alamblak, yam-t ‘moon’ is feminine (and bears <strong>the</strong> feminine suffix -t). In myths, <strong>the</strong><br />
moon is a woman, and <strong>the</strong> sun (mar-r) is her son. The ‘sun’ is masculine and bears<br />
<strong>the</strong> masculine suffix -r (despite its round shape typical for feminines; Bruce 1984: 97).<br />
Similarly, in Tunica, thunder and lightning were personified as mythical men, and<br />
were masculine for this reason. 1 In Dyirbal, <strong>the</strong> sun was believed to be a woman, and<br />
thus belongs to gender 2 which includes females (balan garri). The moon is her<br />
husband, and so <strong>the</strong> noun belongs to gender 1 which includes males (bayi gagara).<br />
Thunder, lightning, and rain are believed to be legendary males, and are also included<br />
in gender 1.<br />
In many Australian languages, <strong>the</strong> moon is conceived of as a man and <strong>the</strong> sun<br />
as a woman. 2 In many languages in New Guinea, <strong>the</strong> moon is believed to be a<br />
mythological man who engages in sexual intercourse with women. The moon is<br />
responsible for women’s menstruation: it is said to inflict ‘moon sickness’ upon<br />
women. Coincidentally, in o<strong>the</strong>r societies, e.g. among <strong>the</strong> Vaupés Indians of northwest<br />
Amazonia, Moon is also held responsible for female menstruation. But, since<br />
Moon is a mythical male (referred to with non-feminine Linguistic <strong>Gender</strong>), it is said<br />
to have sexual intercourse with women when <strong>the</strong>y menstruate (Aikhenvald 2012b,<br />
<strong>How</strong> <strong>Gender</strong> <strong>Shapes</strong> <strong>the</strong> <strong>World</strong>. First edition. Alexandra Y. Aikhenvald.<br />
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.
8.2 The metaphors of Linguistic <strong>Gender</strong> 121<br />
and my own fieldwork). In Manambu and many o<strong>the</strong>r languages of <strong>the</strong> Sepik area,<br />
‘moon’ and all stars belong to <strong>the</strong> feminine gender (we recall, from §3.1, that<br />
feminine gender is associated with round shape, which is also <strong>the</strong> shape of <strong>the</strong><br />
moon). An alternative way of referring to <strong>the</strong> moon is bapa-ta:kw (moon-woman),<br />
literally ‘The Lady moon’. 3<br />
In Asheninca and Ashaninka Campa (Arawak languages from Peru), all inanimates<br />
belong to <strong>the</strong> feminine gender. All animates are treated as masculine. But <strong>the</strong><br />
Moon, cashiri, is masculine: this is a mythical man (and so is his son, <strong>the</strong> Sun, who is<br />
also a male). 4 Similarly, in Palikur, heavenly bodies (sun, moon, stars, planets),<br />
thunder, and lightning belong to masculine gender, because according to traditional<br />
legends <strong>the</strong>y were once men.<br />
Linguistic <strong>Gender</strong> assignment of birds can also be determined by <strong>the</strong>ir role in<br />
myth. The traditional Dyirbal people believed that birds are, as a class, <strong>the</strong> spirits of<br />
dead women. Birds are classed as members of gender 2 (balan) which includes<br />
females. One snake, balan bima ‘death adder’, belongs to gender 2, as it is also a<br />
legendary woman. (More examples were given in §2.2.1.)<br />
In numerous languages of <strong>the</strong> Sepik region of New Guinea, <strong>the</strong> cassowary (a large<br />
non-flying bird) is feminine. In many Sepik cultures, including Manambu, <strong>the</strong><br />
cassowary is a totemic woman. In an important Abau creation story, a cassowary<br />
gives birth to a human child, and is thus an ‘honorary female’. Along similar lines,<br />
‘cassowary’ in Dyirbal is feminine: it appears in <strong>the</strong> shape of a woman in myths<br />
(R. M. W. Dixon, p.c.).<br />
Conversely, <strong>the</strong> Linguistic <strong>Gender</strong> of a noun may be projected onto beliefs,<br />
superstitions, and gender imagery.<br />
8.2 The metaphors of Linguistic <strong>Gender</strong><br />
Linguistic <strong>Gender</strong>s influence beliefs, superstitions, and poetic metaphors revolved<br />
around an inanimate referent, which is <strong>the</strong>n promoted to <strong>the</strong> status of an ‘honorary’<br />
human, or ‘anthropomorphized’. Features of Natural <strong>Gender</strong> (male and female) are<br />
<strong>the</strong>n projected onto <strong>the</strong> items in accordance with <strong>the</strong>ir Linguistic <strong>Gender</strong>s, endowing<br />
<strong>the</strong>m with features typical of Social <strong>Gender</strong>s.<br />
As Jakobson (1959: 117) pointed out, grammatical gender<br />
plays a great role in <strong>the</strong> mythological attitudes of a speech community. In Russian <strong>the</strong> feminine<br />
cannot designate a male person, nor <strong>the</strong> masculine specify a female. Ways of personifying<br />
or metaphorically interpreting inanimate nouns are prompted by <strong>the</strong>ir gender. A test in <strong>the</strong><br />
Moscow Psychological Institute (1915) showed that Russians, prone to personify <strong>the</strong> weekdays,<br />
consistently represented Monday, Tuesday, and Thursday as males and Wednesday, Friday,<br />
and Saturday as females, without realizing that this distribution was due to <strong>the</strong> masculine<br />
gender of <strong>the</strong> first three names as against <strong>the</strong> feminine gender of <strong>the</strong> o<strong>the</strong>rs. ...
122 8 The images of gender<br />
The overtones of Natural <strong>Gender</strong> and associated meanings of Social <strong>Gender</strong> are<br />
often imputed to non-human referents depending on <strong>the</strong>ir Linguistic <strong>Gender</strong>, as a<br />
kind of ‘gender metaphor’. As McConnell-Ginet (1979: 77) puts it, ‘poets and<br />
painters convey messages by incarnating abstract concepts’. In both Russian and<br />
Portuguese, <strong>the</strong> noun ‘death’ belongs to <strong>the</strong> feminine Linguistic <strong>Gender</strong>. Jakobson<br />
continues:<br />
The Russian painter Repin was baffled as to why Sin had been depicted as a woman by German<br />
artists: he did not realize that ‘Sin’ is feminine in German (die Sünde), but masculine in<br />
Russian (greh). Likewise a Russian child, while reading a translation of German tales, was<br />
astounded to find that Death, obviously a woman (Russian smert', fem.), was pictured as an old<br />
man (German der Tod, masc.).<br />
In Portuguese, as in Russian, morte ‘death’ belongs to <strong>the</strong> feminine Linguistic<br />
<strong>Gender</strong> and is often respectfully referred to as Dona Morte ‘Ms Death’. The identification<br />
of śmierć ‘death’ with ‘woman’ in Polish culture and tradition goes back a<br />
long way: a fifteenth-century Polish poem, Rozmowa Mistrza Polikarpa ze śmercią<br />
‘Master Polycarpus’ Colloquy with death’, explicitly states that Master Polycarpus<br />
‘saw a naked person of <strong>the</strong> female sex’ (‘uzrzał człowieka nagiego, przyrodzenia<br />
niewieściego’) (Studzińska 2012). In his 1894 picture, <strong>the</strong> German painter Franz<br />
von Stuck depicted <strong>the</strong> gruesome war as a man—in German der Krieg ‘war’ is<br />
masculine. In Polish wojna ‘war’ belongs to <strong>the</strong> feminine Linguistic <strong>Gender</strong>. The<br />
cycle of pictures ‘War’ (Wojna, feminine Linguistic <strong>Gender</strong>), by <strong>the</strong> Polish painter<br />
Arthur Grottger, contains female images—so much so that Baudouin de Courtenay<br />
described <strong>the</strong>m as representing <strong>the</strong> war as a ‘woman’ (de Courtenay 1929: 246). None<br />
of this makes much sense to an English speaker, and even less so to a speaker of<br />
Estonian or Hungarian, with no Linguistic <strong>Gender</strong>.<br />
Linguistic <strong>Gender</strong> of inanimate objects can be deployed in popular beliefs. As a child<br />
in <strong>the</strong> 1960s I used to spend summer holidays with my grandmo<strong>the</strong>r in a ra<strong>the</strong>r remote<br />
place, away from our native Moscow. Russian has three genders: masculine ‘he’,<br />
feminine ‘she’, and neuter ‘it’. Every noun is assigned to a gender. If I drop a spoon<br />
or a fork, I will be expecting a female visitor: lozhka ‘spoon’ and vilka ‘fork’ are<br />
feminine. And if I drop a knife, this implies a male visitor: nozh ‘knife’ is masculine.<br />
We had no telephone, and <strong>the</strong> post office was miles away. The belief seemed to work<br />
quite well—and was of some help to us as to whom we might expect for lunch that day.<br />
In Portuguese of <strong>the</strong> north-east of Brazil, if one drops a spoon colher (feminine), a<br />
woman (mulher, a word which rhymes with colher and is thus reminiscent of it) will<br />
come; and if one drops a fork garfo (‘masculine’), one expects a man (Eduardo<br />
Ribeiro, p.c.). It goes without saying that such correlations and superstitions are hard<br />
to explain to speakers of o<strong>the</strong>r languages. In German, spoon is masculine (der Löffel),<br />
fork is feminine (die Gabel), and knife is neuter (das Messer)—so, when I tell my<br />
German friends about <strong>the</strong> Russian superstition, <strong>the</strong>y just shrug <strong>the</strong>ir shoulders at
8.2 The metaphors of Linguistic <strong>Gender</strong> 123<br />
what looks to <strong>the</strong>m like a ‘barbarian’ custom. For those who speak Estonian—a<br />
language with no genders, and just one word for ‘he’, ‘she’, and ‘it’—this custom<br />
makes even less sense.<br />
<strong>Gender</strong> metaphors and personification make <strong>the</strong> task of a translator ra<strong>the</strong>r difficult.<br />
Again, in Jakobson’s words, My sister Life, <strong>the</strong> title of a book of poems by Boris<br />
Pasternak, is quite natural in Russian, where life is feminine (zhiznj), but was enough<br />
to reduce to despair <strong>the</strong> Czech poet Josef Hora in his attempt to translate <strong>the</strong>se<br />
poems, since in Czech <strong>the</strong> noun zivot ‘life’ is masculine. 5<br />
Linguistic <strong>Gender</strong> choice may be deployed as a poetic device. Then, <strong>the</strong> Linguistic<br />
<strong>Gender</strong>, and <strong>the</strong> imputed Natural and Social <strong>Gender</strong>, of <strong>the</strong> referent form one<br />
conceptual whole. We can recall, from Chapter 2, that <strong>the</strong> assignment of Linguistic<br />
<strong>Gender</strong> to German nouns is subject to a set of fairly complex morphological and<br />
semantic principles (Zubin and Köpcke 1986). Male and female adults of each species<br />
of domestic and game animals belong to masculine and feminine Linguistic <strong>Gender</strong>s.<br />
Types of cloth, precipitation, and minerals belong to masculine Linguistic <strong>Gender</strong>. In<br />
many instances, a generic term belongs to one gender, and specific terms to ano<strong>the</strong>r.<br />
The generic term for ‘tree’, Baum, is masculine (and so are all tree names containing<br />
-baum as <strong>the</strong>ir part). Specific names for trees (which do not contain -baum) are<br />
feminine, e.g. Palme ‘palm tree’, Linde ‘linden’, Weide ‘willow’, Pappel ‘poplar’. This<br />
linguistic fact has been deployed in poetry, with trees being personified through<br />
interpreting <strong>the</strong>ir Linguistic <strong>Gender</strong> as <strong>the</strong>ir Natural <strong>Gender</strong>, and <strong>the</strong>n endowing<br />
<strong>the</strong>m with feelings and behaviour typical for Social males and Social females.<br />
The famous poem by Heinrich Heine Fichtenbaum ‘Pine-tree’ reads as follows<br />
(Linguistic <strong>Gender</strong>-sensitive forms are in bold type, and glossed in brackets after each<br />
word):<br />
8.1 Ein Fichtenbaum (one:masculine pine-tree) steht einsam<br />
Im hohen Norden auf kahler Höh.<br />
Ihn (to him) schläfert, mit weisser Decke<br />
Umhüllen ihn (him) Eis und Schnee.<br />
Er (he) träumt von einer Palme (one:feminine palm tree),<br />
Die (which:feminine), fern im Morgenland,<br />
Einsam und schweigend trauert<br />
Auf brennender Felsenwand.<br />
The noun ‘pine-tree’ belongs to <strong>the</strong> masculine Linguistic <strong>Gender</strong>, and <strong>the</strong> word for<br />
‘palm-tree’ belongs to <strong>the</strong> feminine Linguistic <strong>Gender</strong>. There are obvious romantic<br />
connotations to <strong>the</strong>se Linguistic <strong>Gender</strong>s—a lonely pine-tree-man is longing for a<br />
lonely and silent palm-tree-woman. This play is completely lost in <strong>the</strong> English<br />
translation, in 8.2: English has no means for <strong>the</strong> same Linguistic <strong>Gender</strong> play. Both<br />
trees (in caps) are referred to as ‘it’ (in a well-known translation by James Thompson<br />
(1834–88)):
124 8 The images of gender<br />
8.2 A PINE-TREE standeth lonely<br />
In <strong>the</strong> North on an upland bare.<br />
It standeth whitely shrouded<br />
With snow, and sleepeth <strong>the</strong>re.<br />
It dreameth of a PALM TREE<br />
Which far in <strong>the</strong> East alone,<br />
In mournful silence standeth<br />
On its ridge of burning stone.<br />
The gender-play effect is equally lost in translations of this same poem into<br />
o<strong>the</strong>r languages. In Russian, sosna ‘pine-tree’ belongs to <strong>the</strong> feminine gender, and<br />
so does paljma ‘palm-tree’. The romantic connotations of <strong>the</strong> original poem are<br />
lost in its most famous Russian translation, by Mikhail Lermontov. The French<br />
translation of <strong>the</strong> same poem, by Amélie Ernst (1873), involves le sapin ‘fir-tree’<br />
and le palmier ‘palm-tree’. Bothbelongto<strong>the</strong>masculineLinguisticgender.The<br />
feeling of loneliness and mutual yearning is <strong>the</strong>re—but <strong>the</strong> subtle romantic innuendo<br />
is lost.<br />
Not that it cannot be captured. Emma Lazarus (1849–87), an American poet,<br />
translated Heine’s poem into English, reflecting <strong>the</strong> German Linguistic <strong>Gender</strong>s<br />
through English anaphoric genders (in bold):<br />
8.3 There stands a lonely pine-tree<br />
In <strong>the</strong> north, on a barren height;<br />
He sleeps while <strong>the</strong> ice and snow flakes<br />
Swa<strong>the</strong> him in folds of white.<br />
He dreameth of a palm-tree<br />
Far in <strong>the</strong> sunrise land,<br />
Lonely and silent longing<br />
On her burning bank of sand.<br />
As Guy Deutscher (2010: 196) puts it, ‘<strong>the</strong> price Lazarus pays for this faithfulness is<br />
that her translation sounds somewhat arch, or at least artificially poetic, since in<br />
English it is not natural to speak of trees in this way’. In his version of <strong>the</strong> poem,<br />
Fyodor Tjutschev kept <strong>the</strong> romantic overtones by using kedr ‘cedar’, of masculine<br />
Linguistic <strong>Gender</strong> in Russian, instead of <strong>the</strong> feminine sosna, ‘pine-tree’. 6<br />
Die Lotusblume, <strong>the</strong> main character of Heinrich Heine’s poem ‘Lotus flower’,<br />
belongs to <strong>the</strong> feminine Linguistic <strong>Gender</strong> in German. Its lover is <strong>the</strong> masculine<br />
der Mond (<strong>the</strong> moon), in 8.4a. The forms relevant for Linguistic <strong>Gender</strong>s are in bold.<br />
8.4a Die Lotusblume ängstigt<br />
Sich vor der Sonne Pracht,<br />
Und mit gesenktem Haupte<br />
Erwartet sie träumend die Nacht.
8.2 The metaphors of Linguistic <strong>Gender</strong> 125<br />
Der Mond, der ist ihr Buhle,<br />
Er weckt sie mit seinem Licht,<br />
Und ihm entschleiert sie freundlich<br />
Ihr frommes Blumengesicht.<br />
Sie blüht und glüht und leuchtet,<br />
Und starret stumm in die Höh;<br />
Sie duftet und weinet und zittert<br />
Vor Liebe und Liebesweh.<br />
The English translation in 8.4b—by Emma Lazarus (1867)—reflects <strong>the</strong> loving<br />
heterosexual relationship between <strong>the</strong> she-flower and <strong>the</strong> he-moon.<br />
8.4b The Lotus-flower trembles<br />
Before <strong>the</strong> sun’s gold light;<br />
And, with her head low drooping,<br />
Waits, dreamily, <strong>the</strong> night.<br />
The Moon, he is her lover,<br />
He wakes her with his light,<br />
And unto him reveals she<br />
Her flower-face so bright.<br />
She blooms and glows and brightens,<br />
And dumbly looks above;<br />
She weeps and sighs and trembles<br />
With love and <strong>the</strong> woes of love. 7<br />
In his translation of this poem, <strong>the</strong> Russian poet Apollon Maikov had to replace lotus<br />
with lily (since lotus belongs to <strong>the</strong> masculine Linguistic <strong>Gender</strong>, and lily to <strong>the</strong><br />
feminine Linguistic <strong>Gender</strong>). Instead of luna ‘moon’ (feminine Linguistic <strong>Gender</strong> in<br />
Russian) he used mesjac ‘young moon, month’, of masculine Linguistic <strong>Gender</strong>.<br />
A popular Russian song features a thin and vulnerable rowan tree (rjabina,<br />
feminine Linguistic <strong>Gender</strong>), in her loneliness yearning for <strong>the</strong> company of a big<br />
strong oak tree (dub, masculine Linguistic <strong>Gender</strong>) across <strong>the</strong> river. 8 As Fodor (1959:<br />
206) puts it, <strong>the</strong> masculine gender of dub ‘determines <strong>the</strong> content as well as <strong>the</strong> mood<br />
of <strong>the</strong> poem, and constitutes a stylistic device, of which genderless languages would<br />
be incapable, <strong>the</strong> same thought and feeling having to be rendered by different means<br />
or in roundabout ways’. 9<br />
The Linguistic <strong>Gender</strong>s offer access to a fixed range of metaphors: <strong>the</strong> masculine<br />
Linguistic <strong>Gender</strong> evokes a man, and <strong>the</strong> feminine gender evokes a woman. In a<br />
famous song by Bulat Okudzhava, <strong>the</strong> author requests forgiveness from his three<br />
protectresses—sisters and wives at <strong>the</strong> same time—Faith, Hope, and Love. The terms<br />
(Vera, Nadezhda, Ljubovj, which also happen to be used as female names) belong to<br />
<strong>the</strong> feminine Linguistic <strong>Gender</strong>. The language imposes <strong>the</strong> female imagery on <strong>the</strong><br />
poet—and he makes use of this. 10
126 8 The images of gender<br />
Morton Bloomfield (1963: 162) commented on a comparative freedom of gender<br />
personification in English, with just anaphoric gender, as compared to o<strong>the</strong>r languages:<br />
The subject is particularly interesting in English, because a writer of English after about 1300<br />
[presumably, after <strong>the</strong> loss of agreement gender], unlike his German or French counterparts,<br />
had more or less freedom in choosing masculine or feminine gender for his personified figures.<br />
The Germans seem to have been particularly fascinated with this freedom of choice denied<br />
<strong>the</strong>m. The gender of <strong>the</strong>ir personifications, and even <strong>the</strong> very possibility of personification, was<br />
determined by forces outside <strong>the</strong>ir control. One might say that languages with grammatical<br />
gender, unlike English, have automatically built-in personification of some sort.<br />
Do <strong>the</strong>se ‘automatically built-in’ associations affect <strong>the</strong> ways in which speakers<br />
perceive, and categorize, <strong>the</strong> world of inanimates?<br />
8.3 Does Linguistic <strong>Gender</strong> affect cognition?<br />
Connections between <strong>the</strong> Linguistic <strong>Gender</strong> of a noun and a female—or a male—<br />
image behind it sparked <strong>the</strong> interest of psychologists. A few have attempted to test<br />
how pervasive <strong>the</strong> associations are. Konishi (1993) conducted an experiment with<br />
two dozen or so German and Spanish speakers who were asked to rate a number of<br />
inanimate nouns for ‘potency’ (among o<strong>the</strong>r features)—a feature associated with<br />
‘masculinity’, or male Natural <strong>Gender</strong>. Half <strong>the</strong> nouns were of masculine Linguistic<br />
<strong>Gender</strong>, and half were of feminine Linguistic <strong>Gender</strong>. A number of nouns used in <strong>the</strong><br />
experiment have different genders in <strong>the</strong> two languages. For instance, die Brücke<br />
‘bridge’ and die Luft ‘air’ are feminine in German, and <strong>the</strong>ir Spanish counterparts el<br />
puente and el aire are masculine. Der Apfel ‘apple’ in German is masculine, and<br />
Spanish la manzana is feminine. Nouns which are masculine in German (such as<br />
chairs and keys) were rated as more potent by German speakers. Nouns which are<br />
masculine in Spanish (such as bridges and clocks) were rated as more ‘potent’.<br />
Sera, Berge, and del Castillo (1994) asked a group of Spanish and French children<br />
to rate pictures of objects as ‘masculine’ or ‘feminine’. Speakers consistently classified<br />
objects in agreement with <strong>the</strong>ir Linguistic <strong>Gender</strong>: masculine as men, feminine as<br />
women. (The effect was even stronger when pictures were accompanied by Spanish<br />
and French labels.) When asked to attribute a man’s or a woman’s voice to each<br />
picture, Spanish-speaking children assigned voices in accordance with <strong>the</strong> Linguistic<br />
<strong>Gender</strong> of each item. So when French speakers saw a picture of a fork, most of <strong>the</strong>m<br />
attributed a woman’s voice to it: la fourchette in French is feminine. In contrast,<br />
Spanish speakers tended to attribute a male voice to <strong>the</strong> Spanish el tenedor ‘fork’, of<br />
masculine Linguistic <strong>Gender</strong>.<br />
To try and determine whe<strong>the</strong>r Linguistic <strong>Gender</strong> influences speakers’ mental<br />
representation of an object, Boroditsky, Schmidt, and Phillips (2003) developed a<br />
memory game for a group of German and Spanish speakers (<strong>the</strong> game was conducted
8.4 What men and women look like 127<br />
in English, to avoid explicit mention of overt Linguistic <strong>Gender</strong>s in <strong>the</strong>se two languages).<br />
Speakers were taught proper names for twenty-four objects; for instance, an<br />
apple would have been called Patrick, or Patricia). Names for all <strong>the</strong> objects had<br />
opposite Linguistic <strong>Gender</strong> in Spanish and in German. The participants were given a<br />
fixed time to remember <strong>the</strong> names associated with objects, and <strong>the</strong>n tested on <strong>the</strong><br />
results. The results were counted, and <strong>the</strong>re appeared to be a strong bias towards a<br />
correlation between <strong>the</strong> name (male or female) and <strong>the</strong> Linguistic <strong>Gender</strong> of <strong>the</strong><br />
noun: participants were better at remembering <strong>the</strong> male name associated with <strong>the</strong><br />
object of masculine Linguistic <strong>Gender</strong>, and <strong>the</strong> female name of an object of<br />
feminine Linguistic <strong>Gender</strong>. That is, German speakers easily remembered <strong>the</strong><br />
name Patrick for an apple (der Apfel, masculine), and Spanish speakers remembered<br />
<strong>the</strong> apple’s namePatricia (la manzana, feminine). Spanish speakers found it<br />
easier to remember <strong>the</strong> name of a bridge (el puente, masculine) if it was Claudio, but<br />
not if it was Claudia. It appeared that <strong>the</strong> Natural <strong>Gender</strong> assignment of inanimates<br />
in <strong>the</strong> game context was influenced by Linguistic <strong>Gender</strong> of <strong>the</strong> language spoken<br />
by <strong>the</strong> participants.<br />
One should keep in mind that all such experiments are artificial—how many times<br />
in real life do speakers of any language have to bo<strong>the</strong>r with female or male names<br />
assigned to bridges and apples? They may, however, point towards an interesting<br />
direction: that some people may attribute ‘male’ and ‘female’ labels to inanimate<br />
objects in agreement with <strong>the</strong>ir Linguistic <strong>Gender</strong> assignment. This is echoed by<br />
spontaneous remarks by people who speak languages with Linguistic <strong>Gender</strong>s outside<br />
Europe. The Manambu state that, in <strong>the</strong>ir language, everything in <strong>the</strong> world is ei<strong>the</strong>r<br />
male or female—thus drawing a parallel between <strong>the</strong> Linguistic <strong>Gender</strong> of an entity and<br />
its Natural <strong>Gender</strong> associations. This anecdotal evidence points toward a certain<br />
metalinguistic awareness of gender (something we mentioned in Chapters 1 and 4). 11<br />
In many languages and cultures, special physical and psychological properties are<br />
associated with males and females. And can ‘typical’ characteristics of men and<br />
women be transferred to any entity in <strong>the</strong> world, following its Linguistic-cum-<br />
Natural-<strong>Gender</strong> association? This is our next topic.<br />
8.4 What men and women look like<br />
The ways in which Natural <strong>Gender</strong> is attributed to inanimates through words for man<br />
and woman reveal physical features associated with males and females. Ilocano, a<br />
Western Austronesian language (Rubino 1997: 75–6), has no Linguistic <strong>Gender</strong>s. The<br />
Natural <strong>Gender</strong> of humans is distinguished lexically, through nouns lalaki ‘boy’, babai<br />
‘girl’. These lexemes are also used to distinguish Natural <strong>Gender</strong> of hybrid nouns, e.g.<br />
kabsat a lalaki ‘bro<strong>the</strong>r’, kabsat a babai ‘sister’. They can also co-occur with inanimates,<br />
highlighting <strong>the</strong> overtones of ‘strength’ and ‘sweetness’ associated with ‘masculinity’
128 8 The images of gender<br />
and ‘femininity’.So,basi a lalaki ‘strong sugar cane wine’ is associated with masculinity;<br />
and basi a babai ‘sweet sugar cane wine’ with femininity. 12<br />
Men’s and women’s activities and <strong>the</strong>ir descriptions can reflect <strong>the</strong> combined<br />
images of Natural <strong>Gender</strong> and of Social <strong>Gender</strong>. In Nungon, a Papuan language<br />
with no Linguistic <strong>Gender</strong>, <strong>the</strong> notion of nging ‘hot, hurtful, powerful, sharp’ relates<br />
to male work and practices. These include hunting with bows and arrows, building<br />
houses, and thatching roofs. The notion of hum ‘cold, pacific, weak, dull’ refers to<br />
what women do (forage in <strong>the</strong> forest, dig up small animals’ burrows, and perform<br />
domestic duties). Back in <strong>the</strong> old days, men used to have taboos and food abstinences<br />
before hunting, so as not to become hum-o ‘dull, powerless’—or women-like. An<br />
important dimension in masculine or feminine gender choice in Yangoru Boiken is<br />
‘hardness’ (an esteemed male quality in metaphorical terms): ‘ironwood is masculine<br />
because it is “hard”, hard-textured yam varieties are masculine, and softer tuber ones<br />
are female’ (Roscoe 2001, p.c.). The hard natural fibres used to make traditional string<br />
bags in Manambu are called du-a-maej ‘men fibres’. These are jokingly contrasted with<br />
takwa-maej, lit. women fibres, to refer to imported wool which is much less strong. 13<br />
Ket, from <strong>the</strong> Yeniseian family, distinguishes three genders: masculine, feminine,<br />
and neuter. <strong>Gender</strong> choice is only partially predictable from <strong>the</strong> meaning of <strong>the</strong> noun<br />
(see also §2.3). In Vajda’s (2004: 16) words, ‘in general, masculine-class items are<br />
perceived as having “greater vital force” than feminine-class nouns’. 14 A growing tree<br />
belongs to <strong>the</strong> masculine class, and a log or a dry tree is feminine. Some trees (e.g. <strong>the</strong><br />
birch) are feminine if crooked and masculine if straight. The meaning of gender<br />
choice shows correlation with ‘cultural relevance, or economic importance, with<br />
terms for animals and things, which are valued higher, or are viewed as having<br />
more cultural or societal significance than o<strong>the</strong>rs’, more often ending up in <strong>the</strong><br />
masculine than in <strong>the</strong> feminine class (Georg 2007: 88–90). Typical features of<br />
feminine Social <strong>Gender</strong> are projected onto entities which have no Natural <strong>Gender</strong>,<br />
in <strong>the</strong> form of Linguistic <strong>Gender</strong> assignment.<br />
<strong>Gender</strong> assignment to inanimates in Amharic is based on size: in Pankhurst’s(1992:<br />
168) words, <strong>the</strong> feminine being used to suggest a smaller, <strong>the</strong> masculine a larger object.<br />
In addition to this, ‘a sexual distinction is conferred on an inanimate object which<br />
reveals <strong>the</strong> gender construction in a society. Thus male soils, yewend merét, are rich and<br />
fertile, whilst female soils, yesét merét, are light and relatively infertile.’ Masculine<br />
Linguistic <strong>Gender</strong> in Oromo is associated with larger size, power, and importance.<br />
Feminine Linguistic <strong>Gender</strong> has <strong>the</strong> opposite connotations. And nouns denoting<br />
undesirable social characters—e.g. a thief, or vagabond—tend to be assigned to <strong>the</strong><br />
feminine gender. 15 This is reminiscent of our discussion in §7.5 and §11.3.<br />
Physical properties of women in traditional stories may be associated with women’s<br />
place in ritual, and specific female ‘pollution’ and impurity due to menstrual<br />
blood. One of <strong>the</strong> myths of <strong>the</strong> Tariana of north-west Amazonia concerns libertine<br />
ancestral women who had ‘misbehaved’ with a smelly mucura-rat (an Amazonian
8.4 What men and women look like 129<br />
marsupial, of genus Didelphus, with a characteristic pungent smell). The women<br />
used to smell nice. As a consequence of sexual intercourse with <strong>the</strong> animal, <strong>the</strong>y<br />
became smelly (i:-peni). As we will see in Chapter 10, many more things are wrong<br />
with women in <strong>the</strong> Tariana lore.<br />
Throughout <strong>the</strong> world, proverbs reflect <strong>the</strong> ‘traditional wisdom’ of maledominated<br />
societies. Small size and relative insignificance are <strong>the</strong> physical properties<br />
attributed to women. An Oromo proverb says, ‘A wife and a plough are best when<br />
shorter than <strong>the</strong> man’. A Kalmyk proverb comments on women’s lack of intelligence—‘though<br />
a girl’s hair be long, her brain is short’. Women are depicted as<br />
passive objects, containers, animals, consumables, mentally and physically weak, and<br />
generally inferior beings. ‘The hen is no bird, <strong>the</strong> woman is no person’, says an<br />
Estonian proverb. An overwhelming majority of proverbs collected by Schipper<br />
(2003) comment on women from a dominant male perspective, defending male<br />
privileges and interests. In Schipper’s (2003: 374) words, ‘proverbial metaphors<br />
reproduce a gender hierarchy as an imagined world in which it is self-evident who<br />
are <strong>the</strong> speakers and who <strong>the</strong> silenced ones’. 16<br />
Directions ‘left’ and ‘right’ may be associated with strength and weakness, and<br />
extend to ‘female’, or‘feminine’, and ‘male’, or‘masculine’. For instance, in Khoe,<br />
‘besides expressing distinctions of natural sex, masculine and feminine have a clear<br />
cut secondary meaning’, with masculine being associated with ‘long, high, narrow,<br />
big, strong items, and <strong>the</strong> right hand side’, and <strong>the</strong> feminine gender with ‘short, small,<br />
round, thick, broad, weak items, and <strong>the</strong> left hand side’. An association of ‘right’, or<br />
‘male’, side associated with being ‘strong’, and left, or female, with ‘being weak’, isa<br />
feature of many languages. 17 The terms ‘left’ and ‘right’ in Manambu are used to<br />
describe sides of people’s bodies. The term ‘left hand’ (aki-ta:b) is also used to refer to<br />
one’s mo<strong>the</strong>r’s side in traditional payments, and <strong>the</strong> term ‘right hand’ (mapa-ta:b) is<br />
used to refer to fa<strong>the</strong>r’s side. (A synonym for <strong>the</strong> left hand is kaykwapa-ta:b, lit. lazy<br />
hand, and for <strong>the</strong> right hand vyavya-ta:b, hitting hand).<br />
Linguistic <strong>Gender</strong> choice for inanimates may be associated with <strong>the</strong> idea of<br />
femininity as ‘mo<strong>the</strong>rhood’. We can recall, from §3.1, that in Sare, from <strong>the</strong> Sepik<br />
region of Papua New Guinea, a mature plant or fruit is assigned to <strong>the</strong> feminine<br />
gender while a young plant or fruit will be masculine. This consistent principle of<br />
Linguistic <strong>Gender</strong> assignment in Sare can be seen as a projection of Natural <strong>Gender</strong><br />
properties (that is, femininity associated with procreation), as formulated by Sumbuk<br />
(1999: 117), a native speaker of <strong>the</strong> language. Similarly, Beachy (2005: 63) suggests<br />
that ‘sun’ and ‘blood’ in Dizin, an Omotic language, are assigned to <strong>the</strong> feminine<br />
gender because <strong>the</strong>y are ‘seen as life-giving’. The association of Linguistic <strong>Gender</strong><br />
with women and mo<strong>the</strong>rhood underlies <strong>the</strong> gender assignment.<br />
A nature-versus-artefact dimension has also been aired, as a correlate of female<br />
and male. In an experiment conducted by Mullen (1990), twenty-two 7-year-old<br />
schoolchildren (all English speakers) were asked to classify names or pictures of
130 8 The images of gender<br />
objects into ‘female’ and ‘male’. It turned out that objects pertaining to ‘nature’—such<br />
as plant, seashell, leaf, and fish—were classified as ‘woman-like’ by <strong>the</strong> majority of<br />
children. In contrast, artefacts—such as street sign, sailboat, television, and clock—<br />
were considered ‘man-like’. The female–nature association was a bit stronger than<br />
<strong>the</strong> male–artefact association. This reflects conventional practices of referring to<br />
nature as ‘she’; as in <strong>the</strong> English expressions ‘mo<strong>the</strong>r nature’ and ‘mo<strong>the</strong>r earth’.<br />
These associations originate in cultural stereotypes of <strong>the</strong> Social <strong>Gender</strong> (especially in<br />
Western societies). As Mullen (1990: 585) puts it, ‘<strong>the</strong> realm of nature, as opposed to<br />
<strong>the</strong> realm of artefacts, is thought to be less controllable, less predictable, less rational,<br />
and more balanced, more whole, more nurturing. Each of <strong>the</strong>se characteristics is<br />
generally associated with females as well. The male–artefact domain is associated<br />
with <strong>the</strong> opposite set of characteristics.’ 18 This whole way of looking at males and<br />
females is a ‘construct of culture ra<strong>the</strong>r than a fact of nature’ (as Ortner 1972: 87) puts<br />
it, and relies on an evaluation of women across many cultures as being somehow<br />
‘second-class citizens’ (p. 68). What can be seen as Natural <strong>Gender</strong> features and<br />
associations reflects <strong>the</strong> ingrained stereotypes of Social <strong>Gender</strong>.<br />
Among <strong>the</strong> Kaluli (a Papuan community in <strong>the</strong> New Guinea Highlands), <strong>the</strong><br />
physique and capabilities of women and men are seen through <strong>the</strong> prism of Social<br />
<strong>Gender</strong> stereotypes. The Kaluli language has no Linguistic <strong>Gender</strong>. As Schieffelin<br />
(1977: 121–7) puts it,<br />
‘men’ epitomise <strong>the</strong> ‘Kaluli’ behaviour, partly because of <strong>the</strong> high visibility of male activity. It is<br />
<strong>the</strong> men who tend to be dramatic and flamboyant, who act publicly, who compel attention.<br />
Most of <strong>the</strong> behavior that <strong>the</strong> Kaluli <strong>the</strong>mselves consider culturally important is performed in<br />
<strong>the</strong> male role...In relation to femaleness, maleness encompasses <strong>the</strong> dominant Kaluli values,<br />
and <strong>the</strong> figure of a man is something of a general cultural ideal.<br />
The Kaluli men characterize <strong>the</strong>mselves as ‘wiry, alert, fast’, and see ‘women as heavy<br />
and slow, unable to travel quickly or far’; <strong>the</strong>y are viewed as ‘weaker and less dynamic’.<br />
These images essentially reflect <strong>the</strong> Social gender, correlating <strong>the</strong>m with features of<br />
Natural <strong>Gender</strong>, so much so that even an outsider (in this case, <strong>the</strong> anthropologist<br />
Edward Schieffelin) sees Kaluli women as ‘being generally bulkier in appearance than<br />
men, and carrying <strong>the</strong>mselves with a certain unwieldiness that is enhanced by <strong>the</strong>ir<br />
swaying skirts and ample breasts’ (p. 122).<br />
The ways in which men and women are talked about reflect what scholars call<br />
‘hegemonic masculinity’—an ideal ‘man’ able to subordinate those who may not be<br />
quite up to it (including gay, disabled, and working-class men, and all women).<br />
Caldas-Coulthard and Moon (2010) examined adjectives occurring with ‘man’,<br />
‘woman’, ‘boy’ and ‘girl’ in British tabloid and broadsheet corpora, and found that<br />
men were more often categorized in terms of age, status, and personality, and for<br />
women <strong>the</strong> focus was more on physical appearance and sexuality—that is, as sexual<br />
‘objects’. Baker (2014) investigated how <strong>the</strong> use of adjectives with men and women
8.4 What men and women look like 131<br />
changed throughout <strong>the</strong> Historical Corpus of American English which covers <strong>the</strong><br />
period 1810–2009. The texts <strong>the</strong>re include fiction, magazines, newspapers, movies,<br />
and play scripts. Throughout <strong>the</strong> nineteenth century, ‘man’ but not ‘woman’ would<br />
be accompanied by adjectives like honest, brave, and good-hearted. Morality was <strong>the</strong><br />
‘province of men’ (p. 48), and a woman would be characterized in terms of looks, age,<br />
and breeding. Towards <strong>the</strong> end of <strong>the</strong> twentieth century, <strong>the</strong> focus shifted to<br />
adjectives relating to ‘bodies’: men appear as able-bodied, burly, and stocky, and<br />
women as good-looking, sweet-faced, comely, and so on. A move from referring to<br />
males and females in terms of moral qualities towards focus on <strong>the</strong>ir bodies may be<br />
linked to men being regarded by women as sexual objects. This may indeed represent<br />
an equalizing trend: men, as well as women, can be seen as sexual objects as women<br />
become more independent and empowered.<br />
The attributes of men and women, and <strong>the</strong>ir behaviour, are strongly linked to <strong>the</strong>ir<br />
Social <strong>Gender</strong> roles, and <strong>the</strong> associated rituals and practices.<br />
For many years, cognitive linguists have been attempting to conduct experiments and<br />
uncover associations between physical properties and Natural <strong>Gender</strong>s (male and<br />
female). Results published by Osgood (1960) showed strong correlations between <strong>the</strong><br />
concept of ‘man’ and ‘woman’ and properties such as angular versus rounded, dark<br />
versus light, colourless/colourful, sharp/blunt, thick/thin, large/small, near/far, and<br />
also vertical/horizontal, heterogeneous/homogenous, and clear/hazy. The experiments<br />
included speakers of English, Navajo, Mexican Spanish, and Japanese. Ervin<br />
(1962) reported fur<strong>the</strong>r support for ‘a tendency to ascribe different connotations to<br />
masculine and feminine words in Italian’. She asked thirty bilinguals in Italian and<br />
English to attribute Linguistic <strong>Gender</strong>s to nonsense words, and <strong>the</strong>n rate <strong>the</strong>m with<br />
regard to four parameters: bello/brutto (beautiful/ugly), buono/cattivo (good/bad),<br />
delicato/forte (weak/strong), piccolo/grosso (small/big). The nonsense words assigned<br />
to feminine Linguistic <strong>Gender</strong> were consistently rated as prettier, weaker, and smaller<br />
than those assigned to <strong>the</strong> masculine Linguistic <strong>Gender</strong>.<br />
Boroditsky, Schmidt, and Phillips (2003) put speakers of Spanish, German, and<br />
English through yet ano<strong>the</strong>r test, aimed at proving whe<strong>the</strong>r Linguistic <strong>Gender</strong><br />
influences <strong>the</strong> way people think about objects. Speakers of Spanish and of German<br />
(all highly proficient in English) were given a list of twenty-four object names and<br />
<strong>the</strong>n asked to write down <strong>the</strong> first three adjectives that came to mind to describe<br />
<strong>the</strong>m. Then a group of speakers were asked to rate <strong>the</strong> adjectives as describing<br />
masculine or feminine properties. As predicted by <strong>the</strong> authors of <strong>the</strong> experiment,<br />
both Spanish and German speakers generated adjectives that were rated as more<br />
‘masculine’ for items with masculine Linguistic <strong>Gender</strong> than for those with feminine.<br />
All object names had opposite genders in Spanish and in German, and so Spanish<br />
and German speakers produced different adjectives to describe <strong>the</strong> same object. The<br />
word for ‘key’ (der Schlüssel) is masculine in German, and feminine in Spanish (la
132 8 The images of gender<br />
llave). So, German speakers described keys as hard, heavy, jagged, metal, serrated, and<br />
useful, while for Spanish speakers <strong>the</strong>y were golden, intricate, little, lovely, shiny, and<br />
tiny. The word for ‘bridge’ is masculine in Spanish—so speakers described bridges as<br />
big, dangerous, long, strong, sturdy, and towering. In German, ‘bridge’ belongs to <strong>the</strong><br />
feminine Linguistic <strong>Gender</strong>—and so bridges were beautiful, elegant, fragile, peaceful,<br />
pretty, and slender. What better proof that people’s thinking about objects is influenced<br />
by <strong>the</strong> grammar of <strong>the</strong> language <strong>the</strong>y speak? Grammatical genders might even<br />
affect <strong>the</strong> design of artefacts—a closer look at German bridges may show that <strong>the</strong>y<br />
differ from Spanish bridges in <strong>the</strong> way <strong>the</strong>y are built. And so on.<br />
This astonishingly neat correlation between Linguistic <strong>Gender</strong> and its Natural<br />
<strong>Gender</strong> associations was tested by Landor (2012). 19 The study involved 1,290 participants<br />
from 24 countries (native speakers of German, Spanish, English, Hebrew, and<br />
Hungarian). They were asked to describe twenty-four nouns (covering humans and<br />
inanimates) with adjectives (‘hard’, ‘soft’, ‘big’, ‘small’, and so on), and <strong>the</strong>n rate <strong>the</strong>m<br />
as ‘male’ or ‘female’. The outcome of <strong>the</strong> study was negative—no positive correlations<br />
to speak of were found between Linguistic <strong>Gender</strong> of nouns and <strong>the</strong>ir Natural <strong>Gender</strong><br />
associations. Many speakers just shrugged <strong>the</strong> questions off with a simple ‘can’t decide’.<br />
Fur<strong>the</strong>r doubt was cast upon <strong>the</strong> whole enterprise of ‘experimental studies’ by Guy<br />
Deutscher (2010: 211–12). In his words,<br />
all <strong>the</strong> experiments described so far suffer from one underlying problem, namely that <strong>the</strong>y forced<br />
<strong>the</strong> participants to exercise <strong>the</strong>ir imaginations. A sceptic could argue with some justification that<br />
<strong>the</strong> only thing <strong>the</strong> experiments proved was that grammatical genders affect associations when <strong>the</strong><br />
participants are coerced unnaturally to dream up properties for various inanimate objects. ... In<br />
o<strong>the</strong>r words, if you force Spanish speakers to be on-<strong>the</strong>-spot poets, and extract properties of<br />
bridges out of <strong>the</strong>m, <strong>the</strong> gender system will indeed affect <strong>the</strong>ir choice of properties. But how can<br />
we tell whe<strong>the</strong>r <strong>the</strong> masculine gender has any influence on speakers’ spontaneous conceptions of<br />
bridges, even outside such exercises in poetry on demand?<br />
And indeed, we cannot. All we can say with a degree of assurance is that Linguistic<br />
<strong>Gender</strong>s may open up an avenue for metaphors, and may reflect <strong>the</strong> roles and images<br />
associated with Social <strong>Gender</strong>s—but do not have to. From behind <strong>the</strong> smoke-screen<br />
of ‘scientific experiment’ bridges in Spain may look slenderer and smaller than<br />
bridges in Germany—a good topic fit only for popular magazines and small talk.<br />
8.5 ‘<strong>Gender</strong>ing’ <strong>the</strong> world: images, metaphors, and cognition<br />
In Jakobson’s (1959: 116) words, ‘languages differ essentially in what <strong>the</strong>y must<br />
convey and not in what <strong>the</strong>y may convey...In its cognitive function, language is<br />
minimally dependent on <strong>the</strong> grammatical pattern...But in jest, in dreams, in magic,<br />
briefly, in what one would call everyday verbal mythology and in poetry above all, <strong>the</strong><br />
grammatical categories carry a high semantic import.’ This is where special meanings
8.5 ‘<strong>Gender</strong>ing’ <strong>the</strong> world 133<br />
of Linguistic <strong>Gender</strong>s—laden with images of Natural and Social <strong>Gender</strong>s—step in.<br />
And this is also where languages with Linguistic <strong>Gender</strong> allow us to see <strong>the</strong> world<br />
through a different lens, creating difficulties for translators.<br />
Linguistic <strong>Gender</strong> is a treasure chest of metaphors—as Guy Deutscher puts it,<br />
‘language’s gift to poets’. Again, in Deutscher’s words, ‘<strong>the</strong> chains of associations<br />
imposed by <strong>the</strong> genders of one’s language are all but impossible to cast off ’. Linguistic<br />
<strong>Gender</strong>s reflect myths and rich traditions of peoples’ heritage embalmed in <strong>the</strong><br />
language. Creole and Pidgin languages with shallow histories and little oral and<br />
written tradition simply lack <strong>the</strong>se. 20<br />
Natural <strong>Gender</strong>-laden images permeate <strong>the</strong> world of myths and beliefs: <strong>the</strong> moon<br />
may be a mythical man or a woman, and so can be <strong>the</strong> sun. Features of Natural and<br />
Social <strong>Gender</strong> roles can be revealed through <strong>the</strong> ways men and women are talked<br />
about. Physical properties typically associated with men and with women may draw<br />
on <strong>the</strong>ir Social <strong>Gender</strong> roles, or cultural norms to do with men and women. For <strong>the</strong><br />
Kaluli (described by a male anthropologist), men are wiry and alert, and women slow<br />
and bulky. Men tend to be conceived of as big, hard, and strong, and women as <strong>the</strong><br />
opposite. Attributes of women—and not so much of men—have negative overtones.<br />
So do words which have an exclusively female reference. We saw in §3.4.2 that<br />
masculine Linguistic <strong>Gender</strong> correlates with ‘value’ and ‘importance’ in many patrilineal<br />
societies where <strong>the</strong> male Social <strong>Gender</strong> is high in status. This reflects traditional<br />
power relationships, to which we return again in subsequent chapters.<br />
There is no hard-core evidence that all inanimate objects are automatically<br />
assigned to a Natural <strong>Gender</strong> by speakers of languages with Linguistic <strong>Gender</strong>. But<br />
Linguistic <strong>Gender</strong> interacts with cognitive mechanisms. In Dan Slobin’s (1996: 91)<br />
words, each language ‘is a subjective orientation to <strong>the</strong> world of human experience,<br />
and this orientation affects <strong>the</strong> ways in which we think while we are speaking’. This is<br />
part of ‘thinking for speaking’—<strong>the</strong> multitude of unconscious decisions a speaker has<br />
to make, in order to express what <strong>the</strong> grammars require.<br />
Having a Linguistic <strong>Gender</strong> in your language forces you to make a choice in gender<br />
forms. And this choice, and <strong>the</strong> meanings of <strong>the</strong> Linguistic <strong>Gender</strong>, in turn, may<br />
reflect <strong>the</strong> ways in which men and women are seen and behave—in <strong>the</strong>ir Natural and<br />
Social <strong>Gender</strong> guises. But <strong>the</strong>re is no reason to believe that <strong>the</strong> overtones of Natural<br />
and Social <strong>Gender</strong> are always projected onto every Linguistic <strong>Gender</strong> choice—<br />
German bridges are no more ‘woman-like’ than bridges in Spanish are ‘man-like’.<br />
Taken to its extreme (and coupled with quasi-scientific experiments), any attempt to<br />
find a perfect one-to-one correspondence between Linguistic <strong>Gender</strong>, Social <strong>Gender</strong>,<br />
and Natural <strong>Gender</strong> verges on <strong>the</strong> grotesque.<br />
To summarize: having Linguistic <strong>Gender</strong> in a language opens up an avenue for<br />
metaphors—deployed in poetry, beliefs, and images. Properties of men and of<br />
women—as seen in adjectives and o<strong>the</strong>r expressions associated with <strong>the</strong>m—reflect<br />
stereotypes of Social <strong>Gender</strong>. As a society changes, men and women may come to be
134 8 The images of gender<br />
seen differently: Linguistic <strong>Gender</strong>s come to reflect social changes and language<br />
reforms. Different patterns of male and female speech fur<strong>the</strong>r reflect differences in<br />
Social <strong>Gender</strong>. This is what we turn to next.<br />
NOTES AND SOURCES<br />
1. See Haas (1941: 65), Mithun (1999: 101) and §2.2.1.<br />
2. See Alpher (1987: 180).<br />
3. The Manambu have no stories which might involve attributing <strong>the</strong> Natural <strong>Gender</strong> to <strong>the</strong><br />
sun. Based on its round shape, it is typically assigned feminine gender. But if it is<br />
particularly strong and hot, it can be referred to as masculine (following <strong>the</strong> principles<br />
in Figure 3.1).<br />
4. See Payne (1989: 130), Romani Miranda (2004).<br />
5. Nissen (2002) offers fur<strong>the</strong>r examples and some hints on how to bridge <strong>the</strong> abyss of ‘nontranslatability’.<br />
6. In Heine’s original, <strong>the</strong> male pine-tree and <strong>the</strong> female palm-tree imagery is thought to<br />
go beyond unrequited romantic love: for a summary, see Deutscher (2010: 196). Nikolaj<br />
Glazkov, a ra<strong>the</strong>r controversial and amusing Russian poet, offered his own version of<br />
Heine’s poem: an ever-green larch (listvennica, feminine Linguistic <strong>Gender</strong>) transplanted<br />
into <strong>the</strong> Far North (evoking Stalin’s concentration camps) is dreaming of a ‘fantastic<br />
sou<strong>the</strong>rly’ oak-tree (dub, masculine) (1976, Voksal. Moscow: Sovetskij pisatelj).<br />
7. Emma Lazarus, title 1: ‘Lotus blossom’, from Poems and translations, published 1867.<br />
Ano<strong>the</strong>r well-known example is a humorous poem by Joachim Ringelnatz about a male<br />
stamp (masculine in German) falling in love with a (female) princess. The sexual Natural<br />
<strong>Gender</strong> imagery is reflected in a translation by Ernest Seemann . The German text is: Ein männlicher Briefmark erlebte | was Schönes,<br />
bevor er klebte | Er war von einer | Prinzessin beleckt | Da war die Liebe in ihm erweckt | Er<br />
wollte sie wiederküssen | Doch hat er verreisen müssen | So liebte er sie vergebens/ | Das ist<br />
die Tragic des Lebens. And <strong>the</strong> English translation is: ‘A postage stamp, male, was elated |<br />
With joy, before he was dated | A princess licked him, by Jove! | Which did awaken his love<br />
| He wanted to kiss her back | But had to go on a trek | His love was thus unavailing | So sad<br />
is often life’s failing.’<br />
8. The song ‘Why are you standing swaying, you thin Rowan’ (Shto stoish, kachajasj, tonkaja<br />
rjabina...) is based on a poem by Ivan Surikov (1864). Fur<strong>the</strong>r illustrative examples from<br />
classical Russian authors and extensive commentary are in Vinogradov (1947: 64–70).<br />
9. The ‘gender symbolism’, with special stylistic effect, can be achieved in languages without<br />
Linguistic <strong>Gender</strong>, by using nouns with Natural <strong>Gender</strong> reference—man and woman. The<br />
young moon is referred to as kuu-poeg ‘moon-son’ in a widely known Estonian lullaby, by<br />
Heljo Mänd (Vikeraare all. Tallinn, Eesti Raamat, 1965, p.26).<br />
10. The two largest cities in Russia translated into ‘a woman’ and ‘a man’, for Nikolaj Gogol:<br />
‘Moscow (Russian Moskva) is of feminine gender, St Petersburg of masculine gender.<br />
Moscow has brides, St Petersburg has bridegrooms.’ He goes on to describe Moscow (of<br />
feminine Linguistic <strong>Gender</strong>) as a bulky woman making pancakes or sitting in her armchair,<br />
never venturing anywhere—in contrast to St Petersburg, a dapper and curious<br />
young man, eager to show himself off to foreigners (Peterburg notes 1836). There are<br />
many more examples of hard-to-translate Linguistic <strong>Gender</strong> associations in poetry—some
8.5 ‘<strong>Gender</strong>ing’ <strong>the</strong> world 135<br />
examples are Charles Baudelaire’s ‘L’homme et la mer’ (<strong>the</strong> man:masculine and <strong>the</strong> sea:<br />
feminine) and Pablo Neruda’s ‘Ode to <strong>the</strong> sea’ where <strong>the</strong> sea (el mar, masculine) strikes a<br />
stone (una piedra, feminine) and ‘caresses her, kisses her, pounds his chest, repeating his<br />
own name’ (see also Deutscher 2010: 215).<br />
11. Having to always mark Linguistic <strong>Gender</strong> agreement imposes a necessity to always make a<br />
Linguistic <strong>Gender</strong> choice. Having to choose a masculine or a feminine Linguistic <strong>Gender</strong><br />
form in Greek conversations helps foreground male or female identity. The grammatical<br />
marking of masculine and feminine Linguistic <strong>Gender</strong>s helps create <strong>the</strong> ‘sociocultural’—or<br />
Social <strong>Gender</strong>—world which participants build step by step in <strong>the</strong>ir interaction. We turn to<br />
this in Chapter 7 (see Alvanoudi 2014, 2015).<br />
12. See Rubino (1997: 75–6).<br />
13. This is reminiscent of Julius Caesar’s description of <strong>the</strong> brave Belgae, who were <strong>the</strong> bravest<br />
in Gaul, since <strong>the</strong>y had less access to objects which tend to ‘effeminate’ <strong>the</strong> mind, Caesar’s<br />
Gallic Wars Book 1: ‘The Belgae are <strong>the</strong> bravest, because <strong>the</strong>y are fur<strong>the</strong>st from <strong>the</strong><br />
civilization and refinement of [our] Province, and merchants least frequently resort to<br />
<strong>the</strong>m, and import those things which tend to effeminate <strong>the</strong> mind’ (‘Horum omnium<br />
fortissimi sunt Belgae, propterea quod a cultu atque humanitate provinciae longissime<br />
absunt, minimeque ad eos mercatores saepe commeant atque ea quae ad effeminandos<br />
animos pertinent important...’).<br />
14. See also Dul'son (1968: 64), Werner (1994, 1997: 88–96).<br />
15. Raga and Woldemariam (2011: 166–7).<br />
16. From Schipper (2003: 19, 27, 386).<br />
17. See, for instance, Raffelsiefen (2011).<br />
18. To what extent such associations are indeed universal as claimed by Mullen (1990) isan<br />
open question. In some non-Western societies, feelings and irrationality are a feature of<br />
women, and rational behaviour is linked to men (see Kulick 1992a, 1992b, on <strong>the</strong> Taiap<br />
people in <strong>the</strong> village of Gapun in <strong>the</strong> Sepik area of Papua New Guinea).<br />
19. Landor (2012: 93–6) contains a description of <strong>the</strong> experiment. Later in <strong>the</strong> <strong>the</strong>sis, he<br />
concludes that <strong>the</strong> Linguistic <strong>Gender</strong> of nouns may have a ‘clouding effect’ on how, and<br />
whe<strong>the</strong>r, objects can be conceptualized as ‘male’ or ‘female’. <strong>Gender</strong>ed conceptualization<br />
of objects: those who speak languages with Linguistic <strong>Gender</strong> cannot decide on an entity’s<br />
male or female attributes, as <strong>the</strong> labels of <strong>the</strong> genders (masculine and feminine) ‘suppress<br />
<strong>the</strong> confidence with which participants ascribe gendered characteristics to <strong>the</strong> objects’<br />
(2012: 193–5). These results go against <strong>the</strong> ‘feel-good’ and suggestive correlations between<br />
Linguistic <strong>Gender</strong> choice and indisputably male or female characteristics of entities. Note<br />
that in <strong>the</strong> majority of ‘experimental’ quantitative studies of cognitive effects of Linguistic<br />
<strong>Gender</strong>s, <strong>the</strong> size of a sample is too small to provide statistically significant results for a<br />
whole population of speakers of German, Spanish, English, and o<strong>the</strong>r languages investigated<br />
(see Dixon 2010: 257–61, on <strong>the</strong> principles of sampling). Landor (2012)—with a<br />
careful statistic back-up—stands out as an exception.<br />
20. No wonder that McWhorter (2014), primarily a Creole scholar, shows little appreciation of<br />
gender as an ‘ornate’ category. Creoles lack many o<strong>the</strong>r linguistic features which reflect social<br />
patterns and peoples’ perception—including evidentials (see Aikhenvald 2014c), possession<br />
(Aikhenvald 2013c), and complex demonstrative systems (Dixon 2012: 189–261). This may<br />
well be why all languages look <strong>the</strong> same for <strong>the</strong> likes of McWhorter—as <strong>the</strong>y are likely to be, if<br />
looked at through <strong>the</strong> lens of Creoles.
9<br />
When women and men speak<br />
differently<br />
Differences between male and female speech vary, from language to language and<br />
from society to society. Ways of speaking differentiate Social, and Natural <strong>Gender</strong>s.<br />
In a number of languages from North and South America, Asia, Australia, and <strong>the</strong><br />
Pacific male and female dialects show regular and salient differences. As Jean Kirton<br />
(1988: 111) put it, <strong>the</strong> ‘separate dialects for <strong>the</strong> men and for <strong>the</strong> women’ in Yanyuwa,<br />
an Australian language, have ‘differences of a kind normally associated with language<br />
dialects in separate locations’. This is what we turn to now.<br />
9.1 ‘Male’ and ‘female’ dialects<br />
Sara Trechter, a major figure in <strong>the</strong> field of male and female dialects in <strong>the</strong> North<br />
American context, describes her experience with <strong>the</strong> Lakhota of <strong>the</strong> North American<br />
south-west:<br />
Lakhota speakers who were unaware of my interest in gender have often told me that women and<br />
men end <strong>the</strong>ir sentences in Lakhota differently. My Pine Ridge ‘permission to record’ agreement<br />
contained a statement alluding to my interest in gendered speech styles. On reading this agreement,<br />
potential consultants directly addressed what <strong>the</strong>y perceived to be my research question.<br />
Their explanation took <strong>the</strong> form of metapragmatic judgements or maxims about appropriate<br />
speech. ...‘Men say yo and women say ye’ or ‘Men say lo and women say le’.(1999: 107)<br />
Similarly to o<strong>the</strong>r Siouan languages, Lakhota has a set of enclitics which express<br />
questions, commands, and emphatic statements. The clitics ye (male) and yo (female)<br />
mark commands, and <strong>the</strong> clitics lo (male) and le (female) mark opinion or emphasis.<br />
Table 9.1 features a selection.<br />
A man will say 9.1 and a woman will say 9.2, telling someone to get up (Trechter<br />
1995: 14):<br />
9.1 kikta ye<br />
get.up female.speech.imperative<br />
‘Get up!’ (woman speaking)<br />
<strong>How</strong> <strong>Gender</strong> <strong>Shapes</strong> <strong>the</strong> <strong>World</strong>. First edition. Alexandra Y. Aikhenvald.<br />
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.
9.1 ‘Male’ and ‘female’ dialects 137<br />
TABLE 9.1. Male and female forms of a selection of enclitics in<br />
Lakhota (Trechter 1995: 57)<br />
MEANING MALE FEMALE<br />
formal question hu̹wo hu̹we (obsolete)<br />
command yo ye<br />
familiar command yet h o nit h o<br />
opinion/emphasis/fact yelo yele (obsolete); ye<br />
emphatic statement ks᷈t ks᷈to<br />
entreaty ye, yee na<br />
surprise/opinion yewa̹ yema̹<br />
9.2 kikta yo<br />
get.up male.speech.imperative<br />
‘Get up!’ (man speaking)<br />
The following fieldwork anecdote shows how real <strong>the</strong> distinction between male<br />
and female registers used to be among <strong>the</strong> Gros Ventre, from <strong>the</strong> Algonquian<br />
language family, at <strong>the</strong> time when Regina Flannery (1946: 134) did her fieldwork:<br />
When an old woman asked me what name I had received in <strong>the</strong> naming ceremony, I repeated<br />
as an old man, my namer, had pronounced it iθénædjæ (woman chief). Whereupon <strong>the</strong> old<br />
woman said indignantly: ‘Your name is iθénækyæ. I knew that woman for whom you are<br />
named. Can’t <strong>the</strong>se men ever do anything right!’<br />
Regina Flannery had repeated <strong>the</strong> name as <strong>the</strong> man had pronounced it, instead of<br />
converting it to <strong>the</strong> form a woman would use.<br />
We look into <strong>the</strong> differences between male and female dialects, <strong>the</strong>ir place in<br />
language history, and correlations with <strong>the</strong> Natural and <strong>the</strong> Social <strong>Gender</strong> of both<br />
speaker and addressee, in §§9.1.1–5.<br />
In two well-known cases, male and female dialects correlate with politeness<br />
registers, speakers’ status, and social identity. In Japanese, men and women employ<br />
different personal pronouns, and markers of politeness and ‘elegance’ of speech. The<br />
choice of forms depends on <strong>the</strong> relationship between <strong>the</strong> speaker and <strong>the</strong> addressee.<br />
As <strong>the</strong> women’s place in Japanese society changes, <strong>the</strong> rules for male and female<br />
speech use transform. Male and female speakers of Thai, <strong>the</strong> national language of<br />
Thailand, use different final particles marking respect for <strong>the</strong> addressee. The choice of<br />
<strong>the</strong> pronoun ‘I’ for men and for women depends on status of <strong>the</strong> speaker and <strong>the</strong><br />
addressee, and <strong>the</strong>ir relationships—see §9.1.6.
138 9 When women and men speak differently<br />
Languages with distinct male and female dialects have been accorded <strong>the</strong> label of<br />
‘gender-exclusive’. Languages in which differences between men’s speech and women’s<br />
speech are more subtle and harder to capture are known as ‘gender-variable’—<strong>the</strong>y<br />
include most familiar European languages, among <strong>the</strong>m English. As Sou<strong>the</strong>rland and<br />
Katamba (1996: 550–1) put it, ‘gender-exclusive differentiation refers to <strong>the</strong> radically<br />
different speech varieties used by men and women in particular societies. In <strong>the</strong>se<br />
societies, a woman or man may, except in special circumstances, not be allowed to<br />
speak <strong>the</strong> variety of <strong>the</strong> o<strong>the</strong>r gender...<strong>Gender</strong>-variable differentiation is much more<br />
common in <strong>the</strong> languages of <strong>the</strong> world than is gender exclusivity. This phenomenon is<br />
reflected in <strong>the</strong> relative frequency with which men and women use <strong>the</strong> same lexical<br />
items or o<strong>the</strong>r linguistic features. If, as is often asserted, female English speakers use<br />
words such as lovely and nice more often than do male speakers we can claim that in<br />
this respect English speakers exhibit gender-variable differentiation.’ 1<br />
A gender-exclusive language imposes <strong>the</strong> choice between <strong>the</strong> male or <strong>the</strong> female<br />
dialect on every speaker. As we will see throughout this section, ei<strong>the</strong>r dialect can<br />
come to be associated with a Social <strong>Gender</strong>, formality, and social status. A gendervariable<br />
language operates <strong>the</strong> o<strong>the</strong>r way round: its linguistic forms are deployed in<br />
different ways, reflecting social conditions, attitudes, and behaviours which <strong>the</strong>n<br />
come to be associated with an image of a man or a woman. These associations are<br />
fluid and hardly obligatory. Unlike gender-exclusive languages, men and women<br />
draw upon <strong>the</strong> same repertoire of forms—see §9.2. As gender-exclusive languages fall<br />
into disuse, <strong>the</strong> boundaries between gender-exclusive and gender-variable languages<br />
become blurred.<br />
9.1.1 <strong>How</strong> male and female dialects differ<br />
Differences between <strong>the</strong> male and <strong>the</strong> female varieties may be purely phonological. In<br />
his grammar of <strong>the</strong> Inuktikut (Eskimo), Boas (1911: 79) remarked:<br />
Among some Eskimo tribes <strong>the</strong> men pronounce <strong>the</strong> terminal p, t, k, and q distinctly, while <strong>the</strong><br />
women always transform <strong>the</strong>se sounds into m, n, ñ and ñ̩.<br />
Men’s speech in Karajá, a Macro-Jê language from Brazil, lacks <strong>the</strong> velar glottal<br />
stop k (which is present in women’s speech). 2 A woman will say ruku, and a man will<br />
say ru ‘night’; a woman will say dıkarə̃, and a man will say dıarə̃‘I’. The principle<br />
affects recent borrowings from Portuguese, <strong>the</strong> main language of <strong>the</strong> country: a<br />
woman will say kawaɾu, and a man will say awaɾu, from Portuguese cavalo ‘horse’.<br />
A woman will say bicikreta, and a man will say bicileta, from Portuguese bicicleta<br />
‘bicycle’. Female speech appears to be more conservative—we return to this in §9.1.3<br />
and <strong>the</strong>n in §10.4.<br />
Male and female dialects of Gros Ventre show systematic differences in <strong>the</strong>ir<br />
phonology. The sounds tc and dj in men’s speech correspond to k and ky in women’s<br />
speech. Men would say itcénibitc ‘his gum’, and women would say ikébinik. 3 That
9.1 ‘Male’ and ‘female’ dialects 139<br />
men’s andwomen’s speech sounds differently in Chukchi, a highly endangered<br />
Chukotko-Kamchatkan language from <strong>the</strong> extreme north-east of <strong>the</strong> Russian Federation,<br />
was first noted by Vladimir Bogoras. There still remain regular correspondences<br />
between sounds in men’s and women’s speech:awomanwouldsay<br />
mcen,andamanmren,for‘mosquito’; a woman would say cajkok and a man sajkok<br />
for ‘teapot’. 4<br />
Systematic differences between male and female speech can extend into morphology,<br />
as we saw for Lakhota in Table 9.1. In her classic paper, Mary Haas (1944)<br />
described differences between male and female speech in imperative and indicative<br />
verbs from Koasati, a Muskogean language from Louisiana. A woman’s form would<br />
end in a nasalized vowel, and a man’s form would replace a nasalized vowel with an<br />
oral vowel followed by s. A woman would say /lakawtakkõ/ and a man would say<br />
/lakawtakkós/ ‘I am not lifting it’.<br />
Ano<strong>the</strong>r prime example of male and female dialects comes from Yana, a now<br />
extinct isolate formerly spoken in nor<strong>the</strong>rn California and described by Sapir<br />
(1929a). Many Yana words had a ‘male’ and a ‘female’ form. A male form used to<br />
be longer, and a female form reduced. The male forms ‘au-na ‘fire’ and ‘au-‘nidja ‘my<br />
fire’ corresponded to female forms ‘au‘ ‘fire’, ‘au-‘nitc‘ ‘my fire’. There were fur<strong>the</strong>r<br />
morphological differences. The demonstratives ending in ‐e would add -‘e in <strong>the</strong> male<br />
forms. A man will say aidje’e ‘that one’, and a woman will say aidje. Men and women<br />
use different forms to mark questions: to form an emphatic interrogative, a man<br />
would use an enclitic nā and a woman an enclitic gā . 5<br />
Male and female speech is a striking feature of Kokama-Kokamilla, a Tupí-<br />
Guaraní language of nor<strong>the</strong>rn Peru. Men use <strong>the</strong> suffix =kana for plural, and<br />
women use =nu; so <strong>the</strong> plural of iwira ‘tree’ will be iwira=kana ‘trees’ for men, and<br />
iwira=nu for women; <strong>the</strong> plural of mesa ‘table’, a recent borrowing from Spanish, will<br />
be mesa=kana for men and mesa=nu for women. Fur<strong>the</strong>r differences lie in personal<br />
pronouns, demonstratives, and linkers—a selection of forms is in Table 9.2. 6<br />
Male and female speech in Awetí, a language from <strong>the</strong> Xingu Indigenous Park in<br />
Brazil closely related to Tupí-Guaraní languages, differs in some of <strong>the</strong> personal<br />
pronouns and prefixes, and all <strong>the</strong> demonstratives. Table 9.3 illustrates some of <strong>the</strong><br />
differences. 7<br />
Just a few lexemes have female and male forms; some male forms tend to have an<br />
initial n and female forms do not, e.g. male nypék, female ypék ‘duck’. The term for<br />
‘duck’ in o<strong>the</strong>r Tupí-Guaraní languages does not have an initial n, e.g. Língua Geral<br />
ipêka ‘pato’. The female dialect appears to be more archaic—we return to this in<br />
§9.1.3 (and <strong>the</strong>n in §10.4).<br />
The female dialect of Yanyuwa has six Linguistic <strong>Gender</strong>s (male, female, masculine,<br />
feminine, food, and arboreal). The men’s dialect classifies male humans, male<br />
animates, and a few o<strong>the</strong>r items into one Linguistic <strong>Gender</strong> (marked with <strong>the</strong> prefix<br />
ki-). The female dialect distinguishes human males (using <strong>the</strong> ‘male’ class prefix nya-)
140 9 When women and men speak differently<br />
TABLE 9.2. Female versus male forms in Kokama-Kokamilla: a selection (Vallejos<br />
2010: 42)<br />
TRANSLATION FEMALE FORM MALE FORM<br />
First person singular pronoun tsa, etse ta<br />
First person exclusive pronoun (we without you) penu tana<br />
Third person plural pronoun inu rana<br />
Proximal demonstrative ‘this’ ajan ikian<br />
Distal demonstrative ‘that’ yama/yamua rama/ramua<br />
also yay riay<br />
<strong>the</strong>re, <strong>the</strong>n yaepe raepe<br />
TABLE 9.3. Female versus male forms in Awetí: a selection<br />
TRANSLATION FEMALE FORM MALE FORM<br />
First person singular pronoun itó atit<br />
Third person singular pronoun ı̃ nã<br />
Third person plural pronoun ta’i tsã<br />
Proximal demonstrative ‘this’ (close to speaker) ujá jatã<br />
Proximal demonstrative ‘this’ (close to addressee) akyʹj kitã<br />
Distal demonstrative ‘that’ (distant from both speaker<br />
and addressee)<br />
akój kujtã<br />
and male animates (using <strong>the</strong> ‘masculine’ class prefix ji-). Pronouns, demonstratives,<br />
and verbal prefixes have different forms in <strong>the</strong> two dialects. 8<br />
Male and female dialects may use different interjections. Among <strong>the</strong> Gros Ventre<br />
greetings wahei (male speakers) and næhæ (female speakers) ‘hello’ would immediately<br />
tell <strong>the</strong> Natural <strong>Gender</strong>s apart. When expressing surprise, a man would say<br />
u . tse, and a woman e . ha, and when answering to a hail, a man would say wei', and a<br />
woman o . . Vocal gestures would also be different. As Flannery (1946: 133) put it, ‘in<br />
order to indicate joy and thankfulness, women give a call which may be described as<br />
rattling <strong>the</strong> tongue, whereas men vocalize and break <strong>the</strong> sounds, not with <strong>the</strong> tongue,<br />
but by striking <strong>the</strong> mouth rhythmically with <strong>the</strong> palm’.
9.1 ‘Male’ and ‘female’ dialects 141<br />
9.1.2 Speakers and addressees of male and female dialects<br />
The choice between male and female speech may depend on <strong>the</strong> Natural <strong>Gender</strong> of<br />
<strong>the</strong> speaker. In Koasati, Kokama-Kokamilla, Awetí, and Yanyuwa men use <strong>the</strong> ‘male’<br />
dialect no matter who <strong>the</strong>y talk to. Or <strong>the</strong> choice of a male or a female dialect may<br />
correlate with <strong>the</strong> Natural <strong>Gender</strong> of <strong>the</strong> person spoken to. ‘Male’ forms in Yana were<br />
used by men speaking to men. ‘Female’ forms were used by women speaking to men<br />
or women, and men speaking to women. 9<br />
Biloxi, an extinct Siouan language, had an elaborate system of male and female<br />
forms of mood particles which would mark a statement, a question, or a command.<br />
There were three sets of forms: <strong>the</strong> same form is used if a man or a woman is<br />
addressing a woman. Men-to-men and women-to-men forms are all different.<br />
A sample is in Table 9.4.<br />
In addition, special singular and plural forms were used by both sexes to address<br />
children: ki ‘carry it (singular)’ and kitu' ‘carry it (plural)’.<br />
Kũr̩ux, a Dravidian language spoken by more than 100,000 people in north India,<br />
has a strict distinction between male and female dialects whose choice depends on<br />
<strong>the</strong> sex of speaker and <strong>the</strong> sex of addressee. The differences are most visible in verbal<br />
paradigms. Male or female forms are also used when talking about a man or a<br />
woman. A selection of forms of <strong>the</strong> verb bar- ‘come’ in present tense is in Table 9.5.<br />
Tunica, a now extinct isolate, is a rare example of a language where <strong>the</strong> choice of<br />
male and female speech is based just on <strong>the</strong> sex of <strong>the</strong> addressee. The differences are<br />
found only in <strong>the</strong> pronominal system (including full pronouns, prefixes, and suffixes,<br />
in three numbers: singular, dual, and plural). An example is in Table 9.6.<br />
Ignaciano (or Mojo), an Arawak language from Bolivia, 10 combines Linguistic<br />
<strong>Gender</strong> with male and female dialectal differences in a ra<strong>the</strong>r curious way. Masculine<br />
gender covers all animate males, feminine gender covers all animate females, and<br />
neuter gender is reserved for all inanimates. Female and male distinctions are made<br />
only in masculine forms of personal pronouns, demonstratives, indefinites, and<br />
possessives: one form is used if a man is speaking, and <strong>the</strong> o<strong>the</strong>r one if <strong>the</strong> speaker<br />
is a woman. This is shown in 9.3, for <strong>the</strong> demonstrative ‘this’.<br />
TABLE 9.4. Male and female speech determined by speaker and addressee in Biloxi<br />
SEX OF SPEAKER SEX OF ADDRESSEE SINGULAR PLURAL MEANING<br />
Man Man ki-aŋko' ki-takaŋko<br />
Man<br />
Woman<br />
ki-tki' ki'-tatki'<br />
Woman<br />
Woman<br />
Woman Man ki-tate' ki-tatute'<br />
‘Carry it!’
142 9 When women and men speak differently<br />
9.3 Singular Plural<br />
juca ‘this:neuter’ (covers inanimate referents)<br />
suca ‘this:feminine’ (animate)<br />
maca ‘this:masculine (animate)’: male speech<br />
ñica ‘this:masculine (animate)’: female speech<br />
juca-na<br />
naca, náca-na<br />
TABLE 9.5. Male and female speech in Kũr̩ux: present tense of <strong>the</strong> verb bar- ‘come’<br />
PERSON/NUMBER<br />
SEX OF SPEAKER AND ADDRESSEE<br />
Man to Man Man to Woman Woman to Woman Woman to Man<br />
1sg bar-d-an barʔ-e-n bar-d-an<br />
1pl exclusive bar-d-am barʔ-e-m bar-d-am<br />
1pl inclusive<br />
bar-d-at<br />
2sg bar-d-ay bar-d-i bar-d-in bar-d-ay<br />
2pl bar-d-ar bar-d-ay bar-d-ar<br />
TABLE 9.6. Male and female speech chosen by <strong>the</strong> sex of <strong>the</strong> addressee in Tunica<br />
SEX OF ADDRESSEE SINGULAR DUAL PLURAL<br />
Pronominal suffixes on semelfactive verbs<br />
Independent pronouns<br />
Man -ʔa -wi'na -wi'ti<br />
Woman -ʔi -hi'na -hi'ti<br />
Man má wi'nima<br />
Woman h h 'ma hi'nima<br />
Different forms for male and for female speech are <strong>the</strong>re only in <strong>the</strong> singular. In <strong>the</strong><br />
plural, <strong>the</strong>re is a straightforward division into animates and inanimates.<br />
9.1.3 Male and female dialects, and language history<br />
Differences between a male and a female dialect may reflect <strong>the</strong> history of <strong>the</strong><br />
speakers. A striking example comes from a ‘mixed’ language of Arawak-Carib origin<br />
in <strong>the</strong> Lesser Antilles which came about just before Columbus’ discovery of <strong>the</strong> West<br />
Indies. The population which originally spoke Iñeri (a dialect of so called Island<br />
Carib), a North Arawak language, were subdued by Carib-speaking invaders. Men<br />
came to speak Carib, with a sprinkling of Iñeri (Arawak) grammatical forms. Women<br />
hung on to <strong>the</strong> original language, of <strong>the</strong> Arawak family. 11
9.1 ‘Male’ and ‘female’ dialects 143<br />
The mixture of Carib-speaking men and Arawak-speaking women brought about<br />
a curious distinction between ‘speech of men’ and ‘speech of women’. Women used<br />
morphemes and lexemes of Arawak origin, while men used lexical items of Carib<br />
origin and grammatical morphemes mostly of Arawak origin. The following example<br />
illustrates <strong>the</strong> female style of <strong>the</strong> Island Carib mixed language. 12 The morphemes of<br />
Arawak origin are in lower case:<br />
9.4 chile-a-tina t-one —Female style<br />
come-perfective-1st.person 3rd.person.feminine-towards<br />
‘I have come to her’<br />
Examples 9.5 and 9.6 come from <strong>the</strong> speech style used by men. Lexical roots, and <strong>the</strong><br />
postposition ‘towards’, are in capitals: <strong>the</strong>se are of Carib origin. Only <strong>the</strong> grammatical<br />
morphemes are of Arawak origin (and in lower case):<br />
9.5 NEMBOUI-a-tina t-IBONAM —Male style<br />
come-perfective-1st.person 3rd.person.feminine-towards<br />
‘I have come to her’<br />
9.6 CHALIMAIN ba-a-(n)na —Male style<br />
paddle 2nd.person-AUX-1st.person<br />
‘Take me across’<br />
The norms regulating <strong>the</strong> use of male and female register were strict: ‘it would be<br />
ridiculous to employ <strong>the</strong> men’s language in speaking to women and vice-versa’. 13<br />
This implies that <strong>the</strong> use of male and female speech forms could have been partially<br />
determined by <strong>the</strong> sex of <strong>the</strong> addressee, and not just <strong>the</strong> speaker.<br />
The two speech styles came about as a result of warfare, and coexisted with an<br />
unusual multilingual situation. The Carib speakers were waging wars against <strong>the</strong><br />
Iñeri (Arawak)-speaking inhabitants of <strong>the</strong> Lesser Antilles. The Caribs won, and <strong>the</strong><br />
Arawak speakers lost. All able-bodied Arawak-speaking men had been killed. The<br />
Carib men took <strong>the</strong> place of <strong>the</strong> Arawak-speaking women’s husbands. But <strong>the</strong>y could<br />
not eliminate <strong>the</strong> existing networks among women and children who continued<br />
speaking <strong>the</strong>ir native Iñeri. At first, <strong>the</strong>y relied on pidgin Carib to speak to <strong>the</strong>ir new<br />
husbands, and fa<strong>the</strong>rs.<br />
Right after <strong>the</strong> Caribs’ military victory, three linguistic codes coexisted: Iñeri, an<br />
Arawak language, was spoken exclusively by women and children, <strong>the</strong> Carib language<br />
by Carib men, and a Pidgin Carib by both parties. As <strong>the</strong> men began to acquire<br />
some knowledge of <strong>the</strong> language of <strong>the</strong>ir wives, new-born children were growing up<br />
with Iñeri and <strong>the</strong> Carib Pidgin—which was known to everyone but considered<br />
‘property’ of <strong>the</strong> men. The society gradually became homogeneous, with everyone<br />
having knowledge of both Iñeri and <strong>the</strong> Carib pidgin. This, by and by, became a<br />
lexical register in <strong>the</strong> Iñeri language.
144 9 When women and men speak differently<br />
Its descendant, known as Island Carib, Black Carib, or Garifuna, still survives in<br />
Belize and Honduras in Central America. It bears numerous traces of Carib influence,<br />
both in its lexicon and in its grammar. There is no longer a systematic<br />
distinction between male and female speech—just a few remnants of ‘men-only’<br />
and ‘women-only’ pairs of synonyms. The most common ones include uogori (male<br />
speech), eieri (female speech) ‘man, male’, uori (male speech), hi̹aru (female speech)<br />
‘woman, female’, goui (male speech), bare (female speech) ‘fishhook’ and abo ie (male<br />
speech), bugabu ia (female speech) ‘come here!’. 14<br />
The female dialect of Island Carib reflects <strong>the</strong> original language of <strong>the</strong> population,<br />
and is more archaic than <strong>the</strong> male dialect—a newly formed mixed language. This<br />
supports a frequently voiced assumption that women are more conservative than<br />
men, in <strong>the</strong>ir speech patterns. Along similar lines, female forms in Karajá keep <strong>the</strong><br />
consonant k of a form, and are thus less innovative than male forms which omit it.<br />
A number of male forms in Awetí appear to be derived from <strong>the</strong> female forms<br />
(Table 9.3 shows this for forms jatã, kitã, and kujtã derived from female forms úja,<br />
akyj, and akoj with a formative -tã: see Drude 2006: 23; 2011). In many instances,<br />
female speech in Koasati uses older forms than male. In o<strong>the</strong>r languages, female<br />
dialects are more innovative than male. Men’s forms in Yana are longer, more<br />
elaborate, and more archaic than women’s. The ways in which male and female<br />
dialects may have come about reflects <strong>the</strong> history of peoples and <strong>the</strong>ir languages of<br />
which not much is known. 15<br />
Phonological differences between male and female dialects can be traced to<br />
women’s practices. Among <strong>the</strong> Suri of south-western Ethiopia, women had to have<br />
a wooden plate inserted into <strong>the</strong>ir lower lip. A plate—made of wood or clay—would<br />
vary in its diameter from one inch to six inches. Having a stretched lip sticking out<br />
makes it difficult for women to articulate bilabial sounds which are replaced by<br />
velars. A man would say mà, and a woman would say ŋà ‘water’. Nowadays, women’s<br />
speech continues to follow <strong>the</strong> same, innovative, patterns even when <strong>the</strong>y take out<br />
<strong>the</strong>ir lip-plates. 16<br />
9.1.4 Male and female speech on <strong>the</strong> way out<br />
The fate of male and female speech varieties reflects <strong>the</strong> history of <strong>the</strong>ir speakers<br />
in yet ano<strong>the</strong>r, unfortunate way. Fewer and fewer people across <strong>the</strong> world continue<br />
speaking small tribal languages. As <strong>the</strong> languages are spoken less and less, <strong>the</strong>ir<br />
linguistic wealth slides into oblivion. And with language attrition, male and female<br />
dialects are in danger.<br />
By <strong>the</strong> time of Mary Haas’s work in <strong>the</strong> 1930s, most Muskogean languages o<strong>the</strong>r<br />
than Koasati had lost <strong>the</strong> male and female speech dialects. Koasati speakers with<br />
whom Geoffrey Kimball started working in 1977 retained just a few phonological<br />
differences between male and female speech, including ‘<strong>the</strong> nasal coloring of men’s<br />
speech contrasted to <strong>the</strong> clarity of women’s’. The male and female speech differences
9.1 ‘Male’ and ‘female’ dialects 145<br />
documented a few decades earlier by Mary Haas were all but gone. 17 Some obsolescence<br />
and mixture of male and female forms in Lakhota was signalled by Trechter<br />
(1995, 1999). In her comprehensive grammar of a highly endangered Osage language,<br />
<strong>the</strong> late Carolyn Quintero (2004: 12–14) could only trace some tendencies<br />
which used to differentiate male and female speech when <strong>the</strong> language used to be<br />
fully spoken. For instance, men would omit a declarative particle after continuative<br />
aspect, and women would tend not to.<br />
Yanyuwa, an Australian language with a rich system of differences for male and<br />
female speech, is on <strong>the</strong> wane: fewer and fewer people still speak it (Bradley 1988). As<br />
is <strong>the</strong> case in many communities across <strong>the</strong> world, Yanyuwa women are more<br />
traditional than men, and are <strong>the</strong> ones who teach young people <strong>the</strong> language. No<br />
wonder that ‘on rare occasions when young people attempt to speak it, <strong>the</strong>y use <strong>the</strong><br />
women’s variety and get chastised’. The male–female speech distinction is dwindling. 18<br />
The Chukchi language is severely endangered, and hardly being learnt by children—<br />
this is <strong>the</strong> fate of most minorities in <strong>the</strong> Russian north. An artificial ‘standard’ language,<br />
promoted throughout <strong>the</strong> Soviet period, was based on <strong>the</strong> men’s dialect. Chukchi<br />
teachers and radio announcers use <strong>the</strong> male variety (female announcers continue using<br />
<strong>the</strong> female forms in private). Discourse about language preservation is framed entirely<br />
in terms of men’s variety.<br />
The destruction of original cultures and traditional institutions may upset <strong>the</strong><br />
status quo of Social <strong>Gender</strong>s, and lead to <strong>the</strong> obsolescence of male and female ways of<br />
speaking. ‘Male’ speech used to be a sign of social malehood among <strong>the</strong> Gros Ventre.<br />
But as <strong>the</strong> language gradually passes into oblivion, <strong>the</strong> difference between male and<br />
female dialects is on <strong>the</strong> wane—in favour of <strong>the</strong> female variety. As Taylor (1982: 305)<br />
puts it, ‘<strong>the</strong> cause for this is probably to be sought in <strong>the</strong> advanced state of cultural<br />
obsolescence in which <strong>the</strong> Gros Ventre find <strong>the</strong>mselves’, and ‘<strong>the</strong> destruction of <strong>the</strong><br />
traditional economy and <strong>the</strong> disappearance of all of <strong>the</strong> social institutions which<br />
supported <strong>the</strong> status of males—war and native religions and <strong>the</strong>ir attendant ceremonies,<br />
age-based societies, and a new way of life which tended to enhance traditional<br />
female roles’. The drastic spread of English, and <strong>the</strong> increasing role of mo<strong>the</strong>rs<br />
in boys’ socialization, fur<strong>the</strong>r contributed to <strong>the</strong> loss of male and female dialects.<br />
The disintegration of traditional social arrangements speeds up <strong>the</strong> loss of male<br />
and female dialects. This is so because male and female dialects mark more than just<br />
sex: <strong>the</strong>y are tokens of Social <strong>Gender</strong>. This is what we turn to now.<br />
9.1.5 Beyond Natural <strong>Gender</strong><br />
If a Kokama-Kokamilla man lapses into ‘feminine’ pronouns, everyone laughs:<br />
mixing male and female dialects is a mark of incompetence. A Gros Ventre man<br />
who uses a female form will be looked upon as effeminate. <strong>How</strong>ever, switching<br />
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<br />
A ‘wrong’ form used on purpose may have a special effect. Kũr̩ux men can address<br />
women ‘as if ’ <strong>the</strong>y were men when <strong>the</strong>y ‘treat <strong>the</strong>m as equals’ (Ekka 1972: 26). On<br />
rare occasions, Yanyuwa men and women used to exploit <strong>the</strong> male/female dialect<br />
differences to draw attention to something unusual or unpleasant. An elderly man in<br />
charge of a funeral ritual was not pleased with <strong>the</strong> performances, and began orating<br />
his displeasure, in <strong>the</strong> female dialect. The inappropriate use of female speech register<br />
highlighted <strong>the</strong> unusual state of affairs, and made people pay attention (Bradley 1988:<br />
131–2). We can recall how Linguistic <strong>Gender</strong> can also be manipulated, to draw<br />
attention to something out of <strong>the</strong> ordinary (§3.4).<br />
Lakhota men use female forms to express affection—if <strong>the</strong> addressee is very much<br />
younger than <strong>the</strong> speaker, or if <strong>the</strong> speaker and <strong>the</strong> addressee are emotionally close<br />
(as a child and a parent would be). One man saw his nephew and said 9.7 to him<br />
using female assertive enclitic: in Sara Trechter’s (1995: 107) words, ‘by imitating<br />
what a woman might say...<strong>the</strong> speaker is indicating a closeness to <strong>the</strong> boy’:<br />
9.7 Wa̹lewa̹ hiyu welê<br />
interjection:male:surprise he:came female:assertive<br />
‘Look who’s come!’<br />
This is reminiscent of how men can address each o<strong>the</strong>r in Amharic using feminine<br />
Linguistic <strong>Gender</strong> forms to express familiarity and affection (see §7.2).<br />
Using male speech in Lahkota signals authority. Some men use female forms<br />
because <strong>the</strong>y feel <strong>the</strong>y are not authoritative enough to speak like ‘proper men’.<br />
Some women may use male speech, which ‘implies a certain masculinity’. A female<br />
speaker of Lakhota ‘who used <strong>the</strong> male clitics regularly was considered by some to<br />
be gay, referred to as a “tomboy” by one speaker, and as a “dyke” or “that woman<br />
with balls” by ano<strong>the</strong>r. Two women who regularly spoke like men shared several<br />
qualities beyond <strong>the</strong>ir speech. They both grew up with a number of bro<strong>the</strong>rs and few<br />
or no sisters; <strong>the</strong>ir biological mo<strong>the</strong>rs died when <strong>the</strong>y were young, and <strong>the</strong>y were in<br />
positions of public authority: one was a judge, <strong>the</strong> o<strong>the</strong>r a police officer.’ 20<br />
In <strong>the</strong> traditional Chukchi society, a person would behave as a representative of <strong>the</strong><br />
opposite sex as part of a shamanic inspiration—changing <strong>the</strong>ir Social <strong>Gender</strong> attributes,<br />
including clothing, social behaviour, work patterns, and also language. 21<br />
Yanyuwa men and women can use <strong>the</strong> dialect of <strong>the</strong> opposite sex in joking<br />
situations, especially those relating to sexual encounters. If a male dancer impersonates<br />
a woman, he may start teasing his bro<strong>the</strong>rs-in-law as if <strong>the</strong>y were his prospective<br />
wives, using <strong>the</strong> female dialect. We can recall a similar pattern of Linguistic <strong>Gender</strong><br />
switch to refer to humans in joking contexts in Machiguenga and in Manambu<br />
(§7.1). Male and female speech are manipulated to mockingly ‘reclassify’ people, as<br />
something overly grotesque.<br />
The male dialect is normally a correlate of male Social <strong>Gender</strong>. Among <strong>the</strong><br />
Yanyuwa, boys used to be initiated into social manhood through a series of rituals
9.1 ‘Male’ and ‘female’ dialects 147<br />
which culminated in circumcision. Only after that were <strong>the</strong>y considered ‘men’ in<br />
<strong>the</strong> social sense, and from that time onward <strong>the</strong>y were expected to speak <strong>the</strong> men’s<br />
dialect.<br />
Male and female dialects can be switched in baby-talk. Karajá men use female<br />
speech forms when talking to <strong>the</strong>ir baby daughters. Women use male forms when<br />
talking to baby boys. 22 Early in his fieldwork with <strong>the</strong> Gros Ventre people, Alan<br />
Taylor was puzzled as to why fluent male speakers insisted on using female forms<br />
when teaching him <strong>the</strong> language. A breakthrough came when <strong>the</strong> speaker explained<br />
that he was talking that way to make things ‘easier’ for Alan Taylor to learn: female<br />
speech among <strong>the</strong> Gros Ventre was used on a par with ‘foreigner talk’. The reason for<br />
this may have to do with child language acquisition. In all likelihood, <strong>the</strong>y were first<br />
taught <strong>the</strong> female forms, by <strong>the</strong>ir mo<strong>the</strong>rs, and only later ‘graduated’ to <strong>the</strong> male form<br />
of speech. In Taylor’s (1982: 304) words, ‘“male” speech is reserved, as it were, for<br />
adult male Gros Ventres, a status acquired only by birth and physical and cultural<br />
maturity’.<br />
The socialization process starts early. Fortune and Fortune (1975: 115) remark that<br />
‘at about age three, mo<strong>the</strong>rs start insisting in <strong>the</strong> Karajá tribe that boys use only men’s<br />
speech. This is <strong>the</strong> beginning of <strong>the</strong> socialization process which continues until <strong>the</strong><br />
adult male is admitted to <strong>the</strong> men’s house where no young initiate would ever use<br />
women’s speech.’<br />
Evidence for male and female speech as indexes of Social <strong>Gender</strong> comes from o<strong>the</strong>r<br />
quarters. Traditionally, <strong>the</strong> Lakhota had at least three Social <strong>Gender</strong> categories: man,<br />
woman, and wi̹kte (woman-potential), or a ‘would-be woman’: a biological man who<br />
possesses <strong>the</strong> spirit of a woman (sometimes called ‘berdache’). A wi̹kte would engage<br />
in traditional feminine activities such as beadwork, would dress as a woman, and<br />
might marry a male. There would also be biological females engaged in traditional<br />
male activities (such as hunting), dressed as a man, and married to a woman (referred<br />
to as kos᷈kalaka ‘young man’). A more encompassing term for kos᷈kalaka and wi̹kte<br />
is ‘two-spirit people’. As Trechter (1995: 7) puts it, <strong>the</strong>ir ‘gendered language ...in<br />
Lakhota is appropriated from or based upon <strong>the</strong> men/women’s speech distinction’.<br />
Male and female Social <strong>Gender</strong> and <strong>the</strong> use of male and female dialect would go<br />
toge<strong>the</strong>r. 23<br />
Male speech may be a token of a formal, and more elevated, style. Sapir (1929a:<br />
212) noted that male forms in Yana were more prestigious than ‘clipped’ female<br />
forms. 24 An analysis, by Luthin (1991), of Yana texts collected by Sapir, showed that<br />
<strong>the</strong> male forms were used in particularly formal and solemn situations (and not just<br />
by men speaking to men). Male forms were used when speaking to a mixed audience.<br />
A man would address his mo<strong>the</strong>r-in-law as if she were a man. The male speech forms<br />
in Koasati may have traditionally been part and parcel of <strong>the</strong> social status of <strong>the</strong><br />
speaker and perhaps <strong>the</strong> stories that only men used to tell. We return to male and<br />
female speech genres in §10.1.
148 9 When women and men speak differently<br />
Male and female forms may be intertwined with politeness, formality, and ultimately<br />
<strong>the</strong> identity of <strong>the</strong> speaker. Japanese and Thai, major languages of major<br />
countries in Asia, are <strong>the</strong> prime examples.<br />
9.1.6 Politeness, identity, and change: male and female dialects<br />
in Japanese and Thai<br />
In Japanese, men and women speak differently. One of <strong>the</strong> most conspicuous distinctions<br />
lies in personal pronouns. Their choice is partly determined by how <strong>the</strong> speaker<br />
and <strong>the</strong> addressee relate to each o<strong>the</strong>r. Personal pronouns are part of <strong>the</strong> well-known<br />
Japanese honorific system which permeates <strong>the</strong> language. Three levels of politeness—<br />
formal, plain, and informal (or deprecatory)—are not reciprocal: a socially superior<br />
person can use ei<strong>the</strong>r formal or informal language, and <strong>the</strong> inferior one is generally<br />
expected to use formal language even if addressed ‘informally’.Table9.7 lists male and<br />
female forms of first and second person pronouns in three politeness (or honorific)<br />
levels. 25<br />
Male and female pronouns differ in <strong>the</strong>ir honorific levels. The pronoun watasi ‘I’<br />
belongs to <strong>the</strong> formal level for men, and <strong>the</strong> plain level for women. Anata ‘you’ is<br />
formal for men, but could be plain or formal for women. Deprecatory second person<br />
pronouns omae and kisama are not used by women. As Ide (1991: 74) puts it,<br />
‘categorical differences in <strong>the</strong> repertoire of personal pronouns lead to women’s automatic<br />
expression of deference and demeanor’. Female forms have a politer flavour than<br />
do male.<br />
Fur<strong>the</strong>r traits contribute to this impression. In Shibatani’s (1990: 374) words, ‘a<br />
conspicuous manifestation of politeness in women’s speech’ is excessive use of prefix<br />
o-, an honorific marker whose meaning is to refer to something with deference. An<br />
TABLE 9.7. Personal pronouns in Japanese: men’s and women’s speech (adapted<br />
from Ide 1991: 73)<br />
PRONOUNS HONORIFIC LEVEL MEN’S SPEECH WOMEN’S SPEECH<br />
First person<br />
Second person<br />
formal<br />
watakusi<br />
watasi<br />
watakusi/atakusi<br />
plain boku watasi/atasi<br />
deprecatory ore Ø<br />
formal anata 26 anata<br />
plain kimi/anta anata/anta<br />
deprecatory<br />
omae<br />
kisama<br />
Ø
9.1 ‘Male’ and ‘female’ dialects 149<br />
object which belongs to a respected person would be marked with o-, e.g. sensei no<br />
o-boosi (teacher genitive HONORIFIC-hat) ‘(respected) teacher’s hat’. Women (and not<br />
men) would use <strong>the</strong> prefix o- to refer to <strong>the</strong>ir own possessions, as in watakusi no<br />
o-saihu (I:female genitive HONORIFIC-purse) ‘my (elegant/lovely) purse’. In its female<br />
usage, o- has gained <strong>the</strong> name of ‘beautification’,or‘elegant’, prefix. The use of o- and<br />
o<strong>the</strong>r honorifics in popular magazines for housewives in Early Modern Japan helped<br />
underscore ‘<strong>the</strong> stances of elegance and grace, both strongly associated with good<br />
demeanour and femininity’, and taught <strong>the</strong> reader how to express <strong>the</strong>se virtues, as<br />
Nerida Jarkey (2015: 197) puts it.<br />
Throughout <strong>the</strong> language, women use more honorific forms than men. Men, but not<br />
women, use words and forms with derogatory overtones. A man, but not a woman, will<br />
say dekee, a derogatory form of dekai ‘big, monstrously huge’. Such ‘male-only’ features<br />
acquire <strong>the</strong> value of what Peter Trudgill (1975: 102) called ‘covert prestige’: <strong>the</strong>y are<br />
tokens of a speaker’s masculinity and an ‘uninhibited quality’.<br />
Sentence-final particles also reflect <strong>the</strong> speaker’s sex. The final particle wa has a<br />
broad meaning of softening a statement, or giving <strong>the</strong> addressee an option to<br />
disagree, or creating empathy. It is exclusively female. Particles ze and zo are only<br />
found in men’s rough speech. A woman, but not a man, will omit <strong>the</strong> copula da when<br />
it is followed by assertion particle yo. A man will say Kirei da yo (pretty COPULA<br />
PARTICLE), and a woman would say Kirei yo (pretty PARTICLE) ‘It is pretty’. 27<br />
A conventional perception of women in many cultures is that <strong>the</strong>y speak more<br />
politely than men. In Japanese this is a rule, and not just a desirable tendency. Katsue<br />
Akiba Reynolds (1991: 133) comments:<br />
As a woman, I am not allowed to say to anybody, even to my younger sibling, Tot-te-kure ‘Get<br />
(it) for me’, using <strong>the</strong> Informal-Benefactive-Imperative. I have to say instead Tot-te. [...]Iam<br />
not allowed to say It-ta ka ‘Did you go?’. I must suppress <strong>the</strong> interrogative marker ka and say<br />
It-ta [...] shifting <strong>the</strong> rising intonation to <strong>the</strong> tense marker. Some rules for gender-marking in<br />
Japanese are categorical, while in English rules are variable.<br />
Only under special circumstances can a man use female forms. A salesman who<br />
deals primarily with female customers may use <strong>the</strong> ‘elegant’ prefix o- ‘as if ’ he were a<br />
woman (Shibatani 1990: 374).<br />
The essence of femininity—as reflected in <strong>the</strong> traditional norms of Japanese—is<br />
softness, politeness, and deference. A female speaker of Japanese is supposed to talk<br />
unassertively, politely, deferentially—<strong>the</strong>se are <strong>the</strong> positive attributes of a proper<br />
woman’s behaviour, called onna-rasiku ‘as expected of women’. This behaviour fitted<br />
in well with <strong>the</strong> Confucian doctrine of ‘men superior, women inferior’—<strong>the</strong> backbone<br />
of <strong>the</strong> hierarchical structure of Japanese society, where women were wives and<br />
mo<strong>the</strong>rs, or, in Pfarr’s(1984: 224) words, ‘<strong>the</strong> lowest rung on <strong>the</strong> entire social ladder,<br />
subordinated within <strong>the</strong> feudal hierarchy and within <strong>the</strong> family hierarchy as well’.<br />
When women and men were given equal rights by <strong>the</strong> new constitution after <strong>the</strong> end
150 9 When women and men speak differently<br />
of <strong>the</strong> Second <strong>World</strong> War, <strong>the</strong> society transformed. Women were no longer confined<br />
to home. Women as supervisors, administrators, teachers, lawyers, doctors, or<br />
colleagues could now—in <strong>the</strong>ory at least—become equal, or superior to men within<br />
<strong>the</strong> work environment.<br />
As social roles changed, so did <strong>the</strong> language. 28 Some women started ‘defeminizing’<br />
<strong>the</strong>ir speech, by—perhaps subconsciously—using male forms. But old habits and<br />
traditions die hard. Reynolds (1991: 138–40) describes <strong>the</strong> difficulties for female<br />
teachers who had to address mixed classes of boys and girls. A 26-year-old female<br />
principal lamented in an interview with a newspaper reporter (<strong>the</strong> Asahi, 18 July<br />
1984), ‘I tend to speak rough language with an imperative tone in spite of my efforts<br />
not to, perhaps, because I am a teacher. I always think regretfully that this is not good<br />
for me.’ The polite, submissive, and differential overtones of female speech in<br />
Japanese seem hardly compatible with a position of authority. A female principal of<br />
a school or a college cannot display many of <strong>the</strong> linguistic traits typical of women—<br />
she ‘must avoid giving an impression in her talk that she is indecisive, indirect or<br />
picky. At <strong>the</strong> same time she cannot be as authoritative as a male counterpart would<br />
be’: this might not be seen as a good example by traditionally minded parents. Every<br />
woman has to face <strong>the</strong> problem of how feminine she wants to sound according to<br />
her own personality, her role and relationships, and <strong>the</strong> situation, and make her own<br />
decision. 29<br />
Overall, <strong>the</strong> strict divide between male and female language appears to be dwindling.<br />
In <strong>the</strong> late 1970s, junior high school girls in Tokyo were referring to <strong>the</strong>mselves<br />
as boku ‘I:male’ in <strong>the</strong> school environment. When interviewed, <strong>the</strong>y explained that<br />
<strong>the</strong>y couldn’t compete with boys in classes, games, and fights if <strong>the</strong>y were to call<br />
<strong>the</strong>mselves watasi ‘I:female’. Ryooko Ozawa, a woman politician elected as a city<br />
council member of a large satellite city of Tokyo in 1971, dismissed female speech as<br />
‘a product of <strong>the</strong> idle life of housewives’. In her own position, when urgent decisions<br />
are to be made, ‘she notes that her own speech tends to become more masculine as<br />
she tries to convey her ideas as precisely and emphatically as possible’. And this is<br />
indeed <strong>the</strong> impression her way of speaking conveys (Reynolds 1985: 41–2).<br />
In her study of a ‘less-feminine speech’ among younger Japanese women,<br />
Okamoto (1995) shows that younger women (between 18 and 34 years of age) no<br />
longer speak <strong>the</strong> way textbook principles of Japanese ‘female speech’ tell <strong>the</strong>m to.<br />
Female college students, professional, and self-employed women are no longer averse<br />
to using male (not female) sentence endings, and even expressions commonly<br />
perceived as strongly masculine, or even vulgar—aitsu ‘that guy’, bakayaroo ‘stupid’,<br />
yabai ‘troublesome’, and dekai ‘big, humongous’. Homemakers lag behind, using<br />
forms traditionally classed as ‘female’. Older men lament <strong>the</strong> changes in female<br />
language—calling it ‘tasteless Japanese’ (<strong>the</strong> title of Okamoto’s paper).<br />
Changes have gone <strong>the</strong> o<strong>the</strong>r way round, too. Younger men in <strong>the</strong> 1970s were not<br />
averse to using female sentence endings and subordinators. And young women at
9.1 ‘Male’ and ‘female’ dialects 151<br />
that time were freely using <strong>the</strong> copula da followed by <strong>the</strong> particle yo—a traditional<br />
feature of male speech.<br />
Modern Japanese women speak differently depending on <strong>the</strong>ir social status,<br />
workplace roles and relationships, surroundings, and <strong>the</strong> impression <strong>the</strong>y want to<br />
convey. Their self-image and <strong>the</strong>ir identity as social beings come through in <strong>the</strong> ways<br />
<strong>the</strong>y speak. Male and female dialects are indicative of Social <strong>Gender</strong>, and are<br />
influenced by it. In Reynolds’s (1991: 141) words, ‘<strong>the</strong> female/male language dichotomy<br />
in Japanese is not a mere differentiation of <strong>the</strong> two sexes but it reflects <strong>the</strong><br />
structure of a society where women were defined as <strong>the</strong> inferior sex’. Growing<br />
equality between male and female Social <strong>Gender</strong>s is enhancing linguistic equality.<br />
The label ‘Japanese women’s language’ no longer fully reflects <strong>the</strong> exact ways in<br />
which modern women speak <strong>the</strong> language. As Okamoto (1995: 317) puts it, ‘Japanese<br />
women strategically use particular speech styles to communicate desired pragmatic<br />
meanings and <strong>the</strong> images of self.’ And some female forms are penetrating <strong>the</strong> speech<br />
of younger males, as men ‘continue to adapt to <strong>the</strong> ongoing changes in Japanese<br />
gender roles and gender ideologies’ (319). Language use is gradually shaped by social<br />
change—we return to this in Chapter 11. 30<br />
Different forms for ‘I’—dichán for women, and phŏm for men—are a notable<br />
feature of female and male speech in Thai, a Tai-Kadai language with a large number<br />
of honorific registers. 31 The pronoun dichán was originally used by royal and highclass<br />
males; it has become restricted to women since <strong>the</strong> early twentieth century.<br />
Phŏm is a high speech level form used by men when talking to someone older or<br />
higher in status than <strong>the</strong> speaker, or as a default term, with no honorific implications.<br />
The female form dichán carries overtones of a social distance between <strong>the</strong> speaker<br />
and <strong>the</strong> addressee: it is avoided in informal situations. The way urban professional<br />
women are using dichán is gradually becoming parallel to <strong>the</strong> way men in general use<br />
phŏm to refer to <strong>the</strong>mselves. Social changes affect o<strong>the</strong>r means of self-reference.<br />
Traditionally, women would use kinship terms or names to refer to <strong>the</strong>mselves in a<br />
womanly, submissive, and humble way. Now men are also using nicknames to refer<br />
to <strong>the</strong>mselves—something only children and women would do in earlier times. But<br />
changes proceed slowly. If a woman uses dichán to refer to herself in a conversation<br />
with senior colleagues, especially females, she may be classed as ‘arrogant’. A young<br />
woman new in her high-ranking position may find it hard to always refer to herself as<br />
dichán. But if she does not, she may get criticized.<br />
Yingluck Shinawatra, Thailand’s first female Prime Minister, referred to herself<br />
with her nickname pu: ra<strong>the</strong>r than with dichán ‘I’, judged more appropriate in a<br />
public speech. She was under fire in <strong>the</strong> news media—some accused her of mixing up<br />
private life with state affairs, that is, representing herself not as <strong>the</strong> government leader<br />
but as a ‘normal’ citizen and a woman.<br />
Social changes in women’s place in society affect male and female dialectal form, as<br />
tokens of changes in Social <strong>Gender</strong>. We now turn to gender-variable ways of speaking.
152 9 When women and men speak differently<br />
9.2 <strong>Gender</strong>-variable skills: men’s and women’s speech practices<br />
Every language has a pool of linguistic devices which reflect <strong>the</strong> speaker’s position in<br />
a society (in terms of class, wealth, or education), <strong>the</strong>ir attitudes, and <strong>the</strong> way <strong>the</strong>y<br />
want to be perceived by o<strong>the</strong>rs. In every society, certain ways of speaking come to be<br />
associated with social status, economic conditions, employment, and educational<br />
opportunities. Each of <strong>the</strong>se may be linked to male and female practices, and ways of<br />
communication. Linguistic forms in gender-variable languages have sets of pragmatic<br />
overtones which can be deployed by men and women depending on <strong>the</strong> image<br />
<strong>the</strong>y want to construct, inseparable from <strong>the</strong>ir societal status. In Sara Mills’s (2003a:<br />
239) words, being male or female ‘cannot simply be correlated with <strong>the</strong> use of<br />
particular linguistic forms or strategies’—in contrast to <strong>the</strong> gender-exclusive languages<br />
we have just discussed.<br />
In a classic paper, Elinor Ochs (1992: 342) makes a similar point, especially<br />
important for gender-variable languages:<br />
Knowledge of how language relates to gender is not a catalogue of correlations between<br />
particular linguistic forms and sex of speakers, referents, addressees and <strong>the</strong> like. Ra<strong>the</strong>r,<br />
such knowledge entails tacit understanding of 1) how particular linguistic forms can be used<br />
to perform particular pragmatic work (such as conveying stance and social action) and 2)<br />
norms, preferences and expectations regarding <strong>the</strong> distribution of this work vis à vis particular<br />
social identities of speakers, referents and addressees.<br />
‘<strong>Gender</strong>-variable’ ways are particularly visible in traditional societies where women’s<br />
and men’s spheres of labour, socialization, and concomitant stereotypes are<br />
relatively clear-cut.<br />
9.2.1 Social status, and women’s speech<br />
In traditional societies, <strong>the</strong> way women choose to speak may correlate with <strong>the</strong>ir roles<br />
and position. In <strong>the</strong> Mayan-speaking community of Tenejapa, women display extremes<br />
of positive and negative politeness repertoire in <strong>the</strong>ir relationships with men, and with<br />
o<strong>the</strong>r women. They use negative questions as commands and offers, sounding mild and<br />
pleading, and frequently stress agreement with <strong>the</strong> addressee. They employ diminutives<br />
and in-group address forms, expressions like ‘you know’ (yaʔwaʔy) and‘you see’<br />
(yaʔwil), and formulate <strong>the</strong>ir requests and offers in a subdued, non-imposing manner.<br />
A woman would inquire in a mild, roundabout manner, ‘You wouldn’t haveany<br />
chickens to sell?’ ra<strong>the</strong>r than commanding someone to sell her <strong>the</strong> chickens. Women’s<br />
way of talking is more deferential. Men’s speechismore‘matter of fact’. One gets <strong>the</strong><br />
impression that women are more sensitive than men to what <strong>the</strong>y are saying and modify<br />
<strong>the</strong>ir speech accordingly.<br />
Three ethnographic facts provide a reason. First, women are vulnerable to men in a<br />
society where women are likely to be beaten if <strong>the</strong>re is any threat to <strong>the</strong>ir reputation,
9.2 <strong>Gender</strong>-variable skills 153<br />
and women are vulnerable to women as possible sources of damage to <strong>the</strong>ir reputation.<br />
Secondly, women may have to be polite to o<strong>the</strong>r women in a household because<br />
women traditionally move to live with <strong>the</strong>ir husbands, and <strong>the</strong>re is a greater social<br />
distance between women than between men within one family. Thirdly, women<br />
speak more cautiously than men: some patterns of behaviour such as talking to<br />
unrelated males are highly face-threatening. This motivates <strong>the</strong> choice of some<br />
speech strategies—for instance, <strong>the</strong> ubiquitous expression denying knowledge or<br />
responsibility, mas᷈kil ‘I don’t know’. In addition, men purposefully exaggerate <strong>the</strong>ir<br />
brusqueness and matter-of-factness as a display of tough masculinity, and women<br />
elaborate <strong>the</strong>ir ‘polite graciousness as a display of feminine (contrasting to masculine)<br />
values’ (Brown 1980: 131).<br />
Women’s careful, polite, and circumspect discourse in Tenejapa does not mean<br />
that <strong>the</strong>y are entirely powerless. In fact, women make a considerable contribution to<br />
<strong>the</strong> household, and men depend on <strong>the</strong>m in domestic contexts. Women are seen as<br />
‘indispensable to <strong>the</strong> order of things’, and ‘in maintaining and guarding society in a<br />
role parallel to men’. The Tenejapan culture is essentially egalitarian, and downplays<br />
all differences in status and power. But a status difference between men and women<br />
remains: husbands directly command wives to do something for <strong>the</strong>m using simple<br />
imperatives. Women do not. While women accord men (especially unrelated males)<br />
‘marked interactional deference, <strong>the</strong> reverse is not <strong>the</strong> case. Politically—men hold <strong>the</strong><br />
positions that are prestigious and publicly visible, and it is men who make <strong>the</strong><br />
decisions affecting <strong>the</strong> community as a whole. Women’s role in decision making,<br />
while very important domestically, is from a society’s view more or less invisible.’ 32<br />
Malagasy men avoid putting someone else ‘in an uncomfortable or unpleasant<br />
position’, and show ‘respect by avoiding this type of confrontation’. It is men who<br />
tend not to express <strong>the</strong>ir sentiments openly, and avoid creating unpleasant face-toface<br />
encounters. Women are <strong>the</strong> opposite—<strong>the</strong>y tend to speak in a more straightforward<br />
manner, and ‘are associated with <strong>the</strong> direct and open expression of anger<br />
towards o<strong>the</strong>rs’, very much unlike Tenejapan women. Again, this difference has an<br />
ethnographic explanation.<br />
In Malagasy, indirectness is a feature of skilful speech associated with men. Directness<br />
and explicitness—a feature of unsophisticated speech—is characteristic of women<br />
(Keenan 1974: 143). This goes with men—not women—being seen as keepers of<br />
proper traditional ways, and women as bringing in destructive innovations. Table 9.8<br />
summarizes <strong>the</strong> principles of female and male speech use among <strong>the</strong> Malagasy.<br />
Malagasy women are excluded from <strong>the</strong> oratory, <strong>the</strong> major formal speech. Men<br />
shy away from <strong>the</strong> type of talk associated with women—gossip and accusations. Male<br />
speech is stereotyped as indirect or respectful, and women’s speech as direct or<br />
confrontative. And both men and women agree that men’s speech is superior to<br />
women’s. Even <strong>the</strong> historical records of <strong>the</strong> Malagasy language contained warnings<br />
against <strong>the</strong> potential ‘evil’ of women’s tongues. 33 Elaborate verbal skills are associated
154 9 When women and men speak differently<br />
TABLE 9.8. Traits of men’s and women’s speech among <strong>the</strong> Malagasy<br />
INDIRECTNESS<br />
MEN<br />
Skilled speech<br />
Traditional speech ways<br />
Malagasy language<br />
DIRECTNESS<br />
WOMEN<br />
Unsophisticated speech<br />
Contemporary speech ways<br />
European languages<br />
with authority and status. The surface difference between men and women has to do<br />
with difference in status, and speech genres associated with each Social <strong>Gender</strong>. 34<br />
Habits of pronunciation and ways of talking can come to project an image of an<br />
affectionate wife or mo<strong>the</strong>r, a submissive and humble being, or a bossy and commanding<br />
man. Features associated with different images create a linguistic stereotype<br />
of a Social <strong>Gender</strong>. As Cameron (1995: 49) put it, ‘feminine’ and ‘masculine’‘are not<br />
what we ARE, nor traits we have, but effects we produce by particular things we DO’.<br />
We now turn to <strong>the</strong> resources which help create such stereotypes.<br />
9.2.2 Tokens of men and women in gender-variable languages<br />
Women and men can sound differently due to inherent differences in <strong>the</strong>ir physique,<br />
or Natural <strong>Gender</strong>. As Labov (1972: 303–4) puts it in his classic study of sociolinguistic<br />
variation in <strong>the</strong> Martha Vineyard variety of American English, ‘<strong>the</strong>re are of<br />
course physical differences between <strong>the</strong> vocal tracts of men and women to be taken<br />
into account’, but <strong>the</strong>re is enough evidence to believe that ‘<strong>the</strong> sexual differentiation<br />
of speakers is...not a product of physical factors alone...but ra<strong>the</strong>r an expressive<br />
posture which is socially more appropriate for one sex or <strong>the</strong> o<strong>the</strong>r...’.<br />
Women’s voices are higher in pitch than men’s. This feature can be deployed by<br />
women, for a special purpose. A Tenejapan Mayan woman would use exaggerated<br />
falsetto voice to express deference when speaking to her husband and o<strong>the</strong>r women,<br />
projecting an image of a circumspect and submissive being compatible with women’s<br />
behaviour in <strong>the</strong> community. A wider range of pitch differences available to women<br />
is deployed for special effects. Russian-speaking women are reported to use high<br />
pitch for positive feelings, and low pitch for <strong>the</strong> opposite. In 9.8, <strong>the</strong> contrast between<br />
<strong>the</strong> speaker’s delight in finding a pretty mushroom and her disgust on looking at it<br />
more closely (and finding that it is poisonous) is partly encoded in pitch level:<br />
9.8 {Oj kakoj xoros᷈en’kij griboc᷈ek} HIGH PITCH<br />
oh what.masc.sg pretty.masc.sg mushroom.dim.sg<br />
{Fu kakaja gadost’!} LOW PITCH<br />
Yuck what.fem.sg nastiness<br />
‘Oh what a pretty little mushroom! Yuck what a nasty thing!’
9.2 <strong>Gender</strong>-variable skills 155<br />
Women’s vocal tracts are shorter than men’s. Pulmonic ingressive airstream is easier<br />
to achieve with <strong>the</strong> smaller female pharynx and larynx. Female speakers of Tohono<br />
O'odham deploy this device in discourse when <strong>the</strong>y take turns in conversation,<br />
clarifications, and repetition, to construct an atmosphere of conversational intimacy. 35<br />
Male speech can sound more harsh and ‘close-mou<strong>the</strong>d’ than female speech—<br />
contributing to a conventional image of men being tougher and more brusque. On<br />
Martha’s Vineyard, men are more ‘close-mou<strong>the</strong>d’ than women, and use more<br />
contracted areas of phonological space; conversely, women in New York City and<br />
Philadelphia use wider ranges of phonological space than men with more extreme<br />
lip-spreading, for vowels. In Japanese, vowels sound less harsh in women’s speech<br />
than when pronounced by men. 36<br />
Phonetic differences between men and women are often closely linked to <strong>the</strong> level<br />
of education and socio-economic level, and sensitivity to <strong>the</strong> ‘norm’. The pronunciation<br />
of vowels and consonants in Norwich, by men and women, varies depending<br />
on <strong>the</strong>ir social background. Trudgill (1974) analysed pronunciation of <strong>the</strong> velar nasal<br />
[ŋ], and concluded that women tend to use a more prestigious and standard [ŋ] no<br />
matter what <strong>the</strong>ir background was. Men—especially, but not exclusively working<br />
class—used <strong>the</strong> stigmatized [n] more often. In o<strong>the</strong>r varieties of English, too, women,<br />
not men, tend to use more prestigious varieties and stick to <strong>the</strong> linguistic norm. In<br />
Glasgow speech, <strong>the</strong> glottal stop is <strong>the</strong> most overtly stigmatized feature—especially<br />
frequent in <strong>the</strong> speech of working-class groups. Lower-middle-class women used<br />
40 per cent fewer glottal stops than men from <strong>the</strong> same class. Among <strong>the</strong> people<br />
interviewed by Macaulay (1977, 1978), a lower-middle-class woman used fewer glottal<br />
stops than any upper-middle-class informant.<br />
Adherence to norm, clearer pronunciation, and prestigious, ra<strong>the</strong>r than stigmatized<br />
forms, is a common feature of women’s speech. 37 Is this because women are<br />
traditionally more involved in rearing children and thus reproducing <strong>the</strong> society<br />
norms? Or that <strong>the</strong>y feel too insecure to deviate from what is accepted and what is<br />
prestigious and eager to maintain <strong>the</strong> status quo? We return to this in §10.4.<br />
Many scholars have described women’s speech in European languages as more<br />
tentative and emotional than men’s. Women who speak Russian and Brazilian<br />
Portuguese use affectionate diminutives—referring to an object as small and cute<br />
and lovely. A Russian woman will tell a child to dry his ‘little face’ (lic᷈iko, diminutive<br />
of litco ‘face’). 38 Jespersen (1922: 249) quotes Lord Chesterfield’s (The <strong>World</strong>, 5<br />
December 1754) comments on how women ‘take a word and change it, like a guinea<br />
into shillings for pocket money, to be employed in <strong>the</strong> several occasional purposes of<br />
<strong>the</strong> day. For instance, <strong>the</strong> adjective vast and its adverb vastly mean anything...A fine<br />
woman...is vastly obliged,orvastly offended, vastly glad,orvastly sorry. Large objects<br />
are vastly great, small ones are vastly little.’ According to Jespersen, ‘<strong>the</strong>re is no doubt<br />
that he has here touched on a distinctive trait: <strong>the</strong> fondness of women for hyperbole<br />
will very often lead <strong>the</strong> fashion with regard to adverbs of intensity’.
156 9 When women and men speak differently<br />
Excessive use of interjections and emotionally expressive words is often relegated<br />
to <strong>the</strong> domain of ‘female’ language. Yokoyama (1999: 408) notes that ‘Russian men<br />
do not verbalize emotions that are “unworthy of men”, such as fear, pity, or worry.<br />
Women, on <strong>the</strong> contrary, do not hesitate to voice <strong>the</strong>m, and <strong>the</strong>y even tend to choose<br />
exaggerated expressions when <strong>the</strong>y do so.’ Only a woman would say Ja z᷈utko<br />
perez᷈ivala (I awfully suffer, worry, go.crazy about+past.fem.sg) ‘I was awfully worried’.<br />
Women are said to swear less than men, and use fewer taboo words. But none<br />
of <strong>the</strong>se features has proved to be exclusively female, on a closer look.<br />
In <strong>the</strong> English-speaking milieu, men are more likely to interrupt women in<br />
conversations, and to try and ‘silence’ women in online and o<strong>the</strong>r debates. Women’s<br />
response to male dominance in different sex-interaction is face-saving and polite—<br />
characterized by hedges, tentative statements, and affective tags. In same sex talk, <strong>the</strong><br />
general tendency for women is to be ‘cooperative’, while men-to-men talk tends to be<br />
‘competitive’. Women’s ways of talking are generally perceived as mild, tentative, and<br />
oriented towards cooperation with <strong>the</strong> addressee, ra<strong>the</strong>r than with a ‘man-like’ selfassertion.<br />
Robin Lakoff, a founder of studies of women’s language, put <strong>the</strong> relative ‘tentativeness’<br />
of women’s speech on <strong>the</strong> map, as a reflection of <strong>the</strong>ir ‘powerless’ talk<br />
(1975). The powerless talk would include intensifiers (‘that’s a really good idea’),<br />
hedges and qualifiers (‘that’s sort of good’), excessive politeness and tag questions<br />
(‘that’s a good idea, isn’t it?’). 39 In Lakoff ’s words, women ‘are socialised to believe<br />
that asserting <strong>the</strong>mselves strongly isn’t nice and ladylike, or even feminine’.<br />
In reality, things are more complex. ‘Powerless’, submissive, and cooperative ways<br />
of speaking are not exclusively associated with women. 40 For one thing, hedges,<br />
qualifiers, and tag questions have many meanings and many functions. Their use is<br />
not a prerogative of ei<strong>the</strong>r Natural or Social <strong>Gender</strong>. A study by Holmes (1986)<br />
showed that women tend to use a hedge you know more often than men if <strong>the</strong> hedge<br />
expresses speaker’s certainty, as in 9.9:<br />
9.9 and that way we’d get rid of exploitation of man by man all that stuff/you know/<br />
you’ve heard it before.<br />
The same you know can express uncertainty—and this is <strong>the</strong> meaning in which<br />
women use it less than men. Coates (1993: 116–17) shows that women would use<br />
hedges ra<strong>the</strong>r frequently when <strong>the</strong>y speak to each o<strong>the</strong>r, to make sure <strong>the</strong> conversation<br />
does not become too face-threatening. The use of a linguistic form depends on<br />
<strong>the</strong> speaker and <strong>the</strong> addressee, and <strong>the</strong>ir relationships, ra<strong>the</strong>r than just <strong>the</strong> gender.<br />
Hedges and intensifiers are not uniformly ‘male’ or ‘female’. Bradac, Mulac, and<br />
Thompson (1995) interviewed 58 male and 58 female students of <strong>the</strong> University of<br />
California (aged between 18 and 25). It turned out that hedges—especially kind of and<br />
fairly—were dominant in male speech. Women did use really more often than men;<br />
o<strong>the</strong>r intensifiers like pretty much were used by both groups. A general stereotyped
9.2 <strong>Gender</strong>-variable skills 157<br />
perception of women’s speech being milder and replete with unnecessary exaggeration<br />
and signs of doubt turned out not to be true to life, in one university context.<br />
Tag questions in English are equally problematic. By using a tag question (e.g. you<br />
are coming, aren’t you?), a speaker avoids committing <strong>the</strong>mself and thus may also<br />
give <strong>the</strong> impression of not being sure of what <strong>the</strong>y are saying. In female speech, a tag<br />
may express a supportive attitude towards <strong>the</strong> addressee. A statement that women<br />
use tags more often than men under any circumstances has not been fully supported<br />
by hands-on studies. 41<br />
Commands and o<strong>the</strong>r directives by women appear to be more tentative than those<br />
by men when women step into social roles traditionally reserved for men. In English,<br />
male doctors will be more likely to use imperatives. Female doctors will use milder<br />
and more roundabout command strategies. A man would say Have my suit cleaned!<br />
and a woman would use longer and politer ways of expression, Would you drop this<br />
by <strong>the</strong> cleaners on your way? 42 In Russian, an abrupt command in an infinitive<br />
form—a feature of <strong>the</strong> army language—appears to be more common in male than<br />
in female speech.<br />
The ways men and women talk, and <strong>the</strong> forms <strong>the</strong>y use, may depend on who <strong>the</strong>y<br />
talk to. Among <strong>the</strong> Manambu, women—and not men—are involved in child-rearing.<br />
In telling children what to do, women use strong prohibitives and imperative forms.<br />
A dependent clause can be used on its own to express a particularly strong command<br />
which would demand immediate compliance, e.g. Təkərəm da-ku! (chair+LOC sit-<br />
SUBORDINATE.SAME.SUBJECT) ‘Sit on <strong>the</strong> chair (or else)’ (literally, having sat on <strong>the</strong><br />
chair!). Men hardly ever use such forms, and comment that this is <strong>the</strong> way women talk.<br />
Manambu women sound more tentative than men under o<strong>the</strong>r circumstances.<br />
Traditional oratories are <strong>the</strong> domain of men, and women are less accustomed to<br />
public speaking. If a woman tells a story, she typically would start it with a verbal<br />
form Wau? (speak+1st.person.command), literally ‘let me talk, may I talk’ (with a<br />
rising intonation typical of a polar question). The same form is used in turn-taking in<br />
conversation. Men who are not accustomed to speak at public ga<strong>the</strong>rings do <strong>the</strong><br />
same. The man mentioned at 7.3 in §7.2 had violated patrilocal practices, and was<br />
considered something of an outcast with little traditional knowledge or oratory<br />
experience. When he was asked to tell a story of his village, he started it with Wau?<br />
A fur<strong>the</strong>r feature of Manambu women’s talk points in a similar direction. In <strong>the</strong><br />
stories and conversations I collected, <strong>the</strong> modal adverb waiwai ‘maybe’ was used<br />
about twice as often by women as by men, to mitigate <strong>the</strong>ir statements. In many<br />
cases, women sound more realistic than men. A man would say Mival yakna ‘plane<br />
will come’. A woman (or an insecure man who is a poor orator) would qualify that by<br />
using waiwai ‘perhaps’. Manambu men tend to flaunt <strong>the</strong>ir knowledge venturing<br />
unqualified statements even if <strong>the</strong>y do not have enough knowledge to support it. The<br />
archetypal difference between men and women using Manambu lies in <strong>the</strong> images<br />
<strong>the</strong>y automatically project, through words <strong>the</strong>y use.
158 9 When women and men speak differently<br />
Differences between male and female ways of talking go back to <strong>the</strong> way children<br />
are brought up. The most salient difference between Kaluli girls and boys is to do<br />
with language: girls consistently employ <strong>the</strong> direct instruction routine, using ɛlɛma<br />
‘say like that’, to <strong>the</strong>ir mo<strong>the</strong>rs and especially to younger children. They replicate <strong>the</strong><br />
ways in which <strong>the</strong>y used to be instructed—as Bambi Schieffelin (1987: 259) puts it,<br />
‘girls engage in verbal routines that are considered within <strong>the</strong> women’s domain and<br />
boys do not’.<br />
What can be perceived as non-aggressive, reconciliatory, cooperative ‘female’ speech<br />
is also used by those who are on <strong>the</strong>ir defensive—for instance, men and women in a<br />
courtroom environment. Linguistic forms and ways of speaking reflect social status<br />
and <strong>the</strong> ways people—men or women—want to be viewed in different situations. If a<br />
woman in a domestic context uses tag questions to facilitate conversation, she would<br />
create an image of someone deferential or subordinate. And if a woman doctor uses <strong>the</strong><br />
same strategy in her consulting rooms, she will be seen as an effective and competent<br />
professional. 43<br />
The differences between what is perceived as male or female language in modern<br />
urban and socially fluid societies in America, Europe, and Australia are best interpreted<br />
as ‘surface reflections of such basic social dimensions as power and solidarity’<br />
(Sherzer 1987: 117). The patterns of Social <strong>Gender</strong> created on <strong>the</strong> basis of o<strong>the</strong>r social<br />
parameters merge and overlap.<br />
When speakers change <strong>the</strong>ir Natural and Social <strong>Gender</strong>, does <strong>the</strong>ir language<br />
change? This is what we turn to next.<br />
9.3 The o<strong>the</strong>r genders<br />
Gay, lesbians, and transgender people deploy linguistic features to construe <strong>the</strong>ir<br />
male or female identity, and create an image of a female or a male. These traits reflect<br />
<strong>the</strong> stereotypes of male and female speech in a society.<br />
Phonetic features which are perceived, by listeners, as characteristic of ‘gay men’s’<br />
speech in English include higher pitch—a suprasegmental feature associated with<br />
female speech. O<strong>the</strong>r ‘gay men’s’ features are hyperstandard pronunciation such as<br />
longer duration of sibilant fricatives, or ‘hissy’ s and z, and <strong>the</strong> release of word-final<br />
stops. In contrast, lesbian speech includes a narrow, male-like, pitch-range and flat<br />
intonation patterns. 44<br />
Standard Brazilian Portuguese requires number agreement on nouns in combination<br />
with a definite article, e.g. a-s coisa-s (<strong>the</strong>.feminine-plural thing-plural) ‘<strong>the</strong><br />
things’. In <strong>the</strong> informal spoken language, <strong>the</strong> plural marker on <strong>the</strong> noun is often<br />
omitted—one hears as coisa ‘<strong>the</strong> things’. These forms are considered substandard.<br />
Gay men and straight women are less likely to delete <strong>the</strong> plural marker than straight<br />
men. This is why gay male speakers of Portuguese of São Paulo in Brazil are said to
9.3 The o<strong>the</strong>r genders 159<br />
speak a ‘better Portuguese’. 45 This agrees with <strong>the</strong> idea that women tend to adhere to<br />
more standard varieties than men—we turn to this in §10.4.<br />
Extensive use of diminutives in Brazilian Portuguese is a feature of straight<br />
women, and also of feminine gay men and non-butch lesbians. In contrast, straight<br />
men, masculine gay men, and butch lesbians avoid diminutives, to assert <strong>the</strong>ir ‘male’<br />
status. Gay men in Tokyo and Osaka use ‘female’ language patterns to assert <strong>the</strong>ir<br />
identity. 46 Transgender male to female speakers of Thai have a broader repertoire of<br />
gendered pronouns than ei<strong>the</strong>r men or women. They tend to use female-specific<br />
pronouns, and final particles. A sentence final particle hâʔ used to be a feature of<br />
female speech; nowadays it ‘seems to be reserved for feminine males’. 47<br />
Resources of Linguistic <strong>Gender</strong> can also be deployed. Brigitte Martel described her<br />
own experience as a transsexual male who changed to being female in <strong>the</strong> 1981<br />
autobiography, with a telling title Né homme, comment je suis devenue femme (Born:<br />
masculine.singular man, how I became (feminine.singular) a woman). French gay<br />
men use feminine forms to address each o<strong>the</strong>r and talk to <strong>the</strong>ir partners, addressing<br />
<strong>the</strong>m as ma petit-e chéri-e (my.fem.sg little-fem.sg darling-fem.sg) ‘my little darling’.<br />
Lesbians will use masculine forms, to project <strong>the</strong> image of masculinity. 48 The use of<br />
Linguistic <strong>Gender</strong> by American gay men has affective overtones: Rudes and Healey<br />
(1979) found that gay men used ‘she’ to refer to someone particularly attractive; ‘he’<br />
was associated with negative attitudes.<br />
Literary fiction reflects stereotypes associated with butch and femme lesbians as<br />
‘manly’ and ‘womanly’ in <strong>the</strong> ways characters talk. Butch lesbians are portrayed as<br />
curt, avoiding expressions of emotion, using sexually explicit vocabulary. Femme<br />
lesbians are portrayed as more ‘lady-like’. All <strong>the</strong>se features do not necessarily reflect<br />
what gay men and lesbian women do. Ra<strong>the</strong>r, <strong>the</strong>y help create recognizable stereotypes<br />
which may <strong>the</strong>n be deployed by gays and lesbians to create a male-like, or a<br />
female-like image in <strong>the</strong> eyes and ears of <strong>the</strong> listener. 49<br />
We can recall, from §9.1.5, that ‘two-spirit’, or transgender people, among <strong>the</strong><br />
Lakhota used to take on <strong>the</strong> dialect of <strong>the</strong>ir newly acquired Social <strong>Gender</strong>. Fletcher<br />
and La Fleche (1972: 132) tell a story of an Osage man who became like a woman and<br />
started ‘unconsciously using <strong>the</strong> feminine terminal of speech’. 50<br />
The hijras of India are perhaps <strong>the</strong> best described instance of a ‘third sex’. 51 They<br />
are born as males and raised as boys. Their female predilections appear later in life.<br />
Hijras—also referred to as ‘hermaphrodites’ or ‘transvestites’—have a special position<br />
in <strong>the</strong> social structure, as <strong>the</strong> ‘third gender’. Their thick male voices are in stark<br />
contrast to <strong>the</strong>ir female clothing and woman-like looks and gait.<br />
Hijras deploy <strong>the</strong> resources of Linguistic <strong>Gender</strong> to play out <strong>the</strong>ir Social <strong>Gender</strong><br />
identity and interpersonal relations. Hindi-speaking hijras refer to <strong>the</strong>mselves as<br />
women would—using <strong>the</strong> feminine (and not <strong>the</strong> masculine) Linguistic <strong>Gender</strong> in<br />
pronouns, adjectives, and habitual, progressive, and o<strong>the</strong>r verb forms. When a hijra<br />
decides to dress up as a man, <strong>the</strong>y switch to masculine Linguistic <strong>Gender</strong>. The use of
160 9 When women and men speak differently<br />
feminine Linguistic <strong>Gender</strong> to talk about <strong>the</strong>mselves is tantamount to <strong>the</strong> projection<br />
of ‘non-masculine’ identity. Some use masculine forms when <strong>the</strong>y talk to men, and<br />
feminine forms when <strong>the</strong>y talk to women. For a hijra, using a feminine form to<br />
ano<strong>the</strong>r hijra may be symbolic of solidarity and familiarity. Masculine forms can<br />
mark respect, if addressed to a man. But using a male name, or a masculine form, by a<br />
hijra to a hijra reflects contempt, and is insulting. Hijras manipulate masculine and<br />
feminine Linguistic <strong>Gender</strong> forms to establish solidarity, power, and social relations<br />
with <strong>the</strong>ir peers and <strong>the</strong> outside world. Linguistic <strong>Gender</strong> helps construct images and<br />
identities of Social <strong>Gender</strong> (in <strong>the</strong> context of partially ambiguous Natural <strong>Gender</strong>,<br />
and <strong>the</strong> in-between status of <strong>the</strong> third ‘sex’).<br />
Male-to-female transsexuals and transvestites in <strong>the</strong> English-speaking world try<br />
and sound like women—<strong>the</strong>y raise <strong>the</strong> pitch of <strong>the</strong>ir voices and use a wide range of<br />
intonation patterns. They use emotionally coloured words <strong>the</strong> way women are<br />
perceived to use <strong>the</strong>m, and adopt female names. Transvestites from sou<strong>the</strong>rn Brazil<br />
use <strong>the</strong> feminine Linguistic <strong>Gender</strong> when <strong>the</strong>y talk to each o<strong>the</strong>r, and refer to<br />
<strong>the</strong>mselves as bicha (a grammatically feminine noun, roughly translatable as ‘fag’)<br />
ra<strong>the</strong>r than travesti ‘transvestite’, a grammatically masculine noun. They would<br />
employ masculine gender to refer to <strong>the</strong>mselves at <strong>the</strong> time <strong>the</strong>ir body transformation<br />
took place and when talking about <strong>the</strong>ir family relationships. A transvestite<br />
explained to <strong>the</strong> researchers that for her mo<strong>the</strong>r, she will always remain a male son,<br />
filho homem. O<strong>the</strong>r transvestites <strong>the</strong>y did not identify with were talked about as male<br />
travestis. Masculine and feminine gender forms are manipulated to create an image<br />
of a Social <strong>Gender</strong>. 52 The French male-to-female transsexual Georgine Noël switches<br />
between masculine and feminine Linguistic <strong>Gender</strong> throughout her biography.<br />
According to Livia (1997), switches from masculine to feminine convey a sense of<br />
success and triumph in Noël’s newly acquired female self. Her switches back to <strong>the</strong><br />
masculine form deploy frustration and failure, in her male body. Linguistic <strong>Gender</strong><br />
has acquired a symbolic value of identity mark.<br />
9.4 Male speech, female speech: a summary<br />
In every language and in every society, Social <strong>Gender</strong> correlates with how men and<br />
women use language. A few languages have, or used to have, ‘gender-exclusive’ male<br />
and female dialects. Men and women would use different phonological systems (as in<br />
Karajá, from Brazil), or even resources from different languages (as in Island Carib).<br />
Male and female dialects would employ different forms of pronouns, various particles,<br />
and o<strong>the</strong>r grammatical forms, as in Yanyuwa (from Australia), Kokama-<br />
Kokamilla (from Peru), Japanese, and Thai. The choice of male or female dialect<br />
may depend on whe<strong>the</strong>r <strong>the</strong> speaker, and <strong>the</strong> addressee, are male or female, as in<br />
Kũrˌux, a Dravidian language, and <strong>the</strong> now extinct Biloxi and Tunica, from <strong>the</strong><br />
USA. The use of ‘gender-exclusive’ dialects could go beyond Natural <strong>Gender</strong> into
9.4 Male speech, female speech 161<br />
<strong>the</strong> domain of Social <strong>Gender</strong>: <strong>the</strong> male variety will be <strong>the</strong> way an initiated man would<br />
speak. And a non-initiated man—incomplete in his manhood—will speak ‘like a<br />
woman’. Male and female gender dialects could be manipulated to achieve special<br />
effects—including jokes. This is similar to how Linguistic <strong>Gender</strong>s can be reversed,<br />
for special purposes. As many minority languages and <strong>the</strong>ir social structures disappear,<br />
so do <strong>the</strong> male and female dialects.<br />
Social changes in <strong>the</strong> status of men and women affect <strong>the</strong> ways in which male and<br />
female dialects are used. With <strong>the</strong> rising equality of men and women in modern<br />
Japan, women no longer use <strong>the</strong> female forms as <strong>the</strong> textbooks say. To assert<br />
<strong>the</strong>mselves, many employ what used to be male patterns. As Okamoto (1995: 309)<br />
puts it, ‘today’s notion of Japanese women’s language can be seen as a lasting legacy’<br />
of <strong>the</strong> late nineteenth-century policy when ‘government officials and intellectuals<br />
sought to standardize <strong>the</strong> language and to discipline women according to <strong>the</strong> ideal of<br />
ryoosai kenbo “good wife and wise mo<strong>the</strong>r”’, based on <strong>the</strong> speech style of traditional<br />
women in <strong>the</strong> middle and upper-middle classes in Tokyo. Alongside o<strong>the</strong>r symbols—<br />
clothing, bearing, and so on—this ideal was constructed, and it is now changing as<br />
<strong>the</strong> society and women’s roles transform. Thai-speaking women have started using<br />
pronouns in <strong>the</strong> same way as men do. Women, and <strong>the</strong>ir verbal expression, are a<br />
barometer of change—we return to this in Chapter 11.<br />
Subtle differences between male and female speech in gender-variable<br />
languages—including English—are more difficult to capture. There is no oneto-one<br />
correspondence between a grammatical or lexical feature and women’s or<br />
men’s speech. On <strong>the</strong> contrary: most features are deployed by women, and by men,<br />
to create an impression and construct social relations (of which Social <strong>Gender</strong><br />
relations are only one part). The resources of Natural <strong>Gender</strong>—such as higher pitch<br />
of women’s voices—can be deployed to express affection or delight. Subdued,<br />
tentative, emotionally charged, affectionate, and hypercorrect speech underlies<br />
a ‘women’s language’ stereotype (see Lakoff 1979: 53). Salient features of this<br />
stereotype—in combination with Linguistic <strong>Gender</strong> if available—can be manipulated<br />
to assert Social <strong>Gender</strong> identity in transgender, gay, and lesbian way of<br />
speaking across <strong>the</strong> world. In Livia’s (1997: 365) words, transgender people reveal<br />
‘resources available in <strong>the</strong> gender system to which more traditional identities have<br />
scant recourse’. 53 Building one’s Social <strong>Gender</strong> identity and constructing one’s<br />
Social <strong>Gender</strong> underlies Linguistic <strong>Gender</strong>s and o<strong>the</strong>r speech distinctions identified<br />
as ‘male’ and ‘female’.<br />
Men’s and women’s activities are clearly and distinctly defined in traditional<br />
societies across Africa, New Guinea, and South America. Then, differences in <strong>the</strong><br />
verbal genres and associated patterns of speaking clearly reflect <strong>the</strong> roles, and thus <strong>the</strong><br />
achieved status, of men and of women. Differences between men and women lie not<br />
just in how <strong>the</strong>y speak, but in what <strong>the</strong>y talk about, and what is off limits to members<br />
of ei<strong>the</strong>r Social <strong>Gender</strong>. This is what Chapter 10 is about.
162 9 When women and men speak differently<br />
NOTES AND SOURCES<br />
1. Competent surveys of male and female dialects are in Bodine (1975), Key (1975),<br />
Günthner (1996), and Trechter (2009); Mithun (1999: 276–80) is a survey of male and<br />
female dialects in North American Indian languages; a brief survey of Amazonian languages<br />
is in Aikhenvald (2012a). Storch (2014: 252–3) discusses male and female dialects<br />
in East African languages. There are many misconceptions about male and female dialects.<br />
Dunn (2014) is a highly flawed and sketchy outline of ‘male’ and ‘female’ dialects, with<br />
numerous mistakes and a peculiar claim that male and female dialects are ‘only attested in<br />
relatively small communities’ (p. 63), backgrounding Japanese and Thai spoken by<br />
millions of people. Male and female dialects are sometimes referred to as ‘genderlects’.<br />
2. Rodrigues (1999: 177), Ribeiro (2009, 2012); Fortune and Fortune (1975, 1987).<br />
3. Flannery (1946: 133–4). What Flannery referred to as tc [ts] was described as c᷈[tʃ] in later<br />
work by Taylor (1982: 101–2).<br />
4. See Bogoras (1922: 665). V. G. (Waldemar) Bogoras was exiled to <strong>the</strong> Russian north for his<br />
revolutionary activities, and spent a substantial amount of time <strong>the</strong>re. Dunn (2000) isa<br />
brief description of what <strong>the</strong> author, as a PhD student, could capture of men’s and<br />
women’s dialects during his brief fieldwork in <strong>the</strong> 1990s.<br />
5. Similarly to Lakhota as described in detail by Sara Trechter, Omaha-Ponca, Quapaw, and<br />
Kansa, extinct members of <strong>the</strong> same Siouan family, had male and female forms for<br />
declarative and also imperative markers: Trechter (1995: 12); John Koontz, p.c. Differences<br />
between male and female speech have been described for Mayrinax, <strong>the</strong> most conservative<br />
dialect of Atayal, an Austronesian language of Taiwan: Li (1982, 1983); see Aalto (1959),<br />
Birtalan (2003) for Kalmyk and Oirat (Mongolic), Rankin (1976) for Moldavian; Camden<br />
(1979: 113) for Tangoa; Hollow (1970: 456) and Mithun (1999: 280) for Mandan, and<br />
LeMaster (1997) for <strong>the</strong> Irish Sign Language.<br />
6. Vallejos (2010: 41–4; 2015), Faust (1963); see also Kroskrity (1983).<br />
7. See Drude (2002, 2006, 2011) on Awetí; Stradelli (1929) for <strong>the</strong> Língua Geral form; Drude<br />
(2006) for <strong>the</strong> relationships within Tupí-Guaraní family and <strong>the</strong> status of Awetí.<br />
8. Kirton (1988: 116).<br />
9. See Haas (1944) for a typology of men’s and women’s speech; Haas (1944: 148–9) on<br />
Koasati, Tunica, and Biloxi; Einaudi (1976: 78–96), Mithun (1999: 280) on Biloxi; Ekka<br />
(1972) onKũr̩ux; Sapir (1929a: 206) on Yana.<br />
10. See Ott and Ott (1983: 26–7), Olza Zubiri, Nuni de Chapi, and Tube (2001: 10); Rose<br />
(2013) discusses only some forms within personal pronominal reference.<br />
11. See Taylor (1977: 89–99); Hoff (1994).<br />
12. Hoff (1994: 162–3).<br />
13. See Hoff (1994: 164), Breton (1978: 16–17, 132), Du Puis (1652: 195–6) on <strong>the</strong> male and<br />
female registers.<br />
14. Douglas Taylor, a major expert in Island Carib, reported a curious switch in Linguistic<br />
<strong>Gender</strong> of some nouns depending on <strong>the</strong> sex of <strong>the</strong> speaker. The term for an ancestral<br />
ritual dogó belongs to <strong>the</strong> masculine Linguistic <strong>Gender</strong> if a woman is speaking, and <strong>the</strong><br />
feminine gender if a man is speaking (Taylor 1951: 44–5, 1959: 204; 1977: 24–43, 72–99).<br />
See also Rat (1898: 311). Note that a partial study based on <strong>the</strong> expatriate community in<br />
<strong>the</strong> USA by Munro (1997) is inconclusive.<br />
15. See also Rankin (1976) on archaic features of female speech in Moldavian and Drude<br />
(2002) on archaic forms in Awetí female speech. Dunn (2000) argues that Chukchi<br />
women’s dialect may have arisen as a result of conscious dialect mixing.
9.4 Male speech, female speech 163<br />
16. Yigezu (1998: 97–9).<br />
17. Kimball (1991: 54, 224–5).<br />
18. Speakers of Kokama-Kokamilla view <strong>the</strong> correct use of genderlects as a marker of linguistic<br />
proficiency (which semi-speakers lack: Vallejos forthcoming).<br />
19. See Flannery (1946) on Gros Ventre, Bradley (1988) on Yanyuwa, Vallejos (2010, 2015) and<br />
also Faust (1963) on Kokama-Kokamilla, Sims and Valiquette (1990) and Maring (1975)on<br />
Keresan, Mithun (1999: 278–8) and Günthner (1996: 453), for a general discussion.<br />
20. Trechter (1995: 106–13).<br />
21. See Bogoras (1901: 98–9). Bradley (1988) reports how Yanyuwa men would use female<br />
dialect forms in <strong>the</strong>ir traditional song cycles. Such occurrences were classed as ‘It’s just <strong>the</strong><br />
dreaming, <strong>the</strong>y are different.’<br />
22. Ribeiro (2012: 149–50).<br />
23. Trechter (1995: 6–7).<br />
24. ‘Possibly <strong>the</strong> reduced female forms constitute a conventionalized symbolism of <strong>the</strong> less<br />
considered ceremonious status of women in <strong>the</strong> community.’<br />
25. See Ide (1991), Shibatani (1990: 371–4), McGloin (1991), Reynolds (1985, 1991). My<br />
discussion here focuses on informal speech in hyoojungo (Standard Japanese) (see<br />
Reynolds 1991, for a brief mention of this issue). In formal speech, while some interesting<br />
differences between male and female speech remain, <strong>the</strong> difference is much less obvious<br />
(Nerida Jarkey, p.c.).<br />
26. Nerida Jarkey commented that ‘even anata is not formal enough for addressing someone’s<br />
superior. It can be used as a polite form to address a stranger.’<br />
27. See Ide (1991: 75–6), McGloin (1991), and also Uyeno (1971: 61–2), Shibatani (1990: 373).<br />
What Shibatani (1990: 373) calls clausal nominalizers no and koto (roughly corresponding<br />
to <strong>the</strong> English subordinator that) may end women’s utterances with an overtone of<br />
surprise and mild admiration. The dubitative ending kasira is also ‘female’ only. Men<br />
and women use different interjections. For instance, a woman would say maa ‘Wow!’ or<br />
ara ‘Oh!’, and a man wouldn’t.<br />
28. Traditional sources date <strong>the</strong> origins of women’s language in Japanese as early as <strong>the</strong> fourth<br />
century. According to Inoue (2002: 393–4, 403–10), <strong>the</strong> consolidation of <strong>the</strong> female<br />
language was due to social changes in modernized Japan in <strong>the</strong> late nineteenth and early<br />
twentieth century. Exclusively female forms were created and imposed as a ‘law’ within <strong>the</strong><br />
context of creating a national language and codifying <strong>the</strong> overall ‘image’ of an ideal<br />
Japanese woman through novels and o<strong>the</strong>r publications. Nakamura (2015) is a comprehensive<br />
history of <strong>the</strong> Japanese female language, its development, and changes in different<br />
social environments.<br />
29. In addition, Ide and Inoue (1992) show that female supervisors in some Japanese workplaces<br />
speak even ‘more’ politely than secretaries, and suggest that this is a strategy for<br />
constructing <strong>the</strong>ir identity as supervisors, and asserting <strong>the</strong>ir authority. Female speech<br />
patterns can be used as a group identity marker, on a par with Tokyo language and o<strong>the</strong>r<br />
‘in-group’ ways of speaking (see Ide 2003).<br />
30. See Gal (1995: 171) and fur<strong>the</strong>r references in Okamoto (1995), on <strong>the</strong> impact of social<br />
changes on Japanese male and female dialects.<br />
31. Anthony Diller (1985) distinguishes at least seven such registers, ranging from a highly<br />
formal ‘official’ to a very informal ‘oral’ one. A fur<strong>the</strong>r form, kraphom ‘I’, is used just by men,<br />
when talking to a superior. An informal form nu ‘I’ is used just by women in informal<br />
contexts (when talking to older relatives). See Haas (1944), Chirasombutti and Diller (1999),<br />
and Attaviriyanupap (2015: 384, 392–4, 397) on <strong>the</strong> usage of male and female forms in Thai.
164 9 When women and men speak differently<br />
32. Brown (1980: 133); see also Brown (1993: 159).<br />
33. Ochs (1992: 57), Hill (1987: 159). To say that women are always more, or less, polite than<br />
men is a simplification.<br />
34. Irvine (1978, 1979). See Gal (1989, 1995: 170–4) on <strong>the</strong> power of silence and verbal skills in<br />
constructing gender stereotypes.<br />
35. See Yokoyama (1999: 403–6), largely based on Zemskaya, Kitajgorodskaja, and Rozanova<br />
(1993); Hill and Zepeda (1999) on Tohono O'odham.<br />
36. See Hagège (2004: 111–12); Labov (1972: 303–4). Yokoyama (1999) gives examples of<br />
vowel leng<strong>the</strong>ning just by women, to express emphasis in Russian.<br />
37. See Coates (1993: 61–86) and Gordon and Heath (1998) for fur<strong>the</strong>r examples.<br />
38. This may be partly role-based—women, and not so much men tend to take care of<br />
children, and thus use diminutives as ‘baby-talk’. We saw a similar association between<br />
diminution and cuteness in feminine Linguistic <strong>Gender</strong>s in languages where feminine<br />
Linguistic <strong>Gender</strong> is associated with size (Chapters 6–7). See Zemskaya, Kitajgorodskaja,<br />
and Rozanova (1993), Yokoyama (1999: 406–7) on Russian; Mendes (2014a) on Brazilian<br />
Portuguese. Sadiqi (2003: 154–5) and Gordon and Heath (1998: 437) mention that Arabic<br />
women tend to use more diminutives than men.<br />
39. See Lakoff (1975: 54); Tannen (1990), and a critical survey in Coates (1993: 116–17).<br />
Trechter (1995: 108) reports that Lakhota women do not use enclitics which imply strong<br />
opinion and assertion: this is not part of <strong>the</strong>ir repertoire.<br />
40. Coates (1993: 132–4; 2012) offers an in-depth analysis of dominance, difference, and<br />
‘constructing’ one’s discourse using gender tokens.<br />
41. Dubois and Crouch (1975) were among <strong>the</strong> first ones to challenge <strong>the</strong> assumption of<br />
women’s overuse of tags as a sign of ‘powerless’ discourse; see Coates (1993: 119–24),<br />
Holmes (1984), and Cameron, McAlinden, and O’Leary (1989).<br />
42. See Key (1975: 76–7); see Coates (1993: 124–6) on female and male commands in<br />
interaction with children and patients.<br />
43. This point was incisively made by Cameron (1992: 24). That women are more verbose<br />
than men also turned out to be untrue, on close examination (Coates 1993). The impression<br />
of female verbosity prominent in what Coates (1993) calls ‘folklinguistics’ comes from<br />
patterns of polite, and over-emotional speech popularly—but not necessarily correctly—<br />
associated with a dubious ideal of ‘standard average women’.<br />
44. See a summary by Barrett (2006), Podesva, Roberts, and Campbell-Kibler (2002), Christ<br />
(1997), Gaudio (1994), Avery and Liss (1996) discuss phonetic features of speech perceived<br />
as ‘gay’; see also Zwicky (1997) and Levon (2006); a useful survey of literature on<br />
gay and lesbian language is in Kulick (2000).<br />
45. Mendes (2014b). See James (1996) for a survey of opinions on how women use more<br />
standard and prestige varieties than do men.<br />
46. See Mendes (2012, 2014b) on diminutives in male and female speech in Brazilian Portuguese<br />
as tokens of female speech and of speech patterns of feminine gay men. The association of<br />
diminutives with ‘female’ or ‘male’ image is also reflected in a reaction straight men have to<br />
those who overuse diminutives. See also Jones (2014) on English-speaking butch lesbians;<br />
Ogawa and Smith (1997) on Japanese-speaking gay men in Tokyo and Osaka.<br />
47. Attaviryanupap (2015: 393–5).<br />
48. Livia (1997: 358–9), Pastre (1997: 372–3).<br />
49. See <strong>the</strong> detailed and entertaining account in Livia (1995). Gay and lesbian norms of<br />
structuring discourse and conversation reflect specific ‘cooperation’ based on <strong>the</strong> affirmation<br />
of gay and lesbian identity ra<strong>the</strong>r than identifiable ‘male’ and ‘female’ stereotypes: see<br />
Barrett (2006: 319–20 and references <strong>the</strong>re).
9.4 Male speech, female speech 165<br />
50. Trechter (1995: 6–7) alludes to this. Berdaches have been documented for a number of<br />
peoples in <strong>the</strong> North American Great Plains, including Lakhota, Arapaho, Mandans,<br />
Poncas, Quapaw, Kiowa, and many o<strong>the</strong>rs. For fur<strong>the</strong>r features of male and female<br />
Berdaches, see Whitehead (1981: 88–93), Lang (1998), Roscoe (n.d.). No fur<strong>the</strong>r details<br />
of female language use are available in <strong>the</strong> literature. The term ‘berdache’ has pejorative<br />
overtones, and is nowadays used to refer to gay people. The traditional berdaches are now<br />
all but gone, as a result of <strong>the</strong> Christian influence.<br />
51. See Hall and O’Donovan (1996) and references <strong>the</strong>re, especially Nanda (1990). Hijras have<br />
an established position in <strong>the</strong> society: <strong>the</strong>y are expected to sing and dance at births and<br />
weddings, and receive payments for this. Some view <strong>the</strong>mselves as nei<strong>the</strong>r men nor<br />
women, and also as ‘deficiently’ masculine and ‘incompletely’ feminine. Despite <strong>the</strong>ir<br />
official recognition in India, Nepal, Pakistan, and Bangladesh, <strong>the</strong>y remain a marginalized<br />
minority.<br />
52. Borba and Ostermann (2007). Kulick (1998: 216) mentions a preference for <strong>the</strong> feminine<br />
term bicha among Brazilian transvestite prostitutes from <strong>the</strong> north-east to refer to<br />
<strong>the</strong>mselves, as part of <strong>the</strong>ir female identity.<br />
53. Besnier (2003) investigates <strong>the</strong> ways in which Tongan fakaleiti, or women-like men, subtly<br />
imitate female speech patterns in <strong>the</strong> language without Linguistic <strong>Gender</strong>. Third gender<br />
(or male to female transsexuals) have been described for numerous societies, but with little<br />
attention to linguistic practices (e.g. Wikan 1977, 1978 on Omani xanīth). Speech of<br />
female to male transsexuals remains a matter for fur<strong>the</strong>r studies (cf. Barrett 2006: 320).
10<br />
The rituals of gender<br />
Social <strong>Gender</strong>s in traditional societies tend to be associated with different domains,<br />
and different speech styles. This is where inequality between <strong>the</strong> male and <strong>the</strong> female<br />
Social <strong>Gender</strong>s comes to light. Special languages and language registers can come to<br />
be used in male-only rituals. A whole set of terms may be forbidden to women.<br />
Women may have to exercise special caution and use deferential forms to address<br />
<strong>the</strong>ir relatives through marriage. Women are often seen as keepers and promoters of<br />
prestigious linguistic norm, and of traditional language. Or <strong>the</strong>y can be viewed as a<br />
dangerous ‘o<strong>the</strong>r’ who lead <strong>the</strong> society in <strong>the</strong> wrong direction.<br />
10.1 Social <strong>Gender</strong>, speech genres, and speech practices<br />
In many traditional societies, public speeches, oratories, and important origin myths<br />
are associated with <strong>the</strong> male domain. 1 We can recall, from §9.2.1, that Malagasy<br />
women are excluded from <strong>the</strong> oratory—a major formal speech performance. Malagasy<br />
men shy away from gossip and mutual accusations—<strong>the</strong> type of talk associated<br />
with women. Elaborate verbal skills are associated with authority and status, and thus<br />
with men. Male speech is described as indirect or respectful, and women’s speech as<br />
direct or confrontational.<br />
Manambu men used to be responsible for culturally important rituals including<br />
<strong>the</strong> mortuary ritual, <strong>the</strong> Yam Harvest ritual, and initiation. All of <strong>the</strong>m assigned to<br />
<strong>the</strong> masculine Linguistic <strong>Gender</strong> and given importance and value (see §3.4.2). Only<br />
some of <strong>the</strong>se are still performed. Traditionally, <strong>the</strong>re used to be (and still are) two<br />
types of songs—laments about foiled marriages and missing or endangered relatives<br />
(namay and sui), sung by men and by women, and mourning songs gra-kudi (crylanguage)<br />
sung exclusively by women after someone’s death and during <strong>the</strong> mortuary<br />
ritual Keketep which may take place a year later. The Manambu language remains in<br />
active use, but all song genres are on <strong>the</strong>ir way out. The totemic knowledge and <strong>the</strong><br />
proficiency in clan names associated with composing songs of any genre are rapidly<br />
decreasing as older people pass away. Women under 60 (and even many of <strong>the</strong> older<br />
women) do not have enough knowledge to perform a mourning song. According to<br />
Kamibau (in her early sixties), <strong>the</strong>y ‘sing’ in men’s fashion: emitting a high-pitched<br />
<strong>How</strong> <strong>Gender</strong> <strong>Shapes</strong> <strong>the</strong> <strong>World</strong>. First edition. Alexandra Y. Aikhenvald.<br />
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.
10.1 Social <strong>Gender</strong>, speech genres, and speech practices 167<br />
whining sound. This is referred to, ra<strong>the</strong>r dismissively, as du-gər (man-cry) ‘crying<br />
men’s fashion’, orkapa-gər (just, for nothing-cry) ‘just crying, crying for nothing’.<br />
This was a special legacy of women within <strong>the</strong> Manambu tradition. Its recent loss is<br />
deplored.<br />
Women-only genres give women ways of expressing <strong>the</strong>mselves. Warao women,<br />
from Eastern Venezuela, are not supposed to speak at public ga<strong>the</strong>rings. The only<br />
important form of social action for <strong>the</strong>m is a ritual wailing genre called sana. When<br />
someone dies, female relatives of <strong>the</strong> deceased compose and sing sana songs until<br />
<strong>the</strong>y come back from <strong>the</strong> graveyards. In <strong>the</strong>se improvised wailing songs women use<br />
words and ways of speaking only men use in public ga<strong>the</strong>rings, and sound aggressive<br />
and self-assertive: <strong>the</strong>ir status as women gives <strong>the</strong>m <strong>the</strong> right to ‘chastise’ <strong>the</strong>ir<br />
relatives. 2<br />
Male and female speech genres among <strong>the</strong> Kuna of Panama are largely complementary.<br />
3 Men perform political and ritual genres such as public speeches, tribal<br />
origin myths, and magic and curing chants. Women perform songs which are closely<br />
linked to life-cycles—lullabies and laments. Important political matters are discussed<br />
in exclusively male ga<strong>the</strong>rings (which women attend only if <strong>the</strong>y are accused of<br />
wrongdoing, accuse someone else, or are called as witnesses). At ga<strong>the</strong>rings attended<br />
by women and men toge<strong>the</strong>r, women are supposed to listen to <strong>the</strong> chanting and<br />
speaking of male chiefs. In recent years, however, women have started holding<br />
ga<strong>the</strong>rings of <strong>the</strong>ir own, with women leading discussion and practising eloquent verbal<br />
skills. Magic and cures can be viewed as predominantly ‘male’ domains; but <strong>the</strong>re are<br />
both men and women who have magic powers to diagnose a disease and cure it.<br />
Kuna puberty rites—held for girls—involve a strict division of labour between men<br />
and women. A girl for whom <strong>the</strong> ritual is held has her hair cut, by a ritual haircutter<br />
and her assistants (all women), in a special enclosure. Meanwhile, <strong>the</strong> ritual director<br />
of <strong>the</strong> puberty rites—a man and his assistants—perform a long (one-, two-, or threeday)<br />
chant to <strong>the</strong> spirit of <strong>the</strong> flute in a large centrally located house, which ensures<br />
<strong>the</strong> proper running of <strong>the</strong> ritual.<br />
Each of <strong>the</strong> Kuna genres—<strong>the</strong> chants of chiefs, <strong>the</strong> speeches of male political<br />
leaders, <strong>the</strong> curing chants of medical men and women, and <strong>the</strong> chants of pubertyrite<br />
directors—have unique linguistic properties. They involve metaphorical language<br />
and special vocabulary which requires special study, and is not intelligible to<br />
lay people. Both men and women have knowledge and understanding of special<br />
properties of <strong>the</strong>se genres.<br />
As Sherzer (1987: 110) puts it,<br />
There is no question that men’s ritual, formal, and public speech is more diversified and<br />
complex than women’s and that men have more access to and control of political authority<br />
through such speaking practices. At <strong>the</strong> same time, women also have ritual, formal and public<br />
speaking roles and <strong>the</strong>se are always positively valued, by men and women alike, and provide<br />
women with a certain access to and significant involvement in Kuna political life and power.
168 10 The rituals of gender<br />
Kuna men may appear more visible than Kuna women in public life. But women<br />
are equally, if not more, important in social and economic affairs: <strong>the</strong>y are <strong>the</strong> ones<br />
who make, wear, and sell molar blouses which are primary markers of Kuna ethnicity<br />
and culture (and an attraction for tourists).<br />
Men’s and women’s knowledge reflects different spheres of competence. In traditional<br />
societies, such as Kaluli of <strong>the</strong> Western Province of Papua New Guinea, girls<br />
learn about gardening, food preparation, and songs—and o<strong>the</strong>r expressive genres—<br />
that <strong>the</strong>ir mo<strong>the</strong>rs teach <strong>the</strong>m. Boys learn about hunting, and <strong>the</strong> genres which<br />
<strong>the</strong>y will be using in male ceremonies in <strong>the</strong>ir later life. Frogs are hunted only by<br />
women of <strong>the</strong> Nungon tribe in Morobe Province of Papua New Guinea: this food<br />
used to be a taboo for men. As a consequence, women are more proficient than men in<br />
names of frogs and <strong>the</strong>ir species. As Schieffelin (1987: 259) aptly puts it, <strong>the</strong> linguistic<br />
differences between men and women, and boys and girls, concern ‘<strong>the</strong> different<br />
distribution of linguistic resources’, associated with Natural <strong>Gender</strong> and Social <strong>Gender</strong><br />
stereotypes.<br />
‘Male’ speech can be associated with ceremonial genres—an exclusive property of<br />
older men—and thus signal a high social status. Among <strong>the</strong> Acoma Keresan speakers in<br />
<strong>the</strong> American south-west, <strong>the</strong> differences between male and female speech lie in <strong>the</strong><br />
vocabulary, and in pronunciation. The perception of native speakers of Acoma Keresan<br />
that <strong>the</strong>re is a major difference between male and female speech ‘isbasedinparton<strong>the</strong><br />
fact that only older males use <strong>the</strong> age-related kiva speech’—a special ceremonial speech<br />
register. 4 We now turn to <strong>the</strong> registers which only men are allowed to use.<br />
10.2 The languages of manhood<br />
In many traditional societies, Social manhood can only be achieved through initiation.<br />
A number of Aboriginal Australian tribes had (or still have) a special speech<br />
style used only by initiated men on ceremonial occasions and taught to boys at <strong>the</strong>ir<br />
initiation.<br />
The most striking example of a secret men-only language is Damin, until recently<br />
used by initiated men of <strong>the</strong> Lardil tribe, on Mornington island in nor<strong>the</strong>rn Australia.<br />
Damin had eleven out of <strong>the</strong> seventeen consonants found in <strong>the</strong> everyday language,<br />
and at least thirteen additional ones. Among <strong>the</strong>m were nasal clicks, an ingressive<br />
lateral fricative, a bilabial trill, and a glottalized or ejective velar stop. This was <strong>the</strong><br />
only language outside Africa to have clicks, and <strong>the</strong> only one with so much phonetic<br />
complexity in its click sounds. Damin used only about 150 lexemes covering <strong>the</strong> full<br />
range of Lardil vocabulary—one term for all vegetable food, one for all bony fish, and<br />
one for all liquids. The everyday language has nineteen pronouns. Damin had only<br />
two, ‘I’ and ‘o<strong>the</strong>r’—highly unusual for a human language.<br />
Damin used to be taught quite explicitly. The youth who was to be initiated was<br />
simply brought forth, and a Damin speaker would shout <strong>the</strong> vocabulary at him, while
10.3 Secrets, avoidance, and taboos 169<br />
ano<strong>the</strong>r repeated <strong>the</strong> terms in <strong>the</strong> everyday Lardil. Damin was used openly in <strong>the</strong><br />
community: women, children, and uninitiated men had <strong>the</strong> opportunity of hearing it,<br />
and even understanding it. This is unlike <strong>the</strong> initiation language among <strong>the</strong> Warlbiri<br />
of Central Australia—known as <strong>the</strong> ‘upside-down Warlbiri’ (or tjiliwiri, ‘<strong>the</strong> antonymous<br />
language’). This used to be ‘spoken by guardians in <strong>the</strong> presence of junior<br />
novices’, that is, by initiated men in <strong>the</strong> Warlbiri men’s rituals (Hale 1971: 473), and<br />
was not supposed to be known to women or uninitiated men, or discussed with <strong>the</strong>m.<br />
Numerous Yanomami-speaking groups in Venezuela and <strong>the</strong> adjacent area of Brazil<br />
use a special archaic variety of Yanomami called <strong>the</strong> ‘Wayamo’ language. This is a kind<br />
of Yanomami lingua franca, ‘spoken in a formalized and chanted manner between<br />
members of different villages’. All Yanomami men speak Wayamo; women and prepuberty<br />
children can understand it, but are not allowed to use it. The Wayamo language<br />
is particularly useful for communication between villages which speak different (and<br />
hardly mutually intelligible) dialects. It is also employed among villages which speak <strong>the</strong><br />
same, or almost <strong>the</strong> same variety, and on certain occasions even by people from <strong>the</strong><br />
same local group. The Wayamo is mostly used at <strong>the</strong> [reaho] or [yãimowei] ga<strong>the</strong>ring,<br />
when surrounding local groups (living from a day’s journey away to as far as fifteen days<br />
away) are invited for a feast. When <strong>the</strong>y arrive, <strong>the</strong>y first camp in a nearby forest, and at<br />
dawn send one or two representatives to <strong>the</strong> communal house in <strong>the</strong> host village to<br />
plead not to have food withheld from <strong>the</strong>m. This is done in Wayamo. 5<br />
This is where linguistic inequality between male and female Social <strong>Gender</strong>s comes<br />
to light. Across <strong>the</strong> world, male secret ritual languages, and languages associated with<br />
male initiation, explicitly exclude women. There have been no reports of secret<br />
languages confined to women’s puberty rites or o<strong>the</strong>r ceremonies in Australia, or<br />
elsewhere. Dixon (1980: 68) suggests that <strong>the</strong> lack of secret language for women may<br />
be explained by differences in men’s and women’s socialization. While ‘<strong>the</strong>re seems<br />
to be a general tendency for <strong>the</strong> details of male rituals to be taught and explained<br />
simply and explicitly’, ‘women are usually expected to assimilate <strong>the</strong> structure of<br />
female rituals by exposure and experience’. Anne Storch (2011: 81–5) offers a fur<strong>the</strong>r<br />
explanation. The Jukun men, from Nigeria, use secret languages ‘to maintain <strong>the</strong>ir<br />
exclusive control over shrines and spirits’. Women are excluded for two reasons—<br />
because <strong>the</strong> men want to dominate <strong>the</strong>m, and also because <strong>the</strong>y are seen ‘as a<br />
potential danger to <strong>the</strong>ir communities’ coherence, in <strong>the</strong> men’s view’. We now turn<br />
to what women are not supposed to know or to mention.<br />
10.3 Secrets, avoidance, and taboos: what women<br />
are not supposed to know<br />
Sacred objects, and associated rituals, which used to be women’s property and had<br />
been appropriated by men, are a taboo for <strong>the</strong>ir previous owners. According to Jorge
170 10 The rituals of gender<br />
Muniz (p.c., in 2012), <strong>the</strong> only able-bodied elder and healer from <strong>the</strong> Tariana community<br />
of Periquitos, women suffer birth-pains as a punishment for <strong>the</strong>m having tried<br />
to hold on to <strong>the</strong> Sacred flutes. The Yurupary flutes in north-west Amazonia used to<br />
be women’s property, and were taken from <strong>the</strong>m by force. These instruments, kept<br />
hidden under water, represent a powerful Yurupary spirit, and <strong>the</strong>ir sound is his voice.<br />
If a woman catches a glimpse of <strong>the</strong> flute, she has to be punished by death (typically, by<br />
poison).<br />
Tariana women are not supposed to hear <strong>the</strong> word piri ‘Yurupary flute’ (let alone<br />
use it), or any o<strong>the</strong>r words which contain this root. In <strong>the</strong> presence of women this<br />
form used to be replaced with ano<strong>the</strong>r one (not necessarily phonetically similar). The<br />
set of forms for women only was called ina: na-pia-nipe (woman:PL 3pl-hide-NOMIN-<br />
ALIZATION), literally ‘what <strong>the</strong>y hide from women’,orina: na-maña-nipe (women 3plcheat-NOMINALIZATION)<br />
‘<strong>the</strong>ir cheating of women’. The original form—off limits to<br />
women—is referred to as mayakani ‘straight one’.<br />
The Tariana language is severely endangered, and a substantial amount of traditional<br />
knowledge has been lost. Catholic missionaries (permanently present in <strong>the</strong><br />
area since 1925) did <strong>the</strong>ir best to eradicate <strong>the</strong> Yurupary cult as something belonging<br />
to <strong>the</strong> devil. As a result, only snippets of <strong>the</strong> old secret register are still remembered.<br />
The overwhelming majority are names of important locations associated with <strong>the</strong><br />
origin myths and travels of <strong>the</strong> Tariana ancestors. Just a few words relating to <strong>the</strong><br />
objects associated with <strong>the</strong> Yurupary cult (but not <strong>the</strong> flutes <strong>the</strong>mselves) have a<br />
‘hidden’ counterpart. As is common for secret and avoidance registers across <strong>the</strong><br />
world, all <strong>the</strong> words with a ‘hidden’ counterparts are noun roots. 6 Some examples are<br />
in Table 10.1.<br />
Two names of sacred instruments associated with <strong>the</strong> Yurupary flute rituals were<br />
used as place names, by men only. They had a counterpart in <strong>the</strong> ‘hidden’ register.<br />
A stone in <strong>the</strong> Vaupés river called Iʧiru ‘a sacred musical instrument’ has a ‘hidden’<br />
counterpart Ukara-da (sardine-CL:ROUND), and a river called Urupema, a name for<br />
ano<strong>the</strong>r sacred instrument, got <strong>the</strong> name of Karaka-pua (rooster-CL:RIVER)—according<br />
to <strong>the</strong> late Cândido Brito, ‘<strong>the</strong>n women will not know what <strong>the</strong>y are not supposed to’.<br />
The only documented instances of lexical replacement for place names which do not<br />
contain <strong>the</strong> root piri are two locations, Ipanoré and Urubuquara. Ipanoré is said by<br />
some Tariana to be <strong>the</strong> ‘navel’ of <strong>the</strong> world; <strong>the</strong> Tariana <strong>the</strong>mselves are said to have<br />
emerged from <strong>the</strong> rapids of Ipanoré. Even nowadays, <strong>the</strong>se rapids are practically<br />
impassable. The Tariana varieties I have been working on since 1991 are spoken<br />
beyond Ipanoré (all <strong>the</strong> original Tariana dialects in Ipanoré <strong>the</strong>re are long forgotten).<br />
We travel by boat to Ipanoré, and <strong>the</strong>n hire a truck to take us, <strong>the</strong> outboard motor, <strong>the</strong><br />
boat itself, and all our belongings, to <strong>the</strong> o<strong>the</strong>r side of <strong>the</strong> rapids—<strong>the</strong> settlement called<br />
Urubuquara. No one could tell me any stories about <strong>the</strong> importance of Urubuquara.<br />
The ‘hidden’ register is now largely a thing of <strong>the</strong> past. Speakers (all of whom<br />
except two were men) were keen on teaching me—a white woman researcher—<strong>the</strong>
10.3 Secrets, avoidance, and taboos 171<br />
TABLE 10.1. ‘Hidden from women’: tabooed nouns employed in traditional<br />
Tariana<br />
Type<br />
‘Straight’ form not<br />
to be heard by<br />
women<br />
A form used to<br />
‘deceive’ women<br />
Meaning<br />
Place names<br />
containing <strong>the</strong><br />
forbidden root piri<br />
O<strong>the</strong>r mythologically<br />
important place<br />
names not containing<br />
<strong>the</strong> forbidden root<br />
Paraphernalia related<br />
to <strong>the</strong> Yurupary cult<br />
(containing <strong>the</strong><br />
forbidden root piri<br />
‘Yurupary flute’)<br />
Piri-dapana<br />
(Yurupary.flute-CL:<br />
HOUSE)<br />
Piri-na (Yurupary.<br />
flute-CL:VERT)<br />
Piri-pua (Yurupary.<br />
flute-CL:RIVER)<br />
Piri-taku (Yurupary.<br />
flute-CL:POINT)<br />
Piri-ya (Yurupary.<br />
flute-CL:SKIN)<br />
Myaka-dapana<br />
(ancient-THEMATIC-<br />
CL:HOUSE)<br />
Eta-dapana (eagle-<br />
CL:HOUSE)<br />
Mapa-dapana (bee-<br />
CL:HOUSE)<br />
Inari-na (mucura.<br />
rat-CL:VERT)<br />
Inari ite (mucura.<br />
rat POSS+NCL:ANIM)<br />
Inari-taku (mucura.<br />
rat-CL:POINT)<br />
Piawa (?)<br />
Pa-hwa-li-dapana<br />
(IMPERS-LIE-NOM-CL:<br />
HOUSE)<br />
Wayuli-dapana<br />
(vulture-CL:HOUSE)<br />
name of a cave<br />
name of a hill<br />
name of a river<br />
name of a<br />
peninsula<br />
name of a flat<br />
stone<br />
Ipanoré<br />
(a Portuguese<br />
place name)<br />
Urubuquara<br />
(a Portuguese<br />
place name)<br />
piruaɾi yakasolo traditional<br />
necklace; apronlike<br />
clothing<br />
‘straight’ language, to make sure I got everything right. The first ‘hidden’ register<br />
form was spontaneously produced, as an afterthought and a curious fact. After that,<br />
I started consistently asking for a ‘hidden’ register equivalent for each term. The<br />
Tariana men were not inhibited by <strong>the</strong> presence of two Tariana women: <strong>the</strong> oldest<br />
speaker’s daughter and my classificatory older sister Olívia and her mo<strong>the</strong>r Maria—<br />
who had a special status. We can recall, from §7.2, that a particularly strong woman<br />
in Tariana can be referred to as ‘he’. This was how Maria’s children referred to her.<br />
All <strong>the</strong> Tariana stories I recorded contained <strong>the</strong> ‘straight’ form of <strong>the</strong> name for<br />
Ipanoré. In contrast, <strong>the</strong> term consistently given for ‘traditional necklace; apron-like<br />
clothing’ was <strong>the</strong> yakasolo, from <strong>the</strong> ‘hidden register’ (and not piruari, a word many<br />
younger Tariana did not seem to know).
172 10 The rituals of gender<br />
The term mayakani ‘straight, right, correct’ is used to refer to <strong>the</strong> ‘original’ form of<br />
nouns and place names available to men, and hidden from women. This fits in with <strong>the</strong><br />
general image of women one gets through <strong>the</strong> Tariana stories and language in general.<br />
We can recall, from §7.2, that <strong>the</strong> notion of ‘manhood’ in Tariana has positive<br />
overtones. Ricardo Brito, an elder from Santa Rosa, told me a moving story about<br />
how he managed to escape from a jaguar, and had said to himself, at a critical moment:<br />
10.1 nuha-misini tsiãri-naka<br />
I-TOO<br />
‘I am also a man’<br />
man-PRES.VIS<br />
This was a statement of self-encouragement: Ricardo was saying to himself that he<br />
was courageous and ‘manly’ enough to overcome an adverse situation. In ano<strong>the</strong>r<br />
story, a male otter was teaching his children to be brave and resilient, telling <strong>the</strong>m to<br />
be ‘men’.<br />
Throughout Tariana lore, women are portrayed as wayward and outright dangerous.<br />
They spoil everything (and are referred to as manihta-kadite (NEG+think/reason-NEG<br />
+NCL:ANIM) ‘<strong>the</strong> one who does not think’). An ancestral woman ate <strong>the</strong> manioc when she<br />
was told not to—as a result manioc got covered with thick peel, and one has to work<br />
hard at peeling it. Since o<strong>the</strong>r ancestral women ‘didn’t listen to advice’ (pa-kalite-ka<br />
mhema-kade-ka (IMPERS-speak-WHEN NEG+listen-NEG-WHEN)), and slept with a smelly<br />
mucura-rat, <strong>the</strong>ir bodies lost <strong>the</strong>ir pleasant odour and are now smelly. If a man dreams<br />
of a woman before a hunting expedition or ano<strong>the</strong>r important event, he is in mortal<br />
danger.<br />
Menstruating women are particularly perilous: <strong>the</strong> late Graciliano Brito explained<br />
to me that <strong>the</strong> pernicious evil spirit Ñamu likes menstrual blood, and so menstruating<br />
women are an easy target and attract attacks. A menstruating woman can be<br />
referred to as inaru puaya alia-ka (woman different/adverse EXIST-WHEN) ‘when a<br />
woman is in a different, adverse state’. The adverb puale ‘<strong>the</strong> o<strong>the</strong>r way; on a different<br />
side’ (cognate with puaya) has overtones of something adverse. And so does every<br />
form with <strong>the</strong> meaning ‘different’. That is, women are a different, strange, and<br />
potentially dangerous ‘o<strong>the</strong>r’. And <strong>the</strong>re is perhaps a reason for this.<br />
The Tariana language is spoken in <strong>the</strong> Vaupés River Basin linguistic area known<br />
for its ‘language-based exogamy’. In Tariana society, and throughout all traditional<br />
communities in <strong>the</strong> Vaupés, women used to come from a language group different<br />
from men. Traditionally, one could only marry someone who speaks a different<br />
language and who belongs to a different tribe. People used to say: ‘My bro<strong>the</strong>rs are<br />
those who share a language with me’ and ‘We don’t marry our sisters.’ That is, in <strong>the</strong><br />
Vaupés River Basin area ethnic identity used to be inextricably linked to linguistic<br />
allegiance. One would identify one’s fa<strong>the</strong>r’s language as one’s identity badge. The<br />
area used to be highly multilingual—women would teach <strong>the</strong>ir own fa<strong>the</strong>rs’ languages<br />
and <strong>the</strong> languages of <strong>the</strong>ir husbands to <strong>the</strong>ir children.
10.3 Secrets, avoidance, and taboos 173<br />
Consistent reference to women as being ‘different’ (puaya) ra<strong>the</strong>r than ‘straight’<br />
(mayakani) suggests that <strong>the</strong>y may have been looked upon as outsiders. Now that<br />
many indigenous languages of <strong>the</strong> Brazilian Vaupés—including Tariana—are rapidly<br />
falling into disuse, women are blamed for not teaching it to <strong>the</strong> children. We return<br />
to this in §10.5.<br />
In a few similar instances, words women are not allowed to pronounce may be<br />
replaced with special synonyms. In Kele, a Bantu language spoken by <strong>the</strong> Lokele tribe<br />
in <strong>the</strong> Democratic Republic of Congo, women cannot utter <strong>the</strong> term libwá ‘nine’<br />
because it is homophonous with <strong>the</strong> curse-word lilwáa in <strong>the</strong> male-only initiation<br />
language Libéli. (The female form for ‘nine’ is is h k h , from <strong>the</strong> stem -s h k- ‘laugh’). 7<br />
This is somewhat similar to what Mary Haas described as ‘interlingual taboo’—<br />
speakers of Creek (a Muskogean language from Oklahoma) and of Thai would avoid<br />
using those words in <strong>the</strong>ir language which bear some phonetic similarity to fourletter<br />
words in English. Creek speakers would avoid <strong>the</strong> word fákki ‘soil, earth, clay’,<br />
and Thai speakers would try and not use fàg ‘sheath’ and fág ‘pumpkin or squash’.<br />
A whole set of terms can be off limits to women. Male body effluvia are kept secret<br />
among <strong>the</strong> Kwami of north-eastern Nigeria. Kwami women are not allowed to ever<br />
mention male private body parts; <strong>the</strong>y have to pretend not to know that men have<br />
genitals, and have to defecate or urinate (which <strong>the</strong>y do secretly in <strong>the</strong> areas off limits<br />
to women). A similar belief in Jukun-speaking societies in <strong>the</strong> same region of Nigeria<br />
has repercussions on <strong>the</strong> grammar of <strong>the</strong> language. Taboo words such as ‘sperm’,<br />
‘urine’, ‘penis’, and ‘testicles’, and also all expressions of body effluvia of adult men,<br />
can never appear with a possessive suffix. That is, forms such as *ʒùr-mii ‘my penis’<br />
or *sùm-mii ‘my (masc.) urine’ are not acceptable.<br />
As Anne Storch (2013: 213) puts it,<br />
That men have genitals and body effluvia is a secret and taboo in many Jukun-speaking<br />
societies. This directly relates to <strong>the</strong> idea of potentially dangerous women. Women were<br />
largely excluded (and often still are) from religious practices and <strong>the</strong> religious knowledge of<br />
<strong>the</strong> men, and <strong>the</strong>y were forced to ‘believe’ that men have non-secreting bodies...Consequently,<br />
men are claimed to have no genitals and produce no body effluvia such as sperm and<br />
faeces, so that <strong>the</strong> male body can be kept as a delicate and protected entity, with all its<br />
secretions and traces. In this sense, it is <strong>the</strong> vulnerable male body in transformation that is not<br />
to be possessed verbally. 8<br />
In Jukun tales, women are often represented as potential witches. No wonder <strong>the</strong><br />
existing taboos serve to protect men from women, and <strong>the</strong>ir dangerous powers.<br />
The asymmetrical relations between men and women are reflected in deferential<br />
registers which only women have to employ. In a number of African societies,<br />
women are not allowed to pronounce <strong>the</strong> names of <strong>the</strong>ir in-laws, or any word that<br />
sounds like an in-law name. If a Sidaama woman from Ethiopia has a fa<strong>the</strong>r-in-law<br />
whose name is Dawasso, she will not be allowed to utter quite a few words starting
174 10 The rituals of gender<br />
with da- (such as daa ‘come’, dancha ‘good’, and darawo ‘peer’); she will have to replace<br />
<strong>the</strong>m with synonyms.<br />
Women-only honorific registers can be viewed as an obligatory mark of respect<br />
towards <strong>the</strong>ir male relatives. Among <strong>the</strong> Big Nambas of North-West Malakula in<br />
Vanuatu, women employ special lexical and grammatical forms when talking to<br />
anyone with a chiefly rank, or to <strong>the</strong>ir own first-born son. So, if a woman has an inlaw<br />
called Sənari (literally, trickster), she cannot use this name, nor can she say, in<br />
ordinary speech, i-sənar-i ‘he tricks him’. She will say instead i-palv-i, with a similar<br />
meaning. 9 Many of <strong>the</strong> forms have been lost, under <strong>the</strong> pressure of Western impact;<br />
never<strong>the</strong>less, many Big Nambas women continue avoiding a mention of <strong>the</strong>ir male<br />
in-law’s name, or any o<strong>the</strong>r word that sounds like it.<br />
In traditional Warlbiri, Kaytej, Warumungo, and o<strong>the</strong>r communities of <strong>the</strong> North<br />
Central Desert in Australia, widows are not allowed to speak during a period of<br />
mourning after <strong>the</strong>ir husband’s death. This period may last for as long as three years.<br />
An elaborate sign language is used by women only to communicate throughout this<br />
period. 10<br />
To evoke <strong>the</strong> name of a deceased is to re-evoke <strong>the</strong> past and <strong>the</strong> memory of grief.<br />
Since women identify most strongly with immediate kin, <strong>the</strong>y are, in a certain<br />
sense, more closely connected to <strong>the</strong> deceased and ‘maybemoreindangerof<br />
reminding o<strong>the</strong>rs of <strong>the</strong> deceased than men, social withdrawal [and with it avoidance<br />
of spoken language] is an obvious solution...Using sign language, thus,<br />
when, as a widow, one is in a dangerously close association with a dead person,<br />
is a way of making oneself relatively safe for o<strong>the</strong>rs’ (Kendon 1988: 450). A taboo—<br />
once again, against women using language—is a way of protecting <strong>the</strong> tribe from<br />
unleashing dangerous forces.<br />
A special avoidance register just for women can be interpreted as a direct<br />
reflection of male supremacy. In Fox’s (1996: 382) words,a‘specialised vocabulary<br />
indicates to what degree <strong>the</strong> women were required to show subservience to men in<br />
<strong>the</strong> days when <strong>the</strong> Big Nambas culture was almost unaffected by external culture<br />
contact. The fact that <strong>the</strong> sub-dialect was accompanied by bowing, hiding oneself<br />
and <strong>the</strong> fear of dreadful happenings if <strong>the</strong> taboos were broken, shows that <strong>the</strong>se<br />
terms were indeed an indication of <strong>the</strong> lesser position of women, and this is still <strong>the</strong><br />
case to <strong>the</strong> present day.’<br />
There is no evidence that any of <strong>the</strong> women-only registers and ways of talking (so<br />
far described) are secret from men. In contrast, men-only languages are kept secret<br />
from women. It is as if men had something to hide and to protect <strong>the</strong>mselves. As<br />
Anne Storch (2011: 84) phrased it, male secret languages are used as strategies ‘by<br />
socially or spiritually inferior groups to gain or maintain power <strong>the</strong>y would not<br />
o<strong>the</strong>rwise have’. The linguistic asymmetry in Social <strong>Gender</strong>s reflects fear and need<br />
for protection—and with this, an open oppression and exclusion of women. Along<br />
similar lines, Dr Samuel Johnson (a major figure in English lexicography and
10.4 Men, women, and language change 175<br />
literature) once said: ‘Men know that women are an over-match for <strong>the</strong>m, and<br />
<strong>the</strong>refore <strong>the</strong>y choose <strong>the</strong> weakest or most ignorant. If <strong>the</strong>y did not think so, <strong>the</strong>y<br />
never could be afraid of women knowing as much as <strong>the</strong>mselves.’ 11<br />
Secret knowledge does not have to go toge<strong>the</strong>r with secret language. Throughout<br />
Aboriginal Australia, numerous men-only sacred rituals and sacred sites are off limits<br />
to women. Women-only sites and rituals are off limits to men. 12 There is no information<br />
about any secret linguistic knowledge confined just to women.<br />
10.4 Men, women, and language change<br />
What is <strong>the</strong> role of men and women in linguistic change? As Jespersen (1922: 242)<br />
put it,<br />
...as a rule women are more conservative than men...<strong>the</strong>y do nothing more than keep to <strong>the</strong><br />
traditional language, which <strong>the</strong>y have learnt from <strong>the</strong>ir parents and hand on to <strong>the</strong>ir children,<br />
while innovations are due to <strong>the</strong> initiative of men. Thus Cicero, in an often-quoted passage says<br />
that when he hears his mo<strong>the</strong>r-in-law Lælia, it is to him as if he heard Plautus or Nævius, for it<br />
is more natural for women to keep <strong>the</strong> old language uncorrupted, as <strong>the</strong>y do not hear many<br />
people’s way of speaking and thus retain what <strong>the</strong>y have first learnt. This however does not<br />
hold good in every respect and in every people. 13<br />
Men, as ‘chief innovators of language’, 14 lead <strong>the</strong> change in casual speech in various<br />
English dialects, including that of Norwich. In gender-exclusive languages—where<br />
male speech is easy to tell apart from female speech—women’s forms are often more<br />
archaic than men’s (see §9.1.3). But <strong>the</strong> reverse can also be <strong>the</strong> case. Gauchat (1905), in<br />
his study of Charmay French patois, discovered that women used innovative linguistic<br />
forms more than men. For instance, women over 40 palatalized l to y, just like younger<br />
speakers (men over 40 did not show this innovative trait). Women leading change in<br />
<strong>the</strong> pronunciation of vowels have been described for many areas in <strong>the</strong> United<br />
States. 15 We can recall, from §9.1.3, that in Yana as described by Edward Sapir female<br />
forms are less archaic than male forms.<br />
The adherence to a standard norm and prestigious forms are often a feature of<br />
women’s speech (as we recall from §9.2.2). Men would use non-standard forms and<br />
stigmatized varieties as tokens of a speaker’s lack of inhibition and tough masculinity,<br />
and also to show solidarity with each o<strong>the</strong>r. 16 And women may choose more<br />
prestigious normative forms to escape <strong>the</strong> bounds of a traditional society where<br />
<strong>the</strong>y are assigned an inferior status.<br />
A salient feature of <strong>the</strong> variety of Spanish spoken in <strong>the</strong> Ucieda village in <strong>the</strong><br />
province of Cantabria is word-final u ra<strong>the</strong>r than o in words like campu/campo ‘field’.<br />
This pronunciation is stigmatized as a symbol of how peasants speak. Holmquist<br />
(1985) discovered that <strong>the</strong> non-standard pronunciation was in use by younger<br />
men—despite <strong>the</strong> fact that <strong>the</strong>y were exposed to standard Spanish variety in <strong>the</strong>ir
176 10 The rituals of gender<br />
work outside <strong>the</strong> community and in military service. Younger women’s speechwas<br />
closer to standard Spanish, and sounded more conservative: <strong>the</strong>y would consistently<br />
say campo ra<strong>the</strong>r than campu. 17 Through a standard pronunciation form,<br />
women were making an effort to try to break away from <strong>the</strong> peasant way of life<br />
which offered <strong>the</strong>m few if any benefits. This resonates with Trudgill’s (1975: 91)<br />
suggestion that in Western societies women are more status-conscious than men,<br />
and <strong>the</strong>y feel <strong>the</strong> necessity to secure <strong>the</strong>ir social status by using <strong>the</strong> normative form.<br />
Language helps women to present <strong>the</strong>mselves as respectable and intellectually<br />
competent people. 18<br />
Women may go with <strong>the</strong> flow more quickly than men. Many modern English<br />
dialects are losing <strong>the</strong>ir specific features, as a result of dialect levelling. The traditional<br />
local forms of Newcastle English are being abandoned in favour of more regionally<br />
widespread—even if non-standard—forms. For instance, a glottalized [tʔ] is shifting<br />
to [ʔ]. Women are leading this change, as communicators with a wider world. 19<br />
Women act as mediators, using a greater variety of forms employed by a majority of<br />
people.<br />
Female speakers of Belizean Creole use <strong>the</strong> English copula be more frequently than<br />
<strong>the</strong> original Creole de in its locative meaning, as in We she de? ‘Where is she?’, thus<br />
shifting to English-like ways of speaking. Men use de as a locative verb more often.<br />
Women, ra<strong>the</strong>r than men, use English forms of <strong>the</strong> past tense (ra<strong>the</strong>r than o<strong>the</strong>r<br />
typical Creole forms). 20 Women tend to adapt to variation more easily than men. But<br />
o<strong>the</strong>r forms in Belizean Creole are used by men and women with equal frequency. It<br />
is perhaps <strong>the</strong> case that with a rise in equality of social position for men and for<br />
women, a wider range of social and linguistic choices becomes available, and people<br />
make personal decisions as to <strong>the</strong> way <strong>the</strong>y speak. Social <strong>Gender</strong> is just one of <strong>the</strong><br />
variables at play in <strong>the</strong> choice of forms; many o<strong>the</strong>r social parameters, including <strong>the</strong><br />
type of community, are equally important.<br />
A number of studies show that women tend to be more conservative than men in<br />
rural settings and more innovative in urban settings. Kera is a Chadic language<br />
spoken in Cameroon and sou<strong>the</strong>rn Chad. Women who live in Kera-speaking villages<br />
continue using three tones, and can easily recognize <strong>the</strong>m. Women are also more<br />
conservative than men in o<strong>the</strong>r respects—<strong>the</strong>y are <strong>the</strong> keepers of ancestral traditions.<br />
Women who have moved to local towns lose tonal distinctions in <strong>the</strong>ir Kera: <strong>the</strong>y are<br />
exposed to French, which does not have phonological tones, and speak Kera less and<br />
less. Men appear to use and recognize tones in rural and urban environment alike.<br />
Urban women promote language change, in all likelihood, under <strong>the</strong> influence of<br />
prestigious French which represents power and status. 21<br />
The asymmetry between men and women in different social settings tends to reflect<br />
women’s sensitivity to social change, values, and ensuing linguistic stereotypes. <strong>How</strong><br />
do men and women fare with regard to language maintenance in communities where<br />
more than one language is spoken?
10.5 Language keepers or language killers? 177<br />
10.5 Language keepers or language killers? Women<br />
and language maintenance<br />
10.5.1 Women as language keepers<br />
In many minority communities women are believed to be <strong>the</strong> keepers of <strong>the</strong><br />
traditional vernacular. Historically, throughout Papua New Guinea, men would<br />
learn fluent Tok Pisin (a lingua franca of <strong>the</strong> country) during lengthy periods of<br />
contracted labour outside <strong>the</strong> villages. Women would learn <strong>the</strong> lingua franca later,<br />
and speak it less well. Women of Gapun, a remote village in <strong>the</strong> East Sepik region of<br />
Papua New Guinea, speak <strong>the</strong> local language, Taiap, and lag behind men in <strong>the</strong>ir shift<br />
to <strong>the</strong> lingua franca, Tok Pisin.<br />
All throughout my fieldwork with <strong>the</strong> Manambu, women were <strong>the</strong> ones who knew<br />
traditional legends best, and used many fewer words from Tok Pisin than did men.<br />
At a traditional ceremony of ‘name debate’, respectable orators would intersperse<br />
<strong>the</strong>ir speeches with Tok Pisin and also English words. There are still some old women<br />
in <strong>the</strong> villages who speak little Tok Pisin, while all men are full bilinguals in Tok Pisin<br />
and Manambu.<br />
Rural women speakers of Tamazight Berber in Morocco are <strong>the</strong> linchpins of<br />
language maintenance; <strong>the</strong>y are seen as ‘heritage vessels’ by men within <strong>the</strong> communities<br />
and also by governments (notwithstanding <strong>the</strong> fact that urban Berber-speaking<br />
men feature much more prominently in <strong>the</strong> indigenous rights movements for valorizing<br />
and preserving Tamazight). This is supported by <strong>the</strong> rhetoric of ‘mo<strong>the</strong>r-tongue’<br />
maintenance which accords women a special symbolic status. 22<br />
But in many cases women, not men, abandon <strong>the</strong> traditional language first, in<br />
search of social status and prestige.<br />
10.5.2 Women leading language shift<br />
About a quarter of <strong>the</strong> population of <strong>the</strong> town of Oberwart in eastern Austria are<br />
bilingual in Hungarian and German; <strong>the</strong> rest are German speakers. Hungarian is<br />
associated with peasant status and peasant way of life, and consequently looked down<br />
upon. The preferred status for young people is a factory worker, not a peasant. The<br />
world of work is German-speaking, and <strong>the</strong> language itself has come to represent a<br />
higher social status. Young women of Oberwart are especially status-conscious, and<br />
very willing to participate in social and linguistic change. They are <strong>the</strong> ones who<br />
speak German in most contexts. Young women feel less allegiance than <strong>the</strong> men to<br />
<strong>the</strong> traditionally male-dominated system of subsistence farming. They have more to<br />
gain by embracing <strong>the</strong> new opportunities of industrial employment, and <strong>the</strong> more<br />
attractive status of a worker and a worker’s life ra<strong>the</strong>r than remaining a ‘backward’<br />
peasant. Young women lead in an already advanced language shift to German. Their<br />
language choice reflects facing a social change, breaking away from <strong>the</strong>ir traditional
178 10 The rituals of gender<br />
inferior social status. 23 This is reminiscent of how young women in Ucienda tend to<br />
adhere to standard Spanish pronunciation norms, moving away from <strong>the</strong>ir ‘peasant’<br />
linguistic roots (see §10.4, and Holmquist 1985).<br />
The Reindeer Saami in Finnish Lapland speak <strong>the</strong>ir own language, alongside <strong>the</strong><br />
national language, Finnish. A drastic shift to Finnish was accelerated by <strong>the</strong> evacuation<br />
of <strong>the</strong> Saami-speaking peoples to Finnish areas during <strong>the</strong> Second <strong>World</strong> War,<br />
and <strong>the</strong> influence of <strong>the</strong> Finnish-language public school system. The traditional<br />
reindeer herding economy was shattered. Speaking Saami at school and in school<br />
dormitories was forbidden; parents were encouraged to speak Finnish to <strong>the</strong>ir<br />
children at all times. As a result, women tend to use <strong>the</strong> majority language much<br />
more than men—who continue being bilingual in both Saami and Finnish. Men use<br />
Saami in what remains of traditional reindeer herding and ritual life. Women’s<br />
sphere nowadays is <strong>the</strong> home and child-rearing (back in <strong>the</strong> old days, both women<br />
and men were involved in reindeer herding). By yielding to <strong>the</strong> pressure of change,<br />
Reindeer Saami women succumb to <strong>the</strong> responsibility for <strong>the</strong>ir children and future—<br />
which lies with <strong>the</strong> majority language, and not with <strong>the</strong> traditional heritage. 24<br />
10.5.3 Women and modernity<br />
The value placed on <strong>the</strong> local language and appreciation of what women do varies<br />
ra<strong>the</strong>r dramatically. Tamazight-speaking women are generally described as ‘lagging<br />
behind men’ in <strong>the</strong>ir language shift. Keeping <strong>the</strong> precious tradition and <strong>the</strong> language<br />
is seen as a positive thing. O<strong>the</strong>r communities view <strong>the</strong> traditional vernacular as an<br />
obstacle to <strong>the</strong> coveted modernization. In <strong>the</strong> village of Gapun, Tok Pisin has become<br />
a symbol of modernity and highly appreciated ‘knowledge’ and ‘understanding’<br />
(termed save, from Tok Pisin save ‘know’), and is associated with men. The local<br />
language, Taiap, has come to be associated with <strong>the</strong> female domain, and a traditional,<br />
ra<strong>the</strong>r than modern, way of life. Its cultural correlate is <strong>the</strong> notion of hed (from Tok<br />
Pisin hed ‘head’), encompassing personal will, autonomy, and potentially unruly bigheadedness.<br />
Speakers of Mexicano, a Uto-Aztecan language of Malinche Volcano,<br />
believe that women ‘lag’ linguistically: <strong>the</strong>y are viewed as speaking less Spanish than<br />
men, and thus appearing backward. Sons nag <strong>the</strong>ir mo<strong>the</strong>rs, ‘Speak Spanish, don’tbe<br />
stupid.’ Men try and control women’s linguistic behaviour, as part of an asymmetry<br />
of Social <strong>Gender</strong>s in <strong>the</strong> community. 25<br />
In o<strong>the</strong>r communities, women can be looked upon as plainly destructive for <strong>the</strong><br />
survival of <strong>the</strong> precious tradition. Malagasy men present <strong>the</strong>mselves as keepers of<br />
proper traditional ways. They are skilled in speech, and employ elaborate ways of<br />
essentially indirect discourse. Women are <strong>the</strong> opposite—as we can recall from<br />
Table 9.8 in Chapter 9. Women speak more directly and are not averse to openly<br />
criticizing <strong>the</strong>ir offenders. Such direct discourse, unsophisticated and contemporary<br />
ways of speaking, and with <strong>the</strong>m European languages, are associated with women as
10.5 Language keepers or language killers? 179<br />
norm breakers. Destructive innovations—brought in by women—involve direct<br />
accusations, and lesser use of proverbs and traditional sayings. These are believed<br />
to be due in part to <strong>the</strong> influence of European languages, especially French. Accordingly,<br />
in commercial interactions, women (ra<strong>the</strong>r than men) are recruited to confront<br />
European buyers as <strong>the</strong>y share directness and matter-of-factness.<br />
10.5.4 Women as language killers<br />
Women can be portrayed as ‘killers’ of an endangered vernacular. Corsican—a<br />
Romance language not mutually intelligible with French—is in decline. Corsican is<br />
associated with predominantly male activities—hunting, shepherding, singing <strong>the</strong><br />
traditional song genre of paghjella, and public politics. Women—especially younger<br />
ones—speak decidedly less Corsican than men. In both popular and academic representations<br />
of women’s roles in language shift towards French, women are described<br />
as more attentive to prestige than men. According to Alexandra Jaffe (1999: 107), in <strong>the</strong><br />
current climate of nostalgic appreciation for Corsican culture and language, women<br />
are said to be ‘buying in to dominant language and culture’,by‘turning <strong>the</strong>ir backs’ on<br />
traditional values and speaking less Corsican than men. Men are viewed more positively,<br />
as preserving important cultural tradition. This is reflected in a poster from Scola<br />
Corsa (Corsican school) which has been in circulation for a few years, where a little<br />
child looks up to <strong>the</strong>ir mo<strong>the</strong>r beseeching her, ‘O Màe, parlami corsu!’ (‘Oh Mom,<br />
speak Corsican to me!’).<br />
Women are overtly blamed for <strong>the</strong> demise of <strong>the</strong> traditional language in Corsica.<br />
This discourse of blame has its roots in traditional attitudes to women. For one thing,<br />
women’s linguistic practices are a priori defined as inferior to men’s. This attitude<br />
underlies ‘scapegoating’ women for <strong>the</strong> loss of traditional values and <strong>the</strong> imposition<br />
of new ones. Also, women—who used to be and to a large extent still are excluded<br />
from access to power—are being blamed for having had too much power to influence<br />
<strong>the</strong> negatively viewed social change. Jaffe (1999: 108) treats this ‘scapegoating’ of <strong>the</strong><br />
least powerful members of <strong>the</strong> society as a way of removing one’s responsibility and<br />
not having to acknowledge one’s powerlessness in <strong>the</strong> face of French domination<br />
leading to an impending language shift. And powerless people are <strong>the</strong> easiest to<br />
blame—<strong>the</strong>y are unlikely to strike back.<br />
We recall, from <strong>the</strong> previous section, that women are portrayed by Tariana men as<br />
wayward and potentially dangerous and destructive. Men are <strong>the</strong> keepers of <strong>the</strong><br />
language as a badge of patrilineal inheritance and identity. And it is women—wives<br />
and mo<strong>the</strong>rs, ra<strong>the</strong>r than missionaries, or colonial employers—who are uniformly<br />
blamed for not teaching <strong>the</strong> language to <strong>the</strong> Tariana children. An anecdote shows<br />
how pervasive this belief is.<br />
Elvia, a teacher in <strong>the</strong> Tariana school in Iauaretê, is a Tucano. She married a<br />
Tariana man (who no longer speaks <strong>the</strong> language). When she got married, her<br />
Tucano fa<strong>the</strong>r said to her: ‘You will now kill his language, this is what women do.’
180 10 The rituals of gender<br />
Many reasons have contributed to <strong>the</strong> demise of Tariana. More than 200 years of<br />
Portuguese presence in north-west Amazonia, and educational policies of Catholics<br />
of <strong>the</strong> Salesian order of Dom Bosco, have taken <strong>the</strong>ir toll. Just as in many o<strong>the</strong>r parts<br />
of <strong>the</strong> world, priests used to take children away from <strong>the</strong>ir parents, so as to educate<br />
<strong>the</strong>m in an appropriate westernized way. While studying with <strong>the</strong> Salesians, children<br />
were supposed to use just one language—Tucano. The ‘hea<strong>the</strong>n’ multilingualism of<br />
<strong>the</strong> Vaupés area was looked upon as something to get rid of. Tucano language—<br />
spoken by a numerical majority in <strong>the</strong> Brazilian Vaupés—was imposed on people as<br />
<strong>the</strong> major language. Men used to be contracted by white people to work for money<br />
(mostly, on rubber extraction). Lengthy absences of men from <strong>the</strong> communities may<br />
have contributed to children’s lesser exposure to <strong>the</strong>ir fa<strong>the</strong>rs’ languages—ironically,<br />
<strong>the</strong> ones <strong>the</strong>y were to identify with. Unlike many o<strong>the</strong>r areas of <strong>the</strong> world, mo<strong>the</strong>r’s<br />
language plays a minor role in <strong>the</strong> identity of <strong>the</strong> Vaupés Indians. One’s badge of<br />
identity is <strong>the</strong> fa<strong>the</strong>r’s language. Fa<strong>the</strong>rs’ languages—Tariana among <strong>the</strong>m—are now<br />
learnt less and less. 26 Just as in <strong>the</strong> Corsican case, women are an easy scapegoat.<br />
Traditionally, mo<strong>the</strong>rs did not aim at maintaining <strong>the</strong> fa<strong>the</strong>r’s language. They<br />
would mainly teach children <strong>the</strong>ir own languages. <strong>How</strong>ever, following <strong>the</strong> language<br />
etiquette of <strong>the</strong> area, women would speak <strong>the</strong> language of <strong>the</strong>ir husbands pretty well,<br />
and still continue using it when speaking to <strong>the</strong>ir husbands and husbands’ male<br />
relatives including <strong>the</strong>ir children (who count as ethnic Tariana). This does not mean<br />
that women or men had to learn <strong>the</strong>ir spouses’ languages from scratch as <strong>the</strong>y started<br />
FIGURE 10.1 Elvia, a Tucano woman married to a Tariana man, was supposed to ‘kill’ his<br />
language. Here she is doing a traditional woman’s task: serving manioc beer to Tariana men at<br />
an Assembly of <strong>the</strong> Tariana in <strong>the</strong> Tariana school (Aikhenvald 2013b)
10.6 Summary 181<br />
living toge<strong>the</strong>r. Typically one would marry someone from <strong>the</strong> same language group<br />
as one’s mo<strong>the</strong>r. It is also important to note that <strong>the</strong> institutionalized multilingualism<br />
in <strong>the</strong> Vaupés area goes back at least a few hundred years. And <strong>the</strong> drastic decline of<br />
<strong>the</strong> language didn’t start until <strong>the</strong> twentieth century. Just as in <strong>the</strong> Corsican case,<br />
women are an easy, but not necessarily <strong>the</strong> right, target for laying blame.<br />
Now <strong>the</strong> rhetoric of accusation in <strong>the</strong> Vaupés area is gradually changing. That<br />
children can hardly speak or write Tariana is considered <strong>the</strong> fault of <strong>the</strong> school<br />
system, and of <strong>the</strong> teachers who are not implementing <strong>the</strong> language programme<br />
properly. Elvia (Figure 10.1) and her colleagues (male and female) will be blamed<br />
again, but in a different context.<br />
10.6 Summary: Social <strong>Gender</strong> through rituals, genres,<br />
and speech practices<br />
Social <strong>Gender</strong>s are a basis for distinctions between speech genres and speech practices.<br />
In traditional societies across <strong>the</strong> world, men’s speech is associated with <strong>the</strong><br />
public domain, oratories, and origin myths. Women-only genres almost universally<br />
include mourning songs and lullabies—<strong>the</strong>y appear to be more involved with<br />
domestic settings and <strong>the</strong> natural life-cycle. The division of labour between men<br />
and women is reflected in what <strong>the</strong>y tend to do, and what <strong>the</strong>y tend to know.<br />
<strong>How</strong>ever, in many societies male and female domains are not mutually exclusive,<br />
and in Joel Sherzer’s (1987: 120) words,<br />
<strong>the</strong> diversity of voices speaks through, reminding us of <strong>the</strong> complexity of language use in<br />
different societies around <strong>the</strong> world, reminding us...that women’s language, women’s speech<br />
and women’s verbal activities are not everywhere socially and culturally inferior, domestic or<br />
polite.<br />
Secret languages, avoidance, and taboos reveal a striking asymmetry, and a gap,<br />
between male and female Social <strong>Gender</strong>s. In many a traditional society, male<br />
initiation and male cults involve special secret language varieties which women are<br />
not supposed to use or to know. And <strong>the</strong>re are overwhelmingly more restrictions on<br />
what women can hear, or say, than on what men can. The Tariana of north-west<br />
Amazonia used to have a special ‘hidden’ register, to avoid women hearing <strong>the</strong> name of<br />
<strong>the</strong> powerful Yurupary flute, women’s former possession which <strong>the</strong>y are now not<br />
allowed to see. Big Nambas women of Vanuatu have to use a special politeness register<br />
when talking to <strong>the</strong>ir sons and in-laws. Similar avoidance principles apply among <strong>the</strong><br />
Kambaata and a few o<strong>the</strong>r groups of Ethiopia and many groups in Sou<strong>the</strong>rn Africa:<br />
women, and not men, have to avoid pronouncing <strong>the</strong> names of <strong>the</strong>ir in-laws. No such<br />
‘women-specific’ registers or languages have ever been documented. It is as if men try<br />
and purposely exclude women, attempting to protect <strong>the</strong>mselves and <strong>the</strong>ir knowledge<br />
from potential intruders who may well show <strong>the</strong>m up.
182 10 The rituals of gender<br />
And indeed, women do—<strong>the</strong>y are <strong>the</strong> ones who tend to implement prestigious<br />
norms. They develop a greater diversity of interactional skills, and adapt to variation<br />
more easily than men, in many cases thanks to <strong>the</strong>ir lower social status which <strong>the</strong>y<br />
are eager to escape from. But a word of caution is in order: in complex industrialized<br />
societies <strong>the</strong> positions of men and women are more fluid and more complex. The<br />
more social variables—age, education, and socio-economic status—are in play, <strong>the</strong><br />
more diffuse <strong>the</strong> patterns are which differentiate men from women in how normative,<br />
or how innovative, <strong>the</strong>y are.<br />
There is no simple answer to a question of whe<strong>the</strong>r women are instrumental in<br />
keeping <strong>the</strong> traditional languages alive, or in orchestrating <strong>the</strong>ir demise. But across<br />
<strong>the</strong> world, <strong>the</strong> attitudes to women as agents of change, tradition, modernity, language<br />
maintenance, or language loss reflect <strong>the</strong> alarming lack of symmetry between <strong>the</strong><br />
male and <strong>the</strong> female Social <strong>Gender</strong>, often viewed through <strong>the</strong> lens of male-centred<br />
norms.<br />
As Jane Hill (1987: 159) put it, ‘women may have <strong>the</strong>ir own norms, quite<br />
independent from men. Women may find that <strong>the</strong> symbolic resources emanating<br />
from an urban elite are a good alternative to male usage patterns, offering benefits<br />
that may be more attractive to women, who are excluded from scarce resources in<br />
<strong>the</strong>ir own communities, than to men.’ This is when women represent a barometer of<br />
change—cultural and linguistic.<br />
We now turn to <strong>the</strong> ways in which women and <strong>the</strong>ir new—and changing—social<br />
roles affect language use.<br />
NOTES AND SOURCES<br />
1. A useful survey of male and female communicative genres is in Günthner (1996).<br />
2. Briggs (1992).<br />
3. See Sherzer (1987: 112–14) for a discussion of <strong>the</strong> Kuna, and fur<strong>the</strong>r examples.<br />
4. Sims and Valiquette (1990: 165).<br />
5. See Dixon (1980: 65–7; 2002: 91–2) on initiation languages in Australia, and references <strong>the</strong>re;<br />
Hale and Nash (1997) on Damin; Migliazza (1978: 568–70) on Wayamo. Numerous examples<br />
of male initiation languages come from Africa. The Lokele tribe (a Bantu-speaking group<br />
of <strong>the</strong> Democratic Republic of Congo) had a special male initiation language called Libéli.<br />
Many words were markedly different from <strong>the</strong> day-to-day language. Their grammars were<br />
similar in all but one point. Kele had three demonstratives which agreed in Linguistic<br />
<strong>Gender</strong> with <strong>the</strong> noun <strong>the</strong>y modified; Libéli had only one demonstrative which showed no<br />
agreement: see Carrington (1947). Storch (2011: 60–81) offers a comprehensive survey of<br />
secret and initiation languages across Africa. A special male initiation language of <strong>the</strong><br />
Gbaya of eastern Cameroon, described by Moñino (1977), was a mixture of several related<br />
languages.<br />
6. See Aikhenvald (2013a); Fleming (2015) on <strong>the</strong> predominance of nouns in many avoidance<br />
registers. All <strong>the</strong> Tariana examples come from my original fieldwork and were collected in<br />
1999–2012. A description of <strong>the</strong> Vaupés Linguistic area (with special attention to Tariana,
10.6 Summary 183<br />
<strong>the</strong> only Arawak language <strong>the</strong>re) is in Aikhenvald (2002, 2012a: 75–82); recent developments,<br />
including breakdowns of linguistic exogamy, are addressed in Aikhenvald (2013b).<br />
The ‘hidden’ register attested in Tariana has not been described for any o<strong>the</strong>r Vaupés<br />
groups; nor has <strong>the</strong> image of women as ‘wayward’. Many tribes in Amazonia and New<br />
Guinea possess magical objects—especially flutes—which used to be owned by women but<br />
were subsequently appropriated by men and cannot be seen by women under <strong>the</strong> penalty<br />
of death or humiliation. If a Waujá woman, from an Arawak-speaking group of <strong>the</strong> Xingu<br />
Indigenous Park in Brazil, saw a magic flute, she would, in former times, have to undergo a<br />
gang-rape. See Villas Bôas and Villas Bôas (1970) on <strong>the</strong> Xingu mythology and <strong>the</strong> magic<br />
flutes; see , on <strong>the</strong> Waujá practices (compiled by<br />
Emilienne Ireland). Ñatabi, one of my teachers of Manambu, from New Guinea, became<br />
blind, on account of her having inadvertently caught a glimpse of a forbidden flute. We<br />
have no information about any special linguistic taboo against women associated with this<br />
sacred knowledge.<br />
7. See note 5 and Carrington (1947: 201) onLibéli; Haas (1951) on interlingual taboo.<br />
8. See Dinslage, Leger, and Storch (2000: 124) on Kwami and Jukun.<br />
9. See Treis (2005) on women’s avoidance register in Kambaata, a Cushitic language, and<br />
similar examples from numerous Cushitic-speaking groups in Ethiopia, including Alaaba,<br />
Sidaama, Libido, and Arsi Oromo, and Teferra (1987: 46) on Sidaama. See Fox (1996: 379)<br />
on Big Nambas. Hlonipha is <strong>the</strong> name of a similar principle of avoidance among <strong>the</strong> Zulu,<br />
<strong>the</strong> Nguni, and <strong>the</strong> Sou<strong>the</strong>rn Sotho of Sou<strong>the</strong>rn Africa: alongside avoiding <strong>the</strong> names of<br />
<strong>the</strong> fa<strong>the</strong>r-in-law and o<strong>the</strong>r male affines, women may not mention words with <strong>the</strong> same<br />
root or root syllables: see Herbert (1990).<br />
10. See Kendon (1988: 449–50). A somewhat similar ban has been described for young<br />
Armenian brides in <strong>the</strong> Baraninski region of Armenia in <strong>the</strong> first half of <strong>the</strong> twentieth<br />
century: <strong>the</strong> brides were not supposed to talk to <strong>the</strong>ir in-laws and used a sign language<br />
instead. See Karbelashvili (1935), and also Fleming (2015), for fur<strong>the</strong>r examples. In <strong>the</strong><br />
Aboriginal communities of Cape York Peninsula, bans on spoken language were not<br />
restricted to women: both widows and widowers were expected to address <strong>the</strong>ir nearest<br />
relatives in signed language following <strong>the</strong> death of <strong>the</strong>ir spouses. Signs were also used in<br />
long-distance communication. See Kendon (1988: 44–7).<br />
11. See Henson (1997: 67) for <strong>the</strong> full reference.<br />
12. See Myers (1986: 64, 85) on <strong>the</strong> Pintupi of Western Australia; Bell (1983) on <strong>the</strong> Warlbiri;<br />
Williams (1986: 51) on <strong>the</strong> Yolngu.<br />
13. Exceptions listed by Jespersen include <strong>the</strong> Botocudo, a Macro-Jê-speaking group from<br />
Brazil, and Japanese.<br />
14. Jespersen (1922: 247).<br />
15. Gauchat (1905: 205, 211, 218–19, 224–6); Labov (1972: 301–3), Trudgill (1971) on<br />
Norwich.<br />
16. Fur<strong>the</strong>r examples are in Coates (1993: 61–86) and Gordon and Heath (1998). See Trudgill<br />
(1975, 1998).<br />
17. See James (1996) for a comprehensive survey; Milroy (1999) on <strong>the</strong> notion of social<br />
networks, <strong>the</strong>ir density, and its role in male and female speech varieties.<br />
18. See Eckert (1989).<br />
19. See Coates (1993: 101–4) on women and vernacular speech, and <strong>the</strong> idea of women as<br />
‘deviants’ in androcentric sociolinguistics. See Milroy (1999) on women and dialect<br />
levelling, and a summary by Brittain (2002: 619).<br />
20. See Escure (2001: 72–81).
184 10 The rituals of gender<br />
21. See Pearce (2013) on Kera. Correlations between <strong>the</strong> use of more-or-less standard forms,<br />
social genders, and linguistic change are often complicated by parameters such as socioeconomic<br />
status, level of education, and age (see e.g. Eckert 2000), making it difficult to<br />
draw simple conclusions.<br />
22. See Hoffman (2006) on Tamazight Berber; Kulick and Stroud (1990) on <strong>the</strong> village of<br />
Gapun; Hill and Hill (1977) on Nahuatl, Gal (1979) for language shift in European<br />
communities, and Aikio (1992: 44) on early reports of monolingualism among women<br />
and multilingualism among men.<br />
23. See a detailed discussion in Gal (1978: 159–71).<br />
24. Aikio (1992). Among linguistic minorities in north-east Russia, women tend to obtain a<br />
higher level of education than men; women work outside <strong>the</strong> communities, while men<br />
stick to <strong>the</strong> traditional activities of hunting and fishing (Kazakevich 1996).<br />
25. Kulick (1992a: 114, 148–9) on Taiap; Hill (1987: 154–5) on Mexicano.<br />
26. See Aikhenvald (2013b). The situation in <strong>the</strong> Colombian part of <strong>the</strong> Vaupés area is<br />
different; but <strong>the</strong>re have never been any Tariana villages <strong>the</strong>re.
11<br />
<strong>Gender</strong> in grammar and society<br />
Language helps construct <strong>the</strong> way speakers perceive <strong>the</strong> world around <strong>the</strong>m. 1 As<br />
Edward Sapir (1929b: 209) put it, ‘The network of cultural patterns of a civilization is<br />
indexed in <strong>the</strong> language which expresses that civilization . . . Language is a guide to<br />
“social reality” . . . The fact of <strong>the</strong> matter is that <strong>the</strong> “real world” is to a large extent<br />
unconsciously built up on <strong>the</strong> language habits of <strong>the</strong> group.’<br />
Linguistic <strong>Gender</strong> is perhaps more of a ‘guide to social reality’ than any o<strong>the</strong>r<br />
linguistic category. Linguistic <strong>Gender</strong> reflects a classification of humans, and is<br />
among <strong>the</strong> lucky categories which are particularly sensitive to <strong>the</strong> reality of social<br />
structures. This is what it shares with o<strong>the</strong>r means of noun categorization, that is,<br />
classifiers of different types. Linguistic <strong>Gender</strong>—as assigned to humans—is a repository<br />
of historical asymmetries between Social <strong>Gender</strong>s. Changes in Linguistic <strong>Gender</strong>s<br />
may accompany external changes which impact Social <strong>Gender</strong>s. Language<br />
reforms target Linguistic <strong>Gender</strong>s. These are <strong>the</strong> topics addressed here.<br />
11.1 What Linguistic <strong>Gender</strong> can tell us about Social <strong>Gender</strong><br />
Meanings of Linguistic <strong>Gender</strong> can reflect roles and functions of Social <strong>Gender</strong>s. In<br />
Kwami, a West Chadic language with feminine and masculine Linguistic <strong>Gender</strong>s,<br />
nouns which refer to <strong>the</strong> domestic sphere are typically feminine, and nouns referring<br />
to objects outside <strong>the</strong> household are usually masculine. We can recall, from §4.1 and<br />
example 4.2, that lòolów-mè (cotton-masculine) is cotton growing on a farm, and<br />
lòolów-jè (cotton-feminine) is cotton which has already been harvested or is at home.<br />
The traditional women’s sphere among <strong>the</strong> Kwami is <strong>the</strong> house and <strong>the</strong> household.<br />
The overwhelming majority of nouns referring to <strong>the</strong> domestic sphere belong to <strong>the</strong><br />
feminine Linguistic <strong>Gender</strong> (<strong>the</strong>se include gàràn ‘cooking place’, fíní ‘room, hut’, and<br />
shémi ‘entrance hut’). Nouns belonging to <strong>the</strong> masculine Linguistic <strong>Gender</strong> tend to<br />
refer to objects outside <strong>the</strong> household, e.g. máála ‘bush, forest’, zírkì ‘woodland’,<br />
bùnké ‘meeting place in <strong>the</strong> bush’. 2<br />
The choice of Linguistic <strong>Gender</strong>s for inanimates in Ket, a Yeniseian language from<br />
Siberia, is partly based on size and shape of <strong>the</strong> object (see §4.1). Terms for large and<br />
economically important species of mammals and fish are masculine, including qàj<br />
<strong>How</strong> <strong>Gender</strong> <strong>Shapes</strong> <strong>the</strong> <strong>World</strong>. First edition. Alexandra Y. Aikhenvald.<br />
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.
186 11 <strong>Gender</strong> in grammar and society<br />
‘elk’, sèl ‘reindeer’, qùr ‘pike’. So are terms for objects of economic or spiritual<br />
importance, including bálbès ‘(Christian) cross’, sùl ‘hook (for holding or fastening)’,<br />
and names for tent parts and snow-sled parts. 3 Men play a major role in fishing and<br />
reindeer herding (<strong>the</strong> major traditional activities of <strong>the</strong> Ket)—and this is what we<br />
discern in <strong>the</strong> Linguistic <strong>Gender</strong> choice.<br />
The meanings of Linguistic <strong>Gender</strong> may reflect positive and negative values, and<br />
<strong>the</strong> relative importance, of Social <strong>Gender</strong>s. Feminine gender choice for non-humans<br />
in Manambu is associated with smaller size, and masculine gender with larger size.<br />
But men’s houses are always masculine, no matter what <strong>the</strong>ir size. We saw in §3.4.2<br />
that terms referring to speech and ceremonies in this language are assigned masculine<br />
gender if <strong>the</strong>y are culturally important, and feminine gender if <strong>the</strong>y are less so.<br />
The term jap ‘thing’ can be masculine or feminine, depending on how big <strong>the</strong> object.<br />
If this same word is used to refer to a traditional valuable object (displayed at malecentred<br />
ceremonies, or used in mortuary payments), it is always masculine. All <strong>the</strong><br />
male-oriented rituals and ceremonial objects in Manambu are assigned masculine<br />
Linguistic <strong>Gender</strong>. So are <strong>the</strong> names for traditional shell valuables and artefacts made<br />
of <strong>the</strong>m (e.g. yu ‘greensnail shell’).<br />
The feminine Linguistic <strong>Gender</strong> in Hamar, an Omotic language, is associated with<br />
<strong>the</strong> importance of an object, and its larger size (see example 3.18). According to Jean<br />
Lydall (1988)—a scholar with an extraordinary knowledge of <strong>the</strong> Hamar language<br />
and culture—this reflects <strong>the</strong> importance of women in agricultural and o<strong>the</strong>r central<br />
tasks in Hamar life.<br />
The effects of <strong>the</strong> reversal of masculine and feminine Linguistic <strong>Gender</strong> choice for<br />
men and for women also point towards attitudes to men and women. We can recall,<br />
from §7.4, that addressing a man with a feminine pronoun in Amharic may imply<br />
belittling him, and is offensive. But addressing a woman with a masculine pronoun is a<br />
way of praising her, by ‘promoting’ her to <strong>the</strong> status of an honorary male. Similar<br />
examples come from Tariana and Jarawara, two South American languages (a summary<br />
is in Table 7.3).<br />
Asymmetries in <strong>the</strong> expression, and meanings, of derivational genders can be<br />
indicative of <strong>the</strong> inequalities in Social <strong>Gender</strong>s.<br />
11.2 Social inequalities through gender asymmetries<br />
The historical asymmetry in male and female Social <strong>Gender</strong>s and <strong>the</strong>ir stereotypes<br />
find <strong>the</strong>ir expression many a time in derogatory and demeaning overtones of female<br />
terms and functions. We can recall, from §5.3, that in Ndyuka (or <strong>the</strong> Eastern<br />
Maroon Creole of Suriname) a number of nouns marked with uman ‘woman’ may<br />
have a negative and over-sexualized connotation: waka-man means ‘traveller’, but<br />
waka-uman is not just a ‘travelling woman’: this has strong connotations of a woman<br />
of easy virtue. 4
11.2 Social inequalities through gender asymmetries 187<br />
Conversely, positive connotations of females in a matrilineal society—such as<br />
Ndyuka—may also be reflected in gendered lexical forms. ‘Mo<strong>the</strong>r’ is a very prestigious<br />
role in Ndyuka society. It is <strong>the</strong> mo<strong>the</strong>r and her family who have <strong>the</strong> primary<br />
authority over children. The line of descent of every lineage is determined by<br />
mo<strong>the</strong>rs: only <strong>the</strong> (male) children of <strong>the</strong> women of elite lineages may carry out <strong>the</strong><br />
community’s main socio-political functions. And we find a number of compounds<br />
where mama ‘mo<strong>the</strong>r’ means ‘important, main’, asinmama liba ‘<strong>the</strong> main river’,<br />
mama kinde ‘native village’, mama mofu ‘important or main message’, and a mama<br />
fu a toli ‘<strong>the</strong> gist of <strong>the</strong> story’. 5<br />
Specific meanings of male and female terms reflect professions and roles reserved<br />
for men and women. In Oriya, an Indo-Aryan language of India, professions<br />
such as carpenter (kaarigara), constable (daarogaa), or driver (draaibhara) have<br />
only a masculine form. At home women do <strong>the</strong> day-to-day cooking; but only men<br />
work as professional chefs; so <strong>the</strong> terms roseiaa and pujaari ‘chef ’ have an<br />
exclusively male reference. A few professions are associated with women—<strong>the</strong>se<br />
include nars ‘nurse’ (from English nurse). The term for ‘male nurse’, aɲɖiraa nars,<br />
contains <strong>the</strong> noun ‘man’. If one wishes to talk about a female policeman, one needs<br />
to use <strong>the</strong> term for woman, maakinaa, asinmaakinaa polis. Female counterparts<br />
of many male occupational terms refer to a wife, and not to a female professional: for<br />
instance, casaa means ‘farmer’ and caasuɲi means ‘farmer’s wife’ (and not a female<br />
farmer).<br />
In this male-dominated society, quite a few terms of abuse refer to women only—<br />
<strong>the</strong>se include daari ‘whore’, baanijhuɲi ‘barren’,orbidhabaa ‘widow’. These reflect a<br />
popular belief that misfortunes—such as barrenness, widowhood, divorce, and<br />
socially unacceptable behaviour such as prostitution or extramarital affairs—are<br />
associated with women. 6<br />
Numerous feminine derivations and compounds referring to women in English<br />
have overtones which go beyond what Baron (1986: 115) calls ‘<strong>the</strong> simple recording<br />
of gender’. A woman appears as distinctly inferior in conventional pairs <strong>the</strong> weaker<br />
sex versus <strong>the</strong> stronger sex. Referring to a woman as <strong>the</strong> fair sex or <strong>the</strong> gentler sex, and<br />
men as <strong>the</strong> superior sex indicates disparity and lack of equality between men and<br />
women. In a few instances, feminine forms have no pejorative connotations—<strong>the</strong><br />
most oft-quoted examples are aristocratic titles in pairs such as baron/baroness, duke/<br />
duchess, prince/princess, and count/countess. But in an overwhelming majority of<br />
cases compounds and derivations referring to women have negative overtones.<br />
We saw, in §7.5, how a feminine derivational form spinster in English has<br />
somewhat pejorative connotations of an unmarried woman who is unwanted and<br />
‘on <strong>the</strong> shelf ’. Its masculine counterpart, bachelor, does not. In English, one can talk<br />
about ‘an eligible bachelor’;but‘an eligible spinster’ sounds like a joke. In English, words<br />
denoting women acquire debased or obscene connotations, e.g. madam or mistress.<br />
The equivalent masculine terms—sir or master—do not. The pair mistress–master is a
188 11 <strong>Gender</strong> in grammar and society<br />
very clear, and oft-quoted, example of disparity in meaning of forms with female and<br />
with male reference.<br />
The term master, ‘a person, predominantly a man, having control and authority’,<br />
was borrowed into Old English (as mægester) about AD 900 from classical Latin<br />
magister. The feminine form mistress was borrowed around 1300, into Middle English,<br />
from Middle French maistresse, first of all in <strong>the</strong> meaning of ‘woman at <strong>the</strong> head of a<br />
household’. Mistress, in its pejorative sense of a woman ‘o<strong>the</strong>r than his wife with whom<br />
a man has a long-lasting sexual relationship’, was first recorded in 1601. There has<br />
never been a corresponding word for a ‘male extramarital lover’—which, in Cheshire’s<br />
(1985: 22) words, ‘is a telling reflection of <strong>the</strong> freedom that society has tolerated for<br />
men, but not for women’. <strong>How</strong>ever, nowadays, as times are changing, <strong>the</strong> word mistress<br />
is often replaced by lover, which can refer to people of any sex.<br />
The noun mistress is still used in compounds such as schoolmistress and headmistress,<br />
as female counterpart of schoolmaster and headmaster. An alternative to ei<strong>the</strong>r<br />
would be schoolteacher and headteacher. <strong>How</strong>ever, as Cheshire (2008) reports in her<br />
follow-up to her 1985 paper, in a sample of 500 BBC web-pages, headmaster appears<br />
to be used more frequently than headteacher, and headmistress is lagging behind both<br />
of <strong>the</strong>m. The pair headmaster–headmistress continue being unequal.<br />
Fur<strong>the</strong>r asymmetry lies in <strong>the</strong> generic use of <strong>the</strong> word master: according to <strong>the</strong><br />
Oxford English Dictionary Online, master started being used in <strong>the</strong> second half of <strong>the</strong><br />
twentieth century to include women, e.g. In her field, she is a master (from Miller and<br />
Swift 1981: 144). Master is also one of <strong>the</strong> words with what R. M. W. Dixon (2014)<br />
calls ‘double duty’: it can be used as a verb, and as a modifier. <strong>How</strong> to master <strong>the</strong> art of<br />
French cooking refers to a person of any sex. Anyone can become a ‘master landscape<br />
architect’, if <strong>the</strong>y are good enough to master <strong>the</strong> art. Miller and Swift (1981: 144)<br />
conclude that <strong>the</strong> word master has ‘outgrown its masculine gender origins’.<br />
In contrast, <strong>the</strong> noun mistress has retained its female-only reference. A few examples<br />
of a verb mistress in <strong>the</strong> sense of ‘master something, by a female’ are offered by <strong>the</strong><br />
Oxford English Dictionary Online: all are said be ‘used humorously or ironically after<br />
master’ (as a verb). The latest example given is from Fiona Cooper’s 1991 novel Jay<br />
loves Lucy: ‘“What’s this?” said Piggy-wig, straddling a chair with <strong>the</strong> triumph of one<br />
who has mistressed a Zanussi automatic.’ The term master in its generic and Natural-<br />
<strong>Gender</strong>-neutral sense appears in numerous derivations and compounds, including<br />
masterful, mastery, masterpiece, mastermind, and mastersinger. 7<br />
Generic masculines—author, actor, poet—cover members of ei<strong>the</strong>r sex. Their<br />
feminine equivalents may acquire unwanted overtones of lesser, feminine-only,<br />
professionals. 8 The generic uses of masculine terms fit in with <strong>the</strong> functionally<br />
unmarked character of masculine Linguistic <strong>Gender</strong> in many languages. Many<br />
feminine forms are derived from, or formed on <strong>the</strong> basis of, <strong>the</strong> erstwhile masculine<br />
ones. The exceptions include professions which are considered traditionally female—<br />
such as nurse or prostitute. A male representative of such a profession will <strong>the</strong>n be
11.2 Social inequalities through gender asymmetries 189<br />
marked—we talk about male nurses and male prostitutes. This is how, in Henley’s<br />
(1989: 60–1) words, ‘language stereotypes women’, and also deprecates <strong>the</strong>m—<br />
reflecting traditional practices, even if on <strong>the</strong>ir way to obsolescence.<br />
Similar stereotyping is a feature of languages with no Linguistic <strong>Gender</strong>. The<br />
covert gender bias in Turkish is reflected in implicit information about genderspecific<br />
domains through terms referring to people: çocuk bakɪcɪsɪ ‘nursery school<br />
teacher’ will be understood as a woman, and futbolcu ‘football player’ as a man. Adam<br />
‘man, human’ is <strong>the</strong> ‘representative member of humanity’ (Braun 2001: 305; see also<br />
§5.3), and women remain on <strong>the</strong> periphery. In Hungarian, stereotypes of Social<br />
<strong>Gender</strong>s are reflected in professional terms: terms for female occupations which<br />
contain <strong>the</strong> feminine derivational marker -nő, and do not have a masculine counterpart.<br />
These include varró-nő(sewing-feminine) ‘seamstress’, apoló-nő‘nurse’ or<br />
takaritó-nő‘cleaning woman’. According to Vasvári (2011: 18), one cannot just say<br />
apoló or apoló férfi ‘male nurse’. The marker -nő‘female’ can be added to a formally<br />
unmarked (and traditionally male) professional term to indicate a woman, e.g. orvos<br />
‘doctor’ versus orvos-nő ‘female doctor’, tanár ‘teacher’ versus tanár-nő ‘female<br />
teacher’. <strong>How</strong>ever, feminine marked forms reflect a somewhat lower status.<br />
As Vasvári (2011: 19), herself a university lecturer, explains: ‘Tanár is a job<br />
traditionally done for a long time by both sexes, and when I am addressed directly,<br />
I am called tanar-nő[teacher-feminine]. <strong>How</strong>ever, if someone were to enquire what<br />
my profession is and I replied tanar-nő[vagyok] “[I am a] female teacher”, <strong>the</strong>y<br />
would assume that I am a secondary school teacher. As a university professor I can<br />
say only egyetemi tanár vagyok, with *egyetemi tanar-nő vagyok impossible. Or<br />
alternately I can be a profeszor [“professor”], but, as one of my former students<br />
reports, whenever she mentions her new yorki profeszor (New York professor) all<br />
listeners always assume . . . that she is speaking of a male professor.’<br />
In Henley’s (1989: 60) words, masculine generics—such as chairman, spokesman,<br />
<strong>the</strong> working man, <strong>the</strong> black man, and <strong>the</strong> man in <strong>the</strong> street are tokens of how ‘language<br />
ignores women’. They have been a target of struggle against ‘sexist language’—we turn<br />
to this in §11.5.2.<br />
An asymmetry between terms for men and for women may have its roots in a<br />
colonial heritage, and differences in male and female roles for colonizers. In Tariana,<br />
nawiki, <strong>the</strong> term for ‘Indian’, or indigenous person as a category, is an epicene: it can<br />
refer to a man or to a woman. Male or female reference can be disambiguated with an<br />
article, or by agreement on verb or adjective, e.g. diha nawiki (he/<strong>the</strong>:masc.sg Indian)<br />
‘<strong>the</strong> Indian (man)’ and duha nawiki (she/<strong>the</strong>:fem.sg Indian) ‘<strong>the</strong> Indian (woman)’.<br />
A number of o<strong>the</strong>r generic terms, including <strong>the</strong> one for evil spirit of <strong>the</strong> jungle ñamu,<br />
can also have male or female reference. In contrast, <strong>the</strong>re are two separate terms for<br />
‘white (or non-Indian) man’ and ‘white (non-Indian) woman’. The word yalana<br />
‘white man’ is an old and fully integrated borrowing from Língua Geral yara<br />
‘master’. 9 This origin of <strong>the</strong> term for ‘white man’ iconically reflects <strong>the</strong> commanding
190 11 <strong>Gender</strong> in grammar and society<br />
role of non-Indian male masters in <strong>the</strong> colonial world. Non-Indian women were not<br />
exactly masters on <strong>the</strong>ir own; <strong>the</strong>y were defined in terms of men and called yalana<br />
i-sa-do (white.man indefinite-spouse/woman-feminine.singular), literally, ‘white man’s<br />
wife/woman’.<br />
As <strong>the</strong> relative status of Social <strong>Gender</strong>s changes, so may <strong>the</strong> composition of<br />
Linguistic <strong>Gender</strong>s and <strong>the</strong> expression of gender with o<strong>the</strong>r means, including classifiers.<br />
This is what we turn to now.<br />
11.3 The value of ‘man’ through gender in lexicon<br />
The asymmetry between male and female Social <strong>Gender</strong>s is especially conspicuous<br />
in <strong>the</strong> lexicon. The positive overtones of ‘malehood’ in contrast to negative overtones<br />
of ‘femalehood’ are reflected in <strong>the</strong> lexicon and derivations in many languages. The<br />
ways masculine Linguistic <strong>Gender</strong> in Hebrew is used as a marker of solidarity,<br />
affection, and protection (discussed in §7.3) may well be based on what Tobin<br />
(2001: 191) calls ‘<strong>the</strong> inherent desirability of masculine attributes’. In o<strong>the</strong>r words,<br />
talking to, or about, a woman as if she were a man has positive connotations. Golda<br />
Meir, during her time in various governments and as Prime Minister of Israel, was<br />
often complimented with male-oriented idioms—such as yesh lah beitsim (exist<br />
to.her balls) ‘she has balls’—and praised as ha-gever ha-yexid ba-memshala (<strong>the</strong>man<br />
<strong>the</strong>-only in.<strong>the</strong>-government) ‘<strong>the</strong> only man in <strong>the</strong> government’. This is reminiscent<br />
of <strong>the</strong> use of <strong>the</strong> phrasal verb Man up! in English when said to a woman, urging<br />
her to get her act toge<strong>the</strong>r. 10<br />
Similarly, <strong>the</strong> noun tsiãri ‘man’ in Tariana has positive overtones of courage,<br />
determination, and reason. The noun inaru ‘woman’ has distinctly opposite overtones.<br />
In order to encourage himself to be brave, a male speaker would say ‘I ama<br />
man’. A turtle who is about to fight a tapir says to himself, as a matter of selfencouragement,<br />
tsiãri-naka inaru-kade-naka nuha (man-PRESENT.VISUAL woman-<br />
NEG-PRESENT.VISUAL I) ‘I am a man, I am not a woman’. This is consistent with a<br />
specially important woman being referred to as ‘he’—as we saw in §7.2. 11 Such<br />
attitudes to men and to women directly reflect <strong>the</strong> Tariana social reality. In terms<br />
of <strong>the</strong> relative position of Social <strong>Gender</strong>s, men occupy a distinctly higher status than<br />
women. Among <strong>the</strong> Tariana, one’s fa<strong>the</strong>r’s language is <strong>the</strong> most valuable asset, and is<br />
<strong>the</strong> badge of one’s identity. Women always come from a different tribe and are held<br />
responsible for language loss (and many o<strong>the</strong>r evils). They are a dangerous and<br />
unruly force to beware of—as we saw in §10.5.4.<br />
Schipper’s(2003) study of proverbs across <strong>the</strong> world reveals a strong association of<br />
women with negative features—weak, unstable, unruly, unreasonable, and better fit<br />
for domestic environment than for dealing with <strong>the</strong> outside world: this is what we<br />
saw in §8.4. Estonian proverbs (analysed by Hasselblatt 2015: 140–1) illustrate an<br />
asymmetry between <strong>the</strong> Social <strong>Gender</strong>s: more often than not <strong>the</strong>y portray women as
11.4 <strong>How</strong> Linguistic <strong>Gender</strong>s reflect social change 191<br />
a negative and disturbing element. Typical examples include Kus on naisi, siel pole<br />
rahu ‘where <strong>the</strong>re are women, <strong>the</strong>re is no peace’, and iga naine on libu oma mehele<br />
‘every women is a bitch to her own husband’. 12<br />
In a few European languages inherently sexist associations between ‘man’ and<br />
‘positive properties’ are embedded in a few derivational forms. Examples of derivations<br />
referring to positive properties of valour and courage associated with ‘man’<br />
include Estonian mehisus (from mees, mehe ‘man’), Russian muzhestvo (from muzh<br />
‘man, husband’), German Mannhaftigkeit, Hungarian ferfiás, all of which mean<br />
‘courage, valour’, literally, ‘manhood’. An Estonian woman-politician can be complimented<br />
as someone who displays ‘statesmanlike’ (riigi-mehe-lik (state-man-AFFIX))<br />
behaviour. The term ‘effeminate’ has a distinctly negative feel to it—as if someone<br />
were coming down in <strong>the</strong> world, from a status of a ‘manly’ and worthy being to a<br />
decadent one. But to what extent are speakers conscious of <strong>the</strong> inherently sexist<br />
character of <strong>the</strong> positive overtones of ‘man’ within established derivations? This is an<br />
issue which has not so far been explored.<br />
Anecdotal evidence indicates that not everyone is aware of any potentially demeaning<br />
connotations of <strong>the</strong> word man in <strong>the</strong> idiom man up. A top-level university<br />
administrator told me how she’d complained to her teenage daughter about how<br />
difficult electronic databases are to access. The daughter replied, ‘Man up, Mum, stop<br />
grumbling.’ When I pointed out that such usage of man up could be considered<br />
sexist—as one would not say to anyone *woman up, or*woman down, <strong>the</strong> administrator<br />
first looked at me with some incredulity, and after a minute or so exclaimed,<br />
‘Oh, I never thought of that.’<br />
We can recall, from §7.3 and §4.1, that Mohawk, an Iroquoian language, offers<br />
two options of referring to a woman—<strong>the</strong> ‘feminine-indefinite form’ and <strong>the</strong><br />
‘feminine-neuter form’. The latter subsumes animals; but, as Mithun (2014: 137)<br />
reports, speakers do not show any awareness of this correlation as ‘demeaning’ for<br />
women. In Lak, a north-east Caucasian language, women have traditionally<br />
belonged to a special feminine gender class II. As <strong>the</strong> language evolved, respected<br />
women (especially those outside <strong>the</strong> immediate family) came to be reclassified<br />
as belonging to gender class III (which also subsumes inanimate objects). In this<br />
unusual case, inanimate gender is used as a mark of respect for women. This does<br />
not in any way imply downgrading anyone. On <strong>the</strong> contrary. Politeness and respect<br />
often involve what Brown and Levinson (1987: 199) call‘conventionalised indirectness’<br />
and distancing. This indirectness is a factor behind <strong>the</strong> use of a more general<br />
erstwhile inanimate category for talking to, or about, those whom one respects. 13<br />
11.4 <strong>How</strong> Linguistic <strong>Gender</strong>s reflect social change<br />
Changes in social structures and attitudes may find <strong>the</strong>ir reflection in how Linguistic<br />
<strong>Gender</strong>s, and o<strong>the</strong>r noun categorization devices, apply to humans. All human beings
192 11 <strong>Gender</strong> in grammar and society<br />
in Setswana, a Bantu language from Botswana, belong to <strong>the</strong> human gender (labelled<br />
1/2: see Table 2.2). Names for some groups traditionally treated as strange and<br />
unusual used to be classified within gender 5/6, toge<strong>the</strong>r with substances, such as<br />
clay or dirt, and abstract nouns, such as foreign institutions. These groups used to<br />
include Bushmen, <strong>the</strong> indigenous minority of Botswana. In <strong>the</strong> modern society, it is<br />
no longer correct to refer to Bushmen with this gender—instead, <strong>the</strong>y have to be<br />
referred to like all o<strong>the</strong>r humans, with <strong>the</strong> human gender: when <strong>the</strong>y are referred to<br />
with <strong>the</strong> non-human class, it is felt that <strong>the</strong>y are being treated like ‘dirt’. 14<br />
We can recall, from §5.1.1, that <strong>the</strong> choice of numeral classifiers in Maonan, a Tai-<br />
Kadai language spoken in China, reflects <strong>the</strong> changing place of women within <strong>the</strong><br />
society. Similarly to a number of languages of South-East Asia, women used to be<br />
counted with <strong>the</strong> classifier tɔ 2 , which also subsumes animals and children. Respected<br />
women were typically counted using <strong>the</strong> human classifier ʔai 1 (NUM.CL:HUMAN). At<br />
present, women who have a professional status—for example teachers in 11.1—are<br />
also counted with <strong>the</strong> ‘human classifier’.<br />
11.1 ja:u 3 fia:k 8 nde 1 mɛ 2 si 5 ʔai 1<br />
inside school 2pl:inclusive have four NUM.CL:HUMAN<br />
ki 2 sɛŋ la:k 8 ʔbjek 8<br />
teacher woman<br />
‘There are four women teachers in our school’<br />
Respected women—counted with <strong>the</strong> human classifier ʔai 1 —include ‘women<br />
government and party officials, teachers, women singers’, and elder relatives such<br />
as aunts and grandmo<strong>the</strong>rs. They also include female fortune tellers, shamans and<br />
matchmakers, and female monsters: it is believed that counting <strong>the</strong>m with <strong>the</strong> animal<br />
classifier might offend <strong>the</strong>m and provoke unwanted results. Women who do not hold<br />
any positions associated with a high status are referred to with <strong>the</strong> classifier tɔ 2<br />
‘animal’.<br />
The choice of <strong>the</strong> human classifier can also be used as a token of a person’s status,<br />
and <strong>the</strong>ir maturity. Children are typically referred to with <strong>the</strong> animal classifier tɔ 2 .<br />
This applies to primary school pupils, as in sa:m tɔ 2 la:k 8 fia:k 8 (three NUM.CL:ANIMAL<br />
pupil) ‘three pupils’. But college students will be referred to with ʔai 1 , e.g. ŋɔ 4 ʔai 1<br />
ta 4 jɔŋ 5 (five NUM.CL:HUMAN college.student) ‘five college students’. 15 The human<br />
classifier in Maonan has been extended to subsume women of professions and<br />
ranks which <strong>the</strong>y had never occupied before. Derivational gender in Russian offers<br />
similar examples.<br />
One of <strong>the</strong> consequences of <strong>the</strong> Bolshevik Revolution in 1917 was, as Rothstein<br />
(1973: 460) put it, ‘<strong>the</strong> relative liberation of women’ which ‘produced a need for new<br />
terminology to refer to women occupying certain jobs or positions’, for example,<br />
frezerovs᷈c᷈ica ‘female milling machine operator’ (a female equivalent of frezerovs᷈c᷈ik)<br />
or syrovarka ‘female cheesemaker’ (a female equivalent of syrovar). Similarly, women
11.4 <strong>How</strong> Linguistic <strong>Gender</strong>s reflect social change 193<br />
of minority groups within Russia acquired access to <strong>the</strong> jobs formerly reserved for<br />
men. In Lak, a north-east Caucasian language from Daghestan, agent nouns such as<br />
xIakin ‘doctor’ used to belong exclusively to <strong>the</strong> masculine gender marked through<br />
agreement on <strong>the</strong> verb, e.g. xIakin ur (doctor masc.sg+exist) ‘<strong>the</strong>re is a (male) doctor’.<br />
They can now be used to refer to women, and trigger feminine agreement, e.g. xIakin<br />
dur (doctor fem.sg+exist) ‘<strong>the</strong>re is a (female) doctor’. 16<br />
Old habits die hard—if <strong>the</strong>y die at all. Language change lags behind <strong>the</strong> changes<br />
in a society. 17 Despite <strong>the</strong> purported equality of men and women during <strong>the</strong> Soviet<br />
period, <strong>the</strong> asymmetry between feminine and masculine derivational forms<br />
remains. 18 Masculine forms of nouns denoting prestigious (and especially traditionally<br />
male) professions continue being used as generics. This is an established feature<br />
of <strong>the</strong> language—as Roman Jakobson (1971: 213) put it, in his description of Russian<br />
gender in terms of its markedness, ‘<strong>the</strong> general meaning of <strong>the</strong> masculine does not<br />
necessarily specify sex’, since <strong>the</strong> masculine form is used to refer to people in general.<br />
A general term for teacher is <strong>the</strong> masculine uchitel'; <strong>the</strong> feminine form uchitelj'-<br />
nica is possible, but it tends to refer to primary school female teachers. The prestigious<br />
title Zasluzhennyj uchitelj' ‘Distinguished teacher’ can be awarded to men and<br />
to women. Making it into Zasluzhennaja uchiteljnica (feminine) would have been<br />
laughable. Along similar lines, awards such as ‘Hero of <strong>the</strong> Soviet Union’ (Geroj<br />
Sovetskogo Sojuza) or‘Hero of Labour’ (Geroj Truda) contain masculine generic<br />
forms. The title Heroine Mo<strong>the</strong>r (matj-geroinja) awarded to mo<strong>the</strong>rs with many<br />
children contains <strong>the</strong> female form heroine. 19 In her Wikipedia entry in Russian, <strong>the</strong><br />
first woman cosmonaut, Valentina Tereshkova, is described exclusively with masculine<br />
terms in <strong>the</strong>ir generic meanings (kosmonavt ‘cosmonaut’, letchik-kosmonavt<br />
‘pilot-cosmonaut’, general-mayor ‘major-general’, etc.).<br />
Overtones of female derivations can be felt to be deprecatory. We can recall, from<br />
§7.5, how Anna Akhmatova made it clear that she was to be referred to as ‘poet’ and<br />
not as ‘poetess’: <strong>the</strong> latter term had, for her, somewhat demeaning overtones of notquite-up-to-scratch<br />
feminine writing. Rothstein (1973: 462) mentions <strong>the</strong> head chef<br />
of a restaurant in St Petersburg (<strong>the</strong>n Leningrad) who burst into tears when an<br />
American tried to compliment her by telling her that she was otlic᷈naja povarixa ‘an<br />
excellent cook’ using <strong>the</strong> feminine form of ‘cook’. This was demeaning for her—she<br />
considered herself a povar, masculine form for ‘cook’. A female author would refer to<br />
herself as avtor ‘author: masculine’. A female form avtorsha would sound out of place<br />
and somewhat derogatory, and <strong>the</strong>re is no neutral feminine counterpart available.<br />
The late Natalja Y. Shvedova, a leading Russian grammarian and lexicographer,<br />
inscribed a copy of her collected papers to me as ot avtora (from author.masculine.<br />
genitive singular). 20<br />
The meanings of derivational gendered forms reflect social asymmetries in <strong>the</strong><br />
position of men and women—in Robin Lakoff ’s (1975: 69) words, ‘linguistic imbalances’<br />
which ‘bring into sharper focus real-world imbalances and inequities’. <strong>How</strong>
194 11 <strong>Gender</strong> in grammar and society<br />
can <strong>the</strong> lack of balance be redressed, by conscious linguistic reform? This is what we<br />
turn to now.<br />
11.5 Thwarting ‘sexist language’<br />
‘Sexist language’ is a term used to describe unequal representation of men and<br />
women across many languages of <strong>the</strong> world. The following features are seen as its<br />
reflections. 21<br />
First, a cross-linguistically common tendency for <strong>the</strong> masculine Linguistic <strong>Gender</strong><br />
to be <strong>the</strong> functionally unmarked choice is seen as a way of subsuming women under<br />
‘men’ and making <strong>the</strong>m invisible (see also §7.7). A major bone of contention in <strong>the</strong><br />
English ‘sexist language’ has been—and partly continues to be—<strong>the</strong> generic use of<br />
masculine singular he to refer to people in general (if <strong>the</strong>ir sex is not known or<br />
irrelevant). The generic pronoun man ‘person, one’ in German has also been <strong>the</strong><br />
target of concern, thanks to its similarity to <strong>the</strong> generic form Mann ‘man, person’.<br />
<strong>Gender</strong> agreement and gendered forms of adjectives and verbs have been targeted in<br />
Romance languages (including French and Spanish). The issue does not arise in<br />
English with no gender agreement, nor in German where masculine, feminine, and<br />
neuter Linguistic <strong>Gender</strong>s are not distinguished in <strong>the</strong> plural. We turn to this in<br />
§11.5.1.<br />
Secondly, a tendency to use terms with male reference as general terms for humans<br />
and for professions has been seen as a sexist way of portraying a man, or <strong>the</strong> male, as<br />
<strong>the</strong> main representative of <strong>the</strong> species, a benchmark for all human beings. The man is<br />
seen as a reference point, or norm. A woman becomes invisible and subsumed under<br />
<strong>the</strong> category of ‘man’. A term for a woman is often derived from that for a man (e.g.<br />
Hebrew is᷈‘man’, is᷈-a ‘woman’). This markedness is seen as showing woman as a<br />
‘deviation’ or ‘exception’ from an essentially male norm. In newspaper reports, one<br />
often finds a woman singled out from <strong>the</strong> ‘human’ category—as in <strong>the</strong> example given<br />
by Cheshire (1985: 25), ‘People would bring <strong>the</strong>ir wives, mo<strong>the</strong>rs and children’, or<br />
‘Sharing our railway compartment were two Norwegians and <strong>the</strong>ir wives’. As we have<br />
seen above, feminine forms and derivations develop pejorative or slighting connotations,<br />
and may sound less prestigious than <strong>the</strong>ir male equivalents—used as generics.<br />
In §11.5.2 we look at <strong>the</strong> attempts to redress <strong>the</strong> balance.<br />
Thirdly, address practices in <strong>the</strong> Western world focus on women’s (but not men’s)<br />
marital status—once again putting <strong>the</strong>m in <strong>the</strong> shadow of a male being. Sexist naming<br />
practices expect women to take on names of <strong>the</strong>ir fa<strong>the</strong>rs and husbands—making <strong>the</strong>m<br />
invisible and <strong>the</strong>ir achievements slighted. This is what we turn to in §11.5.3.<br />
Order of words in coordinations—such as men and women, ladies and gentlemen,<br />
or Bill and Mary—has also come under scrutiny as a potential reflection of gender<br />
bias. <strong>How</strong>ever, many factors come into play—among <strong>the</strong>m a tendency to put a<br />
shorter word first. Considerable variation remains in this area. 22
11.5 Thwarting ‘sexist language’ 195<br />
11.5.1 Masculine bias through Linguistic <strong>Gender</strong>: pronouns and agreement<br />
Personal pronouns in English (she, he, it) are <strong>the</strong> major means for expressing<br />
Linguistic <strong>Gender</strong>. What English lacks is a general indefinite form which could<br />
refer to both men and women without specifying <strong>the</strong>ir sex. A generic non-gender<br />
specific pronoun monn, man, mann, mane, manne, monne (all derived from an<br />
unstressed form of <strong>the</strong> noun man) did exist in Old English and in Middle English;<br />
its latest occurrence goes back to 1500. 23 In earlier stages of English, <strong>the</strong> generic use<br />
of masculine singular pronoun he was an alternative to <strong>the</strong>y covering ‘man’ and<br />
‘woman’ (Curzan 2003: 71). In 11.2, from Old English, generic he refers to a person in<br />
general.<br />
11.2 Swyche a persone ys ful slogh, Be he hygh, or be he logh . . .<br />
‘Such a person is very lazy, be he high or be he low’<br />
(Robert Mannyng of Brunne’s ‘Handlyng Synne’ 161)<br />
And in 11.3, from Old English Alfred’s Introduction to Laws, hie ‘<strong>the</strong>y’ has a singular<br />
reference and can refer to a woman or to a man:<br />
11.3 Gif oxa ofhnite wer oððe wif, Þæt hie dead sien,<br />
sie he mid stanum ofworpod . . .<br />
‘If an ox gores a man or a woman, so that <strong>the</strong>y be dead, may he [<strong>the</strong> ox] be<br />
killed with stones’ (Alfred’s Introduction to Laws 32)<br />
Similar examples are attested in Middle English, as in 11.4:<br />
11.4 Yff man or woman take sekenes that day, <strong>the</strong>i schuld sone recouer . . .<br />
‘If a man or woman take sickness that day, <strong>the</strong>y should soon recover . . .’<br />
(Metham, Physiognomy 149)<br />
That is, <strong>the</strong> option of using a generic form <strong>the</strong>y for singular reference has been<br />
available in <strong>the</strong> written language since <strong>the</strong> earliest times. The generic use of he can be<br />
accounted for by a purely linguistic reason: <strong>the</strong> general word for person mann in Old<br />
English belonged to <strong>the</strong> masculine Linguistic <strong>Gender</strong>. The early spread of <strong>the</strong> generic<br />
masculine form in Old English and <strong>the</strong>n in Middle English—notwithstanding <strong>the</strong><br />
demise of agreement—may have had ano<strong>the</strong>r, social, reason: men, more than<br />
women, were literate in <strong>the</strong>ir majority, and thus were <strong>the</strong> targets of <strong>the</strong> written texts.<br />
The ‘masculinization’ of language had its roots in <strong>the</strong> grammatical system of <strong>the</strong><br />
language where masculine Linguistic <strong>Gender</strong> is <strong>the</strong> functionally unmarked choice.<br />
This is a feature English shares with a number of o<strong>the</strong>r languages, including many<br />
ancient members of <strong>the</strong> Indo-European family. To avoid a plethora of wordy<br />
alternatives, <strong>the</strong> generic he was given <strong>the</strong> support of a law by <strong>the</strong> English Parliament<br />
in 1850: ‘An Act for shortening <strong>the</strong> language used in acts of Parliament’ ordered ‘that<br />
in all acts words importing <strong>the</strong> masculine gender shall be deemed and taken to
196 11 <strong>Gender</strong> in grammar and society<br />
include females, and <strong>the</strong> singular to include plural, and <strong>the</strong> plural <strong>the</strong> singular, unless<br />
<strong>the</strong> contrary as to gender and number is expressly provided.’ 24<br />
The lack of a proper generic pronoun continued to be felt as a ‘gap’—<strong>the</strong> generic<br />
‘he’ was seen as mostly ‘masculine’. Attempts at trying to introduce a new epicene<br />
pronoun into English started in earnest in <strong>the</strong> early nineteenth century. In 1884,<br />
Charles Converse, an American lawyer and hymn-writer, created perhaps <strong>the</strong> widest<br />
known epicene pronoun thon (and possessive thons) as a contraction, or a blend, of<br />
‘that one’. The form never got to be widely used—though it did make its way into a<br />
number of dictionaries, among <strong>the</strong>m Webster’s Second New International Dictionary.<br />
Feminists were not <strong>the</strong> first to feel that <strong>the</strong> generic use of he reflected sexual bias<br />
within <strong>the</strong> language. Otto Jespersen (1894: 27–9) believed that introducing a common<br />
gender pronoun would make sexual bias less prominent. In his own words, ‘ladies<br />
would be spared <strong>the</strong> disparaging implication that <strong>the</strong> leading poets are men’,ifthon as a<br />
generic third person pronoun replaced he in <strong>the</strong> sentence: It would be interesting if each<br />
of <strong>the</strong> leading poets would tell us what he considers his best work. He saw no future for<br />
thon, instead favouring <strong>the</strong> singular <strong>the</strong>y. Many o<strong>the</strong>r alternatives sprang up, among<br />
<strong>the</strong>m borrowings (French on, le, anden), blends (he'er, shem), clippings (e, per), and<br />
new creations (na, ae, ip). 25 Most of <strong>the</strong>m attracted criticism or were neglected, and<br />
none of <strong>the</strong>m made its way into <strong>the</strong> language used by any of <strong>the</strong> English-speaking<br />
communities. The title of Dennis Baron’s (1981) extensive account of epicene pronouns—‘<strong>the</strong><br />
word that failed’—captures <strong>the</strong> sad fate of <strong>the</strong> new coinages.<br />
The most recent ones are <strong>the</strong> attempts to create special pronouns for <strong>the</strong> transgender<br />
community—gender-neutral ze (and its possessive hir) was introduced by<br />
students of Wesleyan University. We are yet to see whe<strong>the</strong>r and when it will get<br />
established: <strong>the</strong> article on ze in The Harvard Crimson (Caputo, 2 November 2005)<br />
ends with a quote from a Wesleyan student saying that ze isn’t ‘exactly ubiquitous’<br />
even in Wesleyan; ‘it is more frequently invoked with irony’.<br />
This is somewhat similar to <strong>the</strong> gap in second person in English: you covers<br />
singular and plural, and may create difficulties in communication. The literary<br />
language offers no solution. But speakers of English come up with various alternatives,<br />
among <strong>the</strong>m y’all or you all, yous, andyou guys—to which we will return<br />
shortly.<br />
The generic use of he came under attack as a direct reflection of male dominance in<br />
<strong>the</strong> second half of <strong>the</strong> twentieth century with <strong>the</strong> rise of <strong>the</strong> feminist movement (<strong>the</strong><br />
major question being whe<strong>the</strong>r women were to be included in <strong>the</strong> scope of he as a<br />
generic pronoun or excluded from it: in <strong>the</strong> latter case, he is seen as ‘pseudo-generic’). 26<br />
Experimental studies showed a tendency to identify ‘he’ as a male. 27<br />
A linguistic issue—using one pronoun as a functionally unmarked option—soon<br />
became political. In November 1971 a number of women students at Harvard<br />
Divinity School called for a ban on <strong>the</strong> use of ‘man’, ‘men’, and masculine pronouns<br />
to refer to God or to people in general. A group of leaders in <strong>the</strong> field of linguistics
11.5 Thwarting ‘sexist language’ 197<br />
reacted to this proposal ‘to recast part of <strong>the</strong> grammar of <strong>the</strong> English language’, ina<br />
letter to The Harvard Crimson (<strong>the</strong> Harvard University periodical) of 16 November<br />
1971. The letter was signed by <strong>the</strong> departmental secretary and sixteen linguists—<br />
many of <strong>the</strong>m household names in linguistics, including Ives Goddard, Michael<br />
Silverstein, Calvert Watkins, Einar Haugen, Jay Jasanoff, Stephen Anderson, Sandra<br />
Chung, Alan Timberlake, and Robert Underhill.<br />
The authors appealed to <strong>the</strong> notion of functional markedness (see §7.7), as ‘one of<br />
<strong>the</strong> fundamental principles which govern <strong>the</strong> organization of <strong>the</strong> internal economics<br />
of all human languages’. In <strong>the</strong>ir own words,<br />
in <strong>the</strong> matter of gender, in some cases <strong>the</strong> feminine is unmarked, in o<strong>the</strong>r cases <strong>the</strong> masculine.<br />
The feminine goose is unmarked—geese can be all male, all female, or of mixed sex, but<br />
ganders are all male. On <strong>the</strong> o<strong>the</strong>r hand <strong>the</strong> masculine lion is unmarked—contrast <strong>the</strong> possible<br />
ranges of meaning of lions and lionesses.<br />
Turning to pronouns and terms for people, <strong>the</strong>y continued:<br />
<strong>the</strong> masculine is <strong>the</strong> unmarked and hence is used as a neutral or unspecified term. This<br />
reflects <strong>the</strong> ancient pattern of <strong>the</strong> Indo-European languages, seen also, for example, in<br />
French: hommes et femmes heureux ‘happy men and women’ (with <strong>the</strong> masculine form of<br />
<strong>the</strong> adjective). Thus we say: All men are created equal. Each student shall discuss his paper<br />
topic with his section man. Madam Chairman, I object. The fact that <strong>the</strong> masculine is <strong>the</strong><br />
unmarked gender in English (or that <strong>the</strong> feminine is unmarked in <strong>the</strong> language of <strong>the</strong><br />
Tunica Indians) is simply a feature of grammar. It is unlikely to be an impediment to any<br />
change in <strong>the</strong> patterns of <strong>the</strong> sexual division of labor towards which our society may wish to<br />
evolve. There is really no cause for anxiety or pronoun-envy on <strong>the</strong> part of those seeking<br />
such changes.<br />
The idea of ‘pronoun-envy’ can be seen in ano<strong>the</strong>r way: R. M. W. Dixon pointed<br />
out that English-speaking women have a better linguistic deal than do men, as <strong>the</strong>y<br />
do have a pronoun of <strong>the</strong>ir own while men have to share <strong>the</strong>ir pronoun with <strong>the</strong> rest<br />
of <strong>the</strong> world. He experienced <strong>the</strong> reverse situation first-hand, during his fieldwork<br />
among <strong>the</strong> Jarawara of sou<strong>the</strong>rn Amazonia. Feminine gender in Jarawara is <strong>the</strong><br />
functionally unmarked option (see §2.3 and §7.7). All pronouns are marked on <strong>the</strong><br />
verb with a feminine agreement suffix: one of <strong>the</strong> difficulties R. M. W. Dixon<br />
encountered in using <strong>the</strong> language was that he had to talk about himself using <strong>the</strong><br />
feminine form (e.g. I walked: feminine). This in no way implies a privileged or<br />
elevated position of women in <strong>the</strong> Jarawara society (which continues to be maledominated).<br />
As we saw in §7.2, a respected Jarawara woman can be referred to with<br />
masculine gender, ‘promoting’ her to <strong>the</strong> status of an honorary male.<br />
This is not <strong>the</strong> way feminist linguists see it—for many, having a feminine pronoun<br />
is an impediment ra<strong>the</strong>r than a gain. But <strong>the</strong> lack of a specifically feminine<br />
pronoun can be felt as a ‘gap’ in <strong>the</strong> language: we can recall, from §5.3, how a new<br />
partly jocular form temake has been recently introduced into Modern Estonian.
198 11 <strong>Gender</strong> in grammar and society<br />
Having this form helps translate expressions like ‘his and hers’ into <strong>the</strong> language—<br />
and we find it used this way in adverts and <strong>the</strong> like.<br />
And yet <strong>the</strong> trend against generic, or pseudo-generic, he—and its sexist overtones,<br />
especially for <strong>the</strong> vociferous milieu of feminists somewhat naive with regard to<br />
general linguistic issues—is winning. Robin Lakoff—writing in 1973 and 1975—<br />
was pessimistic about changing people’s attitudes to <strong>the</strong> pseudo-generic he which<br />
she used throughout her 1975 classic. And now, forty years on, he as a generic is on<br />
<strong>the</strong> wane. 28<br />
Any person starting on a new job in a university in an English-speaking country is<br />
given a set of instructions to use ‘gender-neutral’,or‘gender-inclusive’ language. Staff<br />
and students at <strong>the</strong> University of Melbourne are encouraged to use a gender-neutral<br />
pronoun <strong>the</strong>y (and not he), or reword a sentence to avoid personal pronouns, or<br />
replace he with s/he, she/he, he/she,orhe or she. A dispreferred alternative in 11.4ais<br />
rephrased as 11.4b:<br />
11.4a Where <strong>the</strong> scholar has been pursuing research he shall present to <strong>the</strong> Council<br />
a report embodying <strong>the</strong> results of his work<br />
11.4b Where <strong>the</strong> scholar has been pursuing research <strong>the</strong>y shall present to <strong>the</strong><br />
Council a report embodying <strong>the</strong> results of <strong>the</strong>ir work<br />
And 11.5a should be abandoned in favour of 11.5b:<br />
11.5a The student must present his research to <strong>the</strong> examiner<br />
11.5b The student must present <strong>the</strong> research to <strong>the</strong> examiner 29<br />
As Jenny Cheshire reports (2008: 9–10), ‘it is encouraging that in 2007, in British<br />
university circles at least, it has become unthinkable to use he with an intended<br />
generic sense. Plural subjects are used, or <strong>the</strong> compound he or she. In my own<br />
university [Queen Mary College, University of London], he or she is de rigueur in<br />
speaking as well as in writing. None of my colleagues would dare to use a generic he<br />
pronoun; if one should slip out accidentally, it is soon corrected, by someone else.<br />
This is surely a sign of change.’<br />
The guidelines for European international organizations (such as <strong>the</strong> Council of<br />
Europe) follow similar principles. And so do most publishers. In The Cambridge<br />
Australian English style guide, Peters (1995: 332) states that ‘in ordinary usage he/<br />
his/him seems to be losing its capacity to be common and generic’. The generic<br />
‘unsex’ <strong>the</strong>y is becoming <strong>the</strong> preferred option (with <strong>the</strong> singular reflexive <strong>the</strong>mself no<br />
longer rejected as ‘incorrect’). In her analysis of a variety of sources in British<br />
English, Wales (1996: 125–33) shows how <strong>the</strong>y comes to be used more and more<br />
frequently to refer to nouns whose gender is not specified, in current usage by native<br />
speakers. Similar guidelines for ‘non-discriminatory language use’ are at work for job<br />
advertisements in Welsh: if an advertisement is phrased in <strong>the</strong> singular masculine
11.5 Thwarting ‘sexist language’ 199<br />
and feminine pronoun are used toge<strong>the</strong>r (e.g. iddo ef/iddi hi (for.3sg.masc him/for.3.<br />
fem her) ‘for him/for her’), or a plural form ‘<strong>the</strong>y’ (which does not distinguish<br />
genders) is used. 30<br />
Over <strong>the</strong> last decades we—as lingusts—have witnessed a ra<strong>the</strong>r remarkable change<br />
in <strong>the</strong> grammar of English which has affected <strong>the</strong> closed class of pronouns. The<br />
sphere of use of <strong>the</strong> generic he is now drastically diminished. But this is not to say that<br />
<strong>the</strong> generic he is fully obsolete; one hears this use in casual conversations, and in <strong>the</strong><br />
literature. In a recent novel by Ian McEwan Solar (2011: 71), <strong>the</strong> main character,<br />
Michael Beard, walks into a change room on a ship in <strong>the</strong> Arctic and discovers that<br />
‘someone has put his gear on Bear’s station’. The group of people on <strong>the</strong> ship is mixed<br />
sex: ‘someone’ may have been a man or a woman. In Jenny Cheshire’s experience<br />
(2008: 10), ‘several students still write he in <strong>the</strong>ir written work’, so much so that ‘<strong>the</strong>re<br />
is no guarantee that, if change has occurred amongst some people, it will persist’. For<br />
some, this is still an ideological issue: as Michael Silverstein (1985: 253) put it, <strong>the</strong> use<br />
of generic he (or lack <strong>the</strong>reof) ‘is turned into an index of a certain absence or presence<br />
of ideological solidarity with <strong>the</strong> reformers’.<br />
Despite all this, <strong>the</strong> masculine gender forms in English have not lost <strong>the</strong>ir functionally<br />
unmarked status. They remain, in Goddard et al.’s words (1971), as ‘<strong>the</strong><br />
unmarked and hence [ ...]used as a neutral or unspecified term’. This can be seen in<br />
<strong>the</strong> new plural form for second person in English you guys which contains <strong>the</strong> form<br />
guy with an erstwhile male reference. You guys can be directed to men, women, or to<br />
a mixed group, and is used by women and by men. (I first heard it used by a highprofile<br />
female financial analyst talking to her teenage daughters.) You guys can hardly<br />
be considered ‘sexist language’: <strong>the</strong>re are certain limits beyond which even <strong>the</strong> most<br />
ardent reformers cannot go.<br />
Unlike English, German has a generic pronoun man ‘one, person’ which is<br />
perfectly gender-neutral. <strong>How</strong>ever, it sounds similar to Mann ‘man, person’ (from<br />
which it comes historically). To avoid potential masculine bias, a new pronoun frau<br />
‘generic (woman)’ based on Frau ‘woman’ was introduced. According to a study in<br />
Storjohann (2004: 313) itfirst appeared in <strong>the</strong> Duden dictionary of German in<br />
1997. 31 This indefinite feminine form has partly made it into general usage (though<br />
it does not appear in many dictionaries). It differs from man in that its reference is<br />
exclusively feminine. The pronoun differs from <strong>the</strong> generic man in its meaning, and<br />
in its syntactic behaviour. For instance, in complex sentences, frau ‘she: generic’ can<br />
be anaphorically replaced by sie ‘she’,asin11.6. In contrast, man ‘generic person’ can<br />
only be replaced by man, and not by er ‘he’ or sie ‘she’—as shown in 11.7.<br />
11.6 Wenn frau sich bemüht, kann sie es schaffen<br />
‘If one:feminine makes an effort, she can make it’<br />
11.7 Wenn man sich bemüht, kann man/*er es schaffen<br />
‘If one:generic makes an effort, one (*he) can make it’
200 11 <strong>Gender</strong> in grammar and society<br />
That is, <strong>the</strong> pronoun man has a much less gender-specific meaning than frau. In a<br />
fur<strong>the</strong>r attempt to redress <strong>the</strong> balance between indefinite pronouns, feminist linguists<br />
have coined ano<strong>the</strong>r indefinite mann ‘generic male person’ as a counterpart to frau,<br />
in addition to a number of female-specific alternatives to established generic pronouns<br />
which contain man ‘man’, e.g. jedefrau ‘every woman’ (a counterpart of <strong>the</strong><br />
established form jederman ‘everybody, “every man”’) and jefraud ‘some women’ (as a<br />
counterpart of jemand ‘someone’). It remains to be seen whe<strong>the</strong>r or not <strong>the</strong>se will<br />
become fully established in <strong>the</strong> language beyond feminist literature.<br />
A generic gender-neutral pronoun in Norwegian has been an issue for <strong>the</strong> Norwegian<br />
language council since <strong>the</strong> 1980s. The feminine pronoun ho or hun ‘she’ was<br />
suggested as a replacement for <strong>the</strong> generic masculine han (alongside <strong>the</strong> indefinite<br />
pronoun ein/en ‘one’, replacing man ‘one’ which is obviously derived from mann<br />
‘man’, <strong>the</strong> noun folk ‘people’, or a plural form which does distinguish genders). The<br />
feminine form was finally accepted in children’s schoolbooks in 1989—this was<br />
considered a major victory for feminists. 32<br />
In <strong>the</strong> official guidelines for language use in Swedish, masculine pronoun han is<br />
used for generic or gender-neutral reference. The Language Council of Sweden, a<br />
major authority on language planning, also supports den ‘it’ or han eller hon ‘he or<br />
she’ as a way of avoiding <strong>the</strong> generic masculine. A fur<strong>the</strong>r innovation is a new genderneutral<br />
pronoun hen increasingly used among lesbian, gay, bisexual, and transgender<br />
people. The form may have originated in a blend of three pronouns: han ‘he’, hon<br />
‘she’, and den ‘it’; an alternative hypo<strong>the</strong>sis is that it comes from Finnish hän ‘he/she/<br />
it’. The new form has not been supported by <strong>the</strong> Language Council; never<strong>the</strong>less it is<br />
making its way into <strong>the</strong> language, especially for feminists and members of <strong>the</strong> queer<br />
community. It is as yet hard to predict whe<strong>the</strong>r it will spread beyond <strong>the</strong>se. 33<br />
The issue of masculine gender as a functionally unmarked choice in agreement<br />
on adjectives and verbs does not arise in English or in German. 34 But it continues to<br />
be an issue in French and Spanish. Following <strong>the</strong> recommendations of The French<br />
Council of Europe Guidelines, French occupational terms have to always be given in<br />
a feminine and a masculine form. But agreement follows <strong>the</strong> established grammatical<br />
pattern: only a masculine form will be used with two conjoined nouns, such as un<br />
expert et une expert sera nommé ‘an expert (masc.sg) and an expert (fem.sg) will be<br />
named (masc.sg)’. 35<br />
Similarly, in Spanish a mixed group has to be referred to with masculine gender:<br />
‘one single man in an audience will be sufficient for <strong>the</strong> speaker to address <strong>the</strong><br />
listeners by masculine terms, e.g. vosotros (masculine plural) “you” or los presentes<br />
(masculine plural) “<strong>the</strong> ones present”. To use feminine forms in such cases would be<br />
tantamount to breaking <strong>the</strong> rules of grammar. The same is true with respect to<br />
agreement.’ For instance, one would say Juan (m) y Juana (f) son campesinos<br />
extraordinarios ‘Juan and Juana are extraordinary (masculine pl) peasants (masculine<br />
plural)’, despite <strong>the</strong> fact that Juana on her own will be referred to as campesina
11.5 Thwarting ‘sexist language’ 201<br />
extraordinaria. 36 The European Parliament guidelines for Portuguese reflect such<br />
usage: <strong>the</strong>y explicitly recommend <strong>the</strong> use of generic masculines whenever a group of<br />
different sexes is referred to:<br />
11.8 os antigos primeiros-ministros<br />
<strong>the</strong>:masc.plural former.masc.pl prime.masc.pl-minister.masc.pl<br />
Maria Pintassilgo, Mário Soares, Margaret Thatcher e Tony Blair<br />
Maria Pintassilgo, Mário Soares, Margaret Thatcher and Tony Blair<br />
‘<strong>the</strong> former prime-ministers Maria Pintassilgo, Mário Soares, Margaret Thatcher,<br />
and Tony Blair’<br />
This takes us back to Goddard et al.’s (1971) point: <strong>the</strong> use of masculine forms as<br />
functionally unmarked generics is a fact of grammar which lies beyond <strong>the</strong> reach of<br />
language reformers.<br />
11.5.2 Fighting <strong>the</strong> ‘generic masculine’ throughout <strong>the</strong> language<br />
As women started entering professions previously considered ‘male only’, <strong>the</strong> question<br />
of ‘feminization’ of job titles became ripe. In §7.6, we briefly mentioned <strong>the</strong> issue<br />
of feminine forms for job titles and positions. For <strong>the</strong> first time in <strong>the</strong> history of<br />
France, women constituted more than one-third of ministers in Lionel Jospin’s<br />
government (elected in 1997). The six women ministers explicitly insisted on being<br />
addressed as Madame la ministre. The meeting of <strong>the</strong> Council of Ministers on 17<br />
December 1997 decided that in <strong>the</strong> future women should be addressed with feminine<br />
titles. The proposal was immediately put into practice, and in March 1998 <strong>the</strong> Prime<br />
Minister legalized it by producing a decree Circulaire du 6 mars relative à la<br />
féminisation des noms de métier, fonction, grade ou titre (Jospin 1998). The official<br />
recognition was met with a protest from <strong>the</strong> Académie française, <strong>the</strong> main authority<br />
on French language and culture.<br />
In 1999, <strong>the</strong> guidelines of <strong>the</strong> Institut National de la Langue Française offered an<br />
extensive list of feminized job titles. Feminization of titles and professions in modern<br />
French has achieved a certain measure of success. And yet, as Pastre (1997: 376) put<br />
it, ‘The honor and prestige of masculine forms fascinate many women, who are<br />
happy to accede to honors that have until now been reserved for men . . . When <strong>the</strong><br />
writer Hélène Carrère d’Encausse was accepted into <strong>the</strong> Académie française, for<br />
example, her daughters made it clear that <strong>the</strong>y did not consider her une écrivaine<br />
(a woman writer) and did not approve of <strong>the</strong> feminisation of job titles.’ This author,<br />
now <strong>the</strong> Permanent Secretary of <strong>the</strong> Académie, is referred to with <strong>the</strong> masculine form<br />
as le secrétaire (and not as a feminine la secrétaire) ().<br />
Marking both genders—especially in <strong>the</strong> written language—is an uncanny way of<br />
creating compromise structures. For instance, a feminine marker in upper-case letters
202 11 <strong>Gender</strong> in grammar and society<br />
can be added to a masculine form, to create an all-embracing form which would<br />
subsume men and women, e.g. manifestant-E (demonstrator(masc.sg)-fem.sg ‘demonstrator<br />
(masculine and feminine)’. 37<br />
In o<strong>the</strong>r Romance-speaking countries, <strong>the</strong> situation is different. The only female<br />
minister in <strong>the</strong> ill-fated Collor government in Brazil (1990–2), Zélia Maria Cardoso<br />
de Mello, was always referred to as ministra (feminine form of ministro, minister)<br />
and mentora (feminine of mentor) of Collor himself. Most job titles in Brazil are used<br />
in a masculine and a feminine version with no negative overtones attached to ei<strong>the</strong>r,<br />
e.g. juiz ‘male judge’ versus juiza ‘female judge’, professor ‘male professor, teacher’,<br />
professora ‘female professor, teacher’, orientador ‘male supervisor’ and orientadora<br />
‘female supervisor’, chefe ‘male chief, head (of an organization)’, chefa ‘female chief,<br />
head (of an organization)’. The feminine form of presidente (presidenta) was reintroduced<br />
by law in 2012, and is used to specifically refer to <strong>the</strong> first female president of<br />
Brazil, Dilma Rousseff (see §7.5).<br />
The first official guidelines for non-sexist language use in Spanish (1988) and <strong>the</strong><br />
subsequent UNESCO guidelines explicitly stated <strong>the</strong> ‘propagation of <strong>the</strong> feminisation<br />
of words explicitly referring to women’, and ‘replacement of <strong>the</strong> masculine form for<br />
generic reference by epicene words or by splitted [sic!] constructions’. The sign @ has<br />
been suggested as a way of graphically combining masculine and feminine endings—<br />
so l@s alumn@s will be a blend of los alumnos (masc.pl students.masc.pl) and las<br />
alumnas (fem.pl students.fem.pl). Bengoechea’s (2011: 48) results show that such<br />
blends, and feminine forms of professions, are being more and more accepted. This<br />
echoes Nissen’s (2013) conclusions—that speakers of Spanish in 2005 showed much<br />
less sexist bias in interpretation of generic masculine forms than <strong>the</strong>y did in 1995.<br />
Women have become more visible in public life—and <strong>the</strong> language appears to have<br />
become more inclusive, and perhaps more receptive to feminine forms. 38<br />
Guidelines for non-discriminatory usage in English warn against ‘false generics’—<br />
that is, man and derivations which include -man. This has indeed resulted in general<br />
avoidance of gender-specific forms such as Chairman: we talk about Chair, or<br />
Chairperson. What Goddard et al. (1971) phrased as ‘Madam Chairman, I object’<br />
sounds distinctly obsolete: one will now say ‘Madam Chair’. Guidelines for nongender-biased<br />
use in German, under <strong>the</strong> title ‘More women in language’ (Braun<br />
2000b), focus on gender equality in derivations. The guidelines are being followed—<br />
at least in official contexts. 39<br />
Dictionaries tend to lag behind <strong>the</strong> times in how <strong>the</strong>y represent women. Hampares<br />
(1976) looked at 105 job denominations in The Spanish Royal Academy dictionary<br />
and two Spanish–English dictionaries, and concluded that only a masculine form was<br />
given for high-ranking jobs, such as científico ‘scientist’ and diplomático ‘diplomat’.<br />
Lower-ranking jobs such as sirvienta ‘servant’ or asistenta appeared in <strong>the</strong> feminine<br />
form. Quite a few feminine forms were in <strong>the</strong> dictionary alongside <strong>the</strong>ir masculine<br />
counterpart but were interpreted as ‘<strong>the</strong> wife of . . . ’, e.g. embajador ‘ambassador’,
11.5 Thwarting ‘sexist language’ 203<br />
embajadora ‘wife of ambassador’. There were no feminine forms for jobs such as<br />
mecánico ‘mechanic’ or ingeniero ‘engineer’. Nissen (2002: 266) reports that <strong>the</strong> latest<br />
edition of <strong>the</strong> Academy’s prescriptive dictionary did not show much improvement<br />
on <strong>the</strong> earlier version.<br />
French dictionaries have not fared better. In her critique of English–French<br />
dictionaries (Collins, Harrap, Larousse, and Oxford-Hachette, 1978–2001), Élisabeth<br />
Campbell (2004) pointed out a strong masculine bias: mostly masculine singular<br />
forms are given in French translations, and <strong>the</strong> generic he is prevalent in <strong>the</strong> English<br />
part. Within dictionary entries, <strong>the</strong>re is a strong masculine bias; examples involving<br />
women are focused on domestic activities. Feminine equivalents are given for some<br />
professions but not for o<strong>the</strong>rs: Oxford-Hachette contains <strong>the</strong> word ministre as<br />
masculine and feminine, and <strong>the</strong> feminine form of député ‘delegate’ (députée ‘female<br />
delegate’); feminine forms of o<strong>the</strong>r nouns, including avocate ‘female lawyer’, éditrice<br />
‘female editor’, jurée ‘female member of a jury’, were not included. This in itself may<br />
have nothing to do with being anti-feminist. Dictionaries are prescriptive and<br />
notoriously conservative, reluctant to embrace new terms, and to keep pace with<br />
language change. Indirectly, <strong>the</strong>ir conservatism serves to reinforce <strong>the</strong> masculine bias<br />
and is likely to cause offence. 40<br />
The ‘generic masculine’ is not perceived as sexist in mainstream Hebrew-speaking<br />
society. We can recall, from §2.3 and also §7.7, that in Hebrew and in Arabic, masculine<br />
forms (in singular and in plural) have generic reference. In Hebrew, <strong>the</strong> masculine<br />
singular pronoun ata can be used in general, to talk about a woman or a man, much like<br />
English one or you. A Hebrew-speaking Israel fashion designer explained why she had<br />
become self-employed: ‘One day you (masc.sg) suddenly realize that you (masc.sg) are<br />
(masc.sg) a mo<strong>the</strong>r and all you (masc.sg) wish (masc.sg) to do is stay at home with <strong>the</strong><br />
children.’ 41 Using generic masculine is an unconscious choice. At a meeting of <strong>the</strong><br />
Israeli Parliament (Knesset) committee for <strong>the</strong> promotion of <strong>the</strong> status of women in<br />
March 2004, dedicated to <strong>the</strong> address of a mixed-gender audience, <strong>the</strong> representative of<br />
<strong>the</strong> Hebrew Language Academy—a major regulatory entity for <strong>the</strong> language—opposed<br />
all attempts to introduce feminine forms in <strong>the</strong>ir generic reference. She argued that this<br />
would go against <strong>the</strong> internal structure of <strong>the</strong> language. The point she made was<br />
reminiscent of Goddard et al. (1971) (masculine and feminine endings are in bold<br />
face): ‘When I talk about giving birth, I say “when giving birth (masc.pl) (ksheyoldim)”.<br />
Ican’tsay“when giving birth (fem.pl) (ksheyoldot)”. You will agree with me that <strong>the</strong>se<br />
forms are not masculine, <strong>the</strong>y are unmarked. It is true that historically <strong>the</strong>se were<br />
masculine forms, but essentially <strong>the</strong>y are forms that include <strong>the</strong> female. When I say<br />
‘writing.fem.pl’ (kotvot) I exclude you [<strong>the</strong> addressee was a man], but when I say ‘writing<br />
(masc.pl) (kotvim) [masc.pl], I don’t exclude myself. I am in <strong>the</strong>re. I am telling you as a<br />
woman that it doesn’t disturb me.’<br />
There appears to be little support for <strong>the</strong> feminist language reform in Israeli<br />
society. Jacobs (2004: 237) tells a cautionary tale about Na'ama, an elected official
204 11 <strong>Gender</strong> in grammar and society<br />
on <strong>the</strong> governing council of a major Israeli city. Her term in office was marked by<br />
struggle against institutional practices that limited women’s access to political and<br />
economic power, and her individual attempt to reform <strong>the</strong> language. She attempted<br />
to get city council stationery that would use <strong>the</strong> female title for ‘member of city<br />
council’—xavrat mo'etsa ra<strong>the</strong>r than <strong>the</strong> masculine xaver mo'etsa which used to be<br />
used for males and female alike before her. At first, her request was denied, and only<br />
after repeated struggle, honoured: after two years, <strong>the</strong> municipality ended up printing<br />
official council stationery with <strong>the</strong> feminine title—but Na'ama was <strong>the</strong> only one to<br />
use it. O<strong>the</strong>r female members of <strong>the</strong> council chose not to. As Jacobs (2004: 245) puts<br />
it, this example underscores ‘<strong>the</strong> marginal position of feminists in <strong>the</strong> mainstream<br />
Israeli socio-political context’. The resistance, or ra<strong>the</strong>r, indifference, to <strong>the</strong> generic<br />
masculine forms in Hebrew does not in any way imply that women have a lower<br />
status than men. 42 It is just that <strong>the</strong> feminist movement has not impacted <strong>the</strong><br />
language—social and linguistic change do not have to proceed simultaneously.<br />
11.5.3 Bias in address terms and naming patterns<br />
In most European languages, courtesy titles and honorifics treat women differently<br />
from men. Women are addressed on <strong>the</strong> basis of <strong>the</strong>ir marital status: a married<br />
woman will be traditionally addressed as a Mrs in English, Mevrouw in Dutch,<br />
Madame in French, Senhora in Portuguese, Señora in Spanish, Signora in Italian,<br />
Fru in Danish, and Frue in Norwegian. An unmarried woman will be addressed as<br />
Miss in English, Juffrouw or Mejuffer in Dutch, Mademoiselle in French, Fräulein in<br />
German, Senhorita in Portuguese, Señorita in Spanish, Signorina in Italian, Fruken in<br />
Danish, and Froken in Norwegian. In contrast, <strong>the</strong>re is just one way of addressing a<br />
man; special courtesy titles for young unmarried men, such as Master in English and<br />
Jonge Heer in Dutch, are not often used and more markers of age than of marital<br />
status. Surnames may reflect <strong>the</strong> marital status of a woman. In Lithuanian, an<br />
unmarried girl will be given her fa<strong>the</strong>r’s surname with <strong>the</strong> suffix -aitė, -ytė, -utėʼ or<br />
-tė. A married woman’s surname will be derived from her husband’s name with <strong>the</strong><br />
suffix -ene. There are no markers which would reflect <strong>the</strong> marital status of a man.<br />
The marital status of women has traditionally been flagged—in contrast to that of<br />
men—signalling <strong>the</strong> lack of equality between Social <strong>Gender</strong>s. As Silverstein (2015)<br />
put it, <strong>the</strong> private becomes political (echoing a slogan of <strong>the</strong> second wave feminist<br />
movement). The term Ms in English was registered in <strong>the</strong> Oxford English Dictionary<br />
Online as early as 1901 and was later adopted as an alternative to avoid marking<br />
marital status (though it is not accepted by everyone). 43 The reform in German went<br />
even fur<strong>the</strong>r. Since <strong>the</strong> 1970s, <strong>the</strong> form Frau is <strong>the</strong> only option to address a grown-up<br />
woman (no matter whe<strong>the</strong>r married or not). The term Fräulein ‘miss, unmarried<br />
woman’ is all but obsolete. 44 This is evidence of progress in eliminating male–female<br />
asymmetry in address terms.
11.5 Thwarting ‘sexist language’ 205<br />
Western naming practices have—traditionally—ensured that women remain<br />
invisible. Children in Western societies are given <strong>the</strong> surname of <strong>the</strong> fa<strong>the</strong>r. In<br />
Spanish- and Portuguese-speaking countries, children are given <strong>the</strong> mo<strong>the</strong>r’s and<br />
fa<strong>the</strong>r’s name, but are typically referred to by <strong>the</strong> surname of <strong>the</strong> fa<strong>the</strong>r. The Spanish<br />
and <strong>the</strong> Portuguese traditions differ in <strong>the</strong> order of names: for Spanish speakers, <strong>the</strong><br />
mo<strong>the</strong>r’s name comes last. For instance, Gabriela García Salazar has a fa<strong>the</strong>r with <strong>the</strong><br />
surname García, and a mo<strong>the</strong>r with <strong>the</strong> surname Salazar. She would be referred to as<br />
García—which is her main surname in her native Peru. In Portuguese, <strong>the</strong> order is<br />
<strong>the</strong> opposite: <strong>the</strong> mo<strong>the</strong>r of Rafael da Silva Brito has <strong>the</strong> surname da Silva, and his<br />
fa<strong>the</strong>r’s surname is Brito. He is referred to as Rafael Brito, for short. During her<br />
studies in Brazil, Gabriela Garcia Salazar was known as Gabriela Salazar, following<br />
<strong>the</strong> Brazilian custom. Patronymics in Slavic-speaking traditions reflect <strong>the</strong> name of<br />
<strong>the</strong> fa<strong>the</strong>r—but not of <strong>the</strong> mo<strong>the</strong>r. The maiden name of my fa<strong>the</strong>r’s paternal<br />
grandmo<strong>the</strong>r was Nina Kirillovna Tomasheva (1868–1943). I thus know <strong>the</strong> first<br />
name of her fa<strong>the</strong>r, Kirill, and his surname (Tomashev, <strong>the</strong> masculine version of<br />
Tomasheva). But I have no idea what her mo<strong>the</strong>r’s name was. By keeping this secret,<br />
and by adopting her husband’s name, Nina Kirillovna Aikhenvald hid information<br />
about her aristocratic background—which was by no means an asset under <strong>the</strong> Soviet<br />
regime. The language came to her aid. The woman’s identity, name, and achievements<br />
were lost and obscured. So far, no attempts have been made to thwart <strong>the</strong><br />
practice of patronymics. 45<br />
Traditionally, Western women were expected to adopt <strong>the</strong>ir husbands’ surnames.<br />
At present, this is no longer <strong>the</strong> case; and many professional women keep <strong>the</strong>ir own<br />
name on marriage. As a consequence of <strong>the</strong> liberation of women in <strong>the</strong> Soviet<br />
Union, many educated professional women kept <strong>the</strong>ir maiden names—for some,<br />
giving up <strong>the</strong>ir name was like giving up <strong>the</strong>ir own identity. As my mo<strong>the</strong>r said to<br />
me, ‘if I had changed my surname (Gerlin) to that of my husband’s (Aikhenvald),<br />
I would have stopped being me’. The guidelines for non-discriminatory language<br />
from Melbourne University overtly state that ‘it has become increasingly common<br />
for women to keep <strong>the</strong>ir birth names’. ButPauwels(1998: 59) remarks that ‘women<br />
who do not conform to this practice still face many practical and attitudinal<br />
obstacles’. Mills (2003b) undertook a study of naming practices among professional<br />
women in Britain. Over half of thirty-six subjects did take <strong>the</strong>ir husband’s nameon<br />
marriage—but <strong>the</strong>y gave different reasons for <strong>the</strong> choice. One woman took a<br />
husband’s name because of lack of loyalty to her fa<strong>the</strong>r. Ano<strong>the</strong>r one liked <strong>the</strong><br />
husband’s name better. Naming practices have become negotiable, at least in middleclass<br />
Britain.<br />
A dangerous reverse trend is at work in developing countries which embrace <strong>the</strong><br />
influence of Christianity and Western naming practices. In many societies of New<br />
Guinea, women and children have recently been obliged to take on <strong>the</strong>ir husbands’<br />
and fa<strong>the</strong>rs’ names, following <strong>the</strong> Western colonial practice.
206 11 <strong>Gender</strong> in grammar and society<br />
Personal names are a major asset of <strong>the</strong> traditional Manambu people. In creating a<br />
person’s identity, both maternal and paternal name still play a role. The Manambu<br />
personal names (associated with a totem or a totemic ancestor) are ‘owned’ by<br />
subclans, so that every person carries names of his or her mythological forebears. 46<br />
The first, and <strong>the</strong> main name bestowed upon a person, would be a name belonging to<br />
<strong>the</strong>ir paternal clan. The name is termed ap-a-sə (bone-LINKER-name), literally, ‘<strong>the</strong><br />
bone name’ (we can recall, from §3.4.2, that ‘bone’ is associated with masculinity and<br />
patrilineal inheritance). Numerous paternal names bestowed upon a person by <strong>the</strong>ir<br />
paternal relatives throughout <strong>the</strong>ir lives indicate where people belong. Names<br />
bestowed upon a person by <strong>the</strong>ir maternal relatives protect <strong>the</strong>m from evil spirits<br />
while <strong>the</strong>y were alive. This naming practice is still alive—but only barely so. Many<br />
younger people and children can hardly remember <strong>the</strong>ir traditional names.<br />
Nowadays, children who go to school, and adults who have identity cards and<br />
passports, have to go under <strong>the</strong>ir fa<strong>the</strong>r’s name. Married women are obliged to take<br />
<strong>the</strong>ir husbands’ paternal names (something unheard of in traditional times). The late<br />
Gaialiwag whose husband’s name was Yirvi was a very traditional speaker of Manambu,<br />
with an impressive knowledge of totemic lore and names. She never referred<br />
to herself with her husband’s name until her various stints in <strong>the</strong> hospital in Wewak<br />
(<strong>the</strong> provincial capital) where she was made to follow this now standard practice<br />
for Papua New Guinea. In <strong>the</strong> last story she told me in 2013 she introduced herself<br />
as Wun Gaia, Yirvi du sawa (I Gaia, Yirvi man name+COMITATIVE) ‘I am Gaia, Yirvi<br />
by man’s name’. This was accompanied by an ironic raise of eyebrows and shoulder<br />
shrug—later explained to me as her dismissive reaction to this ‘new’ fashion of<br />
referring to a woman by her husband’s name. Angela Filer, a local politician and a<br />
highly proficient speaker of Kwoma, a neighbouring language, was scathing about <strong>the</strong><br />
westernized imposition of husbands’ names on married women, which background<br />
and ultimately destroy <strong>the</strong>ir identity. 47<br />
Names are just one aspect of a dangerous trend termed ‘repatriarchialization’.<br />
Abbink (2015: 294) comments on how traditional division of labour between men<br />
and women, and special female practices in Suri communities in Ethiopia, weaken<br />
under <strong>the</strong> newly introduced patriarchic practices of Evangelical Christianity; ‘<strong>the</strong><br />
gender egalitarianism that existed ...is undermined’. Colonial influences and <strong>the</strong><br />
destruction of traditional balance between male and female Social <strong>Gender</strong>s among<br />
<strong>the</strong> Igbo of Nigeria led to marginalization of women (Amadiume 2015: 147–61).<br />
11.6 Expression of Linguistic <strong>Gender</strong> and social change: a summary<br />
The meanings of Linguistic <strong>Gender</strong> may reflect <strong>the</strong> status, and <strong>the</strong> occupations, of men<br />
and women. So do o<strong>the</strong>r noun categorization devices applied to humans, including<br />
numeral classifiers. Linguistic <strong>Gender</strong>s, especially <strong>the</strong> use of pronouns and derivational<br />
forms, may mirror social changes through language reforms. The most remarkable
11.6 Expression of Linguistic <strong>Gender</strong> and social change 207<br />
success story for feminist reformers is that of a gradual demise of generic man and<br />
generic he in English, as a way of removing <strong>the</strong> masculine bias. A number of earlier<br />
attempts to create a generic sex-neutral (or ‘epicene’) pronoun, variation in anaphoric<br />
agreement, and o<strong>the</strong>r options—such as a generic <strong>the</strong>y, with singular reference—<br />
prepared a fertile ground for this change. The demise of generic he was not an<br />
innovation: ra<strong>the</strong>r, this was an engineered enhancement of already existing patterns.<br />
If <strong>the</strong> language is a guide to social reality, <strong>the</strong> way Linguistic <strong>Gender</strong> is manipulated<br />
is <strong>the</strong> most salient part of it. Incidentally, <strong>the</strong>se features of Linguistic <strong>Gender</strong> and<br />
<strong>the</strong> ways in which <strong>the</strong>y reflect societal stereotypes and ingrained attitudes to male<br />
and female practices fully debunk <strong>the</strong> myth of gender as a ‘useless’ decoration in a<br />
language.<br />
Scholars of gender should be careful in making <strong>the</strong>ir associations too straightforward.<br />
Having feminine gender as <strong>the</strong> functionally unmarked choice does not mean<br />
that men and women are socially equal, or that women occupy a privileged position.<br />
A conventionalized generic use of an erstwhile feminine pronoun does not in any<br />
way imply a privileged status of women. In Modern Italian, two polite address forms,<br />
Lei and Ella, are identical to third person feminine pronouns, even though <strong>the</strong>y are<br />
used equally in addressing respected males and females. The reason for this rests<br />
with <strong>the</strong> history of <strong>the</strong> language: honorific expressions used to indicate deference,<br />
respect, and admiration—similar to English ‘your honour’, ‘your Excellency’, and<br />
‘your worship’—tended to comprise nouns of feminine Linguistic <strong>Gender</strong>, e.g. la<br />
vostra signoria ‘your lordship’, la vostra eccellenza ‘your Excellency’. These nouns<br />
could also be referred to, anaphorically, with third person feminine pronouns ella<br />
and lei. They gradually became general forms of polite address. 48<br />
Linguistic change never happens quickly enough. Despite numerous substantial<br />
efforts in formulating, and imposing, guidelines for eliminating gender-biased language,<br />
some analysts remain pessimistic. At <strong>the</strong> end of <strong>the</strong>ir comprehensive analysis<br />
of European international organizations and language reform, Teso and Crolley<br />
(2013: 155) remark on ‘a disturbing gap between language policy initiatives and<br />
practice’, leading to ‘<strong>the</strong> lack of success in <strong>the</strong> feminisation of language’, even in <strong>the</strong><br />
languages of Western Europe. A full ‘feminization’ of any language is an impossible<br />
task: it would be linguistically naive to try and override grammatical rules—such as<br />
masculine agreement forms for mixed sex groups—for <strong>the</strong> sake of political struggle to<br />
make men and women equal. And using a formation you-guys to refer to a mixed<br />
group of men and women can hardly be seen as a mark of male supremacy. Linguistic<br />
reforms can only go so far: echoing Goddard et al. (1971), it is important to<br />
remember that grammatical features, such as markedness, can hardly be changed<br />
by guidelines or decrees—no matter how much feminist linguists may lament <strong>the</strong><br />
‘sexist’ nature of masculine agreement for mixed groups.<br />
Using Robin Lakoff ’s (2004: 171) words, ‘language discrepancy makes manifest<br />
social inequality’ in yet ano<strong>the</strong>r way. We can recall how in gender-exclusive
208 11 <strong>Gender</strong> in grammar and society<br />
languages (discussed in §9.1), male and female dialects may correlate with <strong>the</strong> relative<br />
status of <strong>the</strong> Social <strong>Gender</strong>s (see §9.1.5). Along similar lines, ‘male-only’ genres are<br />
often seen as more prestigious and showing more knowledge (§§10.1–3). A covert<br />
male bias may become detrimental to <strong>the</strong> survival of female genderlects. We can<br />
recall, from §9.1.4, that <strong>the</strong> Chukchi language is severely endangered, and hardly<br />
learnt by children. A partly artificial ‘standard’ language, promoted throughout <strong>the</strong><br />
Soviet period, was based on <strong>the</strong> men’s dialect—women were made invisible and not<br />
heard. Chukchi teachers and radio announcers use <strong>the</strong> male variety (female announcers<br />
continue using <strong>the</strong> female forms in private). Discourse about language preservation<br />
is framed entirely in terms of men’s language. The male bias is creeping in.<br />
Social change and growing equality affect <strong>the</strong> use of genderlects in westernized<br />
developed societies. The partial levelling of genderlects in Japanese and in Thai goes<br />
toge<strong>the</strong>r with women getting access to leading positions in society, and phasing out<br />
social asymmetries in Social <strong>Gender</strong>s.<br />
More than just language forms are at stake. We can recall, from §8.4 and §11.2,<br />
how in many cultures of patriarchic type women are stereotyped in predominantly<br />
negative terms. More often than not women, but not men, are portrayed as<br />
sexualized objects. As Cheshire (1985: 24) puts it, ‘newspapers give personal details<br />
about women, but not about men, so that we read about “blonde, shapely, mo<strong>the</strong>rof-three<br />
Mrs Smith”, but not about “bronzed, muscular Mr Jones”’. As demonstrated<br />
by Baker (2014) in his analysis of adjectives used with ‘man’ and ‘woman’ in<br />
<strong>the</strong> Historical Corpus of American English, <strong>the</strong> imbalance appears to be slowly<br />
changing (with men’s bodies being sexualized as much as women’s). As Jenny<br />
Cheshire (2008: 10) puts it in her reappraisal of linguistic sexism, ‘<strong>the</strong>re have been<br />
changes in <strong>the</strong> use of English in <strong>the</strong> last twenty-three years that suggest to me that<br />
our language is freer than before of masculine bias, even if <strong>the</strong> reform is far from<br />
complete. What needs to be determined now is <strong>the</strong> extent to which our thinking is<br />
equally free of masculine bias.’<br />
NOTES AND SOURCES<br />
1. A tight relationship between language and reality has been captured by <strong>the</strong> Sapir–Whorf<br />
hypo<strong>the</strong>sis: see Pauwels (1998: 82–3) and Talbot (2010: 15–16), for feminist approaches.<br />
Excellent discussion is in Salzmann (2004: 42–69).<br />
2. See Leger (1998: 206–7) and Dinslage, Leger, and Storch (2000: 125). Human nouns are<br />
assigned to Linguistic <strong>Gender</strong>s based on <strong>the</strong>ir sex.<br />
3. Vajda (2004: 17–18); Werner (1997: 92–4).<br />
4. Note that, so far, sexist ‘positive’ associations of ‘man’ and negative associations of ‘woman’<br />
have been documented exclusively for patriarchic societies. We lack any information on <strong>the</strong><br />
images of men and women in traditional matriarchies, where women traditionally play a
11.6 Expression of Linguistic <strong>Gender</strong> and social change 209<br />
leading role in inheritance and governance. In her study of matriarchies, Goettner-<br />
Abendroth (2012) does not mention <strong>the</strong> issue of names, naming, or language.<br />
5. Migge (2001: 99–100) on this and o<strong>the</strong>r pairs. Unfortunately, little information is available<br />
on overtones of gendered derivations in o<strong>the</strong>r languages spoken in matrilineal societies.<br />
6. Sahoo (2003: 252–3); see Allan and Burridge for this general trend (2006: 81), and<br />
Gregersen (1979) for feminine terms and insults cross-linguistically.<br />
7. See Barnhart (1988: 639, 668), <strong>the</strong> Oxford English Dictionary Online (entry mistress, sense<br />
7 for an earlier date). See also Lakoff (1975: 59–60), Henley (1989: 60), Cheshire (1985: 22).<br />
8. See Baron (1986: 121–2) for <strong>the</strong> history of <strong>the</strong>se feminine forms in English.<br />
9. See Aikhenvald (2002: 227).<br />
10. In <strong>the</strong> Oxford English Dictionary Online, <strong>the</strong> meaning of man up is given as ‘to demonstrate<br />
manliness, toughness, or courage when faced with a difficult situation; to take<br />
responsibility; to own up’.<br />
11. See also Aikhenvald (2013a). My corpus of Tariana contains several similar examples of<br />
<strong>the</strong> word ‘man’.<br />
12. Hasselblatt (2015: 141–2).<br />
13. Khaidakov (1963: 49–50) on Lak; Mithun (2014: 137–8) on Mohawk.<br />
14. Anderson and Janson (1997: 34–5), Joe Tsonope p.c.<br />
15. Lu(2012: 83–4, 101–2, 115, 119–21) on classifiers in Maonan and <strong>the</strong> correlations between<br />
prestige and <strong>the</strong> use of human classifiers with women. Raised numbers indicate tones.<br />
16. Khaidakov (1963: 50).<br />
17. See also Vinogradov (1947: 62).<br />
18. See Yokoyama (1999: 420–1), Fesenko and Fesenko (1955: 161).<br />
19. In his state-of-<strong>the</strong>-art description of word formation in twentieth-century Russian,<br />
Lopatin (1970: 125) stresses that many female forms are only used if <strong>the</strong> sex of <strong>the</strong> person<br />
needs to be emphasized.<br />
20. See a similar example in note 14 in Rothstein (1973: 465); Yokoyama (1999: 418–20), and<br />
Martynyuk (1990).<br />
21. See Pauwels (1998: 34–80), Cheshire (1985, 2008), Henley (1989: 60), Teso and Crolley<br />
(2013). Investigations of <strong>the</strong> ways women speak tie in with <strong>the</strong> history of feminism (see<br />
Walters 2005, for a brief history; and Mills (2003b) for a comparison of Second-wave<br />
feminism, from <strong>the</strong> 1960s until <strong>the</strong> 1990s, and <strong>the</strong> subsequent Third-wave feminism. See<br />
Lakoff (2004: 205) for a brief reappraisal of methodology used by feminist linguists, and<br />
warnings against <strong>the</strong> limitations of any one methodology—be it <strong>the</strong> Conversation Analysis<br />
or <strong>the</strong> Critical Discourse Analysis.<br />
22. See Hegardty (2014); and fur<strong>the</strong>r discussion in Key (1975).<br />
23. See <strong>the</strong> Oxford English Dictionary Online; <strong>the</strong> form gradually weakened to men and to me.<br />
One of <strong>the</strong> reasons for its demise may have been its similarity to <strong>the</strong> first person form me.<br />
24. Baron (1981: 84), Evans and Evans (1957: 221).<br />
25. See Baron (1981: 88–96) and (1986: 198ff.; 205–9) for <strong>the</strong>ir chronology. Stotko and Troyer<br />
(2007) reported a newly coined generic third person singular pronoun yo in Baltimore<br />
public schools.<br />
26. A history of generic masculine ‘he’ and attempts to replace it with ‘<strong>the</strong>y’ and o<strong>the</strong>r forms is<br />
in Newman (1997), Baron (1986: 191–7), and Curzan (2003: 70–9).<br />
27. Echoing a sweeping statement by Penelope (1990: 94), that ‘in English all persons are<br />
assumed to be male unless o<strong>the</strong>rwise specified’. See §7.6, and also Henley and Abueg<br />
(2003), Henley (1989), Cheshire (1985) and Miller and Swift (1981: 44–60).
210 11 <strong>Gender</strong> in grammar and society<br />
28. In her guidelines to changing sexist language, Pauwels (1998: 127) outlines a number of<br />
alternatives—recasting a sentence in <strong>the</strong> plural, using passive, repeating a generic noun,<br />
replacing he with singular <strong>the</strong>y or he or she, and many fur<strong>the</strong>r variants, such as s/he, she/he,<br />
he/she.Wefind similar guidelines in <strong>the</strong> Handbook of non-sexist writing for writers, editors<br />
and speakers by Miller and Swift (1981).<br />
29. Watch your language: guidelines for non-discriminatory language. University of Melbourne.<br />
30. See Teso and Crolley (2013: 144–5) for details on institutional guidelines. The singular<br />
<strong>the</strong>mself goes back to early Middle English (<strong>the</strong> Oxford English Dictionary Online);<br />
never<strong>the</strong>less, in <strong>the</strong> late 1990s I encountered resistance from some publishers. This echoes<br />
Greenbaum’s(1991: 172) appraisal of <strong>the</strong> singular <strong>the</strong>y (and <strong>the</strong>mself )as‘faulty’. In a later<br />
edition of his English grammar, Greenbaum (1996) predicted that <strong>the</strong> singular <strong>the</strong>y would<br />
become more acceptable. See Awbery, Jones, and Morris (2002: 326–7) for Welsh.<br />
31. See Kochskämper (1999: 464) on masculine overtones of <strong>the</strong> pronoun man. The reader<br />
will notice that nouns in German are written with <strong>the</strong> capital initial letter. Pronouns<br />
(derived from nouns) are written with a small letter.<br />
32. The two forms of <strong>the</strong> feminine pronoun represent two varieties of Norwegian: Nynorsk<br />
ho/Bokmål hun; see Bull and Swan (2002: 244–6).<br />
33. See Hornscheidt (2003: 360–1), Hornscheidt (2006) for attempts to use <strong>the</strong> feminine<br />
pronoun hon as a generic; and Milles (2011: 22–3, 27–8) for a history of reactions to <strong>the</strong><br />
feminist language planning and <strong>the</strong> new generic pronoun.<br />
34. We can recall that gender agreement in German is based on formal assignment of<br />
Linguistic <strong>Gender</strong> in <strong>the</strong> singular; and <strong>the</strong>re is one plural form without any gender<br />
distinctions.<br />
35. Teso and Crolley (2013: 145) lament that ‘<strong>the</strong> female remains invisible in structures with<br />
plural nouns, adjectives and verbs which include both male and female’, in such examples<br />
in French.<br />
36. See Nissen (2002: 257–8) on Spanish. Similar rules apply in Italian (Marcato and Thüne<br />
2002) and Portuguese (Endruschat 2015).<br />
37. See Burr (2003: 128–34) for <strong>the</strong> history of feminization of titles in France; Abbou (2011)on<br />
‘double gender marking’ in <strong>the</strong> written language. See Yaguello (1991: 118–38).<br />
38. The situation in Italian appears to be unstable: see Marcato and Thüne (2002: 206), on<br />
general preference for generic masculine forms.<br />
39. See also Bussmann and Hellinger (2003). See also Vasvári (2015) for Hungarian and<br />
Hasselblatt (2015) for Estonian.<br />
40. The feminine form la soldate ‘female soldier’ (alongside <strong>the</strong> masculine noun le soldat) and<br />
la professeure ‘female professor’ (alongside masculine le professeur) were added to Oxford<br />
Language Dictionaries in 2008. O<strong>the</strong>r dictionaries, including The Collins French Dictionary<br />
and Grammar (2014), contain a negligible number of feminized forms. Lynda Mugglestone<br />
(2011: 74–6) comments on <strong>the</strong> reluctance of dictionaries to embrace innovations, and<br />
include feminized titles. As Cheris Kramarae (1992: 146) put it, dictionaries can be seen<br />
as ‘instruments of social control’ which focus on <strong>the</strong> language of men and male authors<br />
backgrounding women’s voices.<br />
41. Example from Sa'ar (2007: 413). See Sa'ar (2007) on Hebrew and Palestinian Arabic, and<br />
on ata ‘you masculine’ as a generic form translatable with English ‘one’, and Hachimi<br />
(2001) on Arabic.<br />
42. Israeli society does not exclude women from high administrative positions (at least twelve<br />
women have served in Israeli governments since <strong>the</strong> 1970s; Golda Meir has so far been <strong>the</strong><br />
only female Prime Minister (1969–74)).
11.6 Expression of Linguistic <strong>Gender</strong> and social change 211<br />
43. See Pauwels (1998: 58–63), Braun (2000a), and Spender (1980: 27) for <strong>the</strong> criticism of such<br />
practices from a feminist perspective. See Miller and Swift (1981: 128–9) onMs.<br />
44. Duden Dictionary online.<br />
45. Just a handful of peoples have an established tradition of matronyms—or surnames based<br />
on <strong>the</strong> mo<strong>the</strong>r’s name: see Haviland et al. (2014: 130).<br />
46. See Harrison (1990) and Aikhenvald (2008).<br />
47. See Sini (2011: 195–6) on <strong>the</strong> gradual replacement of matrilineally inherited names by<br />
surnames based on fa<strong>the</strong>rs’ names in traditionally matriarchic communities in Malabar<br />
in India.<br />
48. See Maiden and Robustelli (2007: 459); Maiden (1998: 178ff.). The spread of <strong>the</strong>se forms<br />
may have been facilitated by Spanish influence, and copying of Spanish forms like Vuestra<br />
Merced ‘your honour’. Jacqueline Brunet (2003) offers an entertaining description of how<br />
Italian linguists and lay people (including members of <strong>the</strong> National Fascist party in <strong>the</strong><br />
1930s) tried to eradicate <strong>the</strong> feminine pronoun as a formal term of address. In <strong>the</strong>ir view,<br />
this was a lamentable trace of a ‘damned’ Spanish influence. Needless to say, <strong>the</strong>se attempts<br />
have failed. The form Vuestra Merced in Spanish gave rise to <strong>the</strong> polite second person<br />
Usted.
12<br />
The heart of <strong>the</strong> matter: envoi<br />
<strong>Gender</strong> is a wondrous construct, multifaceted in its functions and meanings. It<br />
permeates and shapes <strong>the</strong> world we live in. We can now recapitulate <strong>the</strong> most salient<br />
points discussed so far. The notion of gender has three sides to it.<br />
I. The three faces of <strong>Gender</strong><br />
• First, NATURAL GENDER (or sex) reflects <strong>the</strong> dichotomy between women and men,<br />
or females and males.<br />
• Secondly, SOCIAL GENDER reflects <strong>the</strong> social implications, expectations, and norms of<br />
being a woman or a man. Social <strong>Gender</strong> is a status to be achieved within a society.<br />
• And thirdly, <strong>the</strong> nouns in a language may divide into LINGUISTIC GENDERS. Then<br />
one class will be marked in one way, and ano<strong>the</strong>r class in ano<strong>the</strong>r way. The class<br />
which includes words referring to females will be called ‘feminine’. The one<br />
which includes words for males will be ‘masculine’. Linguistic genders are<br />
defined by <strong>the</strong>ir male and female members, but often extend beyond those.<br />
And this is where <strong>the</strong> trouble begins. Linguistic <strong>Gender</strong>s for objects without<br />
Natural <strong>Gender</strong> can be shrouded in mystery. Linguistic <strong>Gender</strong> of living beings—<br />
animals, birds, insects, and so on—and of natural phenomena may reflect <strong>the</strong>ir role<br />
in legends and myths. In Tunica, <strong>the</strong> sun is believed to be a mythical man, and<br />
belongs to <strong>the</strong> masculine gender. In Dyirbal, it is a mythical woman, and so is<br />
classified as feminine. Every noun in a language will belong to a linguistic gender,<br />
but not every instance will have an explanation. This has earned linguistic genders a<br />
reputation of being a hard nut to crack, among some impatient linguists and hasty<br />
language learners.<br />
We now recapitulate a few home truths about Linguistic <strong>Gender</strong>.<br />
II. Linguistic <strong>Gender</strong> in a nutshell<br />
Linguistic <strong>Gender</strong> is typically expressed on adjectives agreeing with a noun, on<br />
pronouns, and sometimes also on <strong>the</strong> noun itself. We know that <strong>the</strong> noun menina<br />
<strong>How</strong> <strong>Gender</strong> <strong>Shapes</strong> <strong>the</strong> <strong>World</strong>. First edition. Alexandra Y. Aikhenvald.<br />
© Alexandra Y. Aikhenvald 2016. First published in 2016 by Oxford University Press.
12 The heart of <strong>the</strong> matter 213<br />
‘girl’ in Portuguese is feminine because it requires <strong>the</strong> feminine form of an article<br />
(uma ‘a’) and <strong>the</strong> feminine form of an adjective, and also because it ends in ‐a. Every<br />
noun in a language will be assigned a Linguistic <strong>Gender</strong>. There will always be a<br />
meaning to a gender. But not every instance of it will be explainable. Chapters 2, 3,<br />
and 4 focus on what Linguistic <strong>Gender</strong>s mean and what <strong>the</strong>y are good for.<br />
Linguistic <strong>Gender</strong>s can have complex histories. They may develop out of words for<br />
man or woman, or out of o<strong>the</strong>r sources. Children acquire <strong>the</strong>m at an early age. And if<br />
languages are in contact, <strong>the</strong> chances are that <strong>the</strong>ir Linguistic <strong>Gender</strong>s will influence<br />
each o<strong>the</strong>r. Language engineering, and reform, play a role in reshaping Linguistic<br />
<strong>Gender</strong>s. More on this in Chapters 6 and 11, and also in section VI below.<br />
Linguistic <strong>Gender</strong> of women and men is special: it is where <strong>the</strong> three faces of<br />
<strong>Gender</strong> interact. This takes us to <strong>the</strong> next point.<br />
III. <strong>How</strong> <strong>the</strong> three faces of gender relate to each o<strong>the</strong>r<br />
The three faces of <strong>Gender</strong>—Natural, Social, and Linguistic—are inextricably linked.<br />
They are like a chain of elements feeding into each o<strong>the</strong>r: this is what we see in<br />
Scheme 12.1.<br />
Biological men and women acquire <strong>the</strong>ir male and female social status as <strong>the</strong>y<br />
reach maturity. Linguistic <strong>Gender</strong>s reflect biological malehood and femalehood.<br />
They also mirror <strong>the</strong> social status of men and of women. Through Linguistic<br />
<strong>Gender</strong>s, objects can be endowed with features of natural and of social genders.<br />
This is how Linguistic <strong>Gender</strong>s and gender-based metaphors make <strong>the</strong> world come<br />
alive. Heinrich Heine, one of <strong>the</strong> greatest German poets, alluded to a pine-tree (of <strong>the</strong><br />
masculine Linguistic <strong>Gender</strong>) longing for a palm-tree (feminine): <strong>the</strong> Linguistic<br />
<strong>Gender</strong> of plants allowed him to treat trees like human beings (more details and<br />
examples are in §8.2).<br />
Linguistic <strong>Gender</strong>s can be associated with typical shapes—elongated and thin for<br />
males and masculine gender, round for females and feminine gender. Choosing<br />
Linguistic <strong>Gender</strong>s by shape is a way to go for objects which do not have ‘sex’. But<br />
classifying humans by shape can be offensive. Among <strong>the</strong> Manambu of New Guinea,<br />
Natural <strong>Gender</strong><br />
Social <strong>Gender</strong> for humans<br />
Linguistic <strong>Gender</strong> of humans<br />
Social <strong>Gender</strong> features projected onto non-humans<br />
SCHEME 12.1 <strong>How</strong> <strong>the</strong> three faces of <strong>Gender</strong> relate to each o<strong>the</strong>r
214 12 The heart of <strong>the</strong> matter<br />
talking about a man as if he were ‘round’ and feminine means downgrading him to<br />
<strong>the</strong> status of a mere ‘thing’.<br />
Linguistic <strong>Gender</strong>s interact with social values and may reflect <strong>the</strong>m. In many<br />
languages from New Guinea, highly valued objects are masculine. Objects of lesser<br />
value are feminine. Linguistic <strong>Gender</strong>s mirror <strong>the</strong> values of <strong>the</strong> societies where men,<br />
and male cults, are most appreciated. Referring to men as if <strong>the</strong>y were women, and to<br />
women as if <strong>the</strong>y were men, is incongruous. A ‘feminine’ man fails to conform to<br />
Social <strong>Gender</strong> expectations. A ‘masculine woman’ may be too big for her boots.<br />
Alternatively, in some languages talking about a woman as if she were a man implies<br />
promoting her to proper human status. Among <strong>the</strong> Tariana of north-western Brazil,<br />
an important woman is a ‘he’. This is akin to <strong>the</strong> English expression ‘man up!’, in <strong>the</strong><br />
meaning of ‘get your act toge<strong>the</strong>r’. A man (and not a woman) is an epitome of<br />
positive features. The Social <strong>Gender</strong> finds its direct match in Linguistic <strong>Gender</strong>. More<br />
on this in Chapter 7.<br />
Aspects of language o<strong>the</strong>r than Linguistic <strong>Gender</strong> set men and women apart. This<br />
brings us to our fourth point.<br />
IV. Social <strong>Gender</strong>s in speech practices<br />
In a number of societies, men and women have to use different words. A Karajá<br />
woman, from Brazil, will say bicikreta, for ‘bicycle’ (a borrowing from Portuguese<br />
bicicleta). A man will say bicileta, for <strong>the</strong> same thing: men’s speech lacks <strong>the</strong> sound k.<br />
Special forms for men and women (known as genderlects) define Social <strong>Gender</strong>s:<br />
mature men are expected to use <strong>the</strong> male language as a feature of <strong>the</strong>ir social<br />
malehood. The same applies to women, and female language. The trouble is that<br />
fewer and fewer people across <strong>the</strong> world continue speaking minority languages with<br />
fascinating gender differences. The systems become endangered, and harder and<br />
harder to understand. Japanese and Thai are examples of larger communities with<br />
pretty well-defined male and female speech. A Japanese-speaking woman will use a<br />
final particle wa, and a man will not. Recent social changes and growing equality of<br />
men and women have affected genderlects. As Japanese and Thai-speaking women<br />
get access to leading positions in <strong>the</strong> society, asymmetries in speech practices become<br />
less obvious. More on this in §9.1.<br />
In most o<strong>the</strong>r languages, men and women may (but don’t have to) speak somewhat<br />
differently. Speakers have an option of making <strong>the</strong>ir language sound more<br />
male-like, or more female-like, in agreement with <strong>the</strong> stereotypes one has in mind,<br />
and <strong>the</strong> image one wishes to project. So, in Brazilian Portuguese and in Moroccan<br />
Arabic women use many more diminutives than men. Female speech is felt to be<br />
more polite. These are tendencies, and not steadfast rules. The fascinating thing is<br />
that <strong>the</strong>se trends turn into stereotypes of Social <strong>Gender</strong>s. They also play out in<br />
projecting and defining <strong>the</strong> identity of genders o<strong>the</strong>r than plain male or plain female.
12 The heart of <strong>the</strong> matter 215<br />
Gays, lesbians, and transsexuals project <strong>the</strong> image of men, women, or something inbetween,<br />
using <strong>the</strong> gender-variable language resources. More on this in §§9.2–3.<br />
Differences between men and women lie not just in how <strong>the</strong>y speak, but in what<br />
<strong>the</strong>y talk about. Some words, concepts and genres may be plainly off limits to one<br />
Social <strong>Gender</strong>. This takes us to our next point.<br />
V. Social <strong>Gender</strong>s, speech genres, and taboos<br />
In many traditional societies, <strong>the</strong> language reflects a division of labour between Social<br />
<strong>Gender</strong>s. Special male registers and rituals are off limits to women. Men have traditional<br />
secret languages which women are not allowed to know. In <strong>the</strong> Tariana society<br />
of north-western Brazil, women were not supposed to see <strong>the</strong> magic flutes—a woman<br />
who might catch a glimpse of those by pure chance would have to die. Nor was a<br />
woman allowed to hear, let alone use, <strong>the</strong> term piri for magic flute. More on <strong>the</strong>se in<br />
Chapter 10.<br />
Interestingly enough, no women-only registers which would be forbidden to men<br />
have so far been documented. Is it indeed <strong>the</strong> case, as Anne Storch (2011: 84) phrased<br />
it, that male secret languages are strategies ‘by socially or spiritually inferior groups to<br />
gain or maintain power <strong>the</strong>y would not o<strong>the</strong>rwise have’, and all <strong>the</strong>y reflect is fear and<br />
need for protection against women—a potentially powerful competitor?<br />
And this brings up our next point: how can <strong>the</strong> social status of women be seen<br />
through <strong>the</strong> prism of Linguistic <strong>Gender</strong> and ways of saying things? And how does <strong>the</strong><br />
Linguistic <strong>Gender</strong> change as <strong>the</strong> societies transform?<br />
VI. Women through <strong>the</strong> prism of language<br />
In terms of numbers, women are hardly a minority. Yet <strong>the</strong>y have been backgrounded,<br />
downplayed, relegated to <strong>the</strong> domestic sphere, and generally treated as<br />
‘second class citizens’ throughout <strong>the</strong> history of many societies and cultures. Linguistic<br />
<strong>Gender</strong>s clearly reflect <strong>the</strong> historical inequality between men and women, and<br />
<strong>the</strong> asymmetry between Social <strong>Gender</strong>s. In many languages, <strong>the</strong> masculine gender is<br />
used to talk about ‘valuable’ and important objects. A woman can be ‘promoted’ to<br />
social manhood and appreciatively spoken of as ‘aman’. The formidable Golda Meir,<br />
during her time as Israeli Prime Minister, was praised as ha-gever ha-yexid bamemshala<br />
(<strong>the</strong>-man <strong>the</strong>-only in.<strong>the</strong>-government) ‘<strong>the</strong> only man in <strong>the</strong> government’.<br />
In many languages <strong>the</strong> word ‘man’ has overtones of courage and bravery. More on<br />
this in §§11.1.–3.<br />
In contrast, women can be treated as a substandard and somewhat suspicious<br />
‘o<strong>the</strong>r’. Tariana women are dangerous outsiders—<strong>the</strong>y are to blame for many things,<br />
including <strong>the</strong> demise of <strong>the</strong> Tariana language. So are Corsican women, as was mentioned<br />
in §10.5. The historical asymmetry in male and female Social <strong>Gender</strong>s—and
216 12 The heart of <strong>the</strong> matter<br />
suspicious attitude against women—find <strong>the</strong>ir expression in proverbs across <strong>the</strong> world.<br />
Never marry a woman with big feet, says <strong>the</strong> title of Mineke Schipper’s(2003) study of<br />
proverbs all over <strong>the</strong> world: a woman with big feet may try and overrule <strong>the</strong> dominant<br />
man. An Arabic proverb from Lebanon reminds us that ‘a man, even a man of small<br />
size, will be called great in comparison to women’ (Schipper 2003: 20). ‘Where <strong>the</strong>re are<br />
women, <strong>the</strong>re is no peace’, says an Estonian proverb.<br />
Derogatory and demeaning overtones of female terms and functions tell a similar<br />
story. The feminine form spinster in English has unwanted connotations, as was<br />
discussed in §7.5. Its masculine counterpart, bachelor, does not. For centuries and in<br />
many languages, women were subsumed under ‘man’ as a general word for ‘human<br />
being’. Some dictionaries defined woman as ‘subtype of man’. The traditional practice<br />
in English was to use ‘man’ as a general term for human and ‘he’ as a generic pronoun,<br />
to refer to people of ei<strong>the</strong>r sex (though o<strong>the</strong>r options have been available since Old<br />
English times). With <strong>the</strong> rise of feminism and general rise in <strong>the</strong> status of women in<br />
Western societies, <strong>the</strong> question of <strong>the</strong> inherent sexism of generic masculine came to a<br />
head. It became <strong>the</strong> target for linguistic reforms. The political importance of Linguistic<br />
<strong>Gender</strong> was brought to <strong>the</strong> forefront of <strong>the</strong> antisexist struggle. Incidentally, what better<br />
way of proving that Linguistic <strong>Gender</strong> is central to language use, and not an ‘irrational’<br />
excrescence? Linguistic <strong>Gender</strong> has come to be seen by feminists as a way of redressing<br />
<strong>the</strong> balance between Social <strong>Gender</strong>s.<br />
Different languages followed different paths, depending on what resources are<br />
available, and how <strong>the</strong> systems work. The main issue in English has been <strong>the</strong> avoidance<br />
of generic masculine pronoun he and terms involving ‘man’. And a victory has been<br />
achieved: it is no longer advisable to use ‘he’ as a generic. We can recall Jenny<br />
Cheshire’s (2008: 9–10) encouraging report, that none of her colleagues would now<br />
‘dare to use a generic he pronoun’. Addressing a woman chair as Madam Chairman<br />
was fine in 1971 (see Goddard et al. 1971). Nowadays, we will say Madam Chair.<br />
German has gone a different way. Feminists and language reformers targeted<br />
generic forms which involve ‘man’ or masculine forms. Thanks to <strong>the</strong>ir efforts,<br />
feminine and masculine derivational forms are now mostly used in a parallel fashion:<br />
one talks and writes about Kollegen and Kolleginnen (male colleagues and female<br />
colleagues), Studenten and Studentinnen (male students and female students), ra<strong>the</strong>r<br />
than subsuming women under <strong>the</strong> masculine Kollegen or Studenten. Along similar<br />
lines, French feminists and reformers have concentrated on equality in derivations.<br />
Since 1998, using feminine forms for professions and titles has become officially<br />
recognized in France (though it is not yet followed by everyone). Incidentally, <strong>the</strong><br />
feminine forms of many professions and titles in English often have slighting or even<br />
pejorative overtones: a manageress is not equal to a manager, and an authoress is seen<br />
as not quite as prestigious as an author. (O<strong>the</strong>r nouns like colleague or student simply<br />
have no female form.) In all likelihood, this is a reason why English-speaking<br />
feminists never focused on equality in derivations (a point aptly made by Anna
12 The heart of <strong>the</strong> matter 217<br />
Livia in her 2001 book Pronoun envy). The choice of gendered pronouns in German<br />
and in French is determined by <strong>the</strong> Linguistic <strong>Gender</strong> of <strong>the</strong> noun, and is hard to<br />
manipulate. There have been attempts to get rid of <strong>the</strong> German generic pronoun man<br />
‘one, person’ because of its uncanny similarity to <strong>the</strong> masculine Mann ‘man’; but <strong>the</strong><br />
success of this remains to be seen. The strife to redress <strong>the</strong> imbalance in ‘sexist<br />
language’ by feminists across <strong>the</strong> Western world is a living proof of <strong>the</strong> centrality of<br />
Linguistic and Social <strong>Gender</strong> working toge<strong>the</strong>r, with <strong>the</strong>ir uniting and divisive<br />
powers. This is what we saw in §11.5.<br />
The criticism of Linguistic <strong>Gender</strong>s comes from o<strong>the</strong>r quarters. Some feminists<br />
complain that having an obligatory Linguistic <strong>Gender</strong>, and having to always make a<br />
choice between ‘he’ and ‘she’ and masculine and feminine forms is an obstacle for<br />
Social <strong>Gender</strong> equality: <strong>the</strong>re is no way one can escape <strong>the</strong> ‘prison-house’ of <strong>Gender</strong>.<br />
One thing to remember is that <strong>the</strong> presence of a linguistic category hardly makes a<br />
society more or less sexist. In no way does having no Linguistic <strong>Gender</strong> in a language<br />
correlate with gender equality or gender neutrality. Turkish, a language with no<br />
Linguistic <strong>Gender</strong>, does not offer an equal treatment of men and women. Nei<strong>the</strong>r<br />
does Hungarian, nor Estonian. The demeaning stereotypes associated with Social<br />
<strong>Gender</strong>, and <strong>the</strong> male bias, will be reflected in any language—with or without<br />
Linguistic <strong>Gender</strong>—as long as <strong>the</strong>y are relevant for <strong>the</strong> society which speaks it. The<br />
lack of Linguistic <strong>Gender</strong> does not address <strong>the</strong> problem in <strong>the</strong> slightest.<br />
Where to from now? The three faces of <strong>Gender</strong> are <strong>the</strong> life-blood of human<br />
interaction and history. Societies evolve, and language change follows suit. We expect<br />
<strong>the</strong> growing equality of women to find more and more reflection in <strong>the</strong> languages of<br />
<strong>the</strong> world. The linguistic expression of genders o<strong>the</strong>r than plain male or plain female<br />
is ano<strong>the</strong>r fascinating issue to fur<strong>the</strong>r explore, as new studies come to light. We hope<br />
that new developments, and new descriptive studies, will help unravel <strong>the</strong> intricacies<br />
of human cognition and social life.
References<br />
Aalto, Pentti. 1959. ‘Über die kalmückische Frauensprache’. Studia Mongolica 1: 1–8.<br />
Abbink, Jon. 2015. ‘Menstrual synchrony claims among Suri girls (Southwest Ethiopia):<br />
between gender and biology’. Cahiers d’études africaines (2), 218: 279–302.<br />
Abbott, Clifford. 1984. ‘Two feminine genders in Oneida’. Anthropological Linguistics 26:<br />
125–37.<br />
Abbou, Julie. 2011. ‘Double gender marking in French: a linguistic practice of antisexism’.<br />
Current Issues in Language Planning 12: 55–75.<br />
Abondolo, Daniel. 1998. ‘Finnish’, pp. 149–83 of The Uralic languages, edited by Daniel<br />
Abondolo. London and New York: Routledge.<br />
Adams, K. L. 1989. Systems of numeral classification in <strong>the</strong> Mon-Khmer, Nicobarese and Aslian<br />
subfamilies of Austroasiatic. Canberra: Pacific Linguistics.<br />
Adelaar, Willem F. H. 2006. ‘The Quechua impact in Amuesha’, pp. 290–311 of Grammars in<br />
contact: a crosslinguistic typology. Latrobe, Australia: Research Centre for Linguistic Typology,<br />
edited by Alexandra Y. Aikhenvald and R. M. W. Dixon. Oxford: Oxford University Press.<br />
Ahland, Colleen Anne. 2012. ‘A grammar of nor<strong>the</strong>rn and sou<strong>the</strong>rn Gumuz’. PhD dissertation,<br />
University of Oregon.<br />
Aikhenvald, Alexandra Y. 2000. Classifiers: a typology of noun categorization devices. Oxford:<br />
Oxford University Press.<br />
Aikhenvald, Alexandra Y. 2002. Language contact in Amazonia. Oxford: Oxford University Press.<br />
Aikhenvald, Alexandra Y. 2003. A grammar of Tariana, from north-west Amazonia. Cambridge:<br />
Cambridge University Press.<br />
Aikhenvald, Alexandra Y. 2004. ‘<strong>Gender</strong>’, Article 98, pp. 1031–45 of Morfologie/Morphology.<br />
Ein Handbuch zur Flexion und Wortbildung/A handbook on inflection and word formation,<br />
2 Halbband, edited by Geert E. Booij, Christian Lehmann, Joachim Mugdan, and Stavros<br />
Skopetas. Berlin: Mouton de Gruyter.<br />
Aikhenvald, Alexandra Y. 2006. ‘Classifiers and noun classes, semantics’, pp. 463–70 of<br />
Volume 1 (article 1111) ofEncyclopedia of language and linguistics, 2nd edition, edited by<br />
Keith Brown. Oxford: Elsevier.<br />
Aikhenvald, Alexandra Y. 2007. ‘Classifiers in multiple environments: Baniwa of Içana/Kurripako:<br />
a North Arawak perspective’. International Journal of American Linguistics 27:<br />
475–500.<br />
Aikhenvald, Alexandra Y. 2008. The Manambu language, from East Sepik, Papua New Guinea.<br />
Oxford: Oxford University Press.<br />
Aikhenvald, Alexandra Y. 2009. ‘Language contact along <strong>the</strong> Sepik River’. Anthropological<br />
Linguistics 50: 1–66 (dated 2008, published 2009).<br />
Aikhenvald, Alexandra Y. 2010. Imperatives and commands. Oxford: Oxford University Press.<br />
Aikhenvald, Alexandra Y. 2012a. The languages of <strong>the</strong> Amazon. Oxford: Oxford University<br />
Press.
References 219<br />
Aikhenvald, Alexandra Y. 2012b. ‘Round women and long men: shape and size in gender<br />
choice in Papua New Guinea and beyond’. Anthropological Linguistics 54 (1): 33–86.<br />
Aikhenvald, Alexandra Y. 2013a. ‘The language of value and <strong>the</strong> value of language’. Hau: a<br />
journal of ethnographic <strong>the</strong>ory 3.2: 55–73.<br />
Aikhenvald, Alexandra Y. 2013b. ‘Shifting language attitudes in north-west Amazonia’. International<br />
Journal of <strong>the</strong> Sociology of Language 222: 195–216.<br />
Aikhenvald, Alexandra Y. 2013c. ‘Possession and ownership: a typological perspective’,<br />
pp. 1–64 of Aikhenvald and Dixon 2013.<br />
Aikhenvald, Alexandra Y. 2014a. The art of grammar. Oxford: Oxford University Press.<br />
Aikhenvald, Alexandra Y. 2014b. ‘Double talk: parallel structures in Manambu songs, and <strong>the</strong>ir<br />
origin’. Language and Linguistics in Melanesia 32.2: 86–109.<br />
Aikhenvald, Alexandra Y. 2014c. ‘The grammar of knowledge in typological perspective’,<br />
pp. 1–51 of The grammar of knowledge, edited by A. Y. Aikhenvald and R. M. W. Dixon.<br />
Oxford: Oxford University Press.<br />
Aikhenvald, Alexandra Y. 2015a. ‘Body, mind and spirit: what makes up a person in Manambu’.<br />
Studies in Language 39: 85–117.<br />
Aikhenvald, Alexandra Y. 2015b. Oxford Bibliography Online: Classifiers (refereed updatable<br />
resource with summaries and evaluation for each entry). General Editor: Mark Aronoff.<br />
Aikhenvald, Alexandra Y. Forthcoming. ‘A typology of noun categorization devices’, to appear<br />
in The Cambridge handbook of linguistic typology, edited by Alexandra Y. Aikhenvald and<br />
R. M. W. Dixon. Cambridge: Cambridge University Press.<br />
Aikhenvald, Alexandra Y. and R. M. W. Dixon. 2011a. Language at large: essays in syntax and<br />
semantics. Empirical approaches to linguistic <strong>the</strong>ory, 2. Leiden: Brill.<br />
Aikhenvald, Alexandra Y. and R. M. W. Dixon. 2011b. ‘Dependencies between grammatical<br />
systems’, pp. 170–204 of Language at large: essays on syntax and semantics, by Alexandra<br />
Y. Aikhenvald and R. M. W. Dixon. Leiden: Brill.<br />
Aikhenvald, Alexandra Y. and R. M. W. Dixon (eds.). 2013. Possession and ownership. Oxford:<br />
Oxford University Press.<br />
Aikhenvald, Alexandra Y. and Diana Green. 1998. ‘Palikur and <strong>the</strong> typology of classifiers’.<br />
Anthropological Linguistics 40: 429–80.<br />
Aikhenvald, Alexandra Y. and Diana Green. 2011. ‘Palikur and <strong>the</strong> typology of classifiers’.<br />
pp. 394–450 of Language at large: essays on syntax and semantics, by Alexandra Y. Aikhenvald<br />
and R. M. W. Dixon. Leiden: Brill.<br />
Aikio, Marjut. 1992. ‘Are women innovators in <strong>the</strong> shift to a second language? A case study of<br />
Reindeer Sámi women and men’. International Journal of <strong>the</strong> Sociology of Language 94: 43–61.<br />
Alexeyev, M. 1985. Voprosy sravniteljno-istoricheskoj grammatiki lezginskih jazykov. Morfologija.<br />
Sintaksis. (Problems of comparative and historical grammar of Lezghic language.<br />
Morphology. Syntax.) Moscow: Nauka.<br />
Allan, Keith and Kate Burridge. 2006. Forbidden words: taboo and <strong>the</strong> censoring of language.<br />
Cambridge: Cambridge University Press.<br />
Alpher, Barry. 1987. ‘Feminine as <strong>the</strong> unmarked gender: Buffalo girls are no fools’. Australian<br />
Journal of Linguistics 7: 169–87.<br />
Alvanoudi, Angeliki. 2014. Grammatical gender in interaction: cultural and cognitive aspects.<br />
Leiden: Brill.
220 References<br />
Alvanoudi, Angeliki. 2015. ‘The routine achievement of gender in Greek conversation’. <strong>Gender</strong><br />
and Language 9: 11–31.<br />
Amadiume, Ifi. 2015. Male daughters and female husbands: gender and sex in an African<br />
society. London: Zed books.<br />
Ameka, Felix K. 2012. Ewe: its grammatical constructions and illocutionary devices. Munich:<br />
Lincom Europa.<br />
Amha, Azeb. 2001. The Maale language. Amsterdam: Research School of Asian, African, and<br />
Amerindian Studies, Universiteit Leiden.<br />
Amha, Azeb. 2012. ‘Omotic’, pp. 423–504 of Frajzyngier and Shay 2012.<br />
Amha, Azeb. 2013. ‘Male drums and female drums: natural gender and inanimate nouns in<br />
Omotic languages’. Paper presented at <strong>the</strong> International Workshop on <strong>Gender</strong>. LCRC, JCU,<br />
November 2013.<br />
Amha, Azeb. Forthcoming. ‘Omotic languages’, to appear in The Cambridge handbook of<br />
linguistic typology, edited by Alexandra Y. Aikhenvald and R. M. W. Dixon. Cambridge:<br />
Cambridge University Press.<br />
Anderson, L.-G. and T. Janson. 1997. Languages in Botswana: language ecology in Sou<strong>the</strong>rn<br />
Africa. Botswana: Longman.<br />
Armon-Lotem, Sharon and Orit Amiram. 2014. ‘The assignment of gender in L2 Hebrew: <strong>the</strong><br />
role of <strong>the</strong> L1 gender system’, pp. 230–49 of The acquisition of Hebrew phonology and<br />
morphology, edited by Outi Bar-El. Leiden: Brill.<br />
Asher, R. E. 1985. Tamil. London: Croom Helm.<br />
Atanga, Lilian Lem, Sibonile Edith Ellece, Lia Litosseliti, and Jane Sunderland. 2013. <strong>Gender</strong> and<br />
language in Sub-Saharan Africa: tradition, struggle and change. Amsterdam: John Benjamins.<br />
Attaviriyanupap, Korakoch. 2015. ‘The linguistic representation of gender in Thai’, pp. 369–400<br />
of Hellinger and Motschenbacher 2015.<br />
Audring, Jenny. 2013. ‘A pronominal view of gender agreement’. Language Sciences 35: 32–46.<br />
Avery, Jack and Julie Liss. 1996. ‘Acoustic characteristics of less-masculine-sounding male<br />
speech’. Journal of <strong>the</strong> Acoustic Society of America 99: 3738–48.<br />
Awbery, Gwenllian, Kathryn Jones, and Delyth Morris. 2002. ‘The politics of language and<br />
gender in Wales’, pp. 313–30 of Hellinger and Bussmann 2002, Volume 2.<br />
Axenov, Serge. 2006. The Balochi language of Turkmenistan: a corpus-based grammatical<br />
description. Uppsala: Department of Linguistic and Philology.<br />
Baker, Paul. 2014. ‘Two hundred years of American man’, pp. 34–52 of Milani 2014.<br />
Bani, Ephraim. 1987. ‘Masculine and feminine grammatical gender in Kala Lagaw Ya’.<br />
Australian Journal of Linguistics 7: 189–201.<br />
Barnhart, Robert K. 1988. Chambers dictionary of etymology. Edinburgh: Chambers.<br />
Baron, Dennis. 1981. ‘The epicene pronoun: <strong>the</strong> word that failed’. American Speech 56: 83–97.<br />
Baron, Dennis. 1986. Language and gender. New Haven: Yale University Press.<br />
Baron, Naomi S. 1971. ‘A reanalysis of English grammatical gender’. Lingua 27: 113–40.<br />
Barrett, R. 2006. ‘Queer talk’, pp. 316–23 of Encyclopedia of languages and linguistics, edited by<br />
Keith Brown. Oxford: Elsevier.<br />
Bateson, Gregory. 1958. Naven. 2nd edition. Stanford, Calif.: Stanford University Press.<br />
Bauman, R. and J. Sherzer (eds.). 1974. Explorations in <strong>the</strong> ethnography of speaking. Cambridge:<br />
Cambridge University Press.
References 221<br />
Beachy, Marvin Dean. 2005. ‘An overview of Central Dizin phonology and morphology’.<br />
Masters <strong>the</strong>sis in linguistics, University of Texas at Arlington.<br />
Beauvoir, Simone de. 1949. Le deuxième sexe. Paris: Gallimard.<br />
Becker, A. L. 1975. ‘A linguistic image of nature: <strong>the</strong> Burmese numerative classifier system’.<br />
Linguistics 165: 109–21.<br />
Bell, D. 1983. Daughters of <strong>the</strong> dreaming. Melbourne: McPhee Gribble.<br />
Benamara, Hassane. 2013. Dictionnaire amazighe–français. Parler de Figuig et ses régions.<br />
Rabat: IRCAM.<br />
Bengoechea, Mercedes. 2011. ‘Non-sexist Spanish policies: an attempt bound to fail?’ Current<br />
Issues in Language Planning 12: 35–53.<br />
Bergen, John J. 1980. ‘The semantics of gender contrasts in Spanish’. Hispania 63: 48–57.<br />
Berman, Ruth A. 1981. ‘Regularity vs anomaly: <strong>the</strong> acquisition of Hebrew inflectional morphology’.<br />
Journal of Child Language 8: 265–82.<br />
Berman, Ruth A. 1985. ‘The acquisition of Hebrew’, pp. 255–71 of The crosslinguistic study of<br />
language acquisition. Hillsdale, NJ: Lawrence Erlbaum.<br />
Berman, Ruth A. and Dan I. Slobin. 1994. Relating events in narrative: a crosslinguistic<br />
developmental study. Hillsdale, NJ: Lawrence Erlbaum Associates.<br />
Besnier, Niko. 2003. ‘Crossing genders, mixing languages: <strong>the</strong> linguistic construction of<br />
transgenderism in Tonga’, pp. 279–301 of Holmes and Meyerhoff 2003.<br />
Bhatia, Tej K. 1993. Punjabi. London: Routledge.<br />
Birtalan, Ágnes. 2003. ‘Oirat’, pp. 210–28 of Mongolic languages, edited by Juha Janhunen.<br />
London: Routledge.<br />
Bisang, W. 1993. ‘Classifiers, quantifiersandclassnounsinHmong’. Studies in Language 17: 1–51.<br />
Bloomfield, Leonard. 1933. Language. New York: Holt, Rinehart and Winston.<br />
Bloomfield, Morton W. 1963. ‘A grammatical approach to personification allegory’. Modern<br />
Philology 60: 161–71.<br />
Boas, Franz. 1911. ‘Introduction’. Handbook of American Indian languages 1: 5–83.<br />
Boas, Franz. 1938. ‘Language’, pp. 124–45 of General anthropology, edited by Franz Boas.<br />
Boston and New York: D. C. Heath and Company.<br />
Bodine, Ann. 1975. ‘Sex differentiation in language’, pp. 130–51 of Language and sex: difference<br />
and domination, edited by Barrie Thorne and N. Henley. Rowley, Mass.: Newbury House.<br />
Bogoras, Waldemar. 1901. ‘The Chukchi of Nor<strong>the</strong>astern Asia’. American Anthropologist 3:<br />
80–108.<br />
Bogoras, Waldemar. 1922. ‘Chukchee’, pp. 631–903 of Handbook of American Indian languages,<br />
Volume 2, edited by Franz Boas. Washington, DC: Government Printing Office.<br />
Bolinger, Dwight. 1991. ‘Reference and inference: inceptiveness in <strong>the</strong> Spanish preterit’,<br />
pp. 319–34 of his Essays on Spanish: words and grammar. Newark, Del.: Juan de la Cuesta.<br />
Bonfante, G. 1946. ‘Semantics, language’, pp. 847–51 of Encyclopedia of psychology, edited by<br />
P. L. Harriman. New York: ADD.<br />
Bonvillain, Nancy. 1973. A grammar of Akwesasne Mohawk. Mercury Series, Ethnology<br />
Division, Paper 8. Ottawa: National Museum of Man.<br />
Borba, Rodrigo and Ana Christina Ostermann. 2007. ‘Do bodies matter? Travestis’ embodiment<br />
of (trans)gender identity through <strong>the</strong> manipulation of <strong>the</strong> Brazilian Portuguese<br />
grammatical gender system’. <strong>Gender</strong> and Language 1: 131–47.
222 References<br />
Boroditsky, Lera and Lauren A. Schmidt 2000. ‘Sex, syntax, and semantics’. Proceedings of <strong>the</strong><br />
Cognitive Science Society 22: 42–7.<br />
Boroditsky, Lera, Lauren A. Schmidt, and Webb Phillips. 2003. ‘Sex, syntax and semantics’,<br />
pp. 61–80 of Language in mind: advances in <strong>the</strong> study of language and cognition, edited by<br />
Dedre Gentner and Susan Goldin-Meadow. Cambridge, Mass.: MIT Press.<br />
Bossong, Georg. 1985. Empirische Universalienforschung. Differentielle Objekstmarkierung in<br />
den neuiranischen Sprachen. Tübingen: Gunter Narr.<br />
Bossong, Georg. 1991. ‘Differential object marking in Romance and beyond’, pp. 143–70 of<br />
New analyses in Romance linguistics, edited by D. Wanner and D. Kibbee. Amsterdam: John<br />
Benjamins.<br />
Bowden, Ross. 1997. A dictionary of Kwoma, a Papuan language of north-east New Guinea.<br />
Canberra: Pacific Linguistics.<br />
Bradac, J. J., A. Mulac, and S. A. Thompson. 1995. ‘Men’s and women’s use of intensifiers and<br />
hedges in problem-solving interaction: molar and molecular analyses’. Research on Language<br />
and Social Interaction 28.2: 93–116.<br />
Bradley, John. 1988. ‘Yanyuwa: “men speak one way, women speak ano<strong>the</strong>r”’. Aboriginal<br />
Linguistics 1: 126–34.<br />
Brauer, Marcus and Michaël Landry. 2008. ‘Un ministre peut-il tomber enceinte? L’impact du<br />
générique masculin sur les représentations mentales’. Année psychologique 108.2: 243–72.<br />
Braun, Friederike. 1997a. ‘Genuslose Sprache—egalitäre Sprache? Die Behandlung der Geschlechter<br />
in Sprachen mit und ohne Genus’, pp. 39–49 of Jahrbuch für finnisch-deutsche<br />
Literaturbeziehungen 29, edited by Hans Fromm et al. Helsinki: Deutsche Bibliotek.<br />
Braun, Friederike. 1997b. ‘Making men out of people: <strong>the</strong> MAN principle in translating<br />
genderless forms’, pp. 3–30 of Communicating gender in context, edited by H. Kotthof and<br />
R. Wodak. Amsterdam: John Benjamins.<br />
Braun, Friederike. 2000a. Geschlecht in Türkischen. Untersuchungen zum sprachlichen Umgang<br />
mit einer sozialen Kategorie. Wiesbaden: Harrassowitz Verlag.<br />
Braun, Friederike. 2000b. Mehr Frauen in die Sprache. Leitfaden zur geschlechtergerechten<br />
Formulierung. Kiel: Ministerium für Justiz, Frauen, Jugend und Familie des Landes Schleswig-<br />
Holstein.<br />
Braun, Friederike. 2001. ‘The communication of gender in Turkish’, pp. 283–310 of Hellinger<br />
and Bussmann 2001, Volume I.<br />
Braun, Friederike and Geoffrey Haig. 2010. ‘Where are German “girls” feminine? <strong>How</strong> <strong>the</strong><br />
semantics of age influences <strong>the</strong> grammar of gender agreement’, pp. 68–84 of Language in its<br />
socio-cultural context: new explorations in gendered, global and media uses, edited by Markus<br />
Bieswanger, Heiko Motschenbacher, and Susanne Mühleisen. Frankfurt, M. u.a.: Lang.<br />
Braun, Friederike, Sabine Sczesny, and Dagmar Stahlberg. 2005. ‘Cognitive effects of masculine<br />
generics in German: an overview of empirical findings’. Communications 30: 1–21.<br />
Breen, G. 1976a. ‘Wangkumara’, pp. 336–9 of Grammatical categories in Australian languages,<br />
edited by R. M. W. Dixon. Canberra: Australian Institute of Aboriginal Studies.<br />
Breen, G. 1976b. ‘Wagaya’, pp. 340–2 and 590–4 of Grammatical categories in Australian<br />
languages, edited by R. M. W. Dixon. Canberra: Australian Institute of Aboriginal Studies.<br />
Breton, Raymond. 1978. Relations de l’isle de la Guadeloupe, Volume I. Basse-Terre: Société<br />
d’histoire de la Guadeloupe. (Bibliothèque d’histoire antillaise 3.)
References 223<br />
Briggs, C. L. 1992. ‘“Since I am a woman, I will chastise my relatives”: gender, reported speech<br />
and <strong>the</strong> (re)production of social relations in Warao ritual wailing’. American Ethnologist 19:<br />
337–61.<br />
Brittain, David. 2002. ‘Space and spatial diffusion’, pp. 603–37 of The handbook of variation and<br />
change, edited by J. K. Chambers, Peter Trudgill, and Natalie Schilling-Estes. London: Blackwell.<br />
Brosman, P. W. Jr. 1979. ‘The semantics of Hittite gender system’. Journal of Indo-European<br />
Studies 7: 227–36.<br />
Brown, Penelope. 1980. ‘<strong>How</strong> and why are women more polite: some evidence from a Mayan<br />
community’, pp. 111–36 of Women and language in literature and society, edited by<br />
S. McConnell-Ginet, R. Borker, and N. Furman. New York: Praeger.<br />
Brown, Penelope. 1993. ‘<strong>Gender</strong>, politeness and confrontation in Tenejapa’,pp.144–61 of <strong>Gender</strong><br />
and conversational interaction, edited by Deborah Tannen. New York: Oxford University Press.<br />
Brown, Penelope and S. C. Levinson. 1987. Politeness: some universals in language usage.<br />
Cambridge: Cambridge University Press.<br />
Brown, R. 1981. ‘Semantic aspects of some Waris predications’, pp. 93–123 of Syntax and<br />
semantics in Papua New Guinea languages, edited by K. J. Franklin. Ukarumpa: Summer<br />
Institute of Linguistics.<br />
Bruce, Les. 1984. ‘The Alamblak language of Papua New Guinea, East Sepik’. Canberra: Pacific<br />
Linguistics.<br />
Brunet, Jacqueline. 2003. ‘La troisième personne de politesse en italien: fait de langage, fait de<br />
culture’. Franco-British Studies 33/34: 22–34.<br />
Bucholtz, Mary and Kira Hall (eds.). 1995a. <strong>Gender</strong> articulated: language and <strong>the</strong> socially<br />
constructed self. New York: Routledge.<br />
Bucholtz, Mary and Kira Hall. 1995b. ‘Introduction: twenty years after Language and woman’s<br />
place’, pp. 1–24 of Bucholtz and Hall 1995a.<br />
Bucholtz, Mary, Anita C. Liang, and Laurel A. Sutton (eds.). 1999. Reinventing identities: <strong>the</strong><br />
gendered self in discourse. New York: Oxford University Press.<br />
Bull, Tove and Toril Swan. 2002. ‘The representation of gender in Norwegian’, pp. 219–50 of<br />
Hellinger and Bussmann 2002, Volume 2.<br />
Bullock, Alan, Oliver Stallybrass, and Stephen Trombley. 1977. The Fontana dictionary of<br />
modern thought. London: Fontana/Collins.<br />
Bulygina, T. V. and A. D. Shmelev. 1996. ‘Nespecifirovannyj pol i soglasovanie pri anafore’<br />
(‘Unspecified gender and anaphoric agreement’). Moscow Linguistic Journal 2: 98–103.<br />
Burr, Elisabeth. 2003. ‘<strong>Gender</strong> and language politics in France’, pp. 119–40 of Hellinger and<br />
Bussmann 2003, Volume 3.<br />
Bussmann, Hadumod and Marlis Hellinger. 2003. ‘Engendering female visibility in German’,<br />
pp. 141–74 of Hellinger and Bussmann 2003, Volume 3.<br />
Butt, John and Carmen Benjamin. 2004. A new reference grammar of Modern Spanish. 4th<br />
edition. London: Hodder Arnold.<br />
Caldas-Coulthard, Carmen Rosa and Rosamund Moon. 2010. ‘“Curvy, hunky, kinky”: using<br />
corpora as tools for critical analysis’. Discourse and Society 21: 99–133.<br />
Camden, W. G. 1979. ‘Parallels in structure of lexicon and syntax between New Hebrides<br />
Bislama and <strong>the</strong> South Santo language as spoken at Tangoa’, pp. 51–117 of Papers in Pidgin<br />
and Creole Linguistics, 2, edited by Peter Mühlhäusler et al. Canberra: Pacific Linguistics.
224 References<br />
Cameron, Deborah. 1992. ‘Not gender difference but <strong>the</strong> difference gender makes—<br />
explanation in research on sex and language’. International Journal of <strong>the</strong> Sociology of<br />
Language 94: 13–26.<br />
Cameron, Deborah. 1995. Verbal hygiene. London: Routledge.<br />
Cameron, Deborah. 1997. ‘Theoretical debates in feminist linguistics: questions of sex and<br />
gender’,pp.21–36 of <strong>Gender</strong> and discourse, edited by Ruth Wodak. London: Sage publications.<br />
Cameron, Deborah. 2000. ‘Review of Anne Pauwels, Women changing language’. Language in<br />
Society 29: 269–72.<br />
Cameron, Deborah. 2011. ‘<strong>Gender</strong> and language’, pp. 331–3 of The Cambridge encyclopedia of<br />
<strong>the</strong> language sciences, edited by Patrick Hogan. Cambridge: Cambridge University Press.<br />
Cameron, Deborah and Don Kulick. 2003. Language and sexuality. Cambridge: Cambridge<br />
University Press.<br />
Cameron, Deborah, Fiona McAlinden, and Kathy O’Leary. 1989. ‘Lakoff in context: <strong>the</strong> social<br />
and linguistic functions of tag question’, pp. 74–93 of Women in <strong>the</strong>ir speech communities,<br />
edited by Jennifer Coates and Deborah Cameron. London: Longman.<br />
Campbell, Élisabeth. 2004. ‘La représentation des femmes dans les dictionnaires bilingues’.<br />
French Studies 57: 61–76.<br />
Capell, A. and H. E. Hinch. 1970. Maung grammar, texts and vocabulary. The Hague: Mouton.<br />
Caputo, B. Britt. 2005. ‘<strong>Gender</strong> neutrality hits Wesleyan’. Harvard Crimson 2 November.<br />
Carrington, J. F. 1947. ‘The initiation language: Lokele tribe’. African Studies 6: 196–207.<br />
Castellino, G. R. 1975. ‘<strong>Gender</strong> in Cushitic’, pp. 333–59 of Hamito-Semitica, edited by James<br />
and Theodora Bynon. Jauna Linguarum, Series Practica, 200. The Hague: Mouton.<br />
Chafe, Wallace L. 1967. Seneca morphology and dictionary. Washington, DC: Smithsonian Press.<br />
Chafe, Wallace L. 1977. ‘The evolution of third person verb agreement in <strong>the</strong> Iroquoian<br />
languages’, pp. 493–524 of Mechanisms of syntactic change, edited by Charles Li. Austin:<br />
University of Texas.<br />
Chafe, Wallace. 2004. ‘Masculine and feminine in <strong>the</strong> Nor<strong>the</strong>rn Iroquoian languages’,<br />
pp. 99–109 of Ethnosyntax: explorations in grammar and culture, edited by N. J. Enfield.<br />
Oxford: Oxford University Press.<br />
Cheshire, J. 1985. ‘A question of masculine bias’. English Today 1: 22–6.<br />
Cheshire, J. 2002. ‘Sex and gender in variationist research’, pp. 423–43 of The handbook of<br />
language variation and change, edited by J. K. Chambers, P. Trudgill, and N. Schilling-Estes.<br />
London: Blackwell.<br />
Cheshire, J. 2008. ‘Still a gender-biased language? Updates on gender inequalities and <strong>the</strong><br />
English language’. English Today 93: 7–10.<br />
Chirasombutti, Voravudhi and Anthony Diller. 1999. ‘Who am “I” in Thai?—The Thai first<br />
person: self-reference’, pp. 114–33 of <strong>Gender</strong>s and sexualities in modern Thailand, edited by<br />
Peter A. Jackson and Nerida M. Cook. Chiang Mai: Silkworm Books.<br />
Christ, S. 1997. ‘Duration of onset consonants in gay male stereotypes speech’. University of<br />
Pennsylvania Working Papers in Linguistics 4.3: 53–70.<br />
Clackson, James. 2007. Indo-European languages. Cambridge: Cambridge University Press.<br />
Clamons, Cynthia Robb. 1993. ‘<strong>Gender</strong> assignment in Oromo’, pp. 269–80 of Principles and<br />
prediction: <strong>the</strong> analysis of natural language, edited by Mushira Eid and Gregory Iverson.<br />
Amsterdam: John Benjamins.
References 225<br />
Claudi, U. 1985. Zur Entstehung von Genussystemen. Hamburg: Buske.<br />
Coates, Jennifer. 1993. Women, men and language. London and New York: Longman.<br />
Coates, Jennifer. 2012. ‘<strong>Gender</strong> and discourse analysis’, pp. 90–103 of The Routledge handbook<br />
of discourse analysis, edited by James Paul Gee and Michael Handford. London: Routledge.<br />
Coates, Jennifer and Pia Pichler (eds.). 2011. Language and gender: a reader. Oxford: Wiley<br />
Blackwell.<br />
Cobbinah, Alexander. 2010. ‘The Casamance as an area of intense language contact: <strong>the</strong> case of<br />
Baïnounk Gubaher’. Journal of Language Contact 3: 175–201.<br />
Collinder, Björn. 1965. An introduction to <strong>the</strong> Uralic languages. Berkeley: University of<br />
California Press.<br />
Collins French dictionary and grammar. 2014. London: Harper Collins.<br />
Comrie, Bernard and Greville G. Corbett. 1993. ‘Introduction’, pp. 1–19 of The Slavonic<br />
languages, edited by Bernard Comrie and Greville G. Corbett. London: Routledge.<br />
Connors, Kathleen. 1971. ‘Studies in feminine agentives in selected European languages’.<br />
Romance Philology 24: 573–98.<br />
Contini-Morava, Ellen and Marcin Kilarski. 2013. ‘Functions of nominal classification’. Language<br />
Sciences 40: 263–99.<br />
Cooper, Fiona. 1991. Jay loves Lucy. London: Serpent’s tail.<br />
Corbett, Greville G. 1991. <strong>Gender</strong>. Cambridge: Cambridge University Press.<br />
Corbett, Greville (ed.). 2014a. The expression of gender. Berlin: Mouton de Gruyter.<br />
Corbett, Greville. 2014b. ‘<strong>Gender</strong> typology’, pp. 87–130 of Corbett 2014a.<br />
Cowley, Roger, Marvin L. Bender, Charles A. Ferguson, Hailu Fulass, and Getatchew Haile.<br />
1976. ‘The Amharic language’, pp. 77–98 of Language in Ethiopia, edited by M. L. Bender,<br />
J. D. Bowen, R. L. Cooper, and C. A. Ferguson. London: Oxford University Press.<br />
Craig, C. G. 1986a. ‘Jacaltec noun classifiers’. Lingua 70: 241–84.<br />
Craig, C. G. (ed.). 1986b. Noun classes and categorisation. Amsterdam: John Benjamins.<br />
Craik, Brian. 1982. ‘The animate in Cree language and ideology’, pp. 29–35 of Papers of <strong>the</strong><br />
Thirteenth Algonquian Conference, edited by William Cowan. Ottawa: Carleton University.<br />
Creed, Gerald W. 1984. ‘Sexual subordination: institutionalized homosexuality and social<br />
control in Melanesia’. Ethnology 23: 157–76.<br />
Curzan, Anne. 2003. <strong>Gender</strong> shifts in <strong>the</strong> history of English. Cambridge: Cambridge University<br />
Press.<br />
Dahlstrom, Amy. 1995. ‘Motivation vs. predictability in Algonquian gender’, pp. 52–66 of<br />
Papers of <strong>the</strong> Twenty-Sixth Algonquian Conference, edited by David H. Pentland. Winnipeg:<br />
University of Manitoba.<br />
Daley, Karen Ann. 1998. Vietnamese classifiers in narrative texts. University of Texas at<br />
Arlington: Summer Institute of Linguistics.<br />
Darnell, Regna and Anthony L. Vanek. 1976. ‘The semantic basis of <strong>the</strong> animate/inanimate<br />
distinction in Cree’. Papers in Linguistics 9: 159–80.<br />
Dawkins, Richard M. 1916. Modern Greek in Asia Minor: a study of <strong>the</strong> dialects of Silli,<br />
Cappadocia and Phárasa with grammar, texts, translations and glossary. Cambridge: Cambridge<br />
University Press.<br />
de Courtenay, Baudouin. 1929. ‘Einfluss der Sprache auf Weltanschauung und Stimmung’.<br />
Prace filologiczne 14: 185–256.
226 References<br />
De León, Maria de Lourdes Pasquel. 1987. ‘Noun and numeral classifiers in Mixtec and Tzotzil:<br />
a referential view’. PhD <strong>the</strong>sis, University of Sussex.<br />
Demuth, K. A. 1988. ‘Noun classes and agreement in Sesotho acquisition’, pp. 305–22 of<br />
Agreement in natural language: approaches, <strong>the</strong>ories, descriptions, edited by M. Barlow and<br />
C. Ferguson. Stanford, Calif.: Center for <strong>the</strong> Study of Language and Literature.<br />
Demuth, K. A., N. Faraclas, and L. Marchese. 1986. ‘Niger-Congo noun class and agreement<br />
systems in language acquisition and historical change’, pp. 453–71 of Noun classes and<br />
categorisation, edited by C. G. Craig. Amsterdam: John Benjamins.<br />
Denny, J. P. 1976. ‘What are noun classifiers good for?’ Papers from <strong>the</strong> Annual Regional<br />
Meeting of <strong>the</strong> Chicago Linguistic Society 12: 122–32.<br />
Deutscher, Guy. 2010. Through <strong>the</strong> language glass: why <strong>the</strong> world looks different in o<strong>the</strong>r<br />
languages. London: Heineman.<br />
Diakonoff, I. M. 1990. Afrasian languages. Moscow: Nauka.<br />
Diller, A. V. N. 1985. ‘High and low Thai: views from within’, pp. 51–76 of Papers in South-<br />
East Asian Linguistics No. 9: Language policy, language planning and sociolinguistics in<br />
South-East Asia, edited by David Bradley. Canberra: Pacific Linguistics.<br />
Dimmendaal, Gerrit J. 1983. The Turkana language. Dordrecht: Foris Publications.<br />
Dimmendaal, Gerrit J. 2000. ‘Number marking and noun categorization in Nilo-Saharan<br />
languages.’ Anthropological Linguistics 42: 214–61.<br />
Dinslage, Sabine, Rudolf Leger, and Anne Storch. 2000. ‘Space and gender: cultural limitation<br />
of space in two communities of nor<strong>the</strong>astern Nigeria’. Anthropos 95: 121–7.<br />
Dixon, R. M. W. 1972. The Dyirbal language of North Queensland. Cambridge: Cambridge<br />
University Press.<br />
Dixon, R. M. W. 1980. The languages of Australia. Cambridge: Cambridge University Press.<br />
Dixon, R. M. W. 1982. Where have all <strong>the</strong> adjectives gone? And o<strong>the</strong>r essays in semantics and<br />
syntax. Berlin: Mouton.<br />
Dixon, R. M. W. 2002. Australian languages: <strong>the</strong>ir nature and development. Cambridge:<br />
Cambridge University Press.<br />
Dixon, R. M. W. 2004. The Jarawara language of sou<strong>the</strong>rn Amazonia. Oxford: Oxford<br />
University Press.<br />
Dixon, R. M. W. 2010. Basic linguistic <strong>the</strong>ory. Volume 1. Methodology. Volume 2. Grammatical<br />
topics. Oxford: Oxford University Press.<br />
Dixon, R. M. W. 2012. Basic linguistic <strong>the</strong>ory. Volume 3. Fur<strong>the</strong>r grammatical topics. Oxford:<br />
Oxford University Press.<br />
Dixon, R. M. W. 2014. Making new words: morphological derivation in English. Oxford: Oxford<br />
University Press.<br />
Dixon R. M. W. 2015. Edible gender, mo<strong>the</strong>r-in-law style, and o<strong>the</strong>r grammatical wonders:<br />
studies in Dyirbal, Yidiñ, and Warrgamay. Oxford: Oxford University Press.<br />
Dixon R. M. W. 2016. Are some languages better than o<strong>the</strong>rs? Oxford: Oxford University Press.<br />
Dobrin, Lise M. 2012. Concreteness in grammar: <strong>the</strong> noun class systems of <strong>the</strong> Arapesh<br />
languages. Stanford, Calif.: Stanford Studies in Morphology and <strong>the</strong> Lexicon.<br />
Doleschal, Ursula and Sonja Schmid. 2001. ‘Doing gender in Russian: structure and perspective’,<br />
pp. 253–82 of Hellinger and Bussmann 2001, Volume 1.
References 227<br />
Dorian, Nancy C. 1981. Language death: <strong>the</strong> life cycle of a Scottish Gaelic dialect. Philadelphia:<br />
University of Pennsylvania Press.<br />
Drapeau, Lynn and Renée Lambert-Brétière. 2011. ‘Verbal classifiers in Innu’. Anthropological<br />
Linguistics 53: 293–322.<br />
Drude, Sebastian. 2002. ‘Fala masculina e fala feminina em Awetí’, pp. 177–90 of Línguas<br />
indígenas brasileiras: fonologia, gramática e história, edited by Aryon Dall’Igna Rodrigues<br />
and Ana Suelly Arruda Câmara Cabral. Belém: EDUFPA.<br />
Drude, Sebastian. 2006. ‘On <strong>the</strong> position of <strong>the</strong> Awetí language in <strong>the</strong> Tupí family’, pp. 11–45<br />
of Guarani y ‘Maweti-Tupi-Guarani’. Estudios históricos y descriptivos sobre una familia<br />
lingüistica de America del Sur, edited by W. Dietrich and H. Symeonidis. Berlin: LIT Verlag.<br />
Drude, Sebastian. 2011. ‘Aweti in relation with Kamayurá: <strong>the</strong> two Tupian languages of <strong>the</strong><br />
Upper Xingu’, pp. 155–92 of Alto Xingu: uma sociedade multilíngue, edited by Bruna<br />
Franchetto. Rio de Janeiro: Museu do Índio/FUNAI.<br />
Dubois, Betty and Isabel Crouch. 1975. The question of tag questions in women’s speech: <strong>the</strong>y<br />
really don’t use more of <strong>the</strong>m’. Language in Society 4: 289–94.<br />
Dul'son, A. P. 1968. Ketskij jazyk (The Ket language). Tomsk: Izdateljstvo Tomskogo<br />
Universiteta.<br />
Dunn, Michael. 2000. ‘Chukchi women’s language: a historical-comparative perspective’.<br />
Anthropological Linguistics 42: 305–28.<br />
Dunn, Michael. 2014. ‘<strong>Gender</strong> determined dialect variation’, pp. 39–68 of Corbett 2014a.<br />
Du Puis, Mathias. 1652. Relation de l’establissement d’une colonie française dans la Guadeloupe<br />
isle de l’Amérique et des moeurs des Sauvages. Caen: Yvon. (Reissue in 1972, Bibliothèque<br />
d’histoire antillaise 1, Basse Terre: Société d’histoire de la Guadeloupe.)<br />
Eckert, Penelope. 1989. ‘The whole woman: sex and gender differences in variation’. Language<br />
Variation and Change 1: 245–67.<br />
Eckert, Penelope. 2000. Linguistic variation as social practice: <strong>the</strong> linguistic construction of<br />
identity in Belten High. Oxford: Blackwell.<br />
Eckert, Penelope and Sally McConnell-Ginet. 2003. Language and gender. Cambridge: Cambridge<br />
University Press.<br />
Eckert, Penelope and Sally McConnell-Ginet. 2006. ‘Putting communities of practice in <strong>the</strong>ir<br />
place’. <strong>Gender</strong> and Language 1.1: 27–38.<br />
Edelman, D. I. 1966. ‘Yazgulamskij jazyk (Yazgulami language)’, pp. 436–54 of Jazyki narodov<br />
SSSR. Volume 1. Indoevropejskie jazyki (Languages of <strong>the</strong> USSR. Indo-European languages),<br />
edited by V. V. Vinogradov et al. Moscow: Nauka.<br />
Edelman, D. I. 1999. ‘Garvi, or Bashkarik’, pp. 56–8 of Yazyki mira. Dardskie i nuristanskie<br />
jazyki (Languages of <strong>the</strong> world. Dardic and Nuristani languages). Moscow: Indrik.<br />
Edenmyr, Niklas. 2004. ‘The semantics of Hadza gender assignment: a few notes from <strong>the</strong><br />
field’. Africa and Asia 4: 3–19.<br />
Einaudi, Paula. 1976. A grammar of Biloxi. New York: Garland.<br />
Ekka, Francis. 1972. ‘Men’s and women’s speech in Kũrˌux’. Linguistics 81: 25–31.<br />
Emeneau, Murray. 1980. ‘India and linguistic areas’, pp. 126–66 of Language and linguistic<br />
area. Essays by Murray B. Emeneau. Selected and introduced by Anwar S. Dil. Stanford,<br />
Calif.: Stanford University Press.
228 References<br />
Endruschat, Annette. 2015. ‘<strong>Gender</strong> in Portuguese’, pp. 303–32 of Hellinger and Motschenbacher<br />
2015.<br />
Enfield, N. J. 2004. ‘Nominal classification in Lao: a sketch’, pp. 117–43 of Nominal classification,<br />
Special issue of Sprachtypologie und Universalienforschung 57: 2/3, edited by Alexandra<br />
Y. Aikhenvald.<br />
Enfield, N. J. 2007. A grammar of Lao. Berlin: Mouton de Gruyter.<br />
England, N. 1983. A grammar of Mam, a Mayan language. Austin: University of Texas Press.<br />
Ervin, Susan M. 1962. ‘The connotations of gender’. Word 18: 249–61.<br />
Escure, Geneviève. 2001. ‘Belizean Creole: gender, creole, and <strong>the</strong> role of women in language<br />
change’, pp. 53–84 of Hellinger and Bussmann 2001, Volume 1.<br />
Evans, Bergen and Cornelia Evans. 1957. Dictionary of contemporary American usage. New York:<br />
Random House.<br />
Faust, Norma, 1963. ‘El lenguaje de los hombres y mujeres en Cocama’. Perú Indígena 10.22–3:<br />
115–17.<br />
Ferguson, C. A. 1964. ‘Baby talk in six languages’. American Anthropologist 66: 103–14.<br />
Ferrari, F. 2005. ‘A syntactic analysis of <strong>the</strong> nominal system of Italian and Luganda: how nouns<br />
can be formed in syntax’. PhD dissertation, New York University.<br />
Fesenko, Andrej and Tatiana Fesenko. 1955. Russkij jazyk pri sovetakh (Russian language<br />
under <strong>the</strong> Soviets). New York: Rausen Bors.<br />
Flannery, Regina. 1946. ‘Men’s and women’s speech in Gros Ventre’. International Journal of<br />
American Linguistics 12: 133–5.<br />
Fleming, Luke. 2012. ‘<strong>Gender</strong> indexicality in <strong>the</strong> Native Americas: contributions to <strong>the</strong><br />
typology of social indexicality’. Language in Society 41: 295–320.<br />
Fleming, Luke. 2015. ‘Australian exceptionalism in <strong>the</strong> typology of affinal avoidance registers’.<br />
Anthropological Linguistics 56: 155–8.<br />
Fletcher, Alice and Frances La Fleche. 1972. The Omaha tribe. Lincoln: University of Nebraska<br />
Press.<br />
Fodor, I. 1959. ‘The origin of grammatical gender I, II’. Lingua 8: 1–41, 186–214.<br />
Foley, W. A. 1991. The Yimas language of New Guinea. Stanford, Calif.: Stanford University<br />
Press.<br />
Fortune, David and Gretchen Fortune. 1975. ‘Karajá’s men’s-women’s speech differences with<br />
social correlates’. Arquivos de anatomia e antropologia (Instituto de Antropologia Professor<br />
Souza Marques, Rio de Janeiro) 1: 111–34.<br />
Fortune, David and Gretchen Fortune. 1987. ‘Karajá literary acquisition and sociocultural<br />
effects on a rapidly changing culture’. Journal of Multilingual and Multicultural Development<br />
8: 469–91.<br />
Fox, Helen. 1996. ‘An honorific sub-dialect used among Big Nambas women’, pp. 375–82 of<br />
Oceanic studies: proceedings of <strong>the</strong> first international conference on Oceanic linguistics, edited<br />
by John Lynch and Fa’afo Pat. Canberra: Pacific Linguistics.<br />
Frajzyngier, Zygmunt. 1993. A grammar of Mupun. Berlin: Dietrich Reimer.<br />
Fraizyngier, Zygmunt. 2012. ‘Typological outline of <strong>the</strong> Afroasiatic phylum’, pp. 505–624 of<br />
Frajzyngier and Shay 2012.<br />
Fraizyngier, Zygmunt and Erin Shay (eds.). 2012. Afroasiatic languages. Cambridge: Cambridge<br />
University Press.
References 229<br />
Franklin, K. 1981. ‘Existential and pro-verbs in Kewa’, pp. 151–72 of Syntax and semantics in<br />
Papua New Guinea languages, edited by K. J. Franklin. Ukarumpa: Summer Institute of<br />
Linguistics.<br />
Fraurud, Kari. 2000. ‘Proper names and gender in Swedish’, pp. 167–219 of Unterbeck and<br />
Rissanen 2000.<br />
Friedman, Victor. 1993. ‘Macedonian’, pp. 249–305 of The Slavonic languages, edited by<br />
Bernard Comrie and Greville G. Corbett. London: Routledge.<br />
Gal, Susan. 1978. ‘Peasant men can’t get wives: language change and sex roles in a bilingual<br />
community’. Language in Society 7: 1–16.<br />
Gal, Susan. 1979. Language shift. New York: Academic Press.<br />
Gal, Susan. 1989. ‘Between speech and silence: <strong>the</strong> problematics of research on language and<br />
gender’. IPrA Papers in Pragmatics 3: 1–38.<br />
Gal, Susan. 1991. ‘Between speech and silence: <strong>the</strong> problematics of research on language and<br />
gender’, pp. 175–203 of <strong>Gender</strong> at <strong>the</strong> crossroads of knowledge: toward a new anthropology of<br />
gender, ed. M. DiLeonardo. Berkeley: University of California Press.<br />
Gal, Susan. 1995. ‘Language, gender and power: an anthropological review’, pp. 169–82 of<br />
Bucholtz and Hall 1995a.<br />
Garibian, A. S. 1976. ‘Armjanskij jazyk’ (Armenian language), pp. 94–109 of Jazyki Azii i<br />
Afriki. Indo-evropejskie jazyki. Moscow: Nauka.<br />
Gauchat, L. 1905. ‘L’unité phonétique dans le patois d’une commune’, pp. 175–232 of Aus<br />
romanischen Sprachen und Literaturen: Festschrift Heinrich Mort. Halle: Max Niemeyer.<br />
Gaudio, R. 1994. ‘Sounding gay: pitch properties in <strong>the</strong> speech of gay and straight men’.<br />
American Speech 69: 30–7.<br />
Georg, Stefan. 2007. A descriptive grammar of Ket. Part 1: Introduction, phonology and<br />
morphology. Folkestone: Global Oriental Ltd.<br />
Gerdts, Donna and Mercedes Q. Hinkson. 2004. ‘Salish numeral classifiers: a lexical means to a<br />
grammatical end’, pp. 247–79 of Nominal classification, Special issue of Sprachtypologie und<br />
Universalienforschung 57 2/3, edited by Alexandra Y. Aikhenvald.<br />
Goddard, C. 1985. A grammar of Yankunytjatjara. Alice Springs: Institute for Aboriginal<br />
Development.<br />
Goddard, Ives. 2002. ‘Grammatical gender in Algonquian’, pp. 195–231 of Papers of <strong>the</strong> Thirty-<br />
Third Algonquian Conference, edited by H. Christoph Wolfart. Winnipeg: University of<br />
Manitoba.<br />
Goddard, Ives, Michael Silverstein, Calvert Watkins, Einar Haugen, Sandra L. Chung,<br />
T. L. Markey, Robert Underhill, Jay Jasanoff, Engin Sezer, Toeio Sato, Ariene Berman,<br />
Alan H. Timberlake, Michael Szamosl, Anthony Ariotto, Stephen Anderson, Gregory Nagy,<br />
and Janet F. Hinkley. 1971. ‘Pronoun envy’. Harvard Crimson, 16 November.<br />
Goettner-Abendroth, Heide. 2012. Matriarchal societies: studies on indigenous cultures across<br />
<strong>the</strong> globe. New York: Peter Lang.<br />
Gordon, Mat<strong>the</strong>w and Jeffrey Heath. 1998. ‘Sex, sound symbolism, and sociolinguistics’.<br />
Current Anthropology 39.4: 421–49.<br />
Grandi, Nicola. 2002. Morfologie in contatto: le costruzioni valutative nelle lingue del Mediterraneo.<br />
Milan: Franco Angeli.<br />
Green, Ian. 1997. ‘Nominal classification in Marrithiyel’, pp. 229–53 of Harvey and Reid 1997.
230 References<br />
Greenbaum, Sidney. 1991. An introduction to English grammar. London: Longman.<br />
Greenbaum, Sidney. 1996. The Oxford English grammar. Oxford: Oxford University Press.<br />
Greenberg, J. H. 1963. ‘Some universals of grammar with particular reference to <strong>the</strong> order of<br />
meaningful elements’, pp. 58–90 of Universals of language, edited by J. H. Greenberg.<br />
Cambridge, Mass.: MIT Press.<br />
Greenberg, J. H. 1978. ‘<strong>How</strong> does language acquire gender markers’, pp. 241–70 of Universals<br />
of human languages, Volume 3. Word structure, edited by J. H. Greenberg, C. A. Ferguson,<br />
and E. A. Moravcsik. Stanford, Calif.: Stanford University Press.<br />
Gregersen, Edgar A. 1979. ‘Sexual linguistics’, pp. 3–19 of Language, sex and gender: does ‘la<br />
différence’ make a difference?, edited by J. Orasanu, M. K. Slater, and L. L. Adler. New York:<br />
New York Academy of Sciences.<br />
Grünberg, A. L. 1999. ‘Gawar’, pp. 91–5 of Yazyki mira. Dardskie i nuristanskie jazyki<br />
(Languages of <strong>the</strong> world. Dardic and Nuristani languages). Moscow: Indrik.<br />
Gumperz, J. J. and R. Wilson. 1971. ‘Convergence and creolization: a case from <strong>the</strong> Indo-<br />
Aryan/Dravidian border in India’, pp. 151–68 of Pidginization and creolization of languages,<br />
edited by D. Hymes. Cambridge: Cambridge University Press.<br />
Günthner, Susanne. 1996. ‘Male–female speaking practices across cultures’, pp. 447–73 of<br />
Contrastive sociolinguistics, edited by Marlis Hellinger and Ulrich Ammon. Berlin: Mouton<br />
de Gruyter.<br />
Haas, Mary. 1941. Tunica. Extract from HAIL 4: 1–143. New York: J. J. Augustin.<br />
Haas, Mary. 1944. ‘Men’s and women’s speech in Koasati’. Language 20: 142–9.<br />
Haas, Mary. 1951. ‘Interlingual word taboos’. American Anthropologist NS 53: 338–44.<br />
Hachimi, Atiqa. 2001. ‘Shifting sands: language and gender in Moroccan Arabic’, pp. 27–51 of<br />
Hellinger and Bussman 2001, Volume 1.<br />
Hagège, C. 2004. ‘On <strong>the</strong> part played by human conscious choice in language structure and<br />
language evolution,’ pp. 105–17 of Linguistic diversity and language <strong>the</strong>ories, edited by<br />
Z. Frajzyngier, A. Hodges, and D. Rood. Amsterdam: John Benjamins.<br />
Hagman, Roy S. 1977. Nama Hottentot grammar. Bloomington: Indian University Press.<br />
Haiman, J. 1980. Hua: a Papuan language of <strong>the</strong> Eastern Highlands of New Guinea. Amsterdam:<br />
John Benjamins.<br />
Hale, Kenneth. 1971. ‘A note on a Warlbiri tradition of antonymy’, pp. 472–82 of An<br />
interdisciplinary reader in philosophy, linguistics and psychology, edited by Danny<br />
D. Steinberg and Leon A. Jakobovits. Cambridge: Cambridge University Press.<br />
Hale, Kenneth and David Nash. 1997. ‘Damin and Lardil phonotactics’, pp. 247–59 of Boundary<br />
rider: essays in honour of Geoffrey O’Grady, edited by D. T. Tryon and M. Walsh. Canberra:<br />
Pacific Linguistics.<br />
Hall, Kira. 2002. ‘“Unnatural” gender in Hindi’, pp. 133–62 of <strong>Gender</strong> across languages: <strong>the</strong><br />
linguistic representation of women and men, edited by Marlis Hellinger and Hadumod<br />
Bussman. Amsterdam: John Benjamins.<br />
Hall, Kira and Veronica O’Donovan. 1996. ‘Shifting gender positions among Hindi-speaking<br />
hijras’, pp. 228–66 of Rethinking language and gender research: <strong>the</strong>ory and practice, edited by<br />
V. Bergvall et al. London: Longman.<br />
Hamp, Eric. 1965. ‘The Albanian dialect of Mandres’. Die Sprache 11: 137–54.<br />
Hampares, Ka<strong>the</strong>rine. 1976. ‘Sexism in Spanish lexicography?’. Hispania 59: 100–9.
References 231<br />
Harbert, Wayne. 2007. Germanic languages. Cambridge: Cambridge University Press.<br />
Harrison, Simon J. 1990. Stealing people’s names: history and politics in a Sepik river cosmology.<br />
Cambridge: Cambridge University Press.<br />
Harvey, Mark. 1987. ‘The Warray language from Adelaide River’. MA <strong>the</strong>sis, Australian<br />
National University.<br />
Harvey, Mark. 1992. The Gaagudju people and <strong>the</strong>ir language. PhD <strong>the</strong>sis, University of Sydney.<br />
Harvey, Mark. 1997. ‘Nominal classification and gender in aboriginal Australia’, pp. 17–62 of<br />
Nominal classification in Aboriginal Australia, edited by Mark Harvey and Nicholas Reid.<br />
Amsterdam: John Benjamins.<br />
Harvey, Mark. 2002. A grammar of Gaagudju. Berlin: Mouton de Gruyter.<br />
Harvey, Mark and Nicholas Reid (eds.). 1997. Nominal classification in Aboriginal Australia.<br />
Amsterdam: John Benjamins.<br />
Haspelmath, M. 1993. A grammar of Lezgian. Berlin: Mouton de Gruyter.<br />
Hasselblatt, Cornelius. 2015. ‘The representation of gender in Estonian’, pp. 125–52 of<br />
Hellinger and Motschenbacher 2015.<br />
Haugen, Einar Ingvald. 1976. The Scandinavian languages: an introduction to <strong>the</strong>ir history.<br />
Cambridge, Mass.: Harvard University Press.<br />
Hauser-Schäublin, Brigitta. 1977. Frauen in Kararau. Zur Rolle der Frau bei den Iatmul am<br />
Mittelsepik, Papua New Guinea. Basler Beiträge zur Ethnologie, Volume 18. Basel: Wepf.<br />
Hauser-Schäublin, Brigitta. 1995. ‘Puberty rites, women’s Naven, and initiation: women’s<br />
rituals of transition in Abelam and Iatmul culture’, pp. 55–82 of <strong>Gender</strong> rituals: female<br />
initiation in Melanesia, edited by Nancy C. Lutkehaus and Paul B. Roscoe. London:<br />
Routledge.<br />
Haviland, William A., Harald E. L. Prins, Bunny McBride, and Dana Walrath. 2014. Cultural<br />
anthropology: <strong>the</strong> human challenge, 14th edition. Belmont, Calif.: Wadsworth/Cengage<br />
Learning.<br />
Heath, Jeffrey. 1984. Functional grammar of Nunggubuyu. Canberra: Australian Institute of<br />
Aboriginal Studies.<br />
Hegardty, Peter. 2014. ‘Ladies and gentlemen: word order and gender in English’, pp. 69–86 of<br />
Corbett 2014a.<br />
Heine, Bernd. 1982a. ‘African noun class systems’, pp. 189–216 of Apprehension: Das sprachliche<br />
Erfassen von Gegenständen, Teil II: Die Techniken und ihr Zusammenhang in<br />
Einzelsprachen, edited by Hansjakob Seiler and F. J. Stachowiak. Tübingen: Narr Language<br />
Universals Series I/II.<br />
Heine, Bernd. 1982b. The Nubi language of Kibera: an Arabic Creole. Berlin: Dietrich Reimer<br />
Verlag.<br />
Heine, Bernd and Tania Kuteva. 2002. <strong>World</strong> lexicon of grammaticalization. Cambridge:<br />
Cambridge University Press.<br />
Heine, Bernd and Mechthild Reh. 1984. Grammaticalization and reanalysis in African languages.<br />
Hamburg: Helmut Buske Verlag.<br />
Heine, Bernd and Rainer Vossen. 1983. ‘On <strong>the</strong> origin of gender in Eastern Nilotic’, pp. 255–68<br />
of Nilotic studies: proceedings of <strong>the</strong> International Symposium on Languages and History of<br />
<strong>the</strong> Nilotic Peoples, edited by R. Vossen and M. Bechchaus-Gerst. Berlin: Reimer.<br />
Hellenthal, Anne-Christie. 2010. A grammar of Sheko. Amsterdam: LOT.
232 References<br />
Hellinger, Marlis and Hadumod Bussmann (eds.). 2001–3. <strong>Gender</strong> across languages, Volumes<br />
1–3. Amsterdam: John Benjamins.<br />
Hellinger, Marlis and Heiko Motschenbacher. 2015. <strong>Gender</strong> across languages, Volume 4.<br />
Amsterdam: John Benjamins.<br />
Hellinger, Marlis and Anne Pauwels. 2007. ‘Language and sexism’, pp. 651–81 of Handbook of<br />
language and communication: diversity and change, edited by Marlis Hellinger and<br />
Anne Pauwels. Berlin: Mouton de Gruyter.<br />
Henley, Nancy M. 1989. ‘Molehill or mountain? What we know and don’t know about sex bias<br />
in language’, pp. 59–78 of <strong>Gender</strong> and thought: psychological perspectives, edited by Mary<br />
Crawford and Margaret Gentry. New York: Springer Verlag.<br />
Henley, Nancy M. and Joselito Abueg. 2003. ‘A review and syn<strong>the</strong>sis of research on comprehension<br />
of <strong>the</strong> masculine as a generic form in English’. Estudios de Sociolingüística 4:<br />
427–54.<br />
Henson, Eithne. 1997. ‘Johnson and <strong>the</strong> condition of women’, pp. 67–85 of The Cambridge<br />
companion to Samuel Johnson, edited by Greg Clingham. Cambridge: Cambridge University<br />
Press.<br />
Herbert, R. K. 1990. ‘Hlonipha and <strong>the</strong> ambiguous woman’. Anthropos 85: 455–73.<br />
Herbert, R. K. 1991. ‘Patterns in language change, acquisition and dissolution: noun prefixes<br />
and concords in Bantu’. Anthropological Linguistics 33.2: 103–34.<br />
Herdt, Gilbert. 1987. The Sambia: ritual and gender in New Guinea. New York: Holt, Rinehart<br />
and Winston Inc.<br />
Hill, Jane H. 1987. ‘Women’s speech in modern Mexicano’, pp. 121–60 of Philips, Steele, and<br />
Tanz 1987.<br />
Hill, Jane and Kenneth Hill. 1977. ‘Language death and relexification in Tlaxcalan Nahuatl’.<br />
International Journal of <strong>the</strong> Sociology of Language 12: 55–67.<br />
Hill, Jane and Ofelia Zepeda. 1999. ‘Language, gender, and biology: pulmonic ingressive<br />
airstream in women’s speech in Tohono O'odham’. Southwest Journal of Linguistics 18:<br />
15–40.<br />
Hoben, Susan J. 1976. ‘The meaning of <strong>the</strong> second-person pronouns in Amharic’, pp. 281–8 of<br />
Language in Ethiopia, edited by M. L. Bender, J. D. Bowen, R. L. Cooper, and C. A. Ferguson.<br />
London: Oxford University Press.<br />
Hodge, Carleton T. 1958. ‘Non-native Hausa’, pp. 57–69 of Report on <strong>the</strong> Ninth Annual Round<br />
Table Meeting on Linguistics and Language Studies, edited by William M. Austin. Washington,<br />
DC: Georgetown University Press.<br />
Hoff, Berend. 1994. ‘Island Carib, an Arawakan language which incorporated a lexical register<br />
of Cariban origin, used to address men’, pp. 161–8 of Mixed languages: 15 case studies in<br />
language intertwining, edited by P. Bakker and M. Mous. Amsterdam: IFOTT.<br />
Hoffman, Ka<strong>the</strong>rine E. 2006. ‘Berber language ideologies, maintenance and contraction:<br />
gendered varieties in <strong>the</strong> indigenous margins of Morocco’. Language and Communication<br />
26: 144–67.<br />
Hollow, Robert. 1970. A Mandan dictionary. PhD dissertation, University of California,<br />
Berkeley.<br />
Holm, John. 2000. An introduction to pidgins and creoles. Cambridge: Cambridge University<br />
Press.
References 233<br />
Holmes, Janet. 1984. ‘Women’s language: a functional approach’. General Linguistics 24.3:<br />
149–78.<br />
Holmes, Janet. 1986. ‘Functions of you know in women’s and men’s speech’. Language in<br />
Society 15: 1–21.<br />
Holmes, Janet and Miriam Meyerhoff (eds.). 2003. The handbook of language and gender.<br />
Oxford: Blackwell.<br />
Holmquist, Jonathan C. 1985. ‘Social correlates of a linguistic variable: a study in a Spanish<br />
village’. Language in Society 14: 191–203.<br />
Holmquist, Jonathan C. 1991. ‘Semantic features and gender dynamics in Cantabrian Spanish’.<br />
Anthropological Linguistics 33: 57–81.<br />
Hornscheidt, Antje. 2003. ‘Linguistic and public attitudes towards gender in Swedish’,<br />
pp. 339–68 of Hellinger and Bussmann 2003, Volume 3.<br />
Hornscheidt, Antje. 2006. Die sprachliche Benennung von Personen aus konstruktivistischer<br />
Sicht. <strong>Gender</strong>spezifizierung und ihre diskursive Verhandlung im heutigen Schwedisch. Berlin:<br />
de Gruyter.<br />
Houseman, Michael and Carlo Severi. 1998. Naven, or <strong>the</strong> O<strong>the</strong>r Self: a relational approach to<br />
ritual action. Leiden: Brill.<br />
Huntley, David. 1980. ‘The evolution of genitive-accusative animate and personal nouns<br />
in Slavic dialects’, pp. 189–212 of Historical morphology, edited by J. Fisiak. The Hague:<br />
Mouton.<br />
Ibrahim, Muhammad Hasan. 1973. Grammatical gender: its origin and development. The<br />
Hague: Mouton.<br />
Ide, Sachiko. 1991. ‘<strong>How</strong> and why do women speak more politely in Japanese’, pp. 63–80 of Ide<br />
and McGloin 1991.<br />
Ide, Sachiko. 2003. ‘Women’s language as a group identity marker in Japanese’, pp. 227–39 of<br />
<strong>Gender</strong> across languages, Volume 3, edited by Marlis Hellinger and Hadumod Bussmann.<br />
Amsterdam: John Benjamins.<br />
Ide, Sachiko and Miyako Inoue. 1992. ‘Josei kotoba ni miru aidentitii’ (Identities in women’s<br />
language). Gengo 21.10: 46–7.<br />
Ide, Sachiko and Naomi Hanaoka McGloin (eds.). 1991. Aspects of Japanese women’s language.<br />
Tokyo: Kurosio Publishers.<br />
Ikoro, S. M. 1996. The Kana language. Leiden: University of Leiden.<br />
Inoue, Miyako. 2002. ‘<strong>Gender</strong>, language, and modernity: towards an effective history of<br />
Japanese women’s language’. American Ethnologist 29: 392–422.<br />
Irvine, Judith T. 1978. ‘Wolof noun classification: <strong>the</strong> social setting of divergent change’.<br />
Language in Society 7: 37–64.<br />
Irvine, Judith T. 1979. ‘Formality and informality in communicative events’. American<br />
Anthropologist 81: 773–90.<br />
Isac᷈enko, A. V. 1939. Narec᷈je vasi Sele na Roz᷈u. Ljubljana: Uc᷈iteljska tiskarna.<br />
Jacobs, Andrea Michele. 2004. ‘Language reform as language ideology: an examination of<br />
Israeli feminist language practice’. PhD dissertation, University of Texas at Austin.<br />
Jaffe, Alexandra. 1999. Language politics on Corsica. Berlin: Mouton de Gruyter.<br />
Jakobson, Roman O. 1959. ‘On linguistic aspects of translation’, pp. 113–18 of On Translation,<br />
edited by R. A. Brower. Cambridge, Mass.: Harvard University Press.
234 References<br />
Jakobson, Roman O. 1971. ‘Signe zéro’, pp. 211–19 of Roman Jakobson: selected writings,<br />
Volume 2. Word and language. The Hague and Paris: Mouton.<br />
Jakobson, Roman O. 1984. ‘The gender pattern of Russian’, pp. 141–3 of Russian and Slavic<br />
grammar: studies 1931–1981, edited by L. R. Waugh and M. Halle. Berlin: Mouton.<br />
James, Deborah. 1996. ‘Women, men and prestige speech forms: a critical review’, pp. 98–125<br />
of Rethinking language and gender research: <strong>the</strong>ory and practice, edited by Victoria<br />
L. Bergvall, Janet M. Bing, and Alice F. Freed. London: Longman.<br />
Janda, Laura A. 1999. ‘From number to gender, from dual to virile: bridging cognitive<br />
categories’, pp. 73–86 of Between grammar and lexicon, edited by Ellen Contini-Morava<br />
and Yishai Tobin. Amsterdam: John Benjamins.<br />
Janse, Mark. 2002. ‘Aspects of bilingualism in <strong>the</strong> history of <strong>the</strong> Greek language’, pp. 332–90 of<br />
Bilingualism in ancient society: language contact and <strong>the</strong> written text, edited by J. N. Adams,<br />
Mark Janse, and Simon Swain. Oxford: Oxford University Press.<br />
Jarkey, Nerida. 2015. ‘The Housewife’s Companion: identity construction in a Japanese<br />
women’s magazine’, pp. 179–201 of Language and identity across modes of communication,<br />
edited by Dwi Noverini Djenar, Ahmar Mahboob, and Ken Cruickshank. Berlin: De<br />
Gruyter.<br />
Jasanoff, Jay. 1989. ‘Language and gender in <strong>the</strong> Tarim Basin: <strong>the</strong> Tocharian 1sg pronoun’.<br />
Tocharian and Indo-European Studies 3: 125–48.<br />
Jespersen, Otto. 1894. Progress in language with special reference to English. London: Swan<br />
Sonnenschein.<br />
Jespersen, Otto. 1922. Language, its nature, development and origin. London: Allen and Unwin.<br />
Jespersen, Otto. 1949. A modern English grammar on historical principles. Part VII. Syntax.<br />
London: George Allen and Unwin Ltd.<br />
Jespersen, Otto. 1972. Growth and structure of <strong>the</strong> English language. Oxford: Basil Blackwell<br />
(reprint of 1912 edition, Stuttgart: Teubner Verlag).<br />
Jones, Lucy. 2014. ‘Masculinity in lesbian discourse’, pp. 175–196 of Milani 2014.<br />
Joseph, Brian D. and Irene Philippaki-Warburton. 1987. Modern Greek. London: Routledge.<br />
Jospin, Lionel. 1998. ‘Circulaire du 6 mars 1998 relative à la féminisation des noms de<br />
métier, fonction, grade ou titre, NOR: PRMX9802741C’. Journal Officiel 58, 8 March<br />
1998: 3565.<br />
Jungraithmayr, H. 1995. ‘Was ist am Tangale noch tschadisch/hamitosemitisch?,’ pp. 197–205<br />
of Sprachkulturelle und historische Forschungen in Afrika, edited by A. Fleisch and D. Otten.<br />
Cologne: Köppe.<br />
Kahane, Henry and René Kahane. 1948–9. ‘The augmentative feminine in <strong>the</strong> Romance<br />
languages’. Romance Philology 2: 135–75.<br />
Karatsareas, Petros. 2009. ‘The loss of grammatical gender in Cappadocian Greek’. Transactions<br />
of <strong>the</strong> Philological Society 107: 196–230.<br />
Karatsareas, Petros. 2013. ‘On <strong>the</strong> diachrony of gender in Asia Minor Greek: <strong>the</strong> development<br />
of semantic agreement in Pontic’. Language Sciences 43: 77–101.<br />
Karbelashvili, D. P. 1935. Ruchnaia rech na Kavkaze: issledovanie po materialam Baraninskogo<br />
raiona SSR Armenii. Izd. Nauchno-issledovatelʹskogo instituta kavkazovedeniia.<br />
Karmiloff-Smith, A. 1979. A functional approach to child language: a study of determiners and<br />
reference. Cambridge: Cambridge University Press.
References 235<br />
Kastovsky, Dieter. 2000. ‘Inflectional classes, morphological restructuring, and <strong>the</strong> dissolution<br />
of Old English grammatical gender’, pp. 709–27 of Unterbeck and Rissanen 2000.<br />
Kazakevich, Olga A. 1996. ‘The education of <strong>the</strong> Selkups in Russia: teaching <strong>the</strong> mo<strong>the</strong>r tongue<br />
as a foreign language’. International Review of Education 42.4: 388–9.<br />
Keenan, Elinor. 1974. ‘Norm-makers, norm-breakers: uses of speech by men and women in a<br />
Malagasy community’, pp. 125–43 of Bauman and Sherzer 1974.<br />
Kendon, Adam. 1988. Sign languages in Aboriginal Australia: cultural, semiotic, and communicative<br />
perspectives. Cambridge: Cambridge University Press.<br />
Key, Mary Ritchie. 1975. Male/female language, with a comprehensive bibliography. Metuchen,<br />
NJ: The Scarecrow Press, Inc.<br />
Khaidakov, S. M. 1963. ‘The principles of non-class division in Lak’ (in Russian, with a<br />
summary in English). Studia Caucasica 1: 48–55.<br />
Kibrik, A. E. (ed.). 1996. Godoberi. Munich: Lincom Europa.<br />
Kilarski, Marcin. 2007. ‘Algonquian and Indo-European gender in a historiographic perspective’.<br />
Historiographia Linguistica 34: 333–49.<br />
Kilarski, Marcin. 2014. Nominal classification: a history of its study from <strong>the</strong> classical period to<br />
<strong>the</strong> present. Amsterdam: John Benjamins.<br />
Kilian-Hatz, Christa. 2008. A grammar of Modern Khwe (Central Khoisan). Cologne: Rüdiger<br />
Köppe.<br />
Kimball, Geoffrey D. 1987. ‘Men’s and women’s speech in Koasati: a reappraisal’. International<br />
Journal of American Linguistics 53: 30–8.<br />
Kimball, Geoffrey D. 1991. Koasati grammar. Lincoln: University of Nebraska Press.<br />
Kirton, Jean F. 1988. ‘Men’s and women’s dialects’. Aboriginal Linguistics 1: 111–25.<br />
Klenin, E. 1983. Animacy in Russian: a new interpretation. Columbus, Oh.: Slavica.<br />
Klumpp, D. 1990. Piapoco grammar. Colombia: SIL.<br />
Kochskämper, Birgit. 1999. ‘Frau’ und ‘Mann’ im Althochdeutschen. Frankfurt: Peter Lang.<br />
Köhler, O. 1962. ‘Studien zum Genussystem und Verbalbau der zentralen Khoisan-Sprachen’.<br />
Anthropos 57: 529–46.<br />
Konishi, T. 1993. ‘The semantics of grammatical gender: a cross-cultural study’. Journal of<br />
Psycholinguistic Research 22: 519–34.<br />
Kooyers, O. 1974. Washkuk grammar sketch. Workpapers in Papua New Guinea Languages 6:<br />
5–74.<br />
Kossmann, Maarten. 2014. ‘Derivational gender in Moroccan Berber: examples from Ayt<br />
Seghrushen’. Sprachtypologie und Universalienforschung 67.1: 21–33.<br />
Kramarae, Cheris. 1992. ‘Punctuating <strong>the</strong> dictionary’. International Journal of <strong>the</strong> Sociology of<br />
Language 94: 135–54.<br />
Kramarae, Cheris and Paula A. Treichler. 1992. Amazons, bluestockings and crones: a feminist<br />
dictionary. A woman’s companion to words and ideas. London: Pandora Press.<br />
Krishnamurti, Bhadriraju. 2003. The Dravidian languages. Cambridge: Cambridge University<br />
Press.<br />
Krishnamurti, Bhadriraju and John Peter Lucius Gwynn. 1985. A grammar of modern Telugu.<br />
New Delhi: Oxford University Press.<br />
Kroskrity, Paul V. 1983. ‘On <strong>the</strong> male and female speech in <strong>the</strong> Pueblo Southwest’. International<br />
Journal of American Linguistics 49: 88–91.
236 References<br />
Kryk-Kastovsky, Barbara. 2000. ‘Norm vs. use: on gender in Polish’, pp. 729–47 of <strong>Gender</strong> in<br />
grammar and cognition. Part 2: Manifestations of gender, edited by Barbara Unterbeck and<br />
Matti Rissanen. Berlin: Mouton de Gruyter.<br />
Kulick, Don. 1992a. Language shift and cultural reproduction: socialisation, self, and syncretism<br />
in a Papua New Guinean village. Cambridge and New York: Cambridge University Press.<br />
Kulick, Don. 1992b. ‘Anger, gender, language shift and <strong>the</strong> politics of revelation in a Papua<br />
New Guinean village’. Pragmatics 2: 281–96.<br />
Kulick, Don. 1998. Sex, gender, and culture among Brazilian transgendered prostitutes. Chicago:<br />
University of Chicago Press.<br />
Kulick, Don. 2000. ‘Gay and lesbian language’. Annual Review of Anthropology 29: 243–85.<br />
Kulick, Don and C. Stroud. 1990. ‘Code-switching in Gapun: social and linguistic aspects of<br />
language used in a language shifting community’, pp. 205–34 of Melanesian Pidgin and Tok<br />
Pisin, edited by J. W. M. Verhaar. Amsterdam: John Benjamins.<br />
Kuntjara, Es<strong>the</strong>r. 2001. ‘<strong>Gender</strong> in Javanese Indonesian’. <strong>Gender</strong> across Languages 1: 199–227.<br />
Kwee, Jihn B. 1965. Indonesian. London: The English Universities Press Ltd.<br />
Labov, William. 1972. Sociolinguistic patterns. Oxford: Blackwell.<br />
Lakoff, George. 1987. Women, fire and dangerous things: what categories reveal about <strong>the</strong> mind.<br />
Chicago: University of Chicago Press.<br />
Lakoff, Robin Tolmach. 1973. ‘Language and woman’s place’. Language in Society 2: 45–80.<br />
Lakoff, Robin Tolmach. 1975. Language and woman’s place. New York: Harper and Row.<br />
Lakoff, Robin Tolmach. 1979. ‘Stylistic strategies within a grammar of style’, pp. 53–78 of<br />
Language, sex and gender: does ‘la différence’ make a difference?, edited by J. Orasanu,<br />
M. K. Slater, and L. L. Adler. New York: New York Academy of Sciences.<br />
Lakoff, Robin Tolmach. 2004. Language and woman’s place: text and commentaries. Oxford:<br />
Oxford University Press.<br />
Landor, Roland. 2012. ‘Grammatical categories and cognition across five languages: <strong>the</strong> case of<br />
grammatical gender and its potential effects on <strong>the</strong> conceptualisation of objects’. PhD <strong>the</strong>sis,<br />
Griffith University.<br />
Lang, Adrienne. 1975. The semantics of classificatory verbs in Enga and o<strong>the</strong>r Papua New<br />
Guinea languages. Canberra: Pacific Linguistics.<br />
Lang, Sabine. 1998. Men as women, women as men: changing gender in Native American<br />
cultures. Austin: University of Texas Press.<br />
Laycock, D. C. 1965. The Ndu language family (Sepik District, New Guinea). Canberra:<br />
Linguistic Circle of Canberra Publications.<br />
Leap, William. 2012. ‘Queer linguistics, sexuality and discourse analysis’, pp. 558–71 of The<br />
Routledge handbook of discourse analysis, edited by James Paul Gee and Michael Hanford.<br />
London: Routledge.<br />
Lee, J. 1987. Tiwi today: a study of language change in a contact situation. Canberra: Pacific<br />
Linguistics.<br />
Lee, Yun-seok. 2014. Classifiers in Korean. Munich: Lincom Europa.<br />
Leger, Rudolf. 1998. ‘Grammatical gender in some sou<strong>the</strong>rn Bole-Tangale languages—Kwami,<br />
Kupto, Kushi and Piya’, pp. 204–16 of Africa: society, culture, and languages. Institute of<br />
African Studies. Moscow: University Press.
References 237<br />
Lehmann, Thomas. 1993. A grammar of Modern Tamil. Pondicherry: Pondicherry Institute of<br />
Linguistics and Culture.<br />
LeMaster, Barbara. 1997. ‘Sex difference in Irish Sign Language’, pp. 67–86 of The life of<br />
language: papers in linguistics in honor of William Bright, edited by Jane H. Hill, P. J. Mistry,<br />
and Lyle Campbell. Berlin: Mouton de Gruyter.<br />
Levon, Erez. 2006. ‘Hearing “gay”: prosody, interpretation and <strong>the</strong> affective judgement of<br />
men’s speech’. American Speech 81: 56–78.<br />
Levy, Y. 1983a. ‘The acquisition of Hebrew plurals: <strong>the</strong> case of <strong>the</strong> missing gender category’.<br />
Journal of Child Language 10: 107–21.<br />
Levy, Y. 1983b. ‘It’s frogs all <strong>the</strong> way down’. Cognition 15: 75–93.<br />
Lewis, Geoffrey. 2000. Turkish grammar. Oxford: Oxford University Press.<br />
Li, Paul Jen-kuei. 1982. ‘Male and female forms of speech in <strong>the</strong> Atayalic group’. Bulletin of <strong>the</strong><br />
Institute of History and Philology 53.2: 265–304.<br />
Li, Paul Jen-kuei. 1983. ‘Types of lexical derivation of men’s speech in Mayrinax’. Bulletin of<br />
<strong>the</strong> Institute of History and Philology 54.3: 1–18.<br />
Livia, Anna. 1995. ‘“I ought to throw a Buick at you”: functional representations of butch/<br />
femme speech’, pp. 245–78 of Bucholtz and Hall 1995a.<br />
Livia, Anna. 1997. ‘The creation of coherence in coming-out stories’, pp. 287–309 of Livia and<br />
Hall 1997.<br />
Livia, Anna. 2001. Pronoun envy: literary uses of linguistic gender. Oxford: Oxford University<br />
Press.<br />
Livia, Anna and Kira Hall (eds.). 1997. Queerly phrased: language, gender, and sexuality. New<br />
York: Oxford University Press.<br />
Lock, Arnold (Arjen) Hugo. 2011. Abau grammar. Data Papers on Papua New Guinea<br />
languages 57. Ukarumpa: SIL-PNG Academic Publishers.<br />
Lopatin, V. V. 1970. ‘Word formation of nouns’ (Slovoobrazovanie imen suschestviteljnykh),<br />
pp. 46–176 of Grammatika sovremennogo russkogo literaturnogo jazyka, edited by<br />
N. Y. Shvedova. Moscow: Nauka.<br />
Loporcaro, Michele, Vincenzo Faraoni, and Francesco Gardani. 2014. ‘The third gender of Old<br />
Italian’. Diachronica 31: 1–22.<br />
Lu, Tian-Qiao. 2012. Classifiers in Kam-Tai languages: a cognitive and cultural perspective.<br />
Boca Raton, Fla.: Universal Publishers.<br />
Luraghi, Silvia. 2009. ‘The origin of <strong>the</strong> feminine gender in PIE: an old problem in a new<br />
perspective’, pp. 1–13 of Grammatical change in Indo-European languages, edited by Vit<br />
Bubenik, John Hewson, and Sarah Rose. Amsterdam: John Benjamins.<br />
Luraghi, Silvia. 2011. ‘The origin of <strong>the</strong> Proto-Indo-European gender system: typological<br />
considerations’. Folia Linguistica 45.2: 435–63.<br />
Luraghi, Silvia. 2014. ‘<strong>Gender</strong> and word formation: <strong>the</strong> PIE gender system in cross-linguistic<br />
perspective’, pp. 199–232 of Studies on <strong>the</strong> collective and feminine in Indo-European from a<br />
diachronic and typological perspective, edited by Sergio Neri and Roland Schuhmann.<br />
Leiden: Brill.<br />
Luthin, Herbert. 1991. ‘Restoring <strong>the</strong> voice of Yanan traditional narrative: prosody, performance,<br />
and presentational form’. PhD dissertation, University of California, Berkeley.
238 References<br />
Lydall, Jean. 1988. ‘<strong>Gender</strong>, number and size in Hamar’, pp. 77–99 of Cushitic and Omotic:<br />
papers from <strong>the</strong> International Symposium on Cushitic and Omotic Languages, Cologne,<br />
January 6–9, 1986, edited by Marianne Bechhaus-Gerst and Fritz Serzisko. Hamburg:<br />
Helmut Buske Verlag.<br />
Macaulay, R. K. S. 1977. Language, social class and education. Edinburgh: Edinburgh University<br />
Press.<br />
Macaulay, R. K. S. 1978. ‘The myth of female superiority in language’. Journal of Child<br />
Language 5: 353–63.<br />
McConnell-Ginet, Sally. 1979. ‘Prototypes, pronouns and persons’, pp. 63–83 of Ethnology:<br />
Boas, Sapir and Whorf revisited, edited by Madeleine Mathiot. The Hague: Mouton.<br />
McConnell-Ginet, Sally. 2003. ‘“What’s in a name?” Social labeling and gender practices’,<br />
pp. 69–97 of Holmes and Meyerhoff 2003.<br />
McConnell-Ginet, Sally. 2014. ‘<strong>Gender</strong> and its relation to sex: <strong>the</strong> myth of “natural gender”’,<br />
pp. 3–38 of Corbett 2014a.<br />
McEwan, Ian. 2011. Solar. London: Vintage Books.<br />
McGloin, Naomi Hanaoka. 1991. ‘Sex difference and sentence-final particles’, pp. 23–42 of Ide<br />
and McGloin 1991.<br />
McGregor, Donald E. and Aileen R. F. McGregor. 1982. Olo language materials. Canberra:<br />
Pacific Linguistics.<br />
McWhorter, John. 2014. The language hoax: why <strong>the</strong> world looks <strong>the</strong> same in any language.<br />
Oxford: Oxford University Press.<br />
Mahapatra, B. P. 1979. Malto: an ethnosemantic study. Manasagangotro, Mysore: Central<br />
Institute of Indian Languages.<br />
Maiden, Martin. 1998. A linguistic history of Italian. London: Routledge.<br />
Maiden, Martin and Cecilia Robustelli. 2007. A reference grammar of Modern Italian. 2nd<br />
edition. New York: MacGraw Hill.<br />
Mallinson, Graham. 1986. Rumanian. London: Croom Helm.<br />
Marcato, Gianna and Eva-Maria Thüne. 2002. ‘<strong>Gender</strong> and female visibility in Italian’,<br />
pp. 187–18 of Hellinger and Bussmann 2002, Volume 2.<br />
Maring, J. 1975. ‘Speech variation in Acoma Keresan’, pp. 471–85 of Linguistics and anthropology<br />
in honor of C. F. Voegelin. Lisse: de Ridder.<br />
Martel, Brigitte. 1981. Né homme, comment je suis devenue femme. Montréal: Québecor.<br />
Martin, Samuel E. 1975. A reference grammar of Japanese. Tokyo: Charles E. Tuttle Company.<br />
Martin, Samuel E. 1992. A reference grammar of Korean: a complete guide to <strong>the</strong> grammar and<br />
history of <strong>the</strong> Korean language. Tokyo: Tuttle Publishing.<br />
Martynyuk, A. 1990. ‘A contrastive study of male and female occupational terms in English<br />
and Russian’. Papers and Studies in Contrastive Linguistics 26: 103–10.<br />
Masica, Colin. 1991. The Indo-Aryan languages. Cambridge: Cambridge University Press.<br />
Matasović, Ranko. 2004. <strong>Gender</strong> in Indo-European. Heidelberg: Winter.<br />
Mathiot, Madeleine (ed.). 1979a. Ethnology: Boas, Sapir and Whorf revisit(ed). The Hague:<br />
Mouton.<br />
Mathiot, Madeleine. 1979b (assisted by Marjorie Roberts). ‘Sex roles as revealed through<br />
referential gender in American English’, pp. 1–47 of Ethnology: Boas, Sapir and Whorf<br />
revisited, edited by Madeleine Mathiot. The Hague: Mouton.
References 239<br />
Matras, Yaron and Jeannette Sakel (eds.). 2007. Grammatical borrowing in cross-linguistic<br />
perspective. Berlin: Mouton de Gruyter.<br />
Mat<strong>the</strong>ws, P. H. 2010. Oxford concise dictionary of linguistics. Oxford: Oxford University Press.<br />
Mat<strong>the</strong>ws, W. K. 1956. ‘The Latvian element in modern Livonian’, pp. 307–18 of Festschrift für<br />
Max Vasmer zum 70 Geburtstag, edited by Margarete Woltner und Herbert Bräuer. Berlin:<br />
Osteuropa Institut an der freien Universität Berlin.<br />
Meillet, Antoine. 1931. Essai de chronologie des langues indo-européennes. Paris: Librairie<br />
Ancienne Honoré Champion.<br />
Meillet, Antoine. 1964. ‘The feminine gender in <strong>the</strong> Indo-European languages’, p.124 of<br />
Language in culture and society: a reader in linguistics and anthropology, edited by Dell<br />
Hymes. New York: Harper and Row.<br />
Melchert, H. C. 1994. ‘The feminine gender in Anatolian’, pp. 231–44 of Früh-, Mittel-,<br />
Spätindogermanisch: Akten der IX Fachtagung der indogermanischen Gesellschaft. Wiesbaden:<br />
Dr. Ludwig Reichert Verlag.<br />
Mendes, Ronald Beline. 2012. ‘Diminutivos como marcadoes de sexo/gênero’. Revista Linguística<br />
(PGL UFRJ) 8: 113–24.<br />
Mendes, Ronald Beline. 2014a. ‘Diminutives and masculinity in Brazilian Portuguese’,<br />
pp. 117–32 of Milani 2014.<br />
Mendes, Ronald Beline. 2014b. ‘<strong>Gender</strong>ed perceptions of noun agreement in Brazilian Portuguese’.<br />
Revista Internacional de Lingüística Iberoamericana 17: 93–108.<br />
Menn, Lise, Michael P. O’Connor, and Audrey Holland. 1995. Studies in speech pathology and<br />
clinical linguistics: non-fluent aphasia in a multilingual world. Amsterdam: John Benjamins.<br />
Merlan, F. 1994. A grammar of Wardaman: a language of <strong>the</strong> Nor<strong>the</strong>rn Territory of Australia.<br />
Mouton Grammar Library 11. Berlin: Mouton de Gruyter.<br />
Merlan, F., S. P. Roberts, and A. Rumsey. 1997. ‘New Guinea “classificatory verbs” and<br />
Australian noun classification: a typological comparison’, pp. 63–103 of Nominal classification<br />
in Aboriginal Australia, edited by M. Harvey and N. Reid. Amsterdam: John Benjamins.<br />
Migge, Bettina. 2001. ‘Communicating gender in <strong>the</strong> Eastern Maroon Creole of Suriname’,<br />
pp. 84–104 of Hellinger and Bussmann 2001, Volume 1.<br />
Migliazza, Ernest C. 1978. ‘Yanomama diglossia’, pp. 561–80 of Approaches to language, edited<br />
by William McCormack and Stephen A. Wurm. The Hague: Mouton.<br />
Milani, Tommaso N. (ed.). 2014. Language and masculinities: performances, intersections,<br />
dislocations. Hoboken: Taylor and Francis.<br />
Miller, Casey and Kate Swift. 1981. The handbook of non-sexist writing for writers, editors and<br />
speakers. London: The women’s press.<br />
Milles, Karin. 2011. ‘Feminist language planning in Sweden’. Current Issues in Language<br />
Planning 12: 21–33.<br />
Mills, A. E. 1986. The acquisition of gender: a study of English and German. Berlin: Springer.<br />
Mills, Sara. 2003a. <strong>Gender</strong> and politeness. Cambridge: Cambridge University Press.<br />
Mills, Sara. 2003b. ‘Third Wave feminist linguistics and <strong>the</strong> analysis of sexism’. Discourse<br />
analysis online.<br />
Mills, Sara. 2003c. ‘Caught between sexism, anti-sexism and “political correctness”: feminist<br />
women’s negotiations with naming practices’. Discourse and Society 14.1: 87–110.<br />
Mills, Sara. 2008. Language and sexism. Cambridge: Cambridge University Press.
240 References<br />
Milroy, L. 1999. ‘Women as innovators and norm-creators: <strong>the</strong> sociolinguistics of dialect<br />
leveling in a nor<strong>the</strong>rn English city’, pp. 361–76 of Engendering communication: proceedings<br />
of <strong>the</strong> 5th Berkeley women and language conference, edited by S. Wer<strong>the</strong>im, A. C. Bailey, and<br />
M. Corston-Oliver.<br />
Miranda, Rocky V. 1975. ‘Indo-European gender: a study in semantic and syntactic change’.<br />
Journal of Indo-European Studies 3: 199–215.<br />
Mithun, Marianne. 1999. The languages of Native North America. Cambridge: Cambridge<br />
University Press.<br />
Mithun, Marianne. 2014. ‘<strong>Gender</strong> and culture’, pp. 131–60 of Corbett 2014a.<br />
Mladenova, Olga M. 2001. ‘Neuter designations of humans and norms of social interactions in<br />
<strong>the</strong> Balkans’. Anthropological Linguistics 43: 18–53.<br />
Moñino, Yves. 1977. ‘Conceptions du monde et langue d’initiation lá'bì des Gbaya-Kara’,<br />
pp. 115–47 of Langages et cultures africaines: essais d’etnolinguistique, edited by Geneviève<br />
Calame-Griaule. Paris: Maspero.<br />
Moore, Henrietta L. 1994. ‘Understanding sex and gender’, pp. 813–30 of Companion encyclopedia<br />
of anthropology: humanity, culture and social life, edited by Tim Ingold. London:<br />
Routledge.<br />
Morris, Lori. 1999. ‘The grammar of English gender’, pp. 185–203 of Between grammar and<br />
lexicon, edited by Ellen Contini-Morava and Yishai Tobin. Amsterdam: John Benjamins.<br />
Motschenbacher, Heiko. 2010. ‘Redefining genderlect’, pp. 45–59 of Language, gender and<br />
sexual identity, by Heiko Motschenbacher. Amsterdam: John Benjamins.<br />
Mous, Maarten. 2012. ‘Cushitic’, pp. 342–422 of Frajzyngier and Shay 2012.<br />
Mugglestone, Lynda. 2011. Dictionaries: a very short introduction. Oxford: Oxford University<br />
Press.<br />
Mullen, M. K. 1990. ‘Children’s classification of nature and artifact pictures into female and<br />
male categories’. Sex Roles 23: 577–87.<br />
Munro, Pamela. 1997. ‘The Garifuna gender system’, pp. 443–61 of The life of language: papers<br />
in linguistics in honor of William Bright, edited by Jane H. Hill, P. J. Mistry, and Lyle<br />
Campbell. Berlin: Mouton de Gruyter.<br />
Myers, Fred R. 1986. Pintupi country, Pintupi self: sentiment, place, and politics among Western<br />
Desert Aborigines. Washington, DC: Smithsonian Institution; Canberra: Australian Institute<br />
of Aboriginal Studies.<br />
Nakamura, Momoko. 2015. <strong>Gender</strong>, language and ideology: a genealogy of Japanese women’s<br />
language. Amsterdam: John Benjamins.<br />
Nanda, Serena. 1990. Nei<strong>the</strong>r man nor woman: <strong>the</strong> hijras of India. Belmont, Calif.: Wadsworth<br />
Publishing Co.<br />
Nekitel, Otto. 1986. ‘A sketch of nominal concord in Abu' (an Arapesh language)’, pp. 177–205<br />
of Papers in New Guinea linguistics 24, edited by D. C. Laycock et al. Canberra: Pacific<br />
Linguistics.<br />
Newman, Michael. 1997. Epicene pronouns: <strong>the</strong> linguistics of a prescriptive problem. New York:<br />
Garland.<br />
Newman, Paul. 2000. The Hausa language: an encyclopedic reference grammar. New Haven<br />
and London: Yale University Press.<br />
Nissen, Uwe Kjær. 2002. ‘Aspects of translation gender’. Linguistik online 11.2/2: 25–37.
References 241<br />
Nissen, Uwe Kjær. 2013. ‘Is Spanish becoming more gender fair? A historical perspective<br />
on <strong>the</strong> interpretation of gender-specific and gender-neutral expressions’. Linguistik online<br />
58.1/13: 99–117.<br />
Ochs, Elinor. 1992. ‘Indexing gender’, pp. 335–58 of Rethinking context: language and an<br />
interactive phenomenon, edited by Alessandro Duranti and Charles Goodwin. Cambridge:<br />
Cambridge University Press.<br />
Ogawa, Naoko and Janet S. Smith (Shibamoto). 1997. ‘The gendering of <strong>the</strong> gay male sex class<br />
in Japan: a case study based on Rasen no Sobyo’, pp. 402–15 of Livia and Hall 1997.<br />
Okamoto, Shigeko. 1995. ‘“Tasteless” Japanese: less “feminine” speech among young Japanese<br />
women’, pp. 297–325 of Bucholtz and Hall 1995a.<br />
Olza Zubiri, Jesús, Conchita Nuni de Chapi, and Juan Tube. 2001. Gramática Moja Ignaciana<br />
(Morfosintaxis). San Cristóbal: Universidad Católica del Táchira.<br />
Onishi, M. 1994. A grammar of Motuna (Bougainville, Papua New Guinea). PhD <strong>the</strong>sis,<br />
Australian National University.<br />
Onishi, M. 1996. ‘A grammatical summary of Bengali’. Materials for <strong>the</strong> Research Project<br />
‘Universals of Human Languages’, Australian National University/James Cook University.<br />
Ortner, Sherry B. 1972. ‘Is female to male as nature to culture?’, pp. 57–87 of Woman, culture<br />
and society, edited by Michelle Zimbalist Rosaldo and Louise Lamphere. Stanford, Calif.:<br />
Stanford University Press.<br />
Ortner, Sherry B. and Harriet Whitehead. 1981. ‘Introduction: accounting for sexual meanings’,<br />
pp. 1–28 of Sexual meanings: <strong>the</strong> cultural construction of gender and sexuality, edited<br />
by S. Ortner and H. Whitehead. Cambridge: Cambridge University Press.<br />
Osborne, C. R. 1974. The Tiwi language: grammar, myths and dictionary of <strong>the</strong> Tiwi language<br />
spoken on Melville and Bathurst Islands, nor<strong>the</strong>rn Australia. Canberra: Australian Institute<br />
of Aboriginal Studies.<br />
Osgood, Charles E. 1960. ‘The cross-cultural generality of visual-verbal synes<strong>the</strong>tic tendencies’.<br />
Behavioral Science 5: 146–69.<br />
Ott, W. and R. Ott. 1983. Diccionario ignaciano y castellano con apuntes gramaticales.<br />
Cochabamba: SIL.<br />
Oussikoum, Bennaser. 2013. Dictionnaire amazighe–français. Le parler des Ayt Wirra. Moyen<br />
Atlas–Maroc. Rabat : IRCAM.<br />
Owens, Jonathan. 2001. ‘Creole Arabic: <strong>the</strong> orphan of all orphans’. Anthropological Linguistics<br />
43: 348–78.<br />
Owens, Jonathan. 2014. ‘The morphologization of an Arabic Creole’. Journal of Pidgin and<br />
Creole Languages 29: 232–98.<br />
Palmer, W. D. 2009. Kokota grammar. Honolulu: University of Hawaii Press.<br />
Pandharipande, Rajeshwari V. 1997. Marathi. London: Routledge.<br />
Pankhurst, Helen. 1992. <strong>Gender</strong>, development and identity: an Ethiopian study. London: Zed Books.<br />
Pastre, Geneviève. 1997. ‘Linguistic gender play among French gays and lesbians’, pp. 369–79<br />
of Livia and Hall 1997.<br />
Pauwels, Anne. 1998. Women changing language. London: Longman.<br />
Pawley, Andrew K. 2004. ‘Using He and She for inanimate referents in English: questions of<br />
grammar and world view’, pp. 110–37 of Ethnosyntax: explorations in grammar and culture,<br />
edited by N. J. Enfield. Oxford: Oxford University Press.
242 References<br />
Payne, Judith. 1989. Lecciones para el aprendizaje del idioma Asheninca. Serie Lingüística<br />
Peruana 28, Pucalpa, Peru.<br />
Pearce, Mary. 2013. ‘Variation with gender in <strong>the</strong> tonal speech varieties of Kera (Chadic)’,<br />
pp. 79–93 of Atanga et al. 2013.<br />
Penelope, Julia. 1990. Speaking freely. New York: Pergamon Press.<br />
Pensalfini, R. 2003. A grammar of Jingulu, an Aboriginal language of <strong>the</strong> Nor<strong>the</strong>rn Territory.<br />
Canberra: Pacific Linguistics.<br />
Pérez-Pereira, M. 1991. ‘The acquisition of gender: what Spanish children tell us’. Journal of<br />
Child Language 18: 571–90.<br />
Perlak, D. and G. Jarema. 2003. ‘The recognition of gender-marked nouns and verbs in Polishspeaking<br />
aphasic patients’. PubMed 39: 383–403.<br />
Pet, W. J. A. 1987. Lokono Dian, <strong>the</strong> Arawak language of Suriname: a grammatical sketch of its<br />
grammatical structure and lexicon. PhD <strong>the</strong>sis, Cornell University.<br />
Peters, Pam. 1995. The Cambridge Australian style guide. Cambridge: Cambridge University<br />
Press.<br />
Pfarr, S. J. 1984. ‘Status conflict: <strong>the</strong> rebellion of <strong>the</strong> tea pourers’, pp. 214–40 of Conflict in<br />
Japan, edited by E. S. Kraus, T. P. Rohlen, and P. G. Steinhoff. Honolulu: University of<br />
Hawaii Press.<br />
Philips, S. U. 1980. ‘Sex differences in language’. Annual Review of Anthropology 9: 523–44.<br />
Philips, S. U., S. Steele, and C. Tanz (eds.). 1987. Language, gender, and sex in comparative<br />
perspective. Cambridge: Cambridge University Press.<br />
Plank, F. and W. Schellinger. 1997. ‘The uneven distribution of genders over numbers:<br />
Greenberg Nos. 37 and 45’. Linguistic Typology 1: 53–101.<br />
Podesva, R., S. Roberts, and K. Campbell-Kibler. 2002. ‘Sharing resources and indexing<br />
meanings in <strong>the</strong> production of gay styles’, pp. 175–90 of Language and sexuality: contesting<br />
meaning in <strong>the</strong>ory and practice, edited by K. Campbell-Kibler, R. Podesva, S. Roberts, and<br />
A. Wong. Stanford, Calif.: CSLI.<br />
Pope,M.K.1934. FromLatintoModernFrench. Manchester: Manchester University<br />
Press.<br />
Popova, M. I. 1958. ‘Grammatic᷈eskie elementy jazyka v rec᷈i detej preddos᷈kol'nogo vozrasta’<br />
(‘Grammatical elements of language in <strong>the</strong> speech of children of pre-school age’). Voprosy<br />
psixologii 4.3: 106–17. (Russian version of Popova 1973.)<br />
Popova, M. I. 1973. ‘Grammatical elements of language in <strong>the</strong> speech of pre-school children’,<br />
pp. 269–80 of Studies of child language development, edited by C. A. Ferguson and<br />
D. I. Slobin. New York: Holt, Rinehart and Winston.<br />
Poser, William J. 2005. ‘Noun classification in Carrier’. Anthropological Linguistics 47: 143–68.<br />
Posner, Rebecca. 1966. The Romance languages: a linguistic introduction. New York: Anchor<br />
books.<br />
Pountain, C. J. 2005. ‘<strong>Gender</strong> without sex: <strong>the</strong> semantic exploitation of <strong>the</strong> masculine/feminine<br />
opposition in <strong>the</strong> history of Spanish’, pp. 329–48 of Studies on Ibero-Romance linguistics<br />
dedicated to Ralph Penny, edited by R. Wright and R. Ricketts. Newark, Del.: Juan de la<br />
Cuesta.<br />
Priestly, Tom M. S. 1983. ‘On “drift” in Indo-European gender systems’. Journal of Indo-<br />
European Studies 11.3/4: 339–63.
References 243<br />
Puşcariu, Sextil. 1943. Die rumänische Sprache. Ihr Wesen und ihre volkliche Prägung. Leipzig:<br />
Otto Harrassowitz.<br />
Quinn, Conor. 2001. ‘A preliminary survey of animacy categories in Penobscot’, pp. 395–426<br />
of Actes du trente-deuxième congrès des Algonquinistes, edited by John D. Nichols. Winnipeg:<br />
University of Manitoba.<br />
Quintero, Carolyn. 2004. Osage grammar. Lincoln and London: University of Nebraska Press.<br />
Quirk, Randolph and C. L. Wrenn. 1957. An Old English grammar. 2nd edition. London:<br />
Methuen.<br />
Raffelsiefen, Renate. 2011. ‘Reference to body parts in attributive adjective constructions’.<br />
Paper presented at a Conference ‘Body parts’, University of Warsaw.<br />
Raga, Amanuel and Hirut Woldemariam. 2011. ‘<strong>Gender</strong> bias as manifested in <strong>the</strong> grammar of<br />
Afan Oromo’. Journal of Language and Culture 2: 162–73.<br />
Rankin, Robert L. 1976. ‘Latin kw, gw > Rumanian p, b: an explanation’, pp. 14–26 of Current<br />
studies in romance linguistics. Washington, DC: Georgetown University Press.<br />
Rapold, Christian J. 2006. ‘Towards a grammar of Benchnon’. PhD <strong>the</strong>sis, University of Leiden.<br />
Rat, Joseph Numa. 1898. ‘The Carib language as now spoken in Dominica’. Journal of <strong>the</strong><br />
Anthropological Institute of Great Britain and Ireland 27: 293–315.<br />
Reid, Nicholas. 1997. ‘Class and classifier in Ngan.'gityemerri’, pp. 165–228 of Harvey and<br />
Reid 1997.<br />
Repp, Hanna. 1996. ‘Die Genustausch in der Kindersprache des Arabischen und das Komplementärwesen<br />
“Qarîna”—Konzepte des Volksglaubens zur Erklärung eines sprachlichen<br />
Phänomens’. Zeitschrift für Arabische Linguistik 32: 25–56.<br />
Reynolds, Katsue Akiba. 1985. ‘Female speakers of Japanese’. Feminist Issues 5: 13–46.<br />
Reynolds, Katsue Akiba. 1991. ‘Female speakers of Japanese in transition’, pp. 129–46 of Ide<br />
and McGloin 1991.<br />
Reynolds, Katsue Akiba. 2011. ‘Female speakers of Japanese in transition’, pp. 291–9 of Coates<br />
and Pichler 2011.<br />
Ribeiro, Eduardo Rivail. 2009. ‘Macro-Jê’, pp. 665–9 of Concise encyclopedia of languages of <strong>the</strong><br />
world, edited by Keith Brown and Sarah Ogilvie. Oxford: Elsevier.<br />
Ribeiro, Eduardo Rivail. 2012. ‘A grammar of Karajá’. PhD dissertation, University of Chicago.<br />
Richards, Cara B. 1974. ‘Among <strong>the</strong> liberated’, pp. 401–19 of Many sisters: women in crosscultural<br />
perspective, edited by Carolyn J. Matthiasson. New York: Free Press.<br />
Rodina, Yulia. 2014. ‘Variation in <strong>the</strong> input: child and caregiver in <strong>the</strong> acquisition of grammatical<br />
gender in Russian’. Language Sciences 43: 116–32.<br />
Rodrigues, Aryon Dall’Igna. 1999. ‘Macro-Jê’, pp. 165–206 of The Amazonian languages,<br />
edited by R. M. W. Dixon and A. Y. Aikhenvald. Cambridge: Cambridge University Press.<br />
Romaine, Suzanne. 1999. Communicating gender. London: Lawrence Erlbaum Associates.<br />
Romani Miranda, Maggie M. 2004. ‘Toponimia en el Gran Pajonal con especial atención a los<br />
topónimos de afiliación Ashaninka’. PhD dissertation, Universidad Nacional Mayor de San<br />
Marcos, Lima.<br />
Roscoe, Paul B. 2001. ‘“Strength” and “sexuality”: sexual avoidance and masculinity in New<br />
Guinea and Amazonia’, pp. 279–308 of <strong>Gender</strong> in Amazonia and Melanesia: an exploration<br />
of <strong>the</strong> comparative method, edited by Thomas A. Gregor and Donald Tuzin. Berkeley and<br />
Los Angeles: University of California Press.
244 References<br />
Roscoe, Will. n.d. ‘Berdaches’. Encyclopedia of <strong>the</strong> Great Plains. .<br />
Rose, Françoise. 2013. ‘Los generolectos del mojeño’. LIAMES 13: 115–34.<br />
Rosenhouse, J. and N. Dbayyat. 2006. ‘<strong>Gender</strong> switch in female speech of an urbanized Arabic<br />
dialect in Israel’. Anthropological Linguistics 48: 169–86.<br />
Ross, Kristiina. 1996. ‘Naine eesti keele grammatikas’ [The woman in Estonian grammar].<br />
Vikerkaar 11/12: 102–6.<br />
Ross, Kristiina. 2012. ‘Sõbranna—sõbralaadne toode. Kultuurlisi representatsioone meilt ja<br />
mujalt’. [Female friend—a friend resembling product: cultural representations from us and<br />
from elsewhere]. Vikerkaar 1.2: 166–73.<br />
Rothstein, Robert A. 1973. ‘Sex, gender, and <strong>the</strong> October Revolution’, pp. 460–6 of A Festschrift for<br />
Morris Halle, edited by S. R. Anderson and P. Kiparsky. New York: Holt, Rinehart and Winston.<br />
Rubino, C. R. G. 1997. A reference grammar of Ilocano. PhD <strong>the</strong>sis, University of California<br />
(Santa Barbara).<br />
Rudes, B. and B. Healey. 1979. ‘Is she for real? The concepts of femaleness and maleness in <strong>the</strong><br />
gay world’, pp. 49–61 of Mathiot 1979a.<br />
Rushforth, S. 1991. ‘Uses of Bearlake and Mescalero (Athapaskan) classificatory verbs’. International<br />
Journal of American Linguistics 57: 251–66.<br />
Sa'ar, Amalia. 2007. ‘Masculine talk: on <strong>the</strong> subconscious use of masculine linguistic forms<br />
among Hebrew- and Arabic-speaking women in Israel’. Signs 32: 405–29.<br />
Sadiqi, F. 2003. Language and gender in Morocco. Leiden: Brill.<br />
Sahoo, Kalyanamalini. 2003. ‘Linguistic and socio-culural implications of gendered structures<br />
in Oriya’, pp. 239–58 of Hellinger and Bussmann 2003, Volume 3.<br />
Salzmann, Zdenek. 2004. Language, culture and society: an introduction to linguistic anthropology.<br />
Boulder, Colo.: Westview Press.<br />
Sandström, Caroline. 2000. ‘The changing system of grammatical gender in <strong>the</strong> Swedish<br />
dialects of Nyland, Finland’, pp. 793–806 of Unterbeck and Rissanen 2000.<br />
Sapir, Edward. 1929a. ‘Male and female forms of speech in Yana’, pp. 79–85 of Donum<br />
Natalicium Schrijnen, edited by St. W. J. Teeuwen. Nijmegen-Utrecht: N. V. Dekker and<br />
van de Vegt.<br />
Sapir, Edward. 1929b. ‘Status of linguistics as a science’. Language 5: 207–14.<br />
Sarvasy, Hannah. 2016. ‘Sexless babies, sexed grandparents. Nungon gendered person forms’.<br />
International Journal of Language and Culture.<br />
Sasse, Hans-Jürgen. 1974. ‘Notes on <strong>the</strong> structure of Galab’. BSOAS 37: 407–38.<br />
Sasse, Hans-Jürgen. 1985. ‘Sprachkontakt und Sprachwandel: die Gräzisierung der albanischen<br />
Mundarten Griechenlands’. Papiere zur Linguistik 32: 37–95.<br />
Saville-Troike, M. 1988. ‘A note on men’s and women’s speech in Koasati’. International<br />
Journal of American Linguistics 54: 241–2.<br />
Sayers, Dorothy L. 1995. Lord Peter. New York: Harper paperback.<br />
Schane, S. A. 1970. ‘Phonological and morphological markedness’, pp. 286–94 of Progress in<br />
linguistics, edited by M. Bierwisch and K. E. Heidolph. The Hague: Mouton.<br />
Schapper, Antoinette. 2010a. ‘Bunaq, a Papuan language of central Timor’. PhD <strong>the</strong>sis,<br />
Australian National University.
References 245<br />
Schapper, Antoinette. 2010b. ‘Neuter gender in Eastern Indonesia’. Oceanic Linguistics 49:<br />
407–35.<br />
Schaub, W. 1985. Babungo. Croom Helm Descriptive Grammars. London: Croom Helm.<br />
Schenker, Alexander M. 1993. ‘Proto-Slavonic’, pp. 60–121 of The Slavonic languages, edited<br />
by Bernard Comrie and Greville G. Corbett. London: Routledge.<br />
Schieffelin, Bambi B. 1987. ‘Do different worlds mean different words? An example from<br />
Papua New Guinea’, pp. 249–60 of Philips, Steele, and Tanz 1987.<br />
Schieffelin, Edward L. 1977. The sorrow of <strong>the</strong> lonely and <strong>the</strong> burning of <strong>the</strong> dancers. St Lucia:<br />
University of Queensland Press.<br />
Schipper, Mineke. 2003. Never marry a woman with big feet: women in proverbs from around<br />
<strong>the</strong> world. New Haven and London: Yale University Press.<br />
Schmidt, A. 1985. Young People’s Dyirbal: an example of language death from Australia.<br />
Cambridge: Cambridge University Press.<br />
Senft, Gunter. 1986. Kilivila: <strong>the</strong> language of Trobriand Islanders. Berlin: Mouton de Gruyter.<br />
Senft, Gunter. 1996. Classificatory particles in Kilivila. New York: Oxford University Press.<br />
Sera, M., C. Berge, and J. del Castillo. 1994. ‘Grammatical and conceptual forces in <strong>the</strong><br />
attribution of gender by English and Spanish speakers’. Cognitive Development 9: 261–92.<br />
Seyoum, Mulugeta. 2008. A grammar of Dime. Amsterdam: LOT.<br />
Shepard Jr., G. 1997. ‘Noun classification and ethnozoological classification in Machiguenga,<br />
an Arawakan language of <strong>the</strong> Peruvian Amazon’. Journal of Amazonian Languages 1: 29–57.<br />
Sherzer, Joel. 1987. ‘The diversity of voices: men’s and women’s speech in ethnographic<br />
perspective’, pp. 95–120 of Philips, Steele, and Tanz 1987.<br />
Shibatani, Masayoshi. 1990. The languages of Japan. Cambridge: Cambridge University Press.<br />
Silverman, Eric K. 2001. Masculinity, mo<strong>the</strong>rhood and mockery: psychoanalyzing culture and<br />
<strong>the</strong> Iatmul Naven Rite in New Guinea. Ann Arbor: The University of Michigan Press.<br />
Silverstein, Michael. 1972a. ‘Chinook Jargon: language contact and <strong>the</strong> problem of multi-level<br />
generative systems I’. Language 48: 378–406.<br />
Silverstein, Michael. 1972b. ‘Chinook Jargon: language contact and <strong>the</strong> problem of multi-level<br />
generative systems II’. Language 48: 596–625.<br />
Silverstein, Michael. 1985. ‘Language and <strong>the</strong> culture of gender: at <strong>the</strong> intersection of structure,<br />
usage and ideology’, pp. 219–59 of Semiotic meditation: sociocultural and psychological<br />
perspectives, edited by Elizabeth Mertz and Richard J. Parmentier. Orlando, Fla.: Academic<br />
Press.<br />
Silverstein, Michael. 2015. ‘Presupposing demographic sex, entailing sociocultural gender’.<br />
Paper presented at <strong>the</strong> symposium ‘Indexing gender revisited: on <strong>the</strong> non-referential aspects<br />
of gendering’, 14th International Pragmatics Conference, International Pragmatics Association,<br />
26–31 July, Antwerp, Belgium.<br />
Sims, C. and H. Valiquette. 1990. ‘More on male and female speech in (Acoma and Laguna)<br />
Keresan’. International Journal of American Linguistics 56: 162–6.<br />
Singer, Ruth. 2010. ‘Creativity in <strong>the</strong> use of gender agreement in Mawng: how <strong>the</strong> discourse<br />
functions of a gender system can approach those of a classifier system’. Studies in Language<br />
34: 382–416.<br />
Sini, M. 2011. ‘Emergence of patrilineal system and its impact on North Malabar—a study of<br />
Chrakkal village in Kannur district’. PhD dissertation, Kannur University.
246 References<br />
Slobin, Dan I. 1996. ‘From “thought and language” to “thinking for speaking”’, pp. 70–96 of<br />
Rethinking linguistic relativity, edited by John J. Gumperz and Stephen G. Levinson. Cambridge:<br />
Cambridge University Press.<br />
Smith-Stark, S. 1974. ‘The plurality split’. Papers from <strong>the</strong> Annual Regional Meeting of <strong>the</strong><br />
Chicago Linguistic Society 10: 657–71.<br />
Sneddon, James N. 1996. Indonesian reference grammar. Sydney: Allen and Unwin.<br />
Sohn, Ho-Min. 1994. Korean. London: Routledge.<br />
Sontag, Susan. 1973. ‘The third world of women’. Partisan Review 40.2: 181–206.<br />
Sou<strong>the</strong>rland, Ronald H. and Francis Katamba. 1996. ‘Language in social contexts’, pp. 540–90<br />
of Contemporary linguistics: an introduction, edited by William O’Grady, Michael Dobrovolsky,<br />
and Francis Katamba. London: Longman.<br />
Souza, L. 2011. ‘Fala masculina e fala feminina no kadiwéu.’. Anais do VII Congresso Internacional<br />
da Abralin 2488–96.<br />
Speece, R. n.d. Angave grammar. Ukarumpa: Summer Institute of Linguistics.<br />
Spender, Dale. 1980. Man made language. London: Routledge and Kegan Paul.<br />
Spitulnik, D. 1989. ‘Levels of semantic restructuring in Bantu noun classification’, pp. 207–20<br />
of Current approaches to African linguistics, Volume 5, edited by P. Newman and R. D. Botne.<br />
Dordrecht: Foris Publications.<br />
Stebbins, Tonya N. 2005. ‘Nominal classification in Mali’. Anthropological Linguistics 47:<br />
77–131.<br />
Steinhauer, H. 1986. ‘Number in Biak: counterevidence to two alleged language universals a<br />
summary’, pp. 171–3 of FOCAL I: papers from <strong>the</strong> Fourth International Conference on<br />
Austronesian Linguistics, edited by P. Geraghty, L. Carrington, and S. A. Wurm. Canberra:<br />
Pacific Linguistics.<br />
Storch, Anne. 2006. ‘<strong>How</strong> long do linguistic areas last? Western Nilotic grammars in contact’,<br />
pp. 94–113 of Grammars in contact: a cross-linguistic typology, edited by A. Y. Aikhenvald<br />
and R. M. W. Dixon. Oxford: Oxford University Press.<br />
Storch, Anne. 2011. Secret manipulations. New York: Oxford University Press.<br />
Storch, Anne. 2013. ‘Possession in Hone’, pp. 208–23 of Aikhenvald and Dixon 2013.<br />
Storch, Anne. 2014. A grammar of Luwo: an anthropological approach. Amsterdam: John<br />
Benjamins.<br />
Storjohann, Petra. 2004. ‘frau auf dem linguistischen Prüfstand: eine korpusgestützte Gebrauchsanalyse<br />
feministischer Indefinitpronomen’. German Life and Letters 57.3: 309–26.<br />
Stotko, Elaine M. and Margaret Troyer. 2007. ‘A new gender-neutral pronoun in Baltimore,<br />
Maryland: a preliminary study’. American Speech 82: 262–79.<br />
Stradelli, Ermano. 1929. Vocabulários da Língua Geral Portuguez-Nheêngatú e Nheêngatú-<br />
Portuguez. Rio de Janeiro: Revista do Instituto Historico-Geográfico.<br />
Straus, Anne Terry and Robert Brightman. 1982. ‘The implacable raspberry’. Papers in<br />
Linguistics 15: 97–137.<br />
Studzińska, Joanna. 2012. ‘Translating personification across gendered languages’, pp. 259–68<br />
of Cognitive processes in language, edited by K. Kosecki and J. Badio. Bern: Peter Lang.<br />
Sumbuk, Kenneth Memson. 1999. ‘Morphosyntax of Sare’. PhD <strong>the</strong>sis, University of Waikato.<br />
Svartengren, T. Hilding. 1927. ‘The feminine gender for inanimate things in Anglo-American’.<br />
American Speech 3: 83–113.
References 247<br />
Svartengren, T. Hilding. 1954. ‘The use of feminine gender for inanimate things’. Moderna<br />
Sprak 48: 261–92.<br />
Tadmor, Uri. 2007. ‘Grammatical borrowing in Indonesian’, pp. 301–28 of Matras and Sakel<br />
2007.<br />
Taïfi, Miloud. 1991. Dictionnaire tamazight–français. Parlers du Maroc central. Paris:<br />
L’Harmattan.<br />
Talbot, M. 2010. Language and gender. Cambridge: Polity Press.<br />
Tannen, Deborah. 1990. You just don’t understand: women and men in conversation. New<br />
York: William Morrow.<br />
Taylor, Allan R. 1982. ‘“Male” and “female” speech in Gros Ventre’. Anthropological Linguistics<br />
24: 301–7.<br />
Taylor, Douglas M. 1951. ‘Sex gender in Central American Carib’. International Journal of<br />
American Linguistics 17: 102–4.<br />
Taylor, Douglas M. 1959. ‘La catégorie du genre en caraïbe insulaire: un cas d’interférence<br />
linguistique?’. Bulletin de la Société de Linguistique de Paris 54: 201–7.<br />
Taylor, Douglas M. 1977. Languages of <strong>the</strong> West Indies. Baltimore: Johns Hopkins University<br />
Press.<br />
Teferra, Anbessa. 1987. ‘Balishsha: women’s speech among <strong>the</strong> Sidaama’. Journal of Ethiopian<br />
Studies 20: 44–59.<br />
Teso, Elena and Liz Crolley. 2013. ‘<strong>Gender</strong>-based linguistic reform in international organizations’.<br />
Language Policy 12: 139–58.<br />
Thomason, S. G. and T. Kaufman. 1988. Language contact, creolization and genetic linguistics.<br />
Berkeley and Los Angeles: University of California Press.<br />
Thompson, L. C. 1987. A Vietnamese grammar. Seattle: University of Washington Press.<br />
Tobin, Yishai. 2001. ‘<strong>Gender</strong> switch in Modern Hebrew’, pp. 177–98 of Hellinger and Bussmann<br />
2001, Volume 1.<br />
Trask, L. 2000. The dictionary of historical and comparative linguistics. Edinburgh: Edinburgh<br />
University Press.<br />
Trechter, Sara. 1995. ‘The pragmatic functions of gender deixis in Lakhota’. PhD dissertation,<br />
University of Kansas.<br />
Trechter, Sara. 1999. ‘Contextualizing <strong>the</strong> exotic few: gender dichotomies in Lakhota’,<br />
pp. 101–22 of Bucholtz, Liang, and Sutton 1999.<br />
Trechter, Sara. 2009. ‘Native languages of North America, variation in’, pp. 753–60 of Concise<br />
encyclopedia of languages of <strong>the</strong> world, edited by Keith Brown and Sarah Ogilvie. Oxford:<br />
Elsevier.<br />
Treis, Yvonne. 2005. ‘Avoiding <strong>the</strong>ir names, avoiding <strong>the</strong>ir eyes: how Kambaata women respect<br />
<strong>the</strong>ir in-laws’. Anthropological Linguistics 47: 292–320.<br />
Trudgill, Peter. 1971. ‘The social differentiation of English in Norwich’. PhD dissertation,<br />
University of Edinburgh.<br />
Trudgill, Peter. 1974. The social differentiation of English in Norwich. Cambridge: Cambridge<br />
University Press.<br />
Trudgill, Peter. 1975. ‘Sex, covert prestige and linguistic change in <strong>the</strong> urban British English of<br />
Norwich’, pp. 88–104 of Language and sex: difference and dominance, edited by B. Thorne<br />
and N. Henley. Rowley, Mass.: Newbury House.
248 References<br />
Trudgill, Peter. 1977. ‘Creolization in reverse: reduction and simplification in <strong>the</strong> Albanian<br />
dialects of Greece’. Transactions of <strong>the</strong> Philological Society 77: 32–50.<br />
Trudgill, Peter. 1998. ‘Sex and covert prestige’, pp. 20–6 of Language and gender: a reader. 2nd<br />
edition, edited by Jennifer Coates and Pia Picher. London: Wiley-Blackwell.<br />
Trudgill, Peter. 2007. ‘On grammatical gender as an arbitrary and redundant category’,<br />
pp. 24–36 of History of Linguistics 2005: selected papers from <strong>the</strong> Third International<br />
Conference on <strong>the</strong> History of Language Sciences, Urbana-Champaign, Illinois, 1–5 September<br />
2005, edited by Douglas Kibbee. Amsterdam: John Benjamins.<br />
Tsegaye, Mulugeta T., Maarten Mous, and Niels O. Schiller. 2014. ‘Plural as a value in <strong>the</strong><br />
investigation of grammatical gender’, pp. 161–90 of Corbett 2014a.<br />
Unterbeck, Barbara and Matti Rissanen (eds.). 2000. <strong>Gender</strong> in grammar and cognition,<br />
I. Approaches to gender. II. Manifestations of gender. Berlin: Mouton de Gruyter.<br />
Upton, Rebecca. 2012. ‘<strong>Gender</strong>’. Oxford bibliographies online. Oxford: Oxford University Press.<br />
Uyeno, T. 1971. ‘A study of Japanese modality: a performative analysis of sentence particles’.<br />
PhD dissertation, University of Michigan.<br />
Vajda, Edward J. 2004. Ket. Munich: Lincom Europa.<br />
Valentine, J. Randolph. 2001. Nishnaabemwin reference grammar. Toronto: University of<br />
Toronto Press.<br />
Vallejos, Rosa Yopán. 2010. ‘A grammar of Kokama-Kokamilla’. PhD dissertation, University<br />
of Oregon.<br />
Vallejos, Rosa Yopán. 2015. ‘La indexalidad de género en kukama-kukamiria desde una<br />
perspective tipológica’, to appear in Estudios de lenguas ameríndias 3. Escenarios de diversidad<br />
lingüística, edited by Zarina Estrada Fernández, Ana V. Fennándex Garay, and Albert<br />
Alvarez Gonzalez. Hermosillo, Son.: Editorial UNISON.<br />
Vallejos, Rosa Yopán. Forthcoming. ‘Structural outcomes of obsolescence and revitalization:<br />
documenting variation among <strong>the</strong> Kukama-Kukamiria’, to appear in Language documentation<br />
and revitalization in Latin America, edited by Gabriela Perez-Baez, Chris Rogers, and<br />
Jorge Rosés-Labrada. Berlin: De Gruyter Mouton.<br />
van Baarle, Peter and M. A. Sabajo. 1997. Manuel de la langue Arawak. Paris: Éditions du Saule.<br />
van den Berg, René. 2015. ‘The loss of clusivity and <strong>the</strong> rise of gender in West Oceanic<br />
pronominals’. Language and Linguistics in Melanesia 33: 10–47.<br />
Vasvári, Louise O. 2011. ‘Grammatical gender trouble and Hungarian gender[lessness]: Part<br />
I. Comparative linguistic gender.’ AHEA E-Journal 4. .<br />
Vasvári, Louise O. 2015. ‘<strong>Gender</strong> trouble in a grammatically genderless language: Hungarian’,<br />
pp. 203–25 of Hellinger and Motschenbacker 2015.<br />
Villas Bôas, Cláudio and Orlando Villas Bôas. 1970. Xingu: Os índios, seus mitos. Rio de<br />
Janeiro: Zahar.<br />
Vinogradov, V. V. 1947. Russkij Jazyk. (The Russian language.) Moscow: Uchpedgiz.<br />
Wales, Katie. 1996. Personal pronouns in present-day English. Cambridge: Cambridge University<br />
Press.<br />
Walsh, M. 1976. The Murinypata language of north-west Australia. PhD <strong>the</strong>sis, Australian<br />
National University.
References 249<br />
Walsh, M. 1997. ‘Nominal classification and generics in Murrinhpatha’, pp. 255–92 of Harvey<br />
and Reid 1997.<br />
Walters, Margaret. 2005. Feminism: a very short introduction. Oxford: Oxford University Press.<br />
Watkins, L. 1984. A grammar of Kiowa. Lincoln: University of Nebraska Press.<br />
Weber, Doris. 2000. ‘On <strong>the</strong> function of gender’, pp. 495–509 of <strong>Gender</strong> in grammar and<br />
cognition, I. Approaches to gender, edited by Barbara Unterbeck and Matti Rissanen. Berlin:<br />
Mouton de Gruyter.<br />
Wegener, Claudia. 2012. Grammar of Savosavo. Berlin: Mouton de Gruyter.<br />
Welmers, Wm. E. 1973. African language structures. Berkeley: University of California Press.<br />
Werner, Heinrich. 1994. Das Klassensystem in den Jenissei-Sprachen. Wiesbaden: Harrassowitz.<br />
Werner, Heinrich. 1997. Die ketische Sprache. Wiesbaden: Harrassowitz.<br />
Wertz, C. A. 1977. ‘The number of genders in Polish’. Canadian Slavonic Papers 19: 50–63.<br />
Westermann, D. 1947. ‘Pluralbildung und Nominalklassen in einigen afrikanischen Sprachen’,<br />
Abhandlungen der Deutschen Akademie der Wissenschaften zu Berlin 1945/46. Phil.-hist.<br />
Klasse. Nr 1. Berlin.<br />
Whitehead, Harriet. 1981. ‘The bow and <strong>the</strong> burden strap: a new look at institutionalized<br />
homosexuality in native North America’, pp. 80–115 of Sexual meanings: <strong>the</strong> cultural<br />
construction of gender and sexuality, edited by Sherry B. Ortner and Harriet Whitehead.<br />
Cambridge: Cambridge University Press.<br />
Wiedemann, Ferdinand Johann. 1875. Grammatik der ehstnischen Sprache, zunächst wie sie in<br />
Mittelehstland gesprochen wird, mit Berücksichtigung der anderen Dialekte. St-Petersbourg.<br />
Wikan, Unni. 1977. ‘Man becomes woman: transsexualism in Oman as a key to gender roles’.<br />
Man 12: 304–19.<br />
Wikan, Unni. 1978. ‘The Omani Xanith: a third gender role?’ Man 13: 473–5.<br />
Williams, Nancy M. 1986. The Yolngu and <strong>the</strong>ir land: a system of land tenure and <strong>the</strong> fight for<br />
its recognition. Canberra: Australian Institute of Aboriginal Studies.<br />
Winford, Donald. 2005. ‘Contact-induced changes: classification and processes’. Diachronica<br />
22: 373–427.<br />
Woidich, Manfred. 2005. ‘Kindersprache in ilBas᷈andi. Ein text aus der Oase Dakhla’,<br />
pp. 411–31 of Alltagsleben und materielle Kultur in der arabischen Sprache und Literatur.<br />
Festschrift für Heinz Grotzfeld. Abhandlungen für die Kunde des Morgenlandes, Band LVI.<br />
Wolfart, H. Christoph. 1973. Plains Cree: a grammatical study. Transactions of <strong>the</strong> American<br />
Philosophical Society, NS 63–5. Philadelphia: The American Philological Society.<br />
Wołk, Ewa. 2009. ‘Positive and negative emotions encoded in Amharic forms of address’,<br />
pp. 128–36 of Coding and rituals of emotions in Asian and African cultures, edited by Nina<br />
Pawlak. Warsaw: Dom Wydawniczy ELIPSA.<br />
Yadav, Ramawatar. 1996. A reference grammar of Maithili. Berlin: Mouton de Gruyter.<br />
Yaguello, Marina. 1979. Les mots et les femmes. Paris: Payot.<br />
Yaguello, Marina. 1991. Petits faits de langue. Paris: Seuil.<br />
Yigezu, Moges. 1998. ‘Women in society and female speech among <strong>the</strong> Suri of Southwestern<br />
Ethiopia’, pp. 83–101 of Surmic languages and cultures, edited by Gerrit J. Dimmendaal and<br />
Marco Last. Cologne: Rüdiger Köppe Verlag.<br />
Yokoyama, Olga T. 1999. ‘Russian genderlects and referential expressions’. Language in Society<br />
28: 401–29.
250 References<br />
Zavala, Roberto. 1992. El Kanjobal de San Miguel Acatán. Volume 6. Universidad nacional<br />
autónoma de México.<br />
Zavala, Roberto. 2000. ‘Multiple classifier systems in Akatek (Mayan)’, pp. 114–46 of Systems<br />
of nominal classification, edited by Bunter Senft. Cambridge: Cambridge University Press.<br />
Zemskaya, Elena A., Margarita V. Kitajgorodskaja, and Nona N. Rozanova. 1993. ‘Osobennosti<br />
muz᷈skoj i z᷈enskoj rec᷈i’ (Features of male and female speech), pp. 90–157 of Russkij jazyk v<br />
ego funkcionirovanii: Kommunikativno-pragmatic᷈eskij aspekt, edited by Elena A. Zemskaya<br />
and Dmitri N. Shmelev. Moscow: Nauka.<br />
Zgusta, Ladislav. 1971. Manual of lexicography. The Hague: Mouton.<br />
Zhang, Sihong. 2014. ‘A grammar of Ersu’. PhD <strong>the</strong>sis, James Cook University.<br />
Zimman, Lal and Kira Hall. Forthcoming. ‘Language, embodiments, and <strong>the</strong> “third sex”’,to<br />
appear in Language and identities, edited by Dominic Watt and Carmen Llamas. Edinburgh:<br />
Edinburgh University Press.<br />
Zubin, David A. and Klaus-Michael Köpcke. 1984. ‘Affect classification in <strong>the</strong> German gender<br />
system’. Lingua 63: 41–86.<br />
Zubin, David A. and Klaus-Michael Köpcke. 1986. ‘<strong>Gender</strong> and folk taxonomy: <strong>the</strong> indexical<br />
relation between grammatical and lexical categorization’, pp. 139–80 of Noun classes and<br />
categorisation, edited by C. G. Craig. Amsterdam: John Benjamins.<br />
Zwicky, Arnold M. 1997. ‘Two lavender issues for linguists’, pp. 21–34 of Livia and Hall 1997.
Index of languages, linguistic families,<br />
and peoples<br />
For each language, <strong>the</strong> linguistic family or area to which it belongs is noted. Established<br />
subgroups have been added for a number of well-researched families.<br />
Abau (isolate, Papuan area) 121<br />
Abu Arapesh (Arapesh, Papuan area) 23, 31<br />
Acoma Keresan (isolate) 168, see also Keresan<br />
Adamawa (Adamawa-Ubangi) 87<br />
Afan Oromo, see Oromo<br />
African languages 13, 31<br />
Afroasiatic languages 13, 38–9, 44–5, 49–50<br />
Aït Mguild (Berber, Afroasiatic) 103, 109<br />
Aït Wirra (Berber, Afroasiatic) 103, 109<br />
Akatek (Mayan) 78, 95<br />
Alaaba (Cushitic, Afroasiatic) 183<br />
Alamblak (Sepik Hill, Papuan area) 35, 44,<br />
62, 120<br />
Albanian (Indo-European) 23, 92<br />
Albanian, dialect of Mandres 87, 97<br />
Algonquian languages 8, 13, 20–2, 31, 55,<br />
67, 118, 137<br />
Alor-Pantar, Austronesian languages of 97<br />
Amazonian languages 53, 162<br />
American English (Germanic, Indo-European)<br />
109–10, 188, 129–30, 154, 208, see also<br />
English<br />
Amharic (Semitic, Afroasiatic) 44–5, 51, 100,<br />
104–5, 108–9, 117, 128, 146, 186<br />
Amuesha (Arawak) 87<br />
Anatolian languages (Indo-European) 81<br />
Ancient Greek (Indo-European) 13, 89<br />
Angan languages 51<br />
Angave (Angan) 51<br />
Arabic (Semitic, Afroasiatic) 38, 46, 105,<br />
108–9, 117, 164, 203, 210, 216, see also<br />
Egyptian Arabic, Moroccan Arabic,<br />
Omani Arabic, Palestinian Arabic<br />
Arabic Creole (Creole) 75<br />
Arapaho (Algonquian) 165<br />
Arapesh languages (Papuan area) 23, 31<br />
Arawá languages 27, 32<br />
Arawak, see Lokono Dian<br />
Arawak languages 18, 40, 46, 67, 87, 91,<br />
102–4, 121, 141–2, 183<br />
Arawak-Carib mixed language 142–3<br />
Archi (North-east Caucasian) 53<br />
Armenian (North-east Caucasian) 183<br />
Arsi Oromo (Cushitic, Afroasiatic) 183<br />
Arvanitika (Albanian, Indo-European) 92<br />
Ashaninka (Arawak) 121<br />
Ashanti (Kwa) 103<br />
Asheninca (Arawak) 121<br />
Asia Minor Greek (Indo-European) 225,<br />
see also Greek<br />
Assamese (Indo-Aryan, Indo-European) 84<br />
Atayal (Formosan, Austronesian) 162<br />
Athabaskan languages 30, 67–8<br />
Atsina (Algonquian) 8<br />
Australian English (Germanic, Indo-European)<br />
109, 198, see also English<br />
Australian languages 3, 18, 21, 30–2, 41, 58,<br />
62, 66, 69, 75, 80–9, 91, 120, 136, 145<br />
Austroasiatic languages 74<br />
Austronesian languages 18, 55, 63, 71, 97, 127<br />
Avestan (Iranian, Indo-European) 82<br />
Awetí (Tupí) 139–41, 144, 162<br />
Ayt Seghrushen (Berber, Afroasiatic) 50–1<br />
Babungo (Bantu) 21<br />
Baining languages (Papuan area) 41, 44
252 Index of languages, linguistic families, and peoples<br />
Baïnouk (Atlantic) 31<br />
Balochi (Iranian, Indo-European) 82, 87<br />
Baltic languages (Indo-European) 84<br />
Balto-Finnic languages (Finno-Ugric) 87<br />
Baluchi (Iranian, Indo-European) 82,<br />
see also Balochi<br />
Bandjalang (Australian area) 95<br />
Baniwa of Içana (Arawak) 31<br />
Bantu languages 12, 16, 21, 56, 71, 78, 92, 95,<br />
98, 173, 182, 192<br />
Bantu-based Creole, see Fanagalo<br />
Baré (Arawak) 92<br />
Bari (Nilotic) 77<br />
Belizean Creole (Creole) 176<br />
Benchnon (Omotic, Afroasiatic) 38–9,<br />
46–7, 50<br />
Bengali (Indo-Aryan, Indo-European) 83–4<br />
Berber languages (Afroasiatic) 16, 38, 45,<br />
103–4, 109, 117, 177, 184<br />
Biak (Austronesian) 30<br />
Big Nambas (Oceanic, Austronesian) 174,<br />
181, 183<br />
Biloxi (Siouan) 141, 160–2<br />
Black Carib (Arawak) 144<br />
Boiken (Ndu, Papuan area) 44, 49–50, 128<br />
Botocudo (Macro-Jê) 183<br />
Brahui (Dravidian) 87, 97<br />
Brazilian Portuguese 51, 112, 118–19, 155,<br />
158–9, 164, 214, see also Portuguese<br />
Breton (Celtic, Indo-European) 84<br />
Bulgarian (Slavic, Indo-European) 90, 106, 118<br />
Bunaq (Austronesian) 55–6<br />
Burmese (Tibeto-Burman) 64<br />
Camus (Nilotic) 40<br />
Cantabrian Spanish (Romance, Indo-<br />
European) 42–7, 56, 61, 175–6, see also<br />
Spanish<br />
Carib languages 15, 18, 71, 142–3, 160, 162<br />
Carrier (Athabaskan) 67–8, 74<br />
Catalan (Romance, Indo-European) 84<br />
Celtic languages (Indo-European) 84<br />
Central Khoisan 40, 44, 56<br />
Chadic languages 16, 50, 54, 79, 87, 176, 185<br />
Chamicuro (Arawak) 87<br />
Charmay French patois (Romance,<br />
Indo-European) 175, see also French<br />
Chehalis (Salish) 71<br />
Chinese (Sino-Tibetan) 71, 98<br />
Chinook Jargon (a creolized variety of<br />
Chinook) 71, 75<br />
Chinookan languages 71<br />
Chukchi (Chukotko-Kamchatkan) 139,<br />
145–6, 162, 208<br />
Chukotko-Kamchatkan languages 8, 139<br />
Clackamas Chinook (Chinookan) 88<br />
Coatzoquitengo Mixtec (Mixtecan) 79<br />
Cocama, see Kokama-Kokamilla<br />
Cora (Uto-Aztecan) 68<br />
Corsican (Romance, Indo-European)<br />
179–81, 215<br />
Cree (Algonquian) 20, 31<br />
Creek (Muskogean) 173<br />
Creole Arabic, see Arabic Creole<br />
Creole languages 41, 71, see also Arabic<br />
Creole, Belizean Creole, Fanagalo,<br />
French Creole, Guyanais Creole French,<br />
Maroon Creole (or Ndyuka),<br />
Negerhollands Creole Dutch,<br />
Papiamentu, Tok Pisin<br />
Cross-River languages (Benue-Congo) 65,<br />
87, 97<br />
Cushitic languages 16, 22, 31, 39, 44, 50–1, 183<br />
Czech (Slavic, Indo-European) 98, 123<br />
Damin (initiation language of Lardil,<br />
Australian area) 168–9, 182<br />
Dangbon (Australian area) 90<br />
Danish (Germanic, Indo-European) 93,<br />
96–8, 204<br />
Dardic languages (Indo-European) 87<br />
Dasenech (Cushitic, Afroasiatic) 50<br />
Dizin (Omotic, Afroasiatic) 129<br />
Dolakha Newar (Tibeto-Burman) 4<br />
Dravidian languages 13, 18, 31, 68, 83, 87,<br />
90, 141, 160<br />
Dutch (Germanic, Indo-European) 61, 72,<br />
96, 204
Index of languages, linguistic families, and peoples 253<br />
Dyirbal (Australian area) 91, 97<br />
development of gender in 80–1<br />
gender in 7, 13, 57, 117<br />
obsolescence of <strong>the</strong> language 91<br />
semantic basis of gender choice 19–22,<br />
29–31, 35, 59, 69, 120–1, 212<br />
variable gender 53–4<br />
Eastern Maroon Creole of Suriname, see<br />
Maroon Creole<br />
Egyptian Arabic (Semitic, Afroasiatic) 98,<br />
117, see also Arabic<br />
Enga (Engan, Papuan area) 51, 67, 75<br />
English (Germanic, Indo-European) 11–12,<br />
53, 59, 69, 73–4, 81, 91, 94, 105, 123–35,<br />
145, 173, see also American English,<br />
Australian English, English of<br />
Tasmania, Middle English, Old English<br />
acquistion of gender in 92–3<br />
female and male speech patterns in 138,<br />
149, 155–61, 164<br />
gender in 7, 11, 15, 30, 53, 72, 94–7,<br />
108–12, 118–19, 187–90<br />
generic masculine pronoun in 4, 26, 29,<br />
86, 112–16, 174, 195–200, 203–16<br />
history of gender in 6, 85–90, 96–7<br />
language change in 175–7<br />
loans from 12, 24, 71, 187<br />
loss of gender agreement in 6, 15, 85–90<br />
term for gender in 11–12<br />
English of Tasmania 110, see also English<br />
English-based Creoles 41, 72<br />
Ersu (Tibeto-Burman) 67, 74<br />
Estonian (Balto-Finnic, Uralic) 30, 72–5,<br />
88–9, 93, 97, 122–3, 129, 134, 190–1,<br />
197, 210, 216–17<br />
Ewe (Kwa) 69, 75<br />
Fanagalo (a Bantu-based Creole) 71, 75<br />
Farizandi (Iranian, Indo-European) 86<br />
Figuig (Berber, Afroasiatic) 104, 108–9<br />
Finnish (Balto-Finnic, Uralic) 15, 30, 72, 75,<br />
88, 97, 200<br />
French (Romance, Indo-European)<br />
gender in 5, 7, 11, 13–15, 23, 26, 84,<br />
89, 97–8, 103, 113, 118–19, 156–60,<br />
176, 217<br />
loans from 23, 46, 71–2, 179<br />
markedness in gender 26<br />
sexist language in 194–204, 210<br />
translation into 124–6<br />
French Creole 23<br />
Frisian (Germanic, Indo-European) 96<br />
Gaagudju (Australian area) 18, 31<br />
Gala (Ndu, Papuan area) 16, 44, 50, 99<br />
Galab, see Dasenech<br />
Gapun, see Taiap<br />
Garifuna (Arawak) 144<br />
Garvi (Dardic, Indo-European) 87<br />
Gavar (Dardic, Indo-European) 87<br />
Gbaya-Kara (Atlantic-Congo) 182<br />
German (Germanic, Indo-European) 11<br />
acquisition of 92–3, 97<br />
assignment of gender, principles of 13–15,<br />
22–5, 53, 57–8, 210<br />
gender in 6–7, 13–16, 58–9, 85, 113,<br />
122–35, 177, 213, 217<br />
gender resolution in 27<br />
loans from 88<br />
neuter gender in 7, 52–3, 60<br />
sexist language in 191, 194, 199–204,<br />
210, 216<br />
Germanic languages (Indo-European) 72,<br />
88–9, 96–8<br />
Godoberi (North-east Caucasian) 18<br />
Gola (West Atlantic) 58<br />
Greek (Indo-European) 11, 22–3, 52, 81, 88,<br />
93, 113, 118, 135<br />
Greek, Modern, see Greek<br />
Greek dialects (Indo-European) 88<br />
Gros Ventre (Algonquian) 137–40, 145–7, 163<br />
Guyanais Creole French (Creole) 72<br />
Hadza (isolate, Tanzania) 42, 44<br />
Hamar (Omotic, Afroasiatic) 42–4, 49, 186<br />
Hausa (Chadic, Afroasiatic) 13, 22, 71, 83<br />
Hawaiian (Polynesian, Austronesian) 71
254 Index of languages, linguistic families, and peoples<br />
Hebrew (Semitic, Afroasiatic)<br />
acquisition of 92–3<br />
gender in 13, 52, 98, 100, 132, 194<br />
gender reversal in 106, 109, 190<br />
markedness in 26–7, 31–2, 114<br />
masculine generic in 203–4, 210<br />
Hindi-Urdu (Indo-Aryan, Indo-<br />
European) 90<br />
Hittite (Anatolian, Indo-European) 81<br />
Hixkaryana (Carib) 15<br />
Hmong (Hmong-Mien) 68<br />
Hmong-Mien languages 68<br />
Hunde (Bantu) 95<br />
Hungarian (Ugric, Uralic) 30, 73, 122, 132,<br />
177, 189, 191, 210, 217<br />
Huron (Iroquoian) 115<br />
Iatmul (Ndu, Papuan area) 44, 50, 101–3<br />
Igbo (Atlantic-Congo) 61–2, 73, 206<br />
Ignaciano (Arawak) 141–2, see also Mojo<br />
Ilocano (Philippine, Austronesian) 127<br />
Indo-Aryan languages (Indo-European) 15,<br />
22, 83–90, 96, 187<br />
Indo-European languages 6, 10, 13, 16, 24–6,<br />
31, 70, 81, 84, 95–6, 114, 195–6<br />
Indonesian (Western Austronesian,<br />
Austronesian) 30, 64, 71–5, 89, 97<br />
Iñeri (Arawak) 142–3<br />
Innu (Algonquian) 67, 74<br />
Iranian languages (Indo-European) 82,<br />
86–7, 96<br />
Irish sign language 162<br />
Iroquoian languages 32, 56, 107,<br />
115–18, 191<br />
Island Carib (Arawak) 142–44, 160, 162<br />
Island Carib mixed language 142–3<br />
Italian (Romance, Indo-European) 13, 51,<br />
61, 84, 93, 131, 204, 207, 210<br />
Italian dialects 96<br />
Jacaltec (Mayan) 64–6, 68, 74, 95<br />
Japanese (isolate) 67, 74, 131<br />
female and male speech in 8, 137, 148–51,<br />
155, 160–4, 183, 208, 214<br />
Jarawara (Arawá) 4, 27–8, 32, 53, 104, 109,<br />
114–16, 186, 197<br />
Jê languages 15, 18<br />
Jingulu (Australian area) 21<br />
Jukun (Jukunoid, East Benue-Congo) 169,<br />
173, 183<br />
Kaingang (Jê) 15, 18<br />
Kala Lagaw Ya (Australian area) 32<br />
Kalapuyan languages 88<br />
Kalmyk (Tungusic) 129<br />
Kaluli (Kaluli-Kasua, Bosavi) 130, 133,<br />
158, 168<br />
Kambaata (Cushitic, Afroasiatic) 181, 183<br />
Kana (Cross-River, Benue Congo) 86<br />
Kannada (Dravidian) 90<br />
Kansa (Siouan) 162<br />
Kapriman, see Sare<br />
Karajá (Macro-Jê) 138, 144, 147, 160, 214<br />
Katcha (Kadugli-Krongo) 22, 40, 51<br />
Kayardild (Australian area) 69<br />
Kaytej (Australian area) 174<br />
Kele (Bantu) 173<br />
Kera (Chadic, Afroasiatic) 176, 184<br />
Keresan (isolate)163, see also Acoma Keresan<br />
Ket (Yeniseian) 56, 61, 128, 185–6<br />
Kewa (Engan, Papuan area) 74<br />
Kilivila (Oceanic, Austronesian) 68, 75, 78<br />
Kiowa (Kiowa-Tanoan) 82, 165<br />
Kiowa-Tanoan languages 82<br />
Koasati (Muskogean) 8, 12, 69, 139–44,<br />
147, 162<br />
Kokama-Kokamilla (Tupí-Guaraní,<br />
Tupí) 139–40, 145, 160, 163<br />
Kokota (Oceanic, Austronesian) 77<br />
Kolami (Dravidian) 18<br />
Konkani (Indo-Aryan, Indo-European) 83<br />
Korean (isolate) 64, 74, 78, 95<br />
Kristang (Creole) 72<br />
Kru languages 97<br />
Ku Waru (Chimbu-Wahgi, Papuan area) 74<br />
Kuna (Chibchan) 167–8, 182<br />
Kupto (Chadic, Afroasiatic) 10<br />
Kupwar village, languages of 90
Index of languages, linguistic families, and peoples 255<br />
Kurdish (Iranian, Indo-European) 82<br />
Kũr̩ux (Dravidian) 141–2, 146, 162<br />
Kushi (Chadic, Afroasiatic) 10<br />
Kwa languages 69, 70<br />
Kwami (sou<strong>the</strong>rn Bole-Tangale, Chadic,<br />
Afroasiatic) 10<br />
Kwoma (Kwoma-Nukuma, Papuan area) 44,<br />
50, 91, 206<br />
Labwor (Nilotic) 91<br />
Lak (North-east Caucasian) 56, 118,<br />
191–3, 209<br />
Lakhota (Siouan) 136–9, 145–7, 159, 162–5<br />
Lao (Tai-Kadai) 64–5, 68, 74<br />
Lardil (Australian area) 168–9<br />
Latin (Italic, Indo-European) 13, 51, 84, 89, 95<br />
Latvian (Baltic, Indo-European) 84<br />
Lezgian (North-east Caucasian) 70<br />
Libido (Cushitic, Afroasiatic) 183<br />
Língua Geral (Tupí-Guaraní, Tupí) 91–2,<br />
139, 162, 189<br />
Livonian (Balto-Finnic, Uralic) 87<br />
Lokele (Bantu) 173, 182<br />
Lokono Dian (Arawak) 46–7, 104, 108–9, 117<br />
Lower Sepik languages (Papuan area) 23<br />
Lower Chinook (Chinookan) 71<br />
Luwo (Nilotic) 91<br />
Maa (Ubangi, Benue-Congo) 18<br />
Maale (Omotic, Afroasiatic) 38, 50<br />
Maasai (Nilotic) 51<br />
Macedonian (Slavic, Indo-European) 70, 75<br />
Machiguenga (Arawak) 102, 108<br />
Macro-Jê languages 138, 183<br />
Maithili (Indo-Aryan, Indo-European) 15, 17<br />
Malagasy (Western Austronesian,<br />
Austronesian) 153–4, 166, 178<br />
Mali (Baining, Papuan area) 41, 44<br />
Malto (Dravidian) 18, 68<br />
Mam (Mayan) 66, 74, 78, 91<br />
Manambu (Ndu, Papuan area)<br />
cultural associations of gender in 29,<br />
47–50, 113–14, 117, 186<br />
female and male speech in 146–57<br />
gender and physical properties in 33–7,<br />
44, 48, 50<br />
gender in 16, 27, 53, 91–3, 99–104, 213<br />
markedness in gender in 27–8<br />
names in 206<br />
Mandan (Siouan) 162<br />
Marathi (Indo-Aryan, Indo-European) 90,<br />
97, 108–9, 117<br />
Maroon Creole (Creole) 72, 186<br />
Marra (Australian area) 69<br />
Maung (Australian area) 54–5<br />
Mawayana (Arawak) 87<br />
Mayali dialect chain (Australian area) 90<br />
Mayan languages 13, 64–6, 78, 152–4<br />
Mayan-speaking community of<br />
Tenejapa 152–4<br />
Mayrinax dialect of Atayal (Formosan,<br />
Austronesian) 162<br />
Mescalero Apache (Athabaskan) 67<br />
Mexican Spanish (Romance, Indo-European)<br />
131, see also Spanish<br />
Mexicano (Uto-Aztecan) 178, 184<br />
Middle English 86–8, 89, 94, 97, 188, 195,<br />
210, see also English<br />
Middle Irish (Celtic, Indo-European) 84<br />
Mixtecan languages 71, 79, see also<br />
Coatzoquitengo Mixtec<br />
Modern Khwe (Central Khoisan) 40, 44, 51,<br />
56, 61<br />
Mohawk (Iroquioan) 56, 107, 115,<br />
191, 209<br />
Mojo (Arawak) 141–2, see also Ignaciano<br />
Moldavian (Romance, Indo-European) 162<br />
Mongolic languages 162<br />
Moroccan Arabic (Semitic, Afroasiatic) 32,<br />
214, see also Arabic<br />
Motuna (Bougainville, Papuan area) 57–8<br />
Mupun (Chadic, Afroasiatic) 79, 95<br />
Murrinhpatha (Australian area) 32, 66, 74<br />
Muskogean languages 8, 69, 139, 144, 173<br />
Nahuatl (Uto-Aztecan) 184, see also Mexicano<br />
Nama (Khoekhoe) 61<br />
Navajo (Athapaskan) 131
256 Index of languages, linguistic families, and peoples<br />
Ndu languages (Papuan area) 16, 33, 50, 91<br />
Ndyuka, see Maroon Creole<br />
Negerhollands Creole Dutch (Creole) 72<br />
Ngala (Ubangi) 78<br />
Ngandi (Australian area) 69<br />
Ngan.gityemerri (Australian area) 79–81<br />
Nguni (Bantu) 183<br />
Nheêngatú, see Língua Geral<br />
Nilotic languages 40, 44, 58, 77, 91, 95–7<br />
Nootka (Wakashan) 71<br />
Norman French (Romance,<br />
Indo-European) 97<br />
North America, languages of 4, 8, 13, 18, 27,<br />
64, 118, 136, 162, 165<br />
North-east Caucasian languages 18, 53, 56,<br />
70, 191–3<br />
Nor<strong>the</strong>rn Mao (Omotic, Afroasiatic) 87<br />
Norwegian (Germanic, Indo-European) 93,<br />
96–8, 200, 204, 210<br />
Nubi (Creole) 72, 75<br />
Nunggubuyu (Australian area) 58, 62<br />
Nynorsk 93, 210, see Norwegian<br />
Oceanic languages (Austronesian) 68, 77–8, 97<br />
Oirat (Mongolic) 162<br />
Ojibwe (Algonquian) 55<br />
Old English (Germanic, Indo-European) 6,<br />
15, 85–6, 90, 95–6, 188, 195, 216<br />
Old French (Romance, Indo-European) 84, 90<br />
Old Irish (Celtic, Indo-European) 84<br />
Old Norse (Germanic, Indo-European) 93, 97<br />
Old Persian (Iranian, Indo-European) 82<br />
Ollari (Dravidian) 18<br />
Omaha-Ponca (Siouan) 162<br />
Omani Arabic (Semitic, Afroasiatic) 165,<br />
see also Arabic<br />
Omotic languages (Afroasiatic) 38–9, 42–6,<br />
49, 87, 129, 186<br />
Oneida (Iroquoian) 56, 107<br />
Onondaga (Iroquioan) 107, 115<br />
Oriya (Indo-Aryan, Indo-European) 84, 96, 187<br />
Oromo (Cushitic, Afroasiatic) 39, 45–7, 105,<br />
118, 128–9, 183<br />
Osage (Siouan) 145, 159<br />
Palestinian Arabic (Semitic, Afroasiatic) 27,<br />
31, 100, 108, 117, 210, see also Arabic<br />
Palikur (Arawak) 18, 23, 31, 45–7, 67–9,<br />
74, 121<br />
Papago (Uto-Aztecan) 2, 68, 155, 164,<br />
see also Tohono O’odham<br />
Papiamentu (Creole) 72<br />
Papua New Guinea, languages of 23, 71–2,<br />
75, 81, 133, 135<br />
Papuan languages 15, 23, 38, 40, 44, 56–7,<br />
67, 74, 93, 128–30, see also Papua New<br />
Guinea, languages of<br />
Parji (Dravidian) 18<br />
Parthian (Iranian, Indo-European) 82<br />
Persian (Iranian, Indo-European) 82<br />
Pidgin languages 71, 133<br />
Pintupi (Australian area) 183<br />
Piya (Chadic, Afroasiatic) 10<br />
Plains Cree (Algonquian) 31<br />
Polish (Slavic, Indo-European) 70, 82, 92–3,<br />
96–8, 117–18, 122<br />
Portuguese 13–15, 23–6, 45, 52, 59, 84, 113,<br />
118–19, 122, 201, 204–5, 210–13<br />
female and male speech patterns 155,<br />
158–9, 164, 214<br />
loans from 138, 171<br />
Proto-Arawak 31<br />
Proto-Bantu 21<br />
Proto-Germanic 85<br />
Proto-Indo-European 81, 83<br />
Proto-Mixtec 79<br />
Proto-Uralic 69<br />
Punjabi (Indo-Aryan, Indo-European) 22,<br />
87, 97<br />
Quapaw (Siouan) 162<br />
Romance languages (Indo-European) 17, 31,<br />
83–4, 90, 96, 179, 194<br />
Romanian (Romance, Indo-European) 17,<br />
31, 90, 96<br />
Russian 13–16, 26, 31, 51, 54–6, 59, 111–13,<br />
117, 121–5, 134–5, 191–3, 209<br />
acquisition of gender in 92–3, 97–8
Index of languages, linguistic families, and peoples 257<br />
animate declension in 83<br />
endearment and gender 105, 109<br />
female and male speech patterns 154–7, 164<br />
gender, term for 11–12<br />
Saami (Finno-Ugric, Uralic) 178<br />
Salish languages 64<br />
Sanskrit (Indo-Aryan, Indo-European) 72,<br />
84–5, 89<br />
Sardinian (Romance, Indo-European) 84<br />
Sare (Sepik Hill, Papuan area) 36–7, 44,<br />
97, 129<br />
Savosavo (Papuan area) 38, 44, 56<br />
Scottish Gaelic (Celtic, Indo-European) 91<br />
Sele Fara dialect of Slovene (Slavic,<br />
Indo-European) 84<br />
Seneca (Iroquioian) 115<br />
Sepik Hill languages (Papuan area)<br />
35–6<br />
Serbo-Croatian (Slavic, Indo-European) 84<br />
Sesotho (Bantu) 92<br />
Setswana (Bantu) 192<br />
Sheko (Omotic) 38, 50<br />
Sidaama (Cushitic, Afroasiatic) 173, 183<br />
Sinhalese (Indo-Aryan, Indo-European)<br />
84, 96<br />
Siouan languages 18, 136, 141, 162<br />
Slavic languages (Indo-European) 70, 82–4,<br />
89–90, 96, 205<br />
Slovene (Slavic, Indo-European) 84<br />
South Dravidian (Dravidian) 83<br />
Sou<strong>the</strong>rn Sotho (Bantu) 183<br />
Spanish (Romance, Indo-European) 46–7,<br />
51, 61, 84, 98, 113, 126–7, 131–4, 175,<br />
178, 205, see also Cantabrian Spanish,<br />
Mexican Spanish<br />
differential object marking in 70, 82, 97<br />
loans from 133, 211<br />
markedness in gender 25–6, 200–4<br />
sexist language in 113–14, 194, 200<br />
Suri (Surmic) 206<br />
Surmic languages 87<br />
Swedish (Germanic, Indo-European) 93,<br />
96–8, 200<br />
Taiap (Papuan area) 135, 177–8, 184<br />
Tai-Kadai languages 64, 141, 192<br />
Tamazight (Berber, Afroasiatic) 174,<br />
177–8, 184<br />
Tamil (Dravidian) 13, 18, 96<br />
Tangale (Chadic, Afroasiatic) 87, 97<br />
Tariana (Arawak) 105, 109, 128–9, 170–3,<br />
179–86, 189–90, 209, 215<br />
Tat (Iranian, Indo-European) 82<br />
Talysh (Iranian, Indo-European) 82<br />
Telugu (Dravidian) 31<br />
Tenejapa, see Mayan-speaking community<br />
of Tenejapa<br />
Thai (Tai-Kadai) 64–5<br />
female and male spech 8, 137, 148, 151,<br />
159–60, 172, 208, 214<br />
Tiwi (Australian area) 41, 44, 97<br />
Tocharian A (Tocharian, Indo-European)<br />
31, 96<br />
Tohono O’odham (Uto-Aztecan) 2, 155,<br />
164, see also Papago<br />
Tok Pisin (Creole) 41, 72, 177–8<br />
Tongan (Polynesian, Austronesian) 165<br />
Toposa (Nilotic) 77<br />
Tualatin (Kalapuyan) 88<br />
Tucano (East-Tucanoan, Tucanoan) 179–80<br />
Tucanoan languages 105<br />
Tungusic languages 71<br />
Tunica (isolate) 18, 20, 120, 141–2, 160–2,<br />
197, 212<br />
Tupí languages 71<br />
Tupí-Guaraní languages (Tupí) 91, 139, 150,<br />
162, 178, 183<br />
Turkana (Nilotic) 40, 51, 58, 62<br />
Turkic languages 71<br />
Turkish (Turkic) 5, 71–5, 79, 87–9, 95, 97,<br />
189, 217<br />
Uralic languages 69, 71–2, 75<br />
Vietnamese (Austroasiatic) 74<br />
Wagaya (Australian area) 32<br />
Wangkumara (Australian area) 32
258 Index of languages, linguistic families, and peoples<br />
Wära (isolate, Papuan area) 37, 44, 50<br />
Warao (isolate, Venezuela) 167<br />
Wardaman (Australian area) 58, 62<br />
Waris (Papuan area) 51<br />
Warlbiri (Australian area) 169, 174, 183<br />
Warndarran (Australian area) 81<br />
Warray (Australian area) 58, 62<br />
Warumungo (Australian area) 174<br />
Waujá (Arawak) 183<br />
Welsh (Celtic, Indo-European) 84, 198<br />
West Chadic (Afroasiatic) 79<br />
Wolaitta (Omotic, Afroasiatic) 79<br />
Wolof (West-Atlanic) 38, 50<br />
Wosera (Ndu, Papuan area) 44, 50<br />
Yalaku (Ndu, Papuan area) 44, 50<br />
Yana (isolate) 139–41, 144, 162, 147, 175<br />
Yangoru Boiken, see Boiken<br />
Yankunytjatjara (Australian area)<br />
66, 74<br />
Yanomami (Yanomami) 169<br />
Yanyuwa (Australian area) 91, 136, 139–41,<br />
145–6, 160, 163<br />
Yazgulami (Iranian, Indo-European) 86<br />
Yeniseian languages 128, 185<br />
Yidiñ (Australian area) 65, 69, 79<br />
Yimas (Lower Sepik, Papuan area) 23, 31<br />
Yolngu (Australian area) 183<br />
Young People’s Dyirbal (Australian area) 91,<br />
97, see also Dyirbal<br />
Zande (Ubangi) 18, 77–8<br />
Zazaki (Iranian, Indo-European) 82<br />
Zulu (Bantu) 183<br />
!Xun (Central Khoisan) 95
Index of authors<br />
Aalto, Pentti 162<br />
Abbink, Jon 206<br />
Abbott, Clifford 61<br />
Abbou, Julie 210<br />
Abondolo, Daniel 97<br />
Abueg, Joselito 209<br />
Adams, K. L. 74<br />
Adelaar, Willem F. H. 97<br />
Ahland, Colleen Anne 97<br />
Aikhenvald, Yuri A. 111<br />
Aikio, Marjut 184<br />
Akhmatova, Anna A. 111, 193<br />
Alexeyev, M. 75<br />
Allan, Keith 209<br />
Alpher, Barry 32, 112, 134<br />
Alvanoudi, Angeliki 11, 135<br />
Amadiume, Ifi 61–2, 73, 206<br />
Ameka, Felix K. 75<br />
Amha, Azeb 49–50, 97<br />
Amiram, Orit 93<br />
Anderson, L-G. 209<br />
Armon-Lotem, Sharon 93<br />
Asher, R. E. 30<br />
Attaviriyanupap, Korakoch 163<br />
Audring, Jenny 61<br />
Avery, Jack 164<br />
Awbery, Gwenllian 210<br />
Axenov, Serge 95<br />
Baker, Paul 130, 208<br />
Bani, Ephraim 31<br />
Barnhart, Robert K. 209<br />
Baron, Dennis 11, 27, 111–12, 118, 187,<br />
196, 209<br />
Baron, Naomi S. 96–7<br />
Barrett, R. 11, 164–5<br />
Bateson, Gregory 3, 11, 101<br />
Baudelaire, Charles 135<br />
Beachy, Marvin Dean 129<br />
Beauvoir, Simone de 2<br />
Becker, A. L. 74<br />
Bell, D. 183<br />
Benamara, Hassane 117<br />
Bengoechea, Mercedes 202<br />
Benjamin, Carmen 51<br />
Bergen, John J. 51<br />
Berman, Ruth A. 97<br />
Besnier, Niko 165<br />
Bhatia, Tej K. 31, 97<br />
Birtalan, Ágnes 162<br />
Bisang, W. 75<br />
Bloomfield, Leonard 12<br />
Bloomfield, Morton W. 126<br />
Boas, Franz 4, 138<br />
Bodine, Ann 162<br />
Bogoras, Waldemar 139, 162–3<br />
Bolinger, Dwight 10<br />
Bonfante, G. 51<br />
Bonvillain, Nancy 61, 107<br />
Borba, Rodrigo 165<br />
Boroditzky, Lera 126–7, 131–2<br />
Bossong, Georg 75<br />
Bowden, Ross 50<br />
Bradac, J. J. 156<br />
Bradley, John 145–6, 163<br />
Brauer, Marcus 119<br />
Braun, Friederike 11, 16, 73–5, 95–7, 113,<br />
189, 202, 211<br />
Breen, G. 32<br />
Breton, Raymond 162<br />
Briggs, C. L. 182<br />
Brightman, Robert 162<br />
Brittain, David 183<br />
Brosman, P. W. Jr. 95<br />
Brown, Penelope 153, 164, 191<br />
Brown, R. 51
260 Index of authors<br />
Bruce, Les 50, 62, 120<br />
Brunet, Jacqueline 211<br />
Bull, Tove 210<br />
Bullock, Alan 11<br />
Bulygina, T. V. 31<br />
Burridge, Kate 209<br />
Butt, John 51<br />
Caldas-Coulthard, Carmen Rosa 130<br />
Camden, W. G. 162<br />
Cameron, Deborah 154, 164<br />
Campbell, Élisabeth 203<br />
Campbell-Kibler, K. 164<br />
Capell, A. 54<br />
Caputo, B. Britt 196<br />
Carrington, J. F. 182–3<br />
Castellino, G. R. 50<br />
Chafe, Wallace L. 52, 96, 115–19<br />
Cheshire, J. 11, 188, 194, 198–9, 208–9, 216<br />
Chirasombutti, Voravudhi 163<br />
Christ, S. 154<br />
Chung, Sandra L. 197<br />
Clackson, James 95–6<br />
Clamons, Cynthia Robb 39, 48, 118<br />
Claudi, U. 95<br />
Coates, Jennifer 11, 119, 156, 164, 183<br />
Cobbinah, Alexander 31<br />
Collinder, Björn 75<br />
Comrie, Bernard 75<br />
Connors, Kathleen 118<br />
Contini-Morava, Ellen 61<br />
Cooper, Fiona 118<br />
Corbett, Greville G. 11, 30–1, 61, 75, 96–7<br />
Cowley, Roger 45, 51<br />
Craig, C. G. 74, 95<br />
Craik, Brian 31<br />
Creed, Gerald W. 11<br />
Crolley, Liz 207, 209–10<br />
Crouch, Isabel 164<br />
Curzan, Anne 31, 85–6, 96–7, 195, 209<br />
Dahlstrom, Amy 31, 61<br />
Daley, Karen Ann 74<br />
Darnell, Regna 31<br />
Dawkins, Richard M. 97<br />
de Courtenay, Baudouin 122<br />
De León, Maria de Lourdes Pasquel 95<br />
del Castillo, J. 126<br />
Demuth, K. A. 97<br />
Denny, J. P. 31<br />
Deutscher, Guy 62, 124, 132–5<br />
Diakonoff, I. M. 50<br />
Diller, Anthony 74, 163<br />
Dimmendaal, Gerrit J. 30, 51, 58, 62<br />
Dinslage, Sabine 61, 183, 208<br />
Dixon, R. M. W. 30–1, 59, 120, 135, 169<br />
on Australian languages 30, 75, 90–1, 95,<br />
169, 182<br />
on Dyirbal 22, 29, 31, 54, 69, 97, 117, 121<br />
on English 97, 118, 188, 197<br />
on Jarawara 27, 32, 104, 114–15, 197<br />
Doleschal, Ursula 117<br />
Dorian, Nancy C. 91, 97<br />
Drapeau, Lynn 74<br />
Drude, Sebastian 144, 162<br />
Du Puis, Mathias 162<br />
Dubois, Betty 164<br />
Dul’son, A. P. 61, 135<br />
Dunn, Michael 162<br />
Eckert, Penelope 119, 183–4<br />
Edelman, D. I. 86, 97<br />
Edenmyr, Niklas 51<br />
Einaudi, Paula 162<br />
Ekka, Francis 146, 162<br />
Emeneau, Murray 97<br />
Endruschat, Annette 210<br />
Enfield, N. J. 74<br />
England, N. 74, 95<br />
Ernst, Amélie 124<br />
Ervin, Susan M. 51<br />
Escure, Geneviève 183<br />
Evans, Bergen 209<br />
Evans, Cornelia 209<br />
Faraclas, N. 97<br />
Faraoni, Vincenzo 96<br />
Faust, Norma 162–3
Index of authors 261<br />
Ferguson, Charles A. 117<br />
Ferrari, F. 51, 61<br />
Fesenko, Andrej 209<br />
Fesenko, Tatiana 209<br />
Flannery, Regina 137, 140, 162–3<br />
Fleming, Luke 182–3<br />
Fletcher, Alice 159<br />
Fodor, I. 95, 125<br />
Foley, W. A. 31<br />
Fortune, David 147, 162<br />
Fortune, Gretchen 147, 162<br />
Fox, Helen 174, 183<br />
Frajzyngier, Zygmunt 95<br />
Franklin, K. 74<br />
Friedman, Victor 75<br />
Gal, Susan 163–4, 184<br />
Gardani, Francesco 96<br />
Garibian, A. S. 96<br />
Gauchat, L. 175, 183<br />
Gaudio, R. 164<br />
Georg, Stefan 128<br />
Gerdts, Donna 74<br />
Glazkov, Nikolaj 134<br />
Goddard, C. 74<br />
Goddard, Ives 31, 55, 61, 118, 197, 199,<br />
201–3, 207, 216<br />
Goettner-Abendroth, Heide 209<br />
Gogol, Nikolaj 134<br />
Gordon, Mat<strong>the</strong>w 164, 183<br />
Grandi, Nicola 51<br />
Green, Diana 31, 51, 71<br />
Green, Ian 95<br />
Greenbaum, Sidney 210<br />
Greenberg, J. H. 30, 81, 97<br />
Gregersen, Edgar A. 93, 98, 209<br />
Grünberg, A. L. 97<br />
Gumperz, J. J. 90, 97<br />
Günthner, Susanne 162–3, 182<br />
Gwynn, John Peter Lucius 31<br />
Haas, Mary 31, 134, 139, 144–5, 162–3,<br />
173, 183<br />
Hachimi, Atiqa 210<br />
Hagège, C. 98, 164<br />
Hagman, Roy S. 61<br />
Haig, Geoffrey 16, 60<br />
Hale, Kenneth L. 169, 182<br />
Hall, Kira 11, 165<br />
Hamp, Eric 97, 230<br />
Hampares, Ka<strong>the</strong>rine 202<br />
Harbert, Wayne 96–7<br />
Harrison, Simon J. 48, 117, 211<br />
Harvey, Mark 31, 62<br />
Haspelmath, M. 75<br />
Hasselblatt, Cornelius 75, 97, 190,<br />
209–10<br />
Haugen, Einar 97, 197<br />
Hauser-Schäublin, Brigitta 101, 104<br />
Haviland, William A. 211<br />
Healey, B. 159<br />
Heath, Jeffrey 62, 164, 173, 183<br />
Hegardty, Peter 209<br />
Heine, Bernd 30–1, 51, 62, 75, 95, 97<br />
Heine, Heinrich 123–4, 134, 213<br />
Hellenthal, Anne-Christie 50<br />
Hellinger, Marlis 119, 210<br />
Henley, Nancy M. 189, 209<br />
Henson, Eithne 183<br />
Herbert, R. K. 98, 183<br />
Herdt, Gilbert 11<br />
Hill, Jane H. 2, 164, 182, 184<br />
Hill, Kenneth 184<br />
Hinch, H. E. 54<br />
Hinkson, Mercedes Q. 74<br />
Hoben, Susan J. 51, 104–5, 117<br />
Hodge, Carleton T. 75<br />
Hoff, Berend 162<br />
Hoffman, Ka<strong>the</strong>rine E. 184<br />
Holland, Audrey 98<br />
Hollow, Robert 162<br />
Holm, John 75<br />
Holmes, Janet 11, 156, 164<br />
Holmquist, Jonathan C. 46, 51, 61,<br />
175, 178<br />
Hornscheidt, Antje 210<br />
Houseman, Michael 117<br />
Huntley, David 96
262 Index of authors<br />
Ibrahim, Muhammad Hasan 96<br />
Ide, Sachiko 148, 163<br />
Ikoro, S. M. 74<br />
Inoue, Miyako 163<br />
Irvine, Judith T. 164<br />
Isačenko, A. V. 84<br />
Jacobs, Andrea Michele 203–4<br />
Jaffe, Alexandra 179<br />
Jakobson, Roman O. 31, 59, 118, 121–3,<br />
132, 193<br />
James, Deborah 164, 183<br />
Janda, Laura A. 95<br />
Janse, Mark 97<br />
Janson, T. 75, 209<br />
Jarkey, Nerida 75, 149, 163<br />
Jasanoff, Jay 31, 96, 197<br />
Jespersen, Otto 2, 6, 155, 175, 183, 196<br />
Jones, Kathryn 210<br />
Jones, Lucy 164<br />
Joseph, Brian D. 23<br />
Jospin, Lionel 201<br />
Jungraithmayr, H. 97<br />
Kahane, Henry 51<br />
Kahane, René 51<br />
Karatsareas, Petros 97<br />
Karbelashvili, D. P. 183<br />
Karmiloff-Smith, A. 97<br />
Kastovsky, Dieter 97<br />
Kazakevich, Olga A. 138<br />
Keenan, Elinor 184<br />
Kendon, Adam 174, 183<br />
Key, Mary Ritchie 162, 164, 209<br />
Khaidakov, S. M. 56, 118, 209<br />
Kibrik, A. E. 31<br />
Kilarski, Marcin 30–1, 61<br />
Kilian-Hatz, Christa 51, 61<br />
Kimball, Geoffrey D. 144, 163<br />
Kirton, Jean F. 137, 162<br />
Kitajgorodksaja, Margarita V. 111, 117–18, 164<br />
Klenin, E. 95<br />
Klumpp, D. 95<br />
Kochskämper, Birgit 210<br />
Köhler, O. 51<br />
Konishi, T. 126<br />
Kooyers, O. 50<br />
Köpcke, Klaus-Michael 24–5, 52, 62, 123<br />
Kossmann, Maarten 50–1, 117<br />
Kramarae, Cheris 11, 111, 210<br />
Krishnamurti, Bhadriraju 31, 96<br />
Kroskrity, Paul V. 162<br />
Kryk-Kastovsky, Barbara 96<br />
Kulick, Don 135, 164–5, 184<br />
Kuntjara, Es<strong>the</strong>r 73<br />
Kuteva, Tania 95<br />
Kwee, John B. 75<br />
La Fleche, Frances 159<br />
Labov, William 2, 154, 164, 183<br />
Lakoff, George 31<br />
Lakoff, Robin Tolmach 9, 156, 161, 164, 193,<br />
198, 207–8<br />
Lambert-Brétière, Renée 74<br />
Landor, Roland 132, 135<br />
Landry, Michaël 119<br />
Lang, Adrienne 51, 74<br />
Lang, Sabine 165<br />
Laycock, D. C. 117<br />
Lazarus, Emma 124–5<br />
Leap, William 11<br />
Lee, J. 97<br />
Lee, Yun-seok 74, 95<br />
Leger, Rudolf 50, 61, 183, 208<br />
Lehmann, Thomas 30<br />
LeMaster, Barbara 162<br />
Lermontov, Mikhail 124<br />
Levon, Erez 164<br />
Levy, Y. 97<br />
Lewis, Geoffrey 97<br />
Li, Paul Jen-kuei 162<br />
Liss, Julie 164, 220<br />
Livia, Anna 160–1, 164, 217<br />
Lock, Arnold (Arjen) Hugo 50<br />
Lopatin, V. V. 209<br />
Loporcaro, Michele 96<br />
Lu, Tian-Qiao 65, 74<br />
Luraghi, Silvia 95–6
Index of authors 263<br />
Luthin, Herbert 147<br />
Lydall, Jean 42, 49–51, 186<br />
McAlinden, Fiona 164<br />
Macaulay, R. K. S. 155<br />
McConnell-Ginet, Sally 11, 113, 119, 122<br />
McEwan, Ian 199<br />
McGloin, Naomi 163<br />
McGregor, Aileen R. F. 50<br />
McGregor, Donald E. 50<br />
McWhorter, John 135<br />
Mahapatra, B. P. 69<br />
Maiden, Martin 211<br />
Maikov, Apollon 125<br />
Mallinson, Graham 31<br />
Mann, Heinrich 16<br />
Marcato, Gianna 210<br />
Marchese, L. 97<br />
Maring, J. 163<br />
Martel, Brigitte 3, 108, 159<br />
Martin, Samuel E. 74, 95<br />
Martynyuk, A. 209<br />
Masica, Colin 84, 96<br />
Matasovic, Ranko 31, 75, 96–7<br />
Mathiot, Madeleine 110, 118<br />
Mat<strong>the</strong>ws, P. H. 61<br />
Mat<strong>the</strong>ws, W. K. 97<br />
Mänd, Heljo 134<br />
Meillet, Antoine 6, 95<br />
Melchert, H. C. 95<br />
Mendes, Ronald Beline 51, 164<br />
Menn, Lise 98<br />
Merlan, F. 62, 74<br />
Migge, Bettina 72, 209<br />
Migliazza, Ernest C. 182<br />
Milani, Tommaso N. 220<br />
Miller, Casey 188<br />
Milles, Karin 210<br />
Mills, A. E. 92<br />
Mills, Sara 11, 119, 152, 205, 209<br />
Milroy, L. 183<br />
Miranda, Rocky V. 96<br />
Mithun, Marianne 31, 49, 56, 61, 96–7, 107,<br />
115–18, 134, 162–3, 191, 209<br />
Mladenova, Olga M. 118<br />
Moñino, Yves 182<br />
Moon, Rosamund 130<br />
Moore, Henrietta L. 11<br />
Morris, Delyth 210<br />
Morris, Lori 118<br />
Motschenbacher, Heiko 30, 220<br />
Mous, Maarten 31<br />
Mugglestone, Lynda 210<br />
Mulac, A. 156<br />
Mullen, M. K. 128–30, 135<br />
Munro, Pamela 162<br />
Myers, Fred R. 183<br />
Nakamura, Momoko 163<br />
Nanda, Serena 165<br />
Nash, David 182<br />
Nekitel, Otto 24, 31<br />
Neruda, Pablo 135<br />
Newman, Michael 31, 209<br />
Newman, Paul 96<br />
Nissen, Uwe Kjær 113, 134,<br />
202–3, 210<br />
Nuni de Chapi, Conchita 162<br />
O’Connor, Michael P. 98<br />
O’Donovan, Veronica 11, 165<br />
O’Leary, Kathy 164<br />
Ochs, Elinor 152, 164<br />
Ogawa, Naoko 164<br />
Okamoto, Shigeko 150–1, 161, 163<br />
Okudzhava, Bulat 125<br />
Olza Zubiri, Jesús 162<br />
Onishi, M. 62, 96<br />
Ortner, Sherry B. 3, 130<br />
Osborne, C. R. 51<br />
Osgood, Charles E. 131<br />
Ostermann, Ana Christina 165<br />
Ott, R. 162<br />
Ott, W. 162<br />
Oussikoum, Bennaser 117<br />
Owens, Jonathan 75<br />
Oxford English Dictionary Online 11, 97,<br />
111, 188, 204, 209–10
264 Index of authors<br />
Palmer, W. D. 95<br />
Pandharipande, Rajeshwari V. 97<br />
Pankhurst, Helen 51, 100, 104–5, 128<br />
Pastre, Geneviève 164, 201<br />
Pauwels, Anne 119, 205, 208–11<br />
Pawley, Andrew K. 110<br />
Payne, Judith 134<br />
Pearce, Mary 184<br />
Penelope, Julia 209<br />
Pensalfini, R. 31<br />
Pérez-Pereira, M. 97<br />
Perlak, D. 98<br />
Pet, W. J. A. 117<br />
Peters, Pam 214<br />
Pfarr, S. J. 149<br />
Philippaki-Warburton, Irene 23<br />
Philipps, Webb 126, 131<br />
Pichler, Pia 11<br />
Plank, F. 30<br />
Podesva, R. 164<br />
Pope, M. K. 84, 96<br />
Popova, M. I. 97<br />
Poser, William J. 74–5<br />
Posner, Rebecca 96<br />
Pountain, C. J. 51, 61<br />
Priestly, Tom M. S. 51, 84, 96–7<br />
Puşcariu, Sextil 90<br />
Quinn, Conor 31<br />
Quintero, Carolyn 145<br />
Quirk, Randolph 96<br />
Raffelsiefen, Renate 135<br />
Raga, Amanuel 135<br />
Rankin, Robert L. 162<br />
Rapold, Christian J. 50<br />
Rat, Joseph Numa 162<br />
Reh, Mechthild 75, 95<br />
Reid, Nicholas 95<br />
Repp, Hanna 117<br />
Reynolds, Katsue Akiba 149–51, 163<br />
Ribeiro, Eduardo Rivail 122<br />
Richards, Cara B. 115<br />
Ringelnatz, Joachim 134<br />
Roberts, S. 164<br />
Roberts, S. P. 62, 74<br />
Robustelli, Cecilia 211<br />
Rodina, Yulia 92, 97<br />
Rodrigues, Aryon Dall’Igna 162<br />
Romaine, Suzanne 11<br />
Romani Miranda, Maggie M. 134<br />
Roscoe, Paul B. 128<br />
Roscoe, Will 165<br />
Rose, Françoise 162<br />
Rosenhouse, J. 117<br />
Ross, Kristiina 72, 75<br />
Rothstein, Robert A. 31, 117–18, 192–3, 209<br />
Rozanova, Nona N. 111, 117–18, 164<br />
Rubino, C. R. G. 127, 135<br />
Rudes, B. 159<br />
Rumsey, A. 62, 74<br />
Rushforth, S. 74<br />
Sa’ar, Amalia 31, 117, 210<br />
Sabajo, M. A. 46<br />
Sadiqi, F. 32, 164<br />
Sahoo, Kalyanamalini 209<br />
Salzmann, Zdenek 208<br />
Sapir, Edward 139, 147, 162, 175, 185<br />
Sarvasy, Hannah 40<br />
Sasse, Hans-Jürgen 50–1, 97<br />
Saville-Troike, M. 12<br />
Sayers, Dorothy L. 61<br />
Särsa, Riisto 50<br />
Schane, S. A. 31<br />
Schapper, Antoinette 30, 61, 97<br />
Schaub, W. 31<br />
Schellinger, W. 30<br />
Schenker, Alexander M. 75<br />
Schieffelin, Bambi B. 130, 158, 168<br />
Schieffelin, Edward L. 130<br />
Schipper, Mineke 103, 129, 135, 190, 216<br />
Schmid, Sonja 117<br />
Schmidt, A. 97<br />
Schmidt, Lauren 126, 131<br />
Sczesny, Sabine 113<br />
Senft, Gunter 75, 95<br />
Sera, M. 126
Index of authors 265<br />
Severi, Carlo 117<br />
Seyoum, Mulugeta 50<br />
Shepard Jr., G. 102<br />
Sherzer, Joel 102, 158, 167, 181–2<br />
Shibatani, Masayoshi 148–9, 163<br />
Shmelev, A. D. 31<br />
Shvedova, Natalja Y. 193<br />
Silverman, Eric K. 11<br />
Silverstein, Michael 11–12, 75, 197, 188, 204<br />
Sims, C. 163, 182<br />
Singer, Ruth 55<br />
Sini, M. 211<br />
Slobin, Dan I. 133<br />
Smith, Janet S. (Shibamoto) 164<br />
Smith-Stark, S. 75<br />
Sneddon, James N. 64, 74<br />
Sontag, Susan 113<br />
Sou<strong>the</strong>rland, Ronald H. 138<br />
Speece, R. 51<br />
Spender, Dale 211<br />
Spitulnik, D. 31<br />
Stahlberg, Dagmar 113<br />
Stallybrass, Oliver 11<br />
Stebbins, Tonya N. 51<br />
Steinhauer, H. 30<br />
Storch, Anne 61, 97, 162, 169, 173–4, 182–3,<br />
208, 215<br />
Storjohann, Petra 199<br />
Stotko, Elaine M. 209<br />
Stradelli, Ermano 162<br />
Straus, Anne Terry 20, 31<br />
Stroud, C. 184<br />
Studzińska, Joanna 122<br />
Sumbuk, Kenneth Memson 37, 50, 129<br />
Surikov, Ivan 134<br />
Svartengren, T. Hilding 118<br />
Swan, Toril 210<br />
Swift, Kate 188, 209–11<br />
Tadmor, Uri 97<br />
Taïfi, Miloud 117<br />
Talbot, M. 11, 208<br />
Tannen, Deborah 164<br />
Taylor, Allan R. 12, 145–7, 162<br />
Taylor, Douglas M. 162<br />
Teferra, Anbessa 183<br />
Teso, Elena 207–10<br />
Thompson, James 123<br />
Thompson, L. C. 74<br />
Thompson, S. A. 156<br />
Thüne, Eva-Maria 210<br />
Timberlake, Alan H. 197<br />
Tjutschev, Fjodor 124<br />
Tobin, Yishai 27, 31, 106, 190<br />
Trask, L. 95<br />
Trechter, Sara 11, 136–7, 145–7,<br />
162–5<br />
Treichler, Paula A. 11, 111<br />
Treis, Yvonne 183<br />
Trombley, Stephen 11<br />
Troyer, Margaret 209<br />
Trudgill, Peter 97, 149, 155, 176, 183<br />
Tsegaye, Mulugeta T. 31<br />
Tube, Juan 162<br />
Underhill, Robert 197<br />
Upton, Rebecca 11<br />
Uyeno, T. 163<br />
Vajda, Edward J. 128<br />
Valentine, J. Randolph 208<br />
Valiquette, H. 163, 182<br />
Vallejos, Rosa Yopán 140, 163<br />
van Baarle, Peter 46<br />
van den Berg, René 95–7<br />
Vanek, Anthony L. 31<br />
Vasvári, Louise O. 75, 189, 210<br />
Villas Bôas, Cláudio 183<br />
Villas Bôas, Orlando 183<br />
Vinogradov, V. V. 113, 134, 209<br />
Vossen, Rainer 95<br />
Wales, Katie 31, 109, 118, 198<br />
Walsh, M. 32, 74<br />
Walters, Margaret 209<br />
Watkins, Calvert 197
266 Index of authors<br />
Watkins, L. 82<br />
Weber, Doris 61<br />
Wegener, Claudia 50, 61<br />
Welmers, Wm. E. 16<br />
Werner, Heinrich 135, 208<br />
Wertz, C. A. 96<br />
Westermann, D. 62<br />
Whitehead, Harriet 3, 165<br />
Wiedemann, Ferdinand Johann 75, 97<br />
Wikan, Unni 165<br />
Williams, Nancy M. 183<br />
Wilson, R. 90, 97<br />
Winford, Donald 97<br />
Woidich, Manfred 117<br />
Woldemariam, Hirut 135<br />
Wolfart, H. Christoph 31<br />
Wołk, Ewa 104–5, 117<br />
Wrenn, C. L. 96<br />
Yadav, Ramawatar 31<br />
Yaguello, Marina 118, 210<br />
Yigezu, Moges 163<br />
Yokoyama, Olga T. 117–18, 156, 164, 209<br />
Zavala, Roberto 95<br />
Zemskaya, Elena A. 111, 117–18, 164<br />
Zepeda, Ofelia 2, 164<br />
Zgusta, Ladislav 62<br />
Zhang, Sihong 74<br />
Zimman, Lal 11<br />
Zubin, David A. 24–5, 52, 62, 123<br />
Zwicky, Arnold M. 164
Index of subjects<br />
abstract noun 6, 18–19, 21, 23, 67, 81, 122, 191<br />
acquisition of linguistic gender 7, 92–3,<br />
97, 147<br />
address terms and practices 56, 104, 106,<br />
108, 141, 159, 163, 186, 204–9<br />
adjective 14, 15–18<br />
adverb 14, 71, 78<br />
agglutinating language 16<br />
agreement 14–18, 30–1, see also gender,<br />
linguistic<br />
alliterative 23, 31<br />
anaphoric 14–16<br />
hierarchy 30<br />
loss 76–98<br />
origin of 76–98<br />
variability in 33–51, 53–7<br />
agreement gender 53, 70–2, 76–82, 86–8,<br />
91–4, 97, 113, 126<br />
alliterative agreement, see agreement,<br />
alliterative<br />
anaphoric gender, see gender, anaphoric<br />
animacy 5, 13–25, 29–30, 34–51, 55–6, 72–5,<br />
82–7, 118, 121–35, 185–6, 191<br />
and classifiers 63–9<br />
and grammatical relations 69–71<br />
and number 69–71, 82<br />
and pronouns 69–71<br />
animate declension 83–4, 96<br />
aphasia 93–4<br />
areal diffusion 87–92, see also language<br />
contact<br />
article 14–19<br />
assignment of gender 18–25, see also gender<br />
assignment<br />
augmentative 22–3, 38, 51<br />
avoidance style 9, 57, 169–75, 181–3, 202,<br />
215, see also Hlonipha, Mo<strong>the</strong>r-in-law<br />
style, taboo<br />
baby talk 72, 108–9, 147, 164<br />
beliefs 19–20, 120–1, 132–5, see also<br />
myth-and-belief principle in linguistic<br />
gender choice<br />
Berdache 147, 165<br />
bilingualism 131, 177–8, see also<br />
multilingualism<br />
borrowing, see language contact, loans<br />
case 7, 63, 69–71, 82–3, 86, 92–6, see also<br />
differential case marking<br />
child language acquisition, see acquisition of<br />
linguistic gender<br />
classificatory verbs, see verbal classifier<br />
classifier 63, 68<br />
noun classifier 52, 63, 65, 68–9, 94<br />
numeral classifier 52, 63–5, 68–9, 94,<br />
190–2<br />
possessive classifier 61, 68, see also<br />
possession<br />
relational classifier 61<br />
verbal classifier 52, 67–9<br />
cognition 8, 10, 32, 87–8, 126–32<br />
collective 20–1, 51, 81<br />
command 4, 136–7, 141, 152–4, 157, 164, 189<br />
common gender 61, 93, 96, 98, 196<br />
compound 25, 81, 88, 113, 187–8<br />
convergence, see areal diffusion, language<br />
contact<br />
coordination 26–9, 194, 200–1<br />
covert marking of linguistic gender 14–15,<br />
58, 81, see also gender, linguistic<br />
creolization 97<br />
declension 13, 69–70, 83–6, 94, 96<br />
default choice in linguistic gender 28,<br />
114–15, 151, see also markedness,<br />
functional
268 Index of subjects<br />
deferential register 173–4, see also politeness<br />
definiteness and gender 57–9<br />
demonstrative 14–20, 27, 31–3, 54–6, 63,<br />
67–8, 78, 81–5, 88, 95–6, 138–41, 182<br />
deprecatory meanings of linguistic<br />
gender 46, 102–5, 111–12, 193<br />
derivation 22–3, see also derivational gender<br />
derivational gender 5, 13–14, 76, 79–83,<br />
88–90, 94, 104, 112, 117, 186, 192–3,<br />
see also gender, linguistic<br />
differential case marking 70, 83, see also case<br />
diminutives<br />
and linguistic gender 7, 15, 22–3, 38, 41,<br />
45, 51, 60, 73, 78, 90–2, 117<br />
in female speech 152–5, 159, 164, 214<br />
in gay and lesbian speech 159<br />
discourse role of linguistic gender 52, 57–9<br />
dissolution of linguistic gender 7, 92–3<br />
double duty words 188<br />
edible gender 19–20, 57, 80–1, 159<br />
elegant prefix 149, see also honorific<br />
endearment 15, 21, 35, 44–7, 92, 100,<br />
105–9, 117<br />
epicene 53, 189, 196, 202, 207<br />
evidential 35<br />
experimental studies of gender 126–35<br />
female social gender 1–9, see also gender, social<br />
female speech 2, 8, 136–65, 214, see also male<br />
speech<br />
phonetic features of 2, 8, 138–41, 144,<br />
154–5, 158, 161–3, 166<br />
feminine as an unmarked linguistic<br />
gender 27–9, 32<br />
feminine gender 1–9, 13–29, 34–61, 63–75,<br />
83–4, 91–135, 185–211, see also gender,<br />
linguistic<br />
feminine-indefinite in Iroquoian<br />
languages 107, 191<br />
feminization of job titles 112, 118, 120, 201,<br />
207, 210<br />
feminist language reform 93–4, 114, 134,<br />
186, 203–8, 211, 216–17<br />
feminist scholars 9, 59–60, 93, 111–13,<br />
196–7, 200–1, 204–11, 216<br />
formal markedness, see markedness, formal<br />
functional markedness, see markedness,<br />
functional<br />
fusional language 16<br />
gay men’s speech 3, 9, 11, 31–2, 80, 130, 146,<br />
158–9, 161, 164–5, 200, 215<br />
gender as term 1–14, 212–17<br />
gender, biological, see gender, natural<br />
gender, linguistic 1–9, 13–32, 214–17<br />
anaphoric 7, 14–16, 30–1, 60, 72, 76–7,<br />
81–8, 91–6, 124–6<br />
and o<strong>the</strong>r categories 16–17<br />
and social stereotypes 2, 44–58, 71–5,<br />
109–19, 185–206<br />
assignment of 19–25, 83–7, 185–6<br />
covert marking of linguistic gender 14–15,<br />
58, 81<br />
derivational, see derivational gender<br />
functions of 52–61<br />
gender agreement 14–18, 30–1, see also<br />
agreement gender<br />
in Athabaskan linguistic tradition 30<br />
loss of 83–8, 94–7, 126<br />
meanings of 19–22, 52–61, see also<br />
semantic opacity in linguistic gender,<br />
semantic transparency in linguistic<br />
gender<br />
origin of 66, 76–91<br />
variable assignment of 53–7<br />
gender, natural 1–9, 185–217, see also sex<br />
gender, social 1–9, 25, 29–32, 44–51, 71–5,<br />
100–2, 137–84, see also gender<br />
reversal<br />
and linguistic gender 185–217<br />
gender bias 194–208, see also male bias,<br />
masculine bias, sexist language<br />
gender etiquette 2, 180<br />
gender reversal 99–118<br />
gender stereotype 4–5, 44–52, 56, 65, 69–74,<br />
109, 130, 164, 185–211, 214–17<br />
gender switch, see gender reversal
Index of subjects 269<br />
gender-exclusive languages 136–51, 161–2,<br />
see also genderlect, gender-variable<br />
languages<br />
gender-inclusive language, see gender-neutral<br />
language<br />
gender-less languages 5, 7, 30, 60–3, 71–2,<br />
87, 92, 123, 127–8<br />
gender-neutral language 5, 61, 73, 188,<br />
198, 217<br />
gender-neutral pronoun 196–200<br />
gender-variable languages 137, 152–62<br />
genderlect 162–3, 208, 214, see also genderexclusive<br />
languages, gender-variable<br />
languages<br />
generic he 6, 86, 195–9, 203–7<br />
generic masculine 12–14, 27–9, 86, 89,<br />
112–16, 188–204, 207–17, see also male<br />
bias, masculine bias<br />
generic noun 4–6, 15, 24–8, 58, 65–6, 72–81,<br />
112–14, 123, see also generic<br />
masculine<br />
generic pronoun 4, 15, see also generic he,<br />
generic masculine, pseudogeneric,<br />
singular <strong>the</strong>y<br />
gesture 8, 140<br />
grammaticalization 76–81, 88, 95<br />
greetings 100, 140<br />
hedges 156<br />
hijra 3, 8, 159–60, 165<br />
Hlonipha 183<br />
honorific 4, 15, 148–51, 174, 204, 207<br />
humanness 5, 13–14, 19–22, 29–30, 63–75,<br />
90–1, 191–2<br />
hybrid nouns 127<br />
inanimate, see animacy<br />
individuation as a function of gender 52,<br />
58, 61<br />
initiation 2–3, 47–8, 103–4, 166–9, 173,<br />
181–2<br />
insults 102–5<br />
interrogatives 26–7, 47, 69, 81<br />
isolating language 15<br />
joking and gender 8, 54, 72–3, 100–2, 106–8,<br />
128, 146, 161, 187, 197<br />
kinship terms 53, 66, 68, 151<br />
language change 76–98, 175–84<br />
language contact 87–97, see also linguistic<br />
area, loans<br />
language endangerment 92, 139, 145, 166,<br />
170, 179, 208, see also language<br />
obsolescence<br />
language loss, see language endangerment,<br />
language obsolescence<br />
language obsolescence 7, 47, 91–4, 137, 145,<br />
189, 199, 202–4<br />
language planning 4, 93–5, see also feminist<br />
language reform, language reform<br />
language reform 88, 93–5, 114, 134,<br />
201–204, 217<br />
lesbian language 3, 9, 11, 159–61, 164, 200, 215<br />
lexicon 52–3<br />
linguistic area 172, 182, see also language<br />
contact<br />
loans 12, 23–5, 28, 38, 41, 46, 56–7, 72,<br />
88–90, 93–7, 138–9, 188–92, 196, 214<br />
male bias 15, 73–4, 208, 217, see also<br />
masculine bias, sexist language<br />
male social gender 1–9, see also gender,<br />
social<br />
male speech 2, 8, 136–65, 214<br />
markedness, formal 25–29, 31–2<br />
markedness, functional 25–9, 31–2, 114–16,<br />
see also generic he, generic noun, generic<br />
pronoun<br />
masculine bias 15, 73–4, 112–14, 127, 189,<br />
194–208<br />
masculine gender 1–9, 13–29, 34–61, 63–75,<br />
83–4, 91–35, 185–211, see also gender,<br />
linguistic<br />
mass noun 35–6<br />
matrilineal societies 115–16, 187, 209–11<br />
metaphor 2, 5, 7–8, 60, 74, 88, 103, 121–5,<br />
132–6, 167, 213
270 Index of subjects<br />
mixed language, Arawak-Carib 142–4<br />
Mo<strong>the</strong>r-in-law style 57, see also avoidance<br />
style<br />
multilingualism 143, 172, 180–1, 184,<br />
see also bilingualism, language contact,<br />
linguistic area<br />
multiple classifier system 63, 68, 78<br />
myth 19–20, 59, 120–1, 166–8<br />
myth-and-belief principle in linguistic gender<br />
choice 19–20, 91, 120–1<br />
myths about gender in English 11<br />
naming practices 137, 194, 204–6, 209<br />
natural gender 29–30, 56–7, 128–30, 133–51,<br />
see also sex<br />
Naven 3, 101, 117, 220<br />
neuter gender 5, 13–16, 22–5, 28–9, 52–3, 60,<br />
72, 83–4, 91–3, 96, 106, 128, 141<br />
nominal hierarchy 75<br />
non-feminine gender, see gender, linguistic<br />
non-rational gender 13, 18<br />
non-verbal communication, see gesture<br />
noun class 13, 16–17, see gender, linguistic<br />
noun classifier, see classifier<br />
noun compounding, see compound<br />
number 15–18, 26–30, 51, 69–71, 82–3, 141<br />
number agreement 16, 29–30<br />
numeral classifier, see classifier<br />
overt marking of linguistic gender 14–16, 54,<br />
58, 81, see also covert marking of<br />
linguistic gender<br />
patrilineal societies 35, 101–2, 133, 179, 206<br />
patronymic 205<br />
pejorative meanings of linguistic<br />
gender 45–51, 111, see also deprecatory<br />
meanings of linguistic gender<br />
personal name, see naming practices<br />
personal pronoun 14–19, 26–30, 38, 71–8,<br />
83–7, 92–6, 99, 102, 110, 115, 137–9,<br />
141, 148, 195, 198, see also generic he,<br />
generic pronoun, singular <strong>the</strong>y<br />
phonological assignment of linguistic<br />
gender 22–5, 31<br />
pitch 2, 8, 154, 158, 161, 166<br />
poetry, gender in 5, 60, 109–10, 120–35<br />
politeness 4, 56, 118, 137, 152–3, 156–7,<br />
163–4, 181, 191, 207, 214, see also<br />
honorific, respect in Japanese 148–51<br />
possession 4, 27, 54, 61, 64, 68, 71, 101, 110,<br />
135, 141, 173, 183, 196<br />
power and gender 8–9, 21, 29, 48, 55, 61, 68,<br />
105, 109, 114–19, 128, 131–3, 153,<br />
156–60, 164, 173–6, 179–81, 204<br />
pragmatic functions of linguistic<br />
gender 57–61<br />
proper name, see naming practices<br />
proverbs, women in 73, 103, 179, 190, 216<br />
pseudogeneric 196–8<br />
queer linguistics 11, 200<br />
questions, see interrogatives, tag questions<br />
rational gender 13, 18<br />
reanalysis 76–7, 81–2, 88–9, 94–5<br />
reinterpretation 51, 63, 76–7, 80–5, 88, 95–6,<br />
107<br />
residual gender 19, 24–5, 30<br />
respect 4, 27, 64–8, 78, 104–7, 116, 122, 137,<br />
149, 153, 160, 166, 174–7, 191–2, 197,<br />
207, see also honorific<br />
ritual language 146, 166–9, 181–4, see also<br />
initiation<br />
Sapir–Whorf hypo<strong>the</strong>sis 208<br />
secret language 169–75, 181, 215, see also<br />
avoidance style, taboo<br />
semantic functions of linguistic<br />
gender 18–22, 51–61<br />
semantic opacity in linguistic gender 5–6,<br />
18–25<br />
semantic transparency in linguistic<br />
gender 5–6, 18–25, 185–6<br />
sex 3, 69–75, 109–18, see also natural gender<br />
in classifier choice 63–9
Index of subjects 271<br />
in linguistic gender choice 1, 13, 18–19,<br />
25, 91–118<br />
sexism 60, 112–14, 208, 216, see also male<br />
bias, masculine bias, sexist language<br />
sexist language 93–4, 100, 112, 114, 189,<br />
194–205, see also male bias, masculine bias<br />
shape in linguistic gender choice 13–14, 18,<br />
34–51, 185–6<br />
sign languages 162, 174, 183<br />
singular <strong>the</strong>y 196–8<br />
size in linguistic gender choice 13–14, 18,<br />
34–51, 97, 102–7, 128–9, 185–6<br />
social change 191–211<br />
societal hierarchies 4, see also honorific,<br />
respect<br />
socio-cultural change 9, 47–51<br />
socio-cultural gender, see gender, social<br />
solidarity and gender 8, 100, 104–9, 158–60,<br />
175, 190, 199<br />
songs 163, 166–8, 179, 181<br />
speech register, see female speech, male speech<br />
Sprachbund, see linguistic area<br />
taboo 169–75, 215, see also<br />
avoidance style<br />
tag questions 156–8<br />
target gender 31<br />
topicality and linguistic gender 57–8<br />
transgender 3, 11, 108, 158–61,<br />
196, 200<br />
transsexual 3, 7, 159–60, 165, 215<br />
transvestite 3, 9, 101, 159–60, 165<br />
two-spirit people 147, 165<br />
value as a semantic parameter in linguistic<br />
gender 7, 44–51, 56, 118, 128–9, 133,<br />
166–7, 190–1, 213<br />
warfare 143<br />
word order 194<br />
writing system 43<br />
Yurupary ritual 105, 169–71, 181<br />
zero marking, see markedness, formal