Linquistics

Language and Gender

PENELOPE ECKERT

SALLY McCONNELL-GINET

published by the press syndicate of the university of cambridge The Pitt Building, Trumpington Street, Cambridge CB2 1RP, United Kingdom

cambridge university press The Edinburgh Building, Cambridge CB2 2RU, UK 40 West 20th Street, New York, NY 10011-4211, USA 477 Williamstown Road, Port Melbourne, VIC 3207, Australia Ruiz de Alarcón 13, 28014 Madrid, Spain Dock House, The Waterfront, Cape Town 8001, South Africa

http://www.cambridge.org

C© Penelope Eckert and Sally McConnell-Ginet 2003

This book is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press.

First published 2003

Printed in the United Kingdom at the University Press, Cambridge

Typeface Swift 10/14 pt System LATEX2ε [tb]

A catalogue record for this book is available from the British Library

Library of Congress Cataloguing in Publication data

ISBN 0 521 65283 9 hardback ISBN 0 521 65426 2 paperback

Contents

List of illustrations vii Acknowledgments ix

Introduction 1

1 Constructing, deconstructing and reconstructing gender 9

Sex and gender 10 Learning to be gendered 15 Keeping gender: the gender order 32 Masculinities and femininities 47 Gender practice 50

2 Linking the linguistic to the social 52

Changing practices, changing ideologies 53 The social locus of change 55 Linguistic resources 60 Analytic practice 79 A matter of method 84

3 Organizing talk 91

Access to situations and events 92 Speech activities 98 Speech situations and events 103 The pursuit of conversation 109 Conversational styles and conversationalists’

character 122

4 Making social moves 129

Speech act theory 130 Functions of talk and motives of talkers: gender

oppositions 133

v

vi Contents

Speech acts embedded in social action 144 Beyond conversation 156

5 Positioning ideas and subjects 157

‘‘Women’s language’’ and gendered positioning 158 Showing deference or respect? 160 Backing down or opening things up? 167 Who cares?: intensity and engagement 176 Calibrating commitment and enlisting support 183 Speaking indirectly 188

6 Saying and implying 192

Case study 192 Aspects of meaning in communicative practice 195 Presupposing: gender schemas and ideologies 203 Assigning roles and responsibility 207 Making metaphors 213

7 Mapping the world 228

Labeling disputes and histories 228 Category boundaries and criteria 232 Category relations 242 Elaborating marked concepts 246 Genderizing discourse: category imperialism 254 Genderizing processes 259 New labels, new categories 261

8 Working the market: use of varieties 266

Languages, dialects, varieties 266 The linguistic market 271 The local and the global 273 Language ideologies and linguistic varieties 276 Case study: standardization and the Japanese woman 278 Gender and language ideologies 281 Gender and the use of linguistic varieties 282 Access 288 Whose speech is more standard? 292

9 Fashioning selves 305

Stylistic practice 306 Style and performativity 315

vii Contents

Legitimate and illegitimate performances 320 One small step 325 Where are we headed? 330

Bibliography 333 Index 357

Illustrations

7.1 US cuts of beef 235 7.2 French cuts of beef 236 7.3 Polarised oppositions 243 7.4 Default background, marked subcategories 243 8.1 The social stratification of (oh) in New York City (from Labov 1972c,

p. 129) 272 8.2 Percent negative concord in Philadelphia by class and gender (casual

speech) (from Labov 2001, p. 265) 296 8.3 (dh) index in Philadelphia by class and gender (casual speech) (from

Labov 2001, p. 265) 298 8.4 Percent reduced-ing in Philadelphia by class and gender (casual

speech) (from Labov 2001, p. 265) 299 8.5 Raising of /ay/ among jock and burnout boys and girls 301 8.6 Height of /æ/ before /s/ in Philadelphia by class (as represented by

occupational group) and gender (from Labov 2001, p. 298) 301

viii

C H A P T E R 1

Constructing, deconstructing and reconstructing gender

We are surrounded by gender lore from the time we are very small.

It is ever-present in conversation, humor, and conflict, and it is called

upon to explain everything from driving styles to food preferences.

Gender is embedded so thoroughly in our institutions, our actions,

our beliefs, and our desires, that it appears to us to be completely

natural. The world swarms with ideas about gender -- and these ideas

are so commonplace that we take it for granted that they are true,

accepting common adage as scientific fact. As scholars and researchers,

though, it is our job to look beyond what appears to be common sense

to find not simply what truth might be behind it, but how it came to

be common sense. It is precisely because gender seems natural, and

beliefs about gender seem to be obvious truth, that we need to step

back and examine gender from a new perspective. Doing this requires

that we suspend what we are used to and what feels comfortable, and

question some of our most fundamental beliefs. This is not easy, for

gender is so central to our understanding of ourselves and of the world

that it is difficult to pull back and examine it from new perspectives.1

But it is precisely the fact that gender seems self-evident which makes

the study of gender interesting. It brings the challenge to uncover the

process of construction that creates what we have so long thought

of as natural and inexorable -- to study gender not as given, but as

an accomplishment; not simply as cause, but as effect. The results of

failure to recognize this challenge are manifest not only in the popular

media, but in academic work on language and gender as well. As a

result, some gender scholarship does as much to reify and support

existing beliefs as to promote more reflective and informed thinking

about gender.

1 It is easier, though, for people who feel that they are disadvantaged in the social order, and it is no doubt partially for this reason that many recent theories of gender have been developed primarily (though not exclusively) by women. (In some times and places, women have not had the opportunity to develop ‘‘theories’’ of anything.)

9

10 Language and Gender

Sex and gender

Gender is not something we are born with, and not something we

have, but something we do (West and Zimmerman 1987) -- something we perform (Butler 1990). Imagine a small boy proudly following his father. As he swaggers and sticks out his chest, he is doing everything

he can to be like his father -- to be a man. Chances are his father is not swaggering, but the boy is creating a persona that embodies what he is

admiring in his adult male role model. The same is true of a small girl

as she puts on her mother’s high-heeled shoes, smears makeup on her

face and minces around the room. Chances are that when these chil-

dren are grown they will not swagger and mince respectively, but their

childhood performances contain elements that will no doubt surface in

their adult male and female behaviors. Chances are, also, that the girl

will adopt that swagger on occasion as well, but adults are not likely

to consider it as ‘‘cute’’ as her mincing act. And chances are that if the

boy decides to try a little mincing, he won’t be considered cute at all.

In other words, gendered performances are available to everyone, but

with them come constraints on who can perform which personae with

impunity. And this is where gender and sex come together, as society

tries to match up ways of behaving with biological sex assignments.

Sex is a biological categorization based primarily on reproductive

potential, whereas gender is the social elaboration of biological sex.

Gender builds on biological sex, it exaggerates biological difference

and, indeed, it carries biological difference into domains in which it is

completely irrelevant. There is no biological reason, for example, why

women should mince and men should swagger, or why women should

have red toenails and men should not. But while we think of sex as

biological and gender as social, this distinction is not clear-cut. People

tend to think of gender as the result of nurture -- as social and hence

fluid -- while sex is simply given by biology. However, there is no obvious

point at which sex leaves off and gender begins, partly because there

is no single objective biological criterion for male or female sex. Sex is

based in a combination of anatomical, endocrinal and chromosomal

features, and the selection among these criteria for sex assignment is

based very much on cultural beliefs about what actually makes some-

one male or female. Thus the very definition of the biological categories

male and female, and people’s understanding of themselves and others as male or female, is ultimately social. Anne Fausto-Sterling (2000) sums

up the situation as follows:

labeling someone a man or a woman is a social decision. We may use scientific knowledge to help us make the decision, but only our beliefs

11 Constructing gender

about gender -- not science -- can define our sex. Furthermore, our beliefs about gender affect what kinds of knowledge scientists produce about sex in the first place. (p. 3)

Biology offers us up dichotomous male and female prototypes, but it

also offers us many individuals who do not fit those prototypes in a

variety of ways. Blackless et al. (2000) estimate that 1 in 100 babies are born with bodies that differ from standard male or female. These bod-

ies may have such conditions as unusual chromosomal makeup (1 in

1,000 male babies are born with two X chromosomes), hormonal dif-

ferences such as insensitivity to androgens (1 in 13,000 births), or a

range of configurations and combinations of genitals and reproductive

organs. The attribution of intersex does not end at birth -- 1 in 66 girls

experience growth of the clitoris in childhood or adolescence (known

as late onset adrenal hyperplasia).

When ‘‘anomalous” babies are born, surgical and/or endocrinal ma-

nipulations may be used to bring their recalcitrant bodies into closer

conformity with either the male or the female category. Common med-

ical practice imposes stringent requirements for male and female gen-

itals at birth -- a penis that is less than 2.5 centimeters long when

stretched, or a clitoris2 that is more than one centimeter long are

both commonly subject to surgery in which both are reduced to an

‘‘acceptable” sized clitoris (Dreger 1998). As a number of critics have

observed (e.g. Dreger 1998), the standards of acceptability are far more

stringent for male genitals than female, and thus the most common

surgery transforms ‘‘unacceptable” penises into clitorises, regardless of

the child’s other sexual characteristics, and even if this requires fash-

ioning a nonfunctional vagina out of tissue from the colon. In recent

years, the activist organization, the Intersex Society of North America,3

has had considerable success as an advocacy group for the medical

rights of intersex people.

In those societies that have a greater occurrence of certain kinds

of hermaphroditic or intersexed infants than elsewhere,4 there

2 Alice Dreger (1998) more accurately describes these as a phallus on a baby classified as male or a phallus on a baby classified as female. 3 The website of the Intersex Society of North America (http://www.isna.org) offers a wealth of information on intersex. [The publisher has used its best endeavors to ensure that the URLs for external websites referred to in this book are correct and active at the time of going to press. However, the publisher has no responsibility for the websites and can make no guarantee that a site will remain live or that the content is or will remain appropriate.] 4 For instance, congenital adrenal hyperplasia (which combines two X chromosomes with masculinized external genitalia and the internal reproductive organs of a potentially fertile woman) occurs in 43 children per million in New Zealand, but 3,500 per million among the Yupik of Southwestern Alaska (www.isna.org).

12 Language and Gender

sometimes are social categories beyond the standard two into which

such babies can be placed. But even in such societies, categories that

go beyond the basic two are often seen as anomalous.5

It is commonly argued that biological differences between males and

females determine gender by causing enduring differences in capabili-

ties and dispositions. Higher levels of testosterone, for example, are said

to lead men to be more aggressive than women; and left-brain dom-

inance is said to lead men to be more ‘‘rational’’ while their relative

lack of brain lateralization should lead women to be more ‘‘emotional.’’

But the relation between physiology and behavior is not simple, and it

is all too easy to leap for gender dichotomies. It has been shown that

hormonal levels, brain activity patterns, and even brain anatomy can

be a result of different activity as well as a cause. For example research

with species as different as rhesus monkeys (Rose et al. 1972) and fish (Fox et al. 1997) has documented changes in hormone levels as a result of changes in social position. Work on sex differences in the brain is

very much in its early stages, and as Anne Fausto-Sterling (2000) points

out in considerable detail, it is far from conclusive. What is supposed

to be the most robust finding -- that women’s corpus callosum, the link

between the two brain hemispheres, is relatively larger than men’s -- is

still anything but robust. Men’s smaller corpus callosum is supposed to

result in greater lateralization, while women’s larger one is supposed

to yield greater integration between the two hemispheres, at least in

visuo-spatial functions. But given that evidence for sex-linked brain dif-

ferences in humans is based on very small samples, often from sick or

injured populations, generalizations about sex differences are shaky at

best. In addition, not that much is known about the connections be-

tween brain physiology and cognition -- hence about the consequences

of any physiological differences scientists may be seeking or finding.

Nonetheless, any results that might support physiological differences

are readily snatched up and combined with any variety of gender stereo-

types in some often quite fantastic leaps of logic. And the products of

these leaps can in turn feed directly into social, and particularly into

5 There are cultures where what we might think of as more than two adult gender categories are named and otherwise institutionally recognized as well: the berdache of the Plains Indians, the hijras in India. Although details vary significantly, the members of such supernumerary categories are outside the ‘‘normal’’ order of things, and tend to be somewhat feared or devalued or otherwise socially disadvantaged. Nonetheless, there is apparently considerably more tolerance for nonstandard gender categories in some societies than in the western industrial societies most likely to be familiar to readers of this book. An early discussion of social groups with more than two sex and/or gender categories is provided by Martin and Voorhies (1975), ch. 4, ‘‘Supernumerary sexes.’’ More recent contributions on this topic from both historical and cross-cultural perspectives appear in Herdt (1996).

13 Constructing gender

educational, policy, with arguments that gender equity in such ‘‘left-

brain areas’’ as mathematics and engineering is impossible.

The eagerness of some scientists to establish a biological basis for

gender difference, and the public’s eagerness to take these findings

up, points to the fact that we put a good deal of work into emphasiz-

ing, producing, and enforcing the dichotomous categories of male and

female. In the process, differences or similarities that blur the edges

of these categories, or that might even constitute other potential cate-

gories, are backgrounded, or erased. The issue here is not whether there are sex-linked biological differ-

ences that might affect such things as predominant cognitive styles.

What is at issue is the place of such research in social and scientific

practice. Sex difference is being placed at the center of activity, as both

question and answer, as often flimsy evidence of biological difference

is paired up with unanalyzed behavioral stereotypes. And the results

are broadcast through the most august media as if their scientific sta-

tus were comparable to the mapping of the human genome. The mere

fact of this shows clearly that everyone, from scientists to journalists to

the reading public, has an insatiable appetite for sensationalist gender

news. Indeed, gender is at the center of our social world. And any evi-

dence that our social world maps onto the biological world is welcome

evidence to those who would like an explanation and justification for

the way things are.

To whatever extent gender may be related to biology, it does not flow

naturally and directly from our bodies. The individual’s chromosomes,

hormones, genitalia, and secondary sex characteristics do not deter-

mine occupation, gait, or use of color terminology. And while male

pattern baldness may restrict some adult men’s choice of hairdo, there

are many men who could sport a pageboy or a beehive as easily as many

women, and nothing biological keeps women from shaving their heads.

Gender is the very process of creating a dichotomy by effacing similar-

ity and elaborating on difference, and even where there are biological

differences, these differences are exaggerated and extended in the ser-

vice of constructing gender. Actual differences are always paired with

enormous similarities, never dichotomizing people but putting them

on a scale with many women and men occupying the same positions.

Consider our voices. On average, men’s vocal tracts are longer than

women’s, yielding a lower voice pitch. But individuals’ actual conver-

sational voice pitch across society does not simply conform to the size

of the vocal tract. At the age of four to five years, well before puberty

differentiates male and female vocal tracts, boys and girls learn to

differentiate their voices as boys consciously and unconsciously lower

14 Language and Gender

their voices while girls raise theirs. In the end, one can usually tell

whether even a very small child is male or female on the basis of their

voice pitch and quality alone, regardless of the length of their vocal

tract.

Relative physical stature is another biological difference that is elab-

orated and exaggerated in the production of gender. Approximately

half of the women and half of the men in the USA (Kuczmarski et al. 2000) are between 64 and 70 inches tall. With this considerable overlap,

one might expect in any randomly chosen male and female pair that

the woman would run a good chance of being taller than the man.

In actuality, among heterosexual couples, one only occasionally sees

such a combination, because height is a significant factor in people’s

choice of a heterosexual mate. While there is no biological reason for

women to be shorter than their male mates, an enormous majority

of couples exhibit this height relation -- far more than would occur

through a process of selection in which height was random (Goffman

1976). Not only do people mate so as to keep him taller than her, they

also see him as taller than her even when this is not the case. For

example, Biernat, Manis, and Nelson 1991 (cited in Valian 1998) pre-

sented college students with photos of people and asked them to guess

the people’s height. Each photo had a reference item like a doorway

or a desk, making it possible to compare the heights of people across

photos. Although photos of a male of a given height were matched by

photos of a female of the same height (and vice versa), the judges saw

the males as taller than they actually were and the females as shorter

than they actually were.

This book will focus on gender as a social construction -- as the means

by which society jointly accomplishes the differentiation that consti-

tutes the gender order. While we recognize that biology imposes certain

physiological constraints on the average male and female, we treat the

elaboration and magnification of these differences as entirely social.

Readers will come to this book with their own set of beliefs about the

origins and significance of gender. They may have certain understand-

ings of the implications for gender of biological and medical science.

They may subscribe to a particular set of religious beliefs about gen-

der. The notion of the social elaboration of sex is not incompatible

with belief in a biological or divine imperative -- the difference will be

in where one leaves off and the other begins. All we ask of our readers

is that they open-mindedly consider the evidence and arguments we

advance. Our own thinking about gender has developed and changed

over many years of thinking about these issues, and it will undoubt-

edly continue to change as we continue to explore gender issues in our

15 Constructing gender

research and in our lives. We have written this account of gender from

a broadly feminist perspective. As we understand that perspective, the

basic capabilities, rights, and responsibilities of women and men are

far less different than is commonly thought. At the same time, that

perspective also suggests that the social treatment of women and men,

and thus their experiences and their own and others’ expectations for

them, is far more different than is usually assumed. In this book we

offer evidence that these differences in what happens to women and to

men derive in considerable measure from people’s beliefs about sexual

difference, their interpretations of its significance, and their reliance

on those beliefs and interpretations to justify the unequal treatment

of women and men.

Learning to be gendered

Dichotomous beginnings: It’s a boy! It’s a girl!

In the famous words of Simone de Beauvoir, ‘‘Women are not born,

they are made.’’ The same is true of men. The making of a man or

a woman is a never-ending process that begins before birth -- from

the moment someone begins to wonder if the pending child will be a

boy or a girl. And the ritual announcement at birth that it is in fact

one or the other instantly transforms an ‘‘it’’ into a ‘‘he’’ or a ‘‘she’’

(Butler 1993), standardly assigning it to a lifetime as a male or as a

female.6 This attribution is further made public and lasting through

the linguistic event of naming. To name a baby Mary is to do something that makes it easy for a wide range of English speakers to maintain the

initial ‘‘girl’’ attribution. In English-speaking societies, not all names are

sex-exclusive (e.g. Chris, Kim, Pat), and sometimes names change their gender classification. For example, Evelyn was available as a male name in Britain long after it had become an exclusively female name in

America, and Whitney, once exclusively a surname or a male first name in America, is now bestowed on baby girls. In some times and places,

the state or religious institutions disallow sex-ambiguous given names.

Finland, for example, has lists of legitimate female and legitimate male

names that must be consulted before the baby’s name becomes official.

Thus the dichotomy of male and female is the ground upon which we

build selves from the moment of birth. These early linguistic acts set

6 Nowadays, with the possibility of having this information before birth, wanting to know in advance or not wanting to know can become ideologically charged. Either way, the sex of the child is frequently as great a preoccupation as its health.

16 Language and Gender

up a baby for life, launching a gradual process of learning to be a boy or

a girl, a man or a woman, and to see all others as boys or girls, men or

women as well. There are currently no other legitimate ways to think

about ourselves and others -- and we will be expected to pattern all

kinds of things about ourselves as a function of that initial dichotomy.

In the beginning, adults will do the child’s gender work, treating it as a

boy or as a girl, and interpreting its every move as that of a boy or of a

girl. Then over the years, the child will learn to take over its part of the

process, doing its own gender work and learning to support the gender

work of others. The first thing people want to know about a baby is its

sex, and convention provides a myriad of props to reduce the necessity

of asking -- and it becomes more and more important, as the child

develops, not to have to ask. At birth, many hospital nurseries provide

pink caps for girls and blue caps for boys, or in other ways provide some

visual sign of the sex that has been attributed to the baby. While this

may seem quite natural to members of the society, in fact this color

coding points out no difference that has any bearing on the medical

treatment of the infants. Go into a store in the US to buy a present

for a newborn baby, and you will immediately be asked ‘‘boy or girl?’’

If the reply is ‘‘I don’t know’’ or, worse, ‘‘I don’t care,’’ sales personnel

are often perplexed. Overalls for a girl may be OK (though they are

‘‘best’’ if pink or flowered or in some other way marked as ‘‘feminine’’),

but gender liberalism goes only so far. You are unlikely to buy overalls

with vehicles printed on them for a girl, and even more reluctant to

buy a frilly dress with puffed sleeves or pink flowered overalls for a

boy. And if you’re buying clothing for a baby whose sex you do not

know, sales people are likely to counsel you to stick with something

that’s plain yellow or green or white. Colors are so integral to our way

of thinking about gender that gender attributions have bled into our

view of the colors, so that people tend to believe that pink is a more

‘‘delicate’’ color than blue. This is a prime example of the naturalization

of what is in fact an arbitrary sign. In America in the late nineteenth

and early twentieth centuries, Anne Fausto-Sterling (2000) reports, blue

was favored for girls and bright pink for boys.

If gender flowed naturally from sex, one might expect the world to sit

back and simply allow the baby to become male or female. But in fact,

sex determination sets the stage for a lifelong process of gendering,

as the child becomes, and learns how to be, male or female. Names

and clothing are just a small part of the symbolic resources used to

support a consistent ongoing gender attribution even when children

are clothed. That we can speak of a child growing up as a girl or as a boy suggests that initial sex attribution is far more than just a simple

17 Constructing gender

observation of a physical characteristic. Being a girl or being a boy is not a stable state but an ongoing accomplishment, something that is actively

done both by the individual so categorized and by those who interact with it in the various communities to which it belongs. The newborn

initially depends on others to do its gender, and they come through in many different ways, not just as individuals but as part of socially

structured communities that link individuals to social institutions and

cultural ideologies. It is perhaps at this early life stage that it is clearest

that gender is a collaborative affair -- that one must learn to perform

as a male or a female, and that these performances require support

from one’s surroundings.

Indeed, we do not know how to interact with another human being

(or often members of other species), or how to judge them and talk

about them, unless we can attribute a gender to them. Gender is so

deeply engrained in our social practice, in our understanding of our-

selves and of others, that we almost cannot put one foot in front of the

other without taking gender into consideration. Although most of us

rarely notice this overtly in everyday life, most of our interactions are

colored by our performance of our own gender, and by our attribution

of gender to others.

From infancy, male and female children are interpreted differently,

and interacted with differently. Experimental evidence suggests that

adults’ perceptions of babies are affected by their beliefs about the

babies’ sex. Condry and Condry (1976) found that adults watching a

film of a crying infant were more likely to hear the cry as angry if

they believed the infant was a boy, and as plaintive or fearful if they

believed the infant was a girl. In a similar experiment, adults judged

a 24-hour-old baby as bigger if they believed it to be a boy, and finer-

featured if they believed it to be a girl (Rubin, Provenzano and Luria

1974). Such judgments then enter into the way people interact with

infants and small children. People handle infants more gently when

they believe them to be female, more playfully when they believe them

to be male.

And they talk to them differently. Parents use more diminutives

(kitty, doggie) when speaking to girls than to boys (Gleason et al. 1994), they use more inner state words (happy, sad ) when speaking to girls (Ely et al. 1995). They use more direct prohibitives (don’t do that! ) and more emphatic prohibitives (no! no! no! ) to boys than to girls (Bellinger and Gleason 1982). Perhaps, one might suggest, the boys need more

prohibitions because they tend to misbehave more than the girls. But

Bellinger and Gleason found this pattern to be independent of the ac-

tual nature of the children’s activity, suggesting that the adults and

18 Language and Gender

their beliefs about sex difference are far more important here than the

children’s behavior.

With differential treatment, boys and girls eventually learn to be different. Apparently, male and female infants cry the same amount

(Maccoby and Jacklin 1974), but as they mature, boys cry less and less.

There is some evidence that this difference emerges primarily from

differential adult response to the crying. Qualitative differences in be-

havior come about in the same way. A study of thirteen-month-old

children in day care (Fagot et al. 1985) showed that teachers responded to girls when they talked, babbled, or gestured, while they responded

to boys when they whined, screamed, or demanded physical attention.

Nine to eleven months later, the same girls talked more than the boys,

and the boys whined, screamed, and demanded attention more than

the girls. Children’s eventual behavior, which seems to look at least sta-

tistically different across the sexes, is the product of adults’ differential

responses to ways of acting that are in many (possibly most) cases very

similar indeed. The kids do indeed learn to ‘‘do’’ gender for themselves,

to produce sex-differentiated behavior -- although even with consider-

able differential treatment they do not end up with dichotomizing

behavioral patterns.

Voice, which we have already mentioned, provides a dramatic ex-

ample of children’s coming to perform gender. At the ages of four to

five years, in spite of their identical vocal apparatus, girls and boys be-

gin to differentiate the fundamental frequency of their speaking voice.

Boys tend to round and extend their lips, lengthening the vocal tract,

whereas girls are tending to spread their lips (with smiles, for example),

shortening the vocal tract. Girls are raising their pitches, boys lowering

theirs. It may well be that adults are more likely to speak to girls in a

high-pitched voice. It may be that they reward boys and girls for differ-

ential voice productions. It may also be that children simply observe

this difference in older people, or that their differential participation

in games (for example play-acting) calls for different voice productions.

Elaine Andersen (1990, pp. 24--25), for example, shows that children use

high pitch when using baby talk or ‘‘teacher register’’ in role play. Some

children speak as the other sex is expected to and thus, as with other

aspects of doing gender, there is not a perfect dichotomization in voice

pitch (even among adults, some voices are not consistently classified).

Nonetheless, there is a striking production of mostly different pitched

voices from essentially similar vocal equipment.

There is considerable debate among scholars about the extent to

which adults actually do treat boys and girls differently, and many

note that the similarities far outweigh the differences. Research on

19 Constructing gender

early gender development -- in fact the research in general on gender

differences -- is almost exclusively done by psychologists. As a result, the

research it reports on largely involves observations of behavior in lim-

ited settings -- whether in a laboratory or in the home or the preschool.

Since these studies focus on limited settings and types of interaction

and do not follow children through a normal day, they quite possibly

miss the cumulative effects of small differences across many different

situations. Small differences here and there are probably enough for

children to learn what it means in their community to be male or

female.

The significance of the small difference can be appreciated from an-

other perspective. The psychological literature tends to treat children

as objects rather than subjects. Those studying children have tended

to treat others -- parents, other adults, peers -- as the primary social-

izing agents. Only relatively recently have investigators begun to ex-

plore children’s own active strategies for figuring out the social world.

Eleanor Maccoby (2002) emphasizes that children have a very clear

knowledge of their gender (that is, of whether they are classified as

male or female) by the time they are three years old. Given this knowl-

edge, it is not at all clear how much differential treatment children

need to learn how to do their designated gender. What they mainly

need is the message that male and female are supposed to be differ-

ent, and that message is everywhere around them.

It has become increasingly clear that children play a very active role

in their own development. From the moment they see themselves as so-

cial beings, they begin to focus on the enterprise of ‘‘growing up.’’ And

to some extent, they probably experience many of the gendered devel-

opmental dynamics we discuss here not so much as gender-appropriate,

but as grown-up. The greatest taboo is being ‘‘a baby,’’ but the devel- opmental imperative is gendered. Being grown-up, leaving babyhood,

means very different things for boys than it does for girls. And the

fact that growing up involves gender differentiation is encoded in the

words of assessment with which progress is monitored -- kids do not

behave as good or bad people, but as good boys or good girls, and they develop into big boys and big girls.7 In other words, they do not have the option of growing into just people, but into boys or girls. This does not

mean that they see what they’re doing in strictly gendered terms. It is

probable that when boys and girls alter the fundamental frequency of

their voices they are not trying to sound like girls or like boys, but that

7 Thorne (1993) and others have observed teachers urging children to act like ‘‘big boys and girls.’’ Very rarely is a child told ‘‘don’t act like a baby -- you’re a big kid now.’’

20 Language and Gender

they are aspiring for some quality that is itself gendered -- cuteness,

authority. And the child’s aspiration is not simply a matter of reason-

ing, but a matter of desire -- a projection of the self into desired forms

of participation in the social world. Desire is a tremendous force in

projecting oneself into the future -- in the continual remaking of the

self that constitutes growing up.

Until about the age of two, boys and girls exhibit the same play be-

haviors. After that age, play in boys’ and girls’ groups begins to diverge

as they come to select different toys and engage in different activities,

and children begin to monitor each other’s play, imposing sanctions on

gender-inappropriate play. Much is made of the fact that boys become

more agonistic than girls, and many attribute this to hormonal and

even evolutionary differences (see Maccoby 2000 for a brief review of

these various perspectives). But whatever the workings of biology may

be, it is clear that this divergence is supported and exaggerated by the

social system. As children get older, their play habits are monitored

and differentiated, first by adults, and eventually by peers. Parents of

small children have been shown to reward their children’s choice of

gender-appropriate toys (trucks for boys, dolls for girls) (Langlois and

Downs 1980). And while parents’ support of their children’s gendered

behavior is not always and certainly not simply a conscious effort at

gender socialization, their behavior is probably more powerful than

they think. Even parents who strive for gender equality, and who be-

lieve that they do not constrain their children’s behavior along gender

lines, have been observed in experimental situations to do just that.

Learning asymmetry

While it takes a community to develop gender, not all participants in

the community are equally involved in enforcing difference. In research

on early gender socialization, males -- both children and adults -- have

emerged as more engaged in enforcing gender difference than females.

In the research by Rubin et al. cited above, for example, fathers were more extreme than mothers in their gender-based misassessments of

infants’ size and texture. Men are more likely than women to play

rough with boys and gently with girls, fathers use differential language

patterns to boys and girls more than mothers, and men are more likely

than women to reward children for choosing gender-appropriate toys.

There are now books aimed at men who want to become more involved

parents than their own fathers were. But the message is still often that

parenting a girl is quite a different enterprise from parenting a boy. On

a self-help shelf encountered at a tourist shop, How to Be Your Daughter’s

21 Constructing gender

Daddy: 365 Ways to Show Her You Care by Dan Bolin (1993) stood right next to How to Be Your Little Man’s Dad: 365 Things to Do with Your Son by Dan Bolin and Ken Sutterfield (1993).

It is not only that male adults seem to enforce gender more than

female. This enforcement is more intensely aimed at boys than at girls.

Adults are more likely to reward boys for choice of gender-appropriate

toys than girls -- and fathers are more likely to do so for their own

sons than for other boys. Boys, in turn, are more rigid in their toy

preferences than girls, and they are harder on other boys than on girls

for gender-inappropriate play styles. A study of three to five year olds

(Langlois and Downs 1980) showed that while girls tended to be neu-

tral about other girls’ choices, boys responded positively only to boys

with male play styles, and were especially likely to punish their male

peers for feminine choices. The outcome is that while activities and

behaviors labeled as male are treated as appropriate for females as well as for males, those labeled as female are treated as appropriate only for females. One way of looking at this is that female activities and

behaviors emerge as marked -- as reserved for a special subset of the population -- while male activities and behaviors emerge as unmarked or normal. This in turn contributes to the androcentric (male-centered) view of gender, which we will discuss in the following section of this

chapter.

This asymmetry is partially a function of the cultural devaluation of

women and of the feminine. One way or another, most boys and girls

learn that most boy things and boy activities are more highly valued

than girl things and girl activities, and boys are strongly discouraged

from having interests or activities that are associated with girls. Even

where they do not encounter such views formulated explicitly or even

find them denied explicitly, most boys and girls learn that it is pri-

marily men and not women who do ‘‘important’’ things as adults, have

opinions that count, direct the course of events in the public world.

It is hardly surprising then that pressures towards gender conformity

are not symmetrical.

This asymmetry extends to many domains. While females may wear

clothing initially viewed as male, the reverse is highly stigmatized:

western women and girls now wear jeans but their male peers are not

appearing in skirts. Even names seem to go from male to female and

not vice versa. There are girls named Christopher, but no boys named

Christine. A girl may be sanctioned for behaving ‘‘like a boy’’ -- particu-

larly if she behaves aggressively, and gets into fights -- on the grounds

that she is being ‘‘unladylike’’ or ‘‘not nice.’’ But there is a categoriza-

tion of ‘‘tomboy’’ reserved for girls who adopt a male rough and tumble

22 Language and Gender

style of play, who display fearlessness and refuse to play with dolls. And

while in some circles this categorization may be considered negative,

in general in western society it earns some respect and admiration.

Boys who adopt girls’ behaviors, on the other hand, are severely sanc-

tioned. The term ‘‘sissy’’ is reserved for boys who do not adhere strictly

to norms of masculinity (in fact, a sissy is a boy who does not display

those very characteristics that make a girl a tomboy).

A child who’s told she has to do more housework than her brother

because she’s a girl, or that she can’t be an astronaut when she grows

up because she’s a girl,8 is likely to say ‘‘that’s not fair!’’ A boy who is

told he cannot play with dolls because he’s a boy, or that he cannot

be a secretary when he grows up, may find that unfair as well. But

the boy who is told he can’t be a nurse is being told that he is too

good to be a nurse. The girl, on the other hand, is essentially being

told that she is not good enough to be a doctor. This is not to say that

the consequences cannot be tragic for the boy who really wants to play

with dolls or grow up to be a nurse. He will be deprived of a legitimate

sense of unfairness within society’s wider discourses of justice, hence

isolated with his sense of unfairness. But gender specialization does

carry the evaluation that men’s enterprises are generally better than

women’s, and children learn this quite early on.9

Now there are some counterexamples to these general trends, many

of them prompted by the feminist and gay rights movements. Some

men are taking over domestic tasks like diaper-changing and every-

day cookery that were once women’s province. Others wear jewels in

their ears or gold chains around their necks, adornments reserved

for women when we were teenagers. But the dominant pattern that

restricts men in moving into what are seen as women’s realms and

thereby devalued is by no means dead.

Separation

To differing degrees from culture to culture and community to commu-

nity, difference is reinforced by separation. Boys play more with boys;

8 These examples may seem anachronistic, but such explicit messages persist. The first is reported by some of the young women in our classes at Stanford and Cornell (though certainly not by all or even most). And the second message was relayed to astronaut Sally Ride in 2001 by a girl whose teacher had offered her that discouragement. 9 Even a child whose own mother is a physician is sometimes heard saying ‘‘ladies can’t be doctors.’’ Of course kids sometimes get it wrong. An anecdote circulated during Margaret Thatcher’s time as prime minister told of a young English boy asked ‘‘do you want to be prime minister when you grow up?’’ ‘‘Oh no,’’ he replied, ‘‘that’s a woman’s job.’’

23 Constructing gender

girls with girls. And this pattern repeats itself cross-culturally, in nonin-

dustrial societies as well as in industrial societies (Whiting and Edwards

1988). The extent to which individuals in western industrial countries

grow up participating in same-sex playgroups varies tremendously, de-

pending on such things as the genders and ages of their siblings and

their neighbors. Some kids spend more time in same-sex groups at one

stage of their lives, less at other stages. The fact remains that however

much kids may play in mixed-sex groups, there is a tendency to seek

out -- and to be constrained to seek out -- same-sex groups. This con-

straint is stronger for boys -- girls who prefer playing with boys are toler-

ated, perhaps admired, while boys who prefer playing with girls are not.

Psychological research shows that many American children begin to

prefer same-sex playmates as they approach the age of three (Maccoby

1998), which is about the age at which they develop a clear sense of

their own gender, and this preference increases rapidly as they age.

Eleanor Maccoby notes that this preference emerges in institutional

settings -- day care, preschool, and elementary school -- where children

encounter large numbers of age peers. On the same theme, Thorne

(1993) points out that schools provide a sufficiently large population

that boys and girls can separate, whereas in neighborhoods there may

be less choice.

Even though children lean towards same-sex groups in these settings,

they often maintain prior cross-sex friendships formed outside the in-

stitution (Howes 1988). It is important to note that the preference for

same-sex play groups is not absolute, and that in fact children often

play in mixed groups. Maccoby and Jacklin’s study (1987) of individual

children’s choice of playmates in a preschool setting shows four and

a half year olds playing in same-sex groups 47 percent of the time,

mixed groups 35 percent of the time and other-sex groups (i.e., where

the child is the only representative of her or his own sex in the group)

18 percent of the time. While these figures show a good deal of mixing,

the same-sex groups are far greater than random playmate selection

would produce. And at age six and a half, children in the Maccoby

and Jacklin study were playing in same-sex groups 67 percent of the

time. Maccoby (1998, pp. 22--23) suggests that the choice of playmates

in school is a strategy for ensuring safety and predictability in an open

setting, as children seek out others with a recognizable play style. This

presupposes different play styles to begin with, presenting a compli-

cated chicken-and-egg problem. For if sex-segregated play groups fill a

need for predictable play and interaction styles, they are also a poten-

tial site for the production and reproduction of this differentiation. It

has been overwhelmingly established that small boys engage in more

24 Language and Gender

physically aggressive behavior than small girls. However, experimental

and observational evidence puts this differentiation at precisely the

same time that same-sex group preference emerges. Maccoby points

out that this play style reaches its peak among boys at about the age

of four and that it is restricted to same-sex groups, suggesting that

there is a complex relation between the emergence of gendered play

styles and of same-sex play groups.

The separation of children in same-sex play groups has led some

gender theorists to propose a view that by virtue of their separation

during a significant part of their childhoods, boys and girls are social-

ized into different peer cultures. In their same-sex friendship groups, they develop different behavior, different norms, and even different

understandings of the world. Daniel Maltz and Ruth Borker (1982) ar-

gue that because of this separation, boys and girls develop different

verbal cultures -- different ways of interacting verbally and different

norms for interpreting ways of interacting. They argue, further, that

this can result in cross-cultural miscommunication between males and fe- males. Deborah Tannen (1990) has popularized this view, emphasizing

the potential for misunderstanding. The separation of gender cultures

does not necessarily entail male--female misunderstanding, although

it describes the conditions under which such misunderstanding could

develop. Certainly, if girls and boys are segregated on a regular basis,

we can expect that they will develop different practices and different

understandings of the world. The extent to which this actually occurs

depends on the nature of the segregation -- when, in what contexts,

for what activities -- in relation to the actual contact between boys

and girls. In other words, to the extent that there is separation, this

separation is structured -- and it is structured differently in different

communities. This structure will have an important bearing on the na-

ture of differences that will develop. It will also have a bearing on the

extent to which these differences are recognized.

The miscommunication model that Maltz and Borker proposed and

that Tannen has further developed draws on John Gumperz’s work

with ethnically distinct subcultures (e.g. Gumperz 1982). It hypothe-

sizes both that male and female understandings of interaction are in

fact different, and, critically, that they are unaware of these differences,

and believe that they are operating from the same understanding. It is

the unawareness that may be the most problematic assumption for this

approach to gender-based miscommunication (or conflict), since the

gender beliefs that most kids are industriously acquiring in their peer

groups and outside them emphasize difference, to the point sometimes

25 Constructing gender

of absurd exaggeration. Gender segregation in childhood almost cer-

tainly plays some role in the development of gendered verbal practice.

But for understanding gender, separation is never the whole picture.

Gender segregation in western societies is virtually always embedded

in practices that bring the sexes together and that impose difference in

interpretations even where there are great similarities in those actions

or people being interpreted.

As we move farther along in development, the complexity of explain-

ing gender differences increases exponentially. As kids spend more

time with their peers, and as they enter into more kinds of situa-

tions with peers, not only does the balance between adult and peer

influence change, but the nature of peer influence also changes. Peer

society becomes increasingly complex, and at some point quite early

on, explicit ideas about gender enter into children’s choices, prefer-

ences, and opportunities. Whatever the initial factors that give rise to

increasing gender separation, separation itself becomes an activity, and

a primary social issue. Barrie Thorne (1993) notes that public choosing

of teams in school activities constrains gender segregation, hence that

games that involve choosing teams are more likely to be same gender,

while games that simply involve lining up or being there are more

likely to be gender-mixed. Separation can carry over to competitions

and rivalries between boys’ groups and girls’ groups, as in elementary

school activities such as ‘‘girls chase the boys’’ (Thorne 1993). These ac-

tivities can be an important site for the construction of difference with

claims that girls or boys are better at whatever activity is in question.

In this way, beliefs about differences in males’ and females’ ‘‘natural’’

abilities may be learned so young and so indirectly that they appear to

be common sense. It is not at all clear, therefore, to what extent differ-

ences in behaviors and activities result from boys’ and girls’ personal

preference, or from social constraint.

The heterosexual market

Towards the end of elementary school, a highly visible activity of pair-

ing up boys and girls into couples begins to dominate the scene. This

activity is not one engaged in by individual children, and it is not an

activity that simply arises in the midst of other childhood ‘‘business

as usual.’’ Rather, it is the beginning of a social market that forms

the basis of an emerging peer social order (Eckert 1996). And with this

market comes a profound change in the terms of gender separation

and difference.

26 Language and Gender

In childhood, it is primarily adults who attend to children’s behavior.

As the peer social order develops, it takes over much of this function

as it develops the means to organize its own social control. Heterosex-

uality is the metaphor around which the peer social order organizes

itself, and a heterosexual market (Thorne 1993) becomes the center of

the emerging peer social order. While up until now, boys and girls may

have seen themselves as simply different, and perhaps as incompatible,

in the context of the heterosexual market, boys and girls emerge as

complementary and cooperating factions.

The market metaphor is not frivolous, for the heterosexual market is

the first of a series of social markets that the age cohort will engage in

on the way to, for example, the academic market and the job market. It

is here that both girls and boys will come to see themselves as having

a place in a structured system of social evaluation. Kids participating

in the heterosexual market can act as both commodity and as broker --

they can be paired up, or they can engage in negotiating the pairing

up of others. The matches that are made on this market are initially

short-lived -- a pair may remain ‘‘together’’ for a few hours, a few days, a

week, sometimes longer. It is the rapidity of ‘‘trades’’ on the market that

establishes individuals’ value, and that establishes the nature of value.

The rapt attention that the market attracts from those participating

in it and even from many nonparticipating observers is part of the

establishment of gender norms, as people’s worth is recalibrated within

the context of heterosexual attractiveness.

It is important to note that for most participants, this activity pre-

cedes active heterosexual activity -- even dating -- by a year or two,

as these relationships have little to do with attachments between the

members of a pair. The activities establish a system and hierarchy of

desirability prior to the actual onset of overt heterosexual desire and

activity. One’s value on the market is a function of the matches that

are made on one’s behalf -- not so much on the number of matches,

but on the people with whom one is matched. The new and enduring

status system that forms around this market constitutes the core of

the emerging adolescent social order. In this way, the social order is --

fundamentally -- heterosexual, dramatically changing the terms of the

cohort’s gender arrangements. What was appropriate for boys and girls

simply as male and female individuals now defines them with respect

to a social order. Their value as human beings and their relations to

others are based in their adherence to gender norms. And the differen-

tiation of these norms intensifies as differentiation of male and female

merges with engagement between male and female.

27 Constructing gender

Readers who were developing gay male or lesbian identities during

this stage of their lives may think that this account forgets about them.

But the point is not that everyone is active in the heterosexual market,

or that everyone who participates in this market is heterosexual. This

market is the means by which the social order comes to presume hetero- sexuality, marginalizing and rendering deviant any who do not even-

tually participate. Sometimes there are alternative markets on which

to claim worth and value -- the academic market, for example -- but the

heterosexual imperative spreads its umbrella very widely, and because

of its central place in the age cohort, it affects all -- even those quite

averse to any direct participation in it.

There are some cultural contexts where heterosexual coupling is not

so early or so central a part of development. Even in the US the hetero-

sexual market was not apparent among such young kids a couple of

generations back. In almost all cultures though, eventual marriage is a

central social goal that marks adulthood even in cases where the young

people themselves do not play a very active role in forging heterosexual

links. Most cultures have some kinds of institutions that focus on het-

erosexual desire among the young and are linked to plans for eventual

marriage. The Tamang women of Nepal whom Kathryn March (2002)

spoke with, often recalled with great fondness those youthful days in

which they and their young female friends went to gatherings where

they sang songs to groups of young males who responded with songs of

their own. Part of the point of the lyrical exchanges was determining

just who might be available marriage partners.

In the US, gender difference and heterosexuality are deeply embed-

ded (and intertwined) in the institution of adolescence and in the for-

mal institution of the high school that houses the age group. Het-

erosexual couples have a special status in high school -- popularity is

closely linked to heterosexual alliances, and ‘‘famous’’ couples gain ex-

tra visibility and provide theater for their cohort (Eckert 1989). Gender

difference and separation are emphasized by such things as mock

elections that have male and female counterparts for ‘‘most popular,’’

‘‘most likely to succeed,’’ and similar categories. The message in these

polls is that being successful or popular is different for males and

females -- that the terms of these statuses are themselves gendered.

Meanwhile, the institutions of prom and homecoming king and queen

emphasize the importance of heterosexual alliances, elevating such al-

liances to institutional status. And the classic pairing of the cheerleader

and the football player emphasizes the role of the female supporting

the male, as the latter upholds the honor of the institution.

28 Language and Gender

Developing desire

Throughout gender development until the emergence of the hetero-

sexual market, the emphasis has been on difference -- on opposition.

The heterosexual market brings an important change in the nature of

dichotomous thinking, as suddenly, opposites are supposed to attract.

Opposition gains the twist of complementarity, and where before male

and female might have been in conflict, now they are collaborators.

And with this comes the introduction to gender of the conscious ele-

ment of desire.

Everywhere we look, we see images of the perfect couple. (For a still

compelling discussion of the construction of male and female in ad-

vertising along these lines, see Goffman, 1976.) They are heterosexual.

He is taller, bigger, darker than her. They appear in poses in which he

looks straight ahead, confident and direct; she looks down or off into

the distance, often dreamily. Standing or sitting, she is lower than him,

maybe leaning on him, maybe tucked under his arm, maybe looking

up to him. And from the time they are very young, most kids have

learned to desire that perfectly matched partner of the other sex. Girls

develop a desire to look up at a boyfriend. A girl begins to see herself

leaning against his shoulder, him having to lean down to kiss her, or

to whisper in her ear. She learns to be scared so she can have him

protect her; she learns to cry so he can dry her tears. Girls put on

large men’s shirts to emphasize their smallness. This concentration of

desire, or cathexis (Connell 1987), is an extraordinarily powerful force in the maintenance of the gender order. It leads one not simply to

desire those in the other sex class, but to form oneself in a particu-

lar mold as an object of desire by those others. Girls come to want to

feel small and delicate; boys want to feel big and strong. Or at least

these are the dominant socially endorsed images of self, images that

sometimes rest uncomfortably with such developments as the explo-

sion of girls and women in competitive sports requiring strength and

often height or weight. Even the athletic young woman, however, is

instructed to work on making her body desirable to men, as is attested

by advertising and features in such publications as Sports Illustrated for Women. Diets, hairstyling, shaving legs or heads, appetite suppressants, steroids, tattoos, body piercing, makeup: all these and more are in

the service of the desired self.10 Consumption of all kinds is driven by

10 Historian Joan Brumberg (1997) has chronicled the historical development of the contemporary extreme focus in the US on the need for young women to work hard at maintaining and improving their bodies (rather than their souls, which got at least as much or more attention in nineteenth-century America). Indeed, even men are