Dechan Ini
Dechan Ini
Dechan Ini
1 Sociolinguistic
Sociolinguistics is one of the primary fields of linguistics that studies language use in a social
setting. "A field that analyzes the relationship between language and society, between the uses of
language and the social structures in which the users of language live," according to Spolsky
(1998, p. 3). This branch of linguistics believes that human civilization is made up of a variety of
interconnected patterns and behaviors. Sociolinguistics is a broad field that includes study in
discourse analysis, interaction studies, sociology, anthropology, cultural studies, feminism, and
other areas
Our diagnosis is that gender and language studies suffer from the same problem as that
Abstracting gender and language from the social practices that produce their particular forms in
given communities often obscures and sometimes distorts the ways they connect and how those
connections are implicated in power relations, in social conflict, in the production and
reproduction of values and plans. Too much abstraction is often symptomatic of too little
theorizing: abstraction should not substitute for theorizing but be informed by and responsive to
it. Theoretical insight into how language and gender interact requires a close look at social
practices in which they are jointly produced." (Gender, Sexuality, and Meaning: Linguistic
Practice and Politics. Oxford University Press, 2011)The study of language and gender was
initiated in 1975 by three books, the latter two of which have continued to significantly influence
sociolinguistic work: Male/Female Language (Mary Ritchie Key), Language and Women's Place
(Robin Lakoff), and Language and Sex: Difference and Dominance (Barrie Thorne and Nancy
Hedley, Eds.). Overly dichotomous ideas of gender pervade Western society in ways that must
difference does not simply result in women assimilating to male, or mainstream, norms, feminist
scholars must simultaneously document and describe the value of attitudes and behaviors long
considered 'feminine.' In doing so, feminist scholars challenge their exclusive association with
women and point out their value for all people."("Language and Gender." Sociolinguistics and
Language Teaching, ed. by Sandra Lee McKay and Nacy H. Hornberger. Cambridge University
Press, 1996)"Language and gender studies have seen significant expansion to encompass sexual
orientation, ethnicity and multilingualism, and, to some extent, class, involving analyses of
spoken, written, and signed gendered identities." (Language and Gender, 2nd ed. Polity Press,
2010)[Deborah] Tannen's (1990, 1994, 1996, 1999) writing on language and gender in which
general audience book You Just Don't Understand (Tannen, 1990) offers insights into everyday
communication rituals of speakers of both genders. Much like Lakoff's (1975) Language and
Women's Place, Tannen's work has fueled both academic and popular interest in the topic. In
fact, language and gender research 'exploded' in the 1990s and continues to be a topic receiving a
great deal of attention from researchers using various theoretical and methodological
perspectives (Kendall and Tannen, 2001)." ("Gumperz and Interactional Sociolinguistics." The
SAGE Handbook of Sociolinguistics, ed. by Ruth Wodak, Barbara Johnstone, and Paul Kerswill.
SAGE, 2011)
2.3 Critical perspectives on gender identity
identifying mechanisms of oppression. Before change can even be wanted, what appears to be
natural aspects of the everyday lives of women and men has to be exposed as culturally produced
and as disadvantageous to women (and ultimately, because of this, some men would say to
themselves as well). This means beginning with an understanding of how gender is socially
constructed.Critical perspectives on gender identity also have an emancipatory aim. The word
‘critical’ is meant in a specific way; not just ‘being critical’ in the ordinary sense, but examining
something in order to unearth hidden connections, assumptions, and so on. One critical
perspective can be found in critical discourse analysis (CDA). This is an approach to discourse
analysis committed to examining the way language contributes to social reproduction and social
change. I provide a basic outline of CDA in this chapter. The practical aim of CDA is to
conventions have come about as a result of relations of power and power struggle. As an
approach to discourse analysis, it aims to show ‘non-obvious ways in which language is involved
in social relations of power and domination’ (Fairclough 2001: 229). This requires unearthing the
social and historical constitution of naturalized conventions (in other words, the constitution of
ways of doing things that are so apparently natural they are just ‘common sense’). Looking at
language critically is a way of denaturalizing it. So CDA is useful for feminists. It can be
employed in explorations of the social construction of gender. There are numerous branches of
critical enquiry into language and discourse issues that are explicitly feminist. The stated aims of
feminist stylistics, for example, are ‘to ask questions about our commonsense notions of gender
and text and to help to create a productive suspicion of all processes of text interpretation’ (Mills
1995a). Other distinct strands are discursive psychology (e.g. Edley 2001; Wetherell and Edley
2014), feminist conversation analysis (e.g. Kitzinger 2000; Speer 2005), performative theory
(e.g. Butler 1997, 1999, 2010), feminist poststructuralist discourse analysis (Baxter 2003),
feminist media studies (e.g. Gill 2007; Macdonald 1995; Talbot 2007b), feminist pragmatics
(e.g. Christie 2000) and an explicitly feminist CDA (e.g. Lazar 2005a, 2007, 2008, 2014; Talbot
1995b). These critical perspectives differ in method and in theoretical emphasis but they share
the important insight that gender is not pre-given or static but actively constructed and in flux.
Some studies of gender construction place their emphasis on gender as performance. People do
not have pre-fixed, stable gender identities; they perform them continuously. Even when we are
quite unaware of gender – simply taking it for granted as an obvious and invariable part of our
identity, as we do most of the time – even then, we are still engaged in routinely performing
gender. Critical perspectives share both an avoidance of gender polarization and a perception of
gender identity as dynamic. This term has the advantage of giving us another framework
beginning with a ‘d’ to add to the alliterative list: deficit, dominance, difference and dynamic.
I have been using the term ‘discourse’ quite a lot. Up untilnow I have not tried to define it, but
now we need to stopand consider what it actually refers to. The term is widelyand variably used
the term tends to be useddifferently, to refer to stretches of language longer than asentence. I
have been using it in the first sense: as socialinteraction in specific contexts. In Part II, we looked
found in thework of the French philosopher and social theorist MichelFoucault. Discourses, for
Foucault, are structures ofpossibility and constraint: ‘practices that systematicallyform the
objects of which they speak’ (Foucault 1972: 49).For example, medicine is a body of knowledge,
practicesand social identities. Medical discourse defines health andsickness. Take hysteria, for
instance: as an object definedby medical discourse, it consists of everything that hasbeen said or
written about it. The term is meaningless initself. What hysteria is has shifted from one century
distribution ofknowledge. Medical discourse also determines who has thepower to do the
defining. Knowledge does not arise out ofthings and does not reflect their essential truth: it is not
theessence of things in the world. Discourses are constitutedin history and society; what is
included as truth, access tothat truth, who may determine it, all these depend onrelations of
throughdiscourses by creating order; that is, by being the ones whomake boundaries and
property of powerfulgroups – men, the upper class, capitalists or whatever – butas something
throughthe definition of objects and social subjects themselves. AsChris Weedon has pointed
out,It is in the work of Michel Foucault that thepoststructuralist principles of the plurality and
integrated into atheory of language and social power which paysdetailed attention to the
institutional effects ofdiscourse and its role in the constitution andgovernment of individual
subjects. (Weedon 1997: 104)In his work, Foucault attends to discourses of the socialsciences,
which he argues have contributed substantially tomaking us what we are as people. He writes
about how thesocial sciences have impinged physically on people,constructed them as patients,
as legal subjects, sexualsubjects and so on. In other words, he argues that practicesand relations
between people are brought ianto being as aresult of those socially constructed bodies of
knowledgethat we call ‘the social sciences’. He shows that domains ofknowledge like medicine
and the law – which form socialsubjects by taking human beings as their subject-matter –are not
timeless but historical constructions. In The Historyof Sexuality , for instance, his focus is on the
discursiveconstitution of sexual subjects in the juridical system, inmedical texts and so on. He
describes his purpose in thisbook in the following way:to account for the fact that [sex] is spoken
about, todiscover who does the speaking, the positions andviewpoints from which they speak, the
institutionswhich prompt people to speak about it and which storeand distribute the things that
are said. What is at issue,briefly, is the over-all ‘discursive fact’, the way in whichsex is ‘put into
discourse’. (Foucault 1990: 11)What Foucault does in his work is to examine the
including how they define social identities.This power is disciplinary. The eighteenth century,
Foucaultobserves, was a time of population growth that broughtabout problems of poor housing
conditions, poor health andsuchlike. Indeed, the notion of a country’s having a‘population’ to be
managed, cared for and controlled stemsfrom this period. In contrast with a ‘people’ –
acomparatively indeterminate mass that a monarch couldlevy tax from – a population was an
economic and politicalproblem. It had to be scrutinized:At the heart of this economic and
political problem ofpopulation was sex: it was necessary to analyse thebirth-rate, the age of
marriage, the legitimate andillegitimate births, the precocity and frequency ofsexual relations, the
ways of making them fertile orsterile, the effects of unmarried life or of theprohibitions, the
impact of contraceptive practices …the sexual conduct of the population was taken both asan
object of analysis and as a target of intervention.(Foucault 1990: 25–6)In this scrutiny, exercised
through the authorities of churchand state, a new regime of discourses was formed. It wasonly at
this point that notions of ‘sexual perversion’,‘deviance’ and ‘unnatural act’ became possible.
masturbating child and the perverseadult (p. 105). By the nineteenth century, forms of
Among others, thehomosexual was born. As Foucault remarks, the ‘sodomitehad been a
temporary aberration; the homosexual was nowa species’ (p. 43).Fortunately, however, the
bringing about social change,such as feminism. The value of Foucault’s conception ofdiscourses
lies in its historical and social account of theirdefinition, delimitation and control. His approach
can besummarized in the following questions he poses at the endof ‘What is an author?’: ‘What
are the modes of existence ofthis discourse? Where has it been used, how can itcirculate, and
who can appropriate it for himself ? What arethe places in it where there is room for possible
Discourse as social practiceCritical discourse analysts use the term ‘discourse’ both inthe
linguistics sense of social interaction in specificsituations and in the Foucauldian sense. One
practitioner tocombine the two is the linguist Gunther Kress, who refersto the work of Foucault
in his characterization of discourseas sociocultural practice. Kress is drawing on Foucault inthis
organised sets ofstatements which give expression to the meaning andvalues of an institution.
Beyond that, they define,describe and delimit what it is possible to say and notpossible to say
(and by extension – what it is possible todo or not to do) with respect to the area of concern
statements about agiven area, topic, object, process that is to be talkedabout. In that it provides
7)Discourses are historically constituted bodies of knowledgeand practice that shape people,
giving positions of power tosome but not to others. But they can only exist in socialinteraction in
specific situations. So discourse is bothaction and convention. It is never just one or the
other.Figure 7.1 is a representation of this conception ofdiscourse and brings together the two
language use as a form of social practice. In thecentre is the text; it can be spoken as well as
written. Thetext contains formal features, nothing else (grammar,vocabulary and so on). These
features are traces of how thetext was produced: that is, how it was written or spoken.They are
also cues for how it can be interpreted: that is,read or heard. How a particular listener or reader
actuallyreads or hears a text depends on what resources she hasaccess to for interpreting it. (A
very obvious example ofresources people need for interpretation is that they musthave
knowledge of the appropriate language or languages.)The point of emphasizing this need for
resources is that itstops us from thinking of texts as though they actually havefixed meanings
independently of the social worlds theycirculate in. The text’s meaning is not already there,
exceptas meaning potential . It gets its meaning when peopleinterpret it (and the meaning it gets
will not be the samefor everybody). This view of the text will probably befamiliar to students of
literature. It does not imply thattexts can mean anything, that interpretation is some sort offree-
for-all. A text’s formal features do impose constraintson what it can mean, how it can be
interpreted.In CDA, textual analysis has both linguistic and intertextualcomponents. Intertextual
analysis requires focus on thegenres and discourses that are drawn upon. As it isunderstood in
CDA, intertextual analysis is influenced byFoucault’s work on discourse (briefly discussed in the
lastsection), and also by Bakhtin’s (1981) theory of dialogism.Hybridity is normal; that is to say,
there will virtuallyalways be traces or cues of more than one discourse andgenre in a given
represented in Figure 7.1 asinside discursive practices. Discursive practices involveboth texts
and the processes by which people produce andinterpret them. Consider a very simple example.
Whenspoken, the words seller and cellar sound exactly thesame; but this is unlikely to confuse,
say, stockbrokers atwork, because of the knowledge they bring with them as aresource in
interpretation. The kind of discursive practicethey are engaged in effectively imposes one
Fairclough 1992: 73The way people respond to texts depends on their socialbackground, hence
on what they bring to the text. Textsconstruct reading or listening positions that actual readersor
listeners have to negotiate with, although whether theyaccept them depends very much on who
they are. This canbe particularly clear in the case of mass mediapublications, which tend to be
targeted quite precisely atspecific audiences. Think of reading the following in ateenage girls’
magazine: ‘When you’re trying your hardestto impress that hunk in the sixth form …’ It contains
textualcues to presupposed ideas – namely that there are suchthings as ‘hunks in the sixth form’
and that ‘you’ try veryhard to impress them. The reader is set up as someone whoalready
entertains these ideas. A girl of about twelve mightaccept this reading position; but a woman of
forty, probablynot!In Figure 7.1 discursive practices are themselves placedwithin social practice.
This is to indicate that discourse is aform of social practice, that language use is not just
anindividual’s activity but a social act. Looking at discourseas social practice makes a
involve attention to ‘the relationship between texts,processes, and their social conditions, both
the immediateconditions of the situational context and the more remoteconditions of institutional
and social structures’ (Fairclough2001: 26). It means paying a lot more attention to thesociety
and history in which the discourse takes place thandiscourse analysts usually do. From the
structures. For a CDA informed byqueer theory, patriarchy itself is ‘one aspect of a
developed in CDA provides avaluable framework for studying language and gender. Aperennial
problem for language and gender researchers isovercoming the sense of ordinariness and
obviousness thatso much everyday language gives, and the accompanyingdanger of treating
everyday experiences as though theysomehow occur independently of society. With the model
ofdiscourse as social practice that is used in CDA, we cannotjust forget the social nature of all
discourse. It helps tocounteract the tendency for the discourse in which weperform our gender
identities to be naturalized.
Gender identity and subj ect Positioning A single individual is placed in a wide range of
positions as a social subject. These are known as subject positions. They are set up in discourse.
A person does not exist independently of them; she is constituted as a person in the act of
working within various discourses. From the beginning of her entry into social life she is
positioned within varied institutional and societal structures, which bestow upon her specific
positions bestowed by different discourses. Social subjects take up positions in activities within
institutions and social formations. In this sense they are effects of discourse rather than its
producers. For example, it is only by entering into medical discourse that a person becomes a
doctor. Her subject position of doctor is an effect of medical discourse. She is not its author. She
didn’t invent it. People enter into different subject positions in discourses. These can shift in an
individual’s lifetime, or indeed within the course of an hour. Consider the tensions experienced
by a woman who is responsible for the daily care of an elderly relation while at the same time
being a full-time wage earner. What she is expected to be, and to do, as a gendered person is
clearly not constant. An individual’s subjectivity is not fixed, invariant and ‘unitary’; it is
diversified and potentially contradictory. Recall your own changing experiences, perhaps as you
were in transition from school to college or from school to work, or returning to education as a
mature student after years of full-time employment, or maybe in transition from bringing up a
family and having part-time employment to becoming a full- time student at university. We all
experience shifts during our lifetime, taking on different gender identities in different
objectives impinge upon us and shape us. These contradictions are part of our gendered
identities. We all harbour contradictory desires and aims. Consider the contradictions
surrounding bodily demands on women as workers. What people can see and what they believe
to be true or right don’t always match up. Many women have contradictory ideas about what
they can and should do as domestic workers (such as cleaning up after incontinent elderly
relations, or carrying heavy children about) and what they are capable of as paid workers. As
domestic social subjects – as wives, mothers, daughters – women are expected to do whatever
work is necessary in the maintenance of their families, regardless of how arduous and unpleasant
it may be. But when it comes to the job market, arduousness and unpleasantness are used as
reasons for excluding women from doing what is traditionally men’s work, on the basis of beliefs
about femininity. The contradiction does not exist only in the minds of these women. It is present
as a consequence of real relations within the family and the economic world. The women may be
unaware of the contradiction between the two subject positions imposed on them; or they may be
aware but feel powerless to change the social conditions that brought it about.People come into
existence as social subjects whose identities appear as self-evident to them. This self-
evidentness gives us the illusion of self-determination, of being able to pull ourselves up by our
own bootstraps – an illusion a French discourse analyst, Michel Pê cheux, calls the ‘Munchausen
effect’ (from Baron Munchausen’s amazing feat of jumping across a crevasse while holding
himself aloft by his own pigtail). It goes against the grain, perhaps, to think of ourselves as
processes, as being constructed by elements we are mostly not even aware of But we are, since
we are defined and delimited in discourse (as patients, students, fathers …) both by being talked
about and written about, and as talkers and listeners, writers and readers. We are ‘subjugated
knowers’ and constrained actors. Our sense of self, of our autonomy as thinking individuals who
illusion, this does not mean that people are passively shaped. People are not just acted upon; they
are active participants in their own construction. They are busily involved in the building of
gender identities, especially their own. They perform their gender identities. If we think of
gender as performance, it can help us to steer clear of a false impression that people are just
passively put together by discourse, like robots on a production line. A gendered identity is, if
you like, ‘a performative accomplishment’ (Butler 1999: 179). Reflecting on her work on
performativity, Butler explains (p. xv): The view that gender is performative sought to show that
what we take to be an internal essence of gender is manufactured through a sustained set of acts,
posited through the gendered stylization of the body. In this way, it showed that what we take to
be an ‘internal’ feature of ourselves is one that we anticipate and produce through certain bodily
acts. Butler’s ideas are particularly valuable in discussion of sexualities and I return to them in
chapter 11. The concept of subject positions is central to all approaches to discourse analysis
positioning within and across conversations: Subject positions can be defined quite simply as
‘locations’ within a conversation. They are the identities made relevant by specific ways of
talking and because those ways of talking can change both within and between conversations
(i.e. as different discourses or interpretative repertoires are employed), then, in some sense at
least, so too do the identities of the speakers. However … we must remember that people are also
the masters of language, the creators of texts. (Edley 2001: 210) The construction of identities
may be far from passive, as the following study of an American TV panel discussion amply
demonstrates (Bucholtz 1996). This study looks at the language patterns of two African
American women who were panellists. The panel discussion was broadcast shortly after, and in
direct response to, the nationwide civil disturbances in the United States in 1992 (this civil unrest
followed a jury’s acquittal of four Los Angeles police officers who had been charged with
brutally beating Rodney King, an African American man). Mary Bucholtz argues that the panel
discussion, as a genre, is itself a combination of the genres of interview and conversation. There
is tension between the conventions of each. The two women challenge the interview-like
conventions being imposed by the moderator (LF). They subvert his institutional authority,
imposing more egalitarian, conversation-like conventions (see chapter 5). They do this in various
ways, including by breaking down the one-way question–answer format. The short extract below
shows one of the women (EH) initiating this disruption. Notice that she begins by conforming to
the interview norms that have been operating so far. By initially asking permission, she is
acknowledging (and perhaps pointing out) that she does not have the right to ask questions
within the existing format: Bucholtz argues that, by breaking down the one-way question–answer
these women are recasting the discussion into a format more congenial to them as African
American women. She points to black feminist theorists who have argued that there is a
distinctive black feminist epistemology or theory of knowledge (Collins 1990; hooks 1984). In
this epistemology knowledge is arrived at by means of open discussion – something that the
panel moderator’s approach is making impossible. The two women also use some elements of
African American vernacular in place of standard American English, the latter being the variety
conventionally used in formal settings such as TV panel discussions. They sometimes use
simplification of consonant clusters (so that j ust becomes j us’, for instance). They also use
some vocabulary elements recognizable as African American vernacular forms, such as brother
and cool out in the sense of ‘black man’ and ‘withdraw’ respectively (Bucholtz 1996: 279). By
means of these and other strategies, the women impose their own kind of cultural space. ‘These
strategies’, Bucholtz observes, ‘allow the speakers to subvert their own imposed position in the
interaction by constructing social identities and patterns of alignment for themselves that do not
conform to the roles assigned to them by the institutional norms of the discourse’ (p. 278). The
two women are acting politically (of course, the changes they succeed in bringing about are
effective only locally and temporarily). In doing so, they are asserting their social identities as
African American women. They publicly perform these social identities. A study of discourse in
gender. She looks at verbal interaction among engineering students, examining how gender
identities are constructed and enacted and focusing particularly on the discourse in classes and on
small group discussions. Engineering is traditionally a masculine domain and, although women
now study to become engineers too, it is still a highly androcentric area, both in education and
elsewhere. Traditional notions of gender identities prevail. Bergvall observes that this creates big
problems for female engineering students. They are caught up in conflicting demands to which
they must respond. On the one hand, there is a social need for them to behave in stereotypically
‘feminine’ ways, if they wish to take part in heterosexual social and sexual relationships (in other
words, the men will only date ‘proper’ women). On the other hand, if they are going to succeed
in their studies, they must assert themselves and their views, which is liable to put them in
competition with fellow students. This involves assertive, competitive behaviour, which is
perceived as ‘masculine’. So the female engineering students perform ‘being feminine’ for their
male fellow students – by conforming toexpectations that they should use speech patterns
displaying supportiveness, cooperativeness, tentativeness and the like – and also present
themselves as assertive and competitive. That is, they contrive to do both. Bergvall observes that
fixed categories of masculine and feminine are really no help at all in accounting for the speech
patterns of the female engineering students, patterns that suggest they are responding to the
competing stereotypical gender roles of asserting and facilitating. Within the ‘difference’
framework we examined in Part II, these women could be accounted for only as aberrations.
Such a range of ‘masculine’ and ‘feminine’ behaviour from women is only explicable in terms of
gender as performative. The performance model does not focus on ‘dichotomous differences
expected under polarized, categorical roles of feminine and masculine, but on the fluid
enactment of gender roles in specific social situations’ (Bergvall 1996: 175). The female
identities for themselves, apparently without being aware of doing so. Before going on to an
extended example in the next section, I will finish this one with an observation about exposing
gender ideology. For feminist CDA, studying language and gender is a political act; its agenda is
people into categories, as ‘women’ and ‘men’: ‘Based upon sexual difference, the gender
structure imposes a social dichotomy of labor and human attributes for women and men.’ The
details vary from one society to another, but ‘the “ sex- gender” system is always intertwined
with political and economic factors in each society’ (Lazar 2014: 186). Gender ideology and the
hierarchies it imposes generally appear to be natural, obvious and a matter of simple common
sense – that is to say, hegemonic. A key objective is to expose the working of ideology, to
denaturalize it. So, for example, a study of domestic violence in Columbia exposes the
association of authority with masculinity, of suffering with femininity, and of abuse with love
(Tolton 2011). Or, in another example, a study of managerial discourse in New Zealand
workplaces examines the tensions between authority and femininity that ‘women in charge’ have
to contend with (Holmes 2006). I will present some details of this workplace study in chapter 10
providecare for a pregnant woman and her unborn babythroughout pregnancy so that both of
them are healthy atdelivery. Antenatal care involves regular tests on bothwoman and foetus,
screening for defects, diseases andother possible hazards. It also routinely includes provisionof
advice on diet, exercise and the like, as well asinstruction on what to expect when the big day
finallyarrives. Antenatal care discourse occurs in a wide varietyof genres. Genres that women in
visits. Language plays a more central role in someof these genres than it does in others.Why look
social identities aretied up in their childbearing capacity. Medical discourse isan important site of
struggle over the domain of childbirthand the formation of women’s identities as mothers.
Themedical discourse on pregnancy and childbirth is not theonly one, but it is more powerful
than any other, and morewidespread. Antenatal care discourse is upheld by thepower behind the
discursive construction ofmaternity was conducted by Sarah Kiær, a Danish scholar,when she
was both a graduate student at LancasterUniversity and an expectant mother (Kiær 1990).
Thecontrast between her identities as scholar and as ‘mum-to-be’ was quite startling. Her
research was triggered off byher experience of being addressed differently once shebecame
‘officially’ pregnant; in other words, once she hadseen her physician and started attending a
clinic. She hadan uneasy sense of being turned into someone else by thesocial interaction side of
pregnancy. In her study, sheexamines the articulation of antenatal care discourseacross a range of
genres, both spoken and written. Hermain sources are the antenatal classes she attended
andmeetings with her physician in his surgery (both of whichshe audiotaped, as well as being a
participant-observer inthem). She refers to various printed materials as well, suchas a pregnancy
book handed out at the antenatal clinic. Herwork includes a history of the medicalization of
focuses on is the space that pregnant womentake up in the discourse of antenatal care. She
attends tothe way pregnancy is constructed as an illness, womenbeing regarded as patients and
labour as a series ofmedical procedures. She investigates the power relationsacted out in the
different genres, the attributes assigned tomothers in antenatal care discourse, and how
thesecontribute to the perpetuation of unequal power relations.The routine consultations she had
with her physician hadan interview format that placed the doctor in control of thetalk all
throughout. He typically produced a string ofstatements and questions intended to elicit brief
yes/noresponses from her and designed to check on whether thebaby was moving, whether her
fingers were swelling up,and so on. This format gave her the limited role ofrespondent and made
it difficult for her to ask anyquestions herself. Topics discussed in these consultationswere her
physical functions (such as blood pressure) andthe baby. This kind of depersonalized talk
constructs themother as a patient ‘suffering from the pathology ofpregnancy’ (Oakley 1984: 213)
and as an object carrying ababy and a set of symptoms.In antenatal classes, mothers are
practitioners, not by the mothers themselves.The focus is on hospital procedure, which pregnant
womenneed to be prepared for if things are to run smoothly. Toput it another way, antenatal
difference between them. Inthe doctor–patient interviews, the pregnant woman isconstructed as a
patient. The talk is likely to remainrecognizably medical, of a kind you might encounter in
medical discourse (which is impersonal) anda discourse of motherhood and the family (which is
morepersonal, about community and shared experience).Both types of discourse are part of
antenatal care; they arenaturalized within the medical institution. The medicalprofession uses
them to exercise power through consent(rather than by force, as they would if women were
legallyobliged to submit to antenatal care). Pregnant womenparticipate: it’s the natural thing to
do. For antenatal careto succeed, women have to be convinced of the need formedical
constructed as an illness (as in theparody in Monty Python’s The Meaning of Life : ‘Don’tworry,
we’ll soon have you cured!’). Women in labour areconstrued as patients (who must not do
anythingthemselves, because they are ‘not qualified!’). In factMarjorie Tew, a medical historian,
relates that thetwentieth-century trend towards hospitalizing mothers mayhave had as much to do
with the professional rivalrybetween obstetricians and midwives as with the needs ofthe mother
and child: ‘The policy of the increasinghospitalization of birth advocated by doctors, allegedly
toimprove the welfare of mothers and babies, was in fact avery effective means of gaining
competitive advantage byreducing the power and status of midwives and confirmingthe doctors’
ascendancy over their professional rivals’ (Tew1990: 7). It was a gendered power grab.
Midwifery,historically, was an exclusively female profession. Theemergent medical profession,
themedical profession has been that women are no good atgiving birth. They need an expert to
take charge. A deliveryis conducted by a trained medical practitioner, not by thepregnant woman
homeconfinements: ‘You will … need a midwife to deliver you athome.’ The woman producing
the baby is neverrepresented as the person doing the delivering; she isnever the grammatical
subject of the verb deliver. Itsaction is represented either as something the midwife doesfor her,
as we’ve seen, or as something that happens to heras an agentless process. ‘During the third stage
Pregnant mothers are patients,ignorant about their condition and a potential danger tothemselves
and their babies. According to Claire Clews inan article for a Royal College of Midwives
in labour are often still used in manymaternity units today’ (Clews 2013). In the course of
of the state, and one giving anunprecedented degree of licence over the bodies andapproved life-
styles of women’ (Oakley 1984: 2). Themedical profession has the right to define what it is to be
agood mother and to determine what women need to knowabout childbirth.In the British
them.These are informational and instructive texts, issuingcommands to readers. For example, a
leaflet entitled‘Reduce the risk of cot death’ produced by the Departmentof Health in 2004
addresses the reader with imperativessuch as these: ‘If your baby is unwell, seek prompt
distributed by the maternity clinictend to be highly normative. Kiær found that, as well astelling
a pregnant woman how she is going to feel, theseprinted materials tended to assume a particular,
at Royal LancasterInfirmary had a heading ‘Can my husband watch the scan?’,which assumes of
course that mothers are married. In thebaby book from the same health authority, a section
titled‘How you will look’ details changes to hair, skin and nailsand recommends ‘using a good
hand lotion and rubbergloves to protect your hands and nails when you are doingthe household
chores’. This presupposes that you do thehousehold chores, and will continue to do so
throughoutpregnancy (even though you can scarcely bend over to tieyour own shoelaces…). On
similar lines are suggestionsabout asking your husband to ‘help’ with the housework.These
publications are also normative by assuming that youfeel positive about pregnancy and at the
prospect of thebaby itself. They invariably open with ‘Congratulations!’,for example. Many,
perhaps most, women’s feelings aboutbeing pregnant are actually quite mixed and, even if
theyhave decided to go ahead with it, they may not bealtogether sure whether congratulations are
authority) is less overtlynormative in its assumptions about the kind of familystructure that
pregnant women are part of. It avoidsassuming that couples are married, frequently using
unmarriedcouples are not, in contrast with husbands (‘Can myhusband watch the scan?’),
explicitly excluded (‘onlyfathers/partners are able to see the scan’). Indeed, if thecouples in the
parents-to-be feature, briefly, in a leafletdetailing state benefit entitlements – and nowhere else.
Butthere are signs of some effort to use language that does notalienate large numbers of women
(in other words, thepamphlet producers are responding to pressures forlanguage reform). A
(very) small proportion of the printedmaterials handed out to expectant mothers are written
inHindi, Gujarati, Punjabi and Urdu, as well as in English,taking into account the ethnic diversity
of the Preston area,which the Sharoe Green Maternity Unit served.The care about terms of
reference in printed materials thatI noted above suggests that a change is taking place in theway
the medical profession defines motherhood. Mothersare not stereotyped as housewives, for
example, a leaflet produced by theToxoplasmosis Trust states that ‘it is important forpregnant
farmers to be aware that toxoplasmosis can becaught from sheep at lambing time’. Being married
and ahousewife no longer seems to be central to the definition ofa good mother in medical
discourse.As well as having the right to define what it is to be a goodmother, the medical
profession has the right to determinewhat counts as a ‘normal’ childbirth and what women
needto know about it. Clews points to the increasingnormalization of caesarean section (CS) and
observes thatwomen can now ‘request a CS in the absence of medicalneed under the guise of “
choice” , constructing CS birth asan alternative to vaginal birth’ (Clews 2013, citing Douchéand
Carryer 2011). What women need to know aboutchildbirth, as Kiær has observed, is largely a
matter oftraining as patients-to-be. In the medical world, thehandling of large numbers of people
consists of routines.Expectant mothers, like other patients, are passiverecipients of care who
must comply with these routines.They need to be familiarized with childbirth as a series
variable,which can make it all the more difficult to perceive. Thegoals of obstetric
In a pamphlet describing the three stages oflabour and produced by physiotherapists, they
arerecommending coping strategies for the woman in labourat each stage. In strong contrast with
antenatal carediscourse elsewhere, this pamphlet represents the midwifeand the birthing woman
as collaborators. Moreover, thewoman’s body knows what it’s doing: ‘Be aware of whatyour
body is telling you to do.’ This sensible advice is likelyto encourage self-confidence. It is not
echoed by othermedical practitioners who, as we have seen, set themselvesup as the ones telling
the mother what to do. Tew is deeplycritical of the severe limits set on women’s
this more true than in childbirth where thephysiological processes are so intimately dependent
onpsychological states. In the sphere of maternity carethe obstetricians’ objective was to make
theirprofession the sole repository of confidence…. Itinspired the confidence of the lay public,
but mostcritically, it destroyed the confidence of mothers intheir own reproductive efficiency.
(Tew 1990: 10)Another relatively active role for the expectant mother inantenatal care discourse
decisionmaker (making her feel involved, as well it might!). Hermain decision, though, is
probably what pain relief to haveand one of the multiple choices – transcutaneous electricalnerve
stimulation (TENS) – is only a choice for theeconomically advantaged, since it has to be paid for
by theindividual (unlike most healthcare in Britain, which is paidfor through taxation). All in all,
choice may be more illusorythan real. It is more a matter of medical procedure beingpresented as
if there were choices.We have been looking at a particular area of medicaldiscourse, focusing on
the construction of pregnancy andbirth as medical conditions. Mothers are shaped aspatients,
whose docility, obedience and overall passivity arerequired for care to succeed. But medical
discourse doesnot just exert its shaping power over patients. It structurespractitioners too.
medicaldiscourse. After all, not just anyone is permitted to ‘deliver’babies or conduct obstetric
court action has been taken against fathersfor practising illegal midwifery, that being the
midwiferyprofession’s view of unregistered home births attended bythe partner alone: see
Oakley 1984: 214). The entryrequirements for the medical profession are rigorous.
than for patients.Finally, antenatal care discourse is not a static body ofknowledge, practices and
subject positions. It undergoeschange. Changes happen, not just in the rather familiarway of
advances in medicine (a view of change solely interms of scientific progress, which is itself
ideologicallycharged), but also as a result of struggle within and acrosssocial institutions. I have
already noted a shift towards therepresentation of care in the National Health Service interms of
patient choice – a change brought about underThatcherism (the Patients’ Charter). Changes in
medicalprocedure happen as a result of pressure groups, bothwithin the National Health Service
(in the case of Britain)and from other bodies, for example the National ChildbirthTrust. This sort
of change comes about because of struggleover the defining and determining powers that
medicaldiscourse bestows upon practitioners. In the case ofantenatal care, this may include what
goes into women’straining as patients when they become pregnant,determining what women
may bementioned in the outline for a birth plan as something to be‘discussed’ with the midwife,
after an assertion that it isnot routinized. These procedures, which used to be amatter of routine,
were not even mentioned in earlierpublications purportedly preparing women for theirexperience
or ordeal.
How might antenatal care be otherwise? Perhaps it’s hardto imagine alternatives, but on
inspection I think it is clearthat antenatal care constructs the people involved inspecific ways. So
there must be alternatives. Tew refers to‘birthing attendants’ rather than midwives, a
rewordingthat alters our perception of the relationship betweenwomen in childbirth and their
women to feel a particular way about theprospect of having a baby. But is that the only way to
feel?In the case of young teenage pregnancies, arecongratulations always in order? In fact, there
are otherkinds of talk about pregnancy, in which it is taken forgranted that you don’t want to be
pregnant.In addressing these kinds of issues, we are examining thesocial construction of common
sense that surroundsmaternity in an industrialized society. CDA can help us toget past the
the cultivation of‘productive suspicion’ about texts (Mills 1995a: 21). Here is a useful set of
questions posed by Kress (1985: 7) forovercoming the commonsensical quality of single texts:1.
Why is this topic being written about?2. How is this topic being written about?3. What other
ways of writing about the topic are there?In attempting to break through the barrier of
be fruitful to look forshifts across discourses and genres, as I did to some extentin examining
how maternity is discursively constituted.Consider also the generic tensions in the panel
discussionstudy looked at in this chapter.Critical discourse analysis often has its starting point
for issues, textual featuresand so on worthy of scrutiny. In them, I am not examiningwhat women
do and what men do, but the production ofpeople as men and as women. In this chapter and in
Part IIIas a whole I give a lot of attention to CDA, but I am nottrying to impose a single
case. People who identify what they do asCDA want to align themselves with an existing,
somewhatdiverse body of work. Within this body of work some aimsare held in common,
namely the objective of exposingpower relations, attention to broader social context and,with
specific regard to gender studies, absence ofpolarization into ‘men do this, women do that’.I will
finish the chapter with one more addition to themodel that I’ve been building: intersectionality.
A person’sethnic identity does not exist separately from that person’sgender identity or age. Just
as ethnicity, gender and age donot function separately; so racism, sexism and ageisminterlock.
Identities and institutional oppressions are notlike layers on a cake. The term intersectionality
was firstcoined by Kimberlé Crenshaw (1989, 1991), American legaltheorist and civil rights
advocate. It emerged in the 1980sas a prominent field in black feminist thought, though theterm
itself was not always used (e.g. Dill 1983). Ithighlights the way in which identity and
oppressioninterlock and the complexity of identities and operations ofpower, thus explicitly
categories such as sexuality, social class, able-bodiedness, religion and nationality; and it is
nowemployed in a wide range of studies, including work onmarginalized and vilified sexualities
Although not conceptualized in quite the way we are proposing, the insight that idea and subject
positioning are interconnected and are both implicated in gender construction is really what
launched language and gender studies. In the early 1970s, American linguist Robin Lakoff
proposed that American women were constrained to soften and attenuate their expression of
• rising intonation on declaratives (A: ‘‘When will dinner be ready?” B: ‘‘Six o’clock?”)
• the use of various kinds of hedges (‘‘That’s kinda sad”or ‘‘it’s probably dinnertime”)
• indirection (saying ‘‘Well, I’ve got a dentist appointment then” in order to convey a
reluctance to meet at some proposed time and perhaps to request that the other person propose an
alternative time)
• diminutives (panties)
• euphemism (avoiding profanities by using expressions like piffle, fudge, or heck; using
circumlocutions like go to the bathroom to avoid ‘‘vulgar’’ or tabooed expressions such as pee or
piss)
• conventional politeness, especially forms that mark respect for the addressee
There were other elements in the picture she painted of ‘‘women’slanguage,’’ but the main focus
was on its ‘‘powerlessness,’’ seen as de-riving from the ‘‘weak’’ stance or position those women
(and others)were assuming. (See esp. Lakoff 1975.)Overall, Lakoff proposed, a distinctive part
of speaking ‘‘as a woman’’is speaking tentatively, side stepping firm commitment and theof
theorists and others who find the traditional notion of a unitary andcoherent self problematic.
Although our own thinking is informed by feminist andpostmodern theorizing, our focus as
linguists is on grounding the abstract notions ofdiscourse and of subject positions in concrete
linguistic practices. Finally, we adopt theterm ‘‘positioning’’ because it brings together stance
towards ideas and towards others.Goffman’s notion of ‘‘footing’’ (1979) is very similar to what
we’re calling idea positioning. appearance of strong opinions. Women are disempowered by
beingconstrained to use ‘‘powerless’’ language, ways of speaking that sim-ply are not very
effective in getting others to think or do what thespeaker wants them to. She was arguing that in
positioning themselvesas women, in taking up a certain place in the gender order, those
whomade use of the various resources she identified were also positioningthemselves as
powerless, were rejecting positions of authority fromwhich they might successfully launch their
meanings into discoursewith a reasonable hope for their success. 2Reading Lakoff’s work, many
drew the moral that women could beempowered by changing their modes of speech, assuming
more au-thoritative positions as speakers. As Mary Crawford (1995, ch. 4) ex-plains, lots of
people jumped on the ‘‘assertiveness bandwagon” dur-ing the late 1970s and the 1980s,
proposing to train women to speakmore assertively, to move away from the positions Lakoff had
others have argued, such moves wrongly assume thatit is deficits in individual women that
explain their relative power-lessness. Promoting compensatory training for individual women,
theysuggest, obscures the social arrangements that keep women’s wagesfar below men’s (in the
twenty-first century, US women still earn lessthan three-quarters of what their male counterparts
do) and assigndisproportionate social and political power to men. 3Other readers of Lakoff
pointed to the fact that the positioning de-vices she described as constitutive of ‘‘speaking as a
woman’’ are actu-ally multifunctional. Many resources that she characterizes as evincinga weak
position for the speaker, a lack of force behind the main mes-sage, may do other things. A tag,
for example, can both indicate awillingness to entertain alternative positions beyond that which
the main clause conveys (thus, the absence of unshakeable conviction) and
also serve to connect the speaker more firmly to others. Establishingsuch connections may
ultimately strengthen a speaker’s position byenlisting social support for the speaker and their
ideas and projects.As we have already stressed, the multifunctionality of linguistic formsis an
important theme in language and gender research of the pastcouple of decades. The work on tags
and on intonation that we discussbelow centers on the point that forms that can be interpreted as
signal-ing the speaker’s position with respect to the content expressed, canalso position the
speaker with respect to other folks: not only thosedirectly addressed but often also overhearers or
those spoken of.Lakoff’s proposals had the salutary effect of directing attention to ahost of
linguistic minutiae that usually are at best minimally noticed inthe flow of conversational
interaction. A flurry of studies followed, pro-ducing somewhat mixed results. William O’Barr
and Kim Atkins (1980),for example, looked at courtroom testimony and found that
speakers’overall social status as well as their familiarity with the courtroomsetting better
predicted use of many of these devices than speakers’sex. They suggested that what Lakoff had
identified as ‘‘women’s’’ lan-guage really was ‘‘powerless’’ language in the sense of being used
bythose with relatively little power, but it was not necessarily gendered.They also tested Lakoff’s
claim that many of these linguistic strategiesmight render language ‘‘powerless’’ in the sense of
rendering it ineffec-tive. They played alternative versions of essentially the same testimonyfor
mock jurors and found that jurors were more likely to believe thattestimony if it were delivered
in the more direct, less hedged, style as-sociated with people in authority. (Men in this study
were overall heardas more credible than women.)It is easy to criticize Lakoff’s specific claims
about gender and theuse of particular forms, but her pioneering work had the importanteffect of
directing attention to the critical issues of power in the interac-tion of language and gender. She
also focused attention on some kindsof linguistic resources that might be central to constructing
genderedidentities and relations and, most importantly for our present pur-poses, gendered
discourse positions. In the remainder of this chapter,we will say something about how gender
interacts with the productionand interpretation of these and other positioning resources.
Gender and the use of linguistic varietiesUse of any variety requires, first of all, access to
participation in thecommunities of practice in which the variety is used, and the right to actually
use it in situations. Use also requires desire. Speakers willnot accept linguistic influence from
people they do not value -- theirlinguistic varieties indicate movement in the direction of desired
iden-tities, of communities of practice in which they desire to participate.Given that our lives
certain amount of hetero-geneity. We may use one language in one situation, another in an-other;
we may use a stronger local accent in one situation, a weakerone in another. Depending on where
we are, what we’re doing, who isin our audience, what we’re talking about, how we’re feeling
about thesituation -- and any other of a number of things -- we call upon re-sources to adapt our
variety to our immediate needs.An important aspect of the gendered use of varieties is the
includes employment outside of the home and, for some,marriage to a partner with good
variety ofsubtle ways. Jobs often require particular kinds of language skills --whether it’s simply
because of the community in which they are lo-cated, or because of the actual kind of work. And
the jobs themselvesmay differentially attract women or men because the work is gender-specific
or because there are local or temporary reasons for womenor men to be attracted to them. The
gendered availability of employ-ment works on linguistic norms in more ways than one. On the
onehand, the actual work may not require specific language skills, butbeing in the workplace
may provide greater access to certain varieties.On the other hand, the differential linguistic
requirements of jobs at-tracting (or specifying) male or female employees may motivate menand
women to develop different linguistic skills. In either case, the ef-fect on community norms
could be profound, as the anticipation of en-tering gendered jobs may motivate differential
a means of advancement. To have a chance of marryinga prince, a young girl had better be able
to talk like a princess.The role of industrialization and urbanization has been powerfulin
language change and language shift, as people have moved awayfrom their local agricultural
communities into larger towns and cities,leaving small farming for salaried work, particularly in
factories. Thiscommonly entails a shift from one’s local variety to a more globalvariety --
whether a regional or a national standard. The gender pat-tern of this shift depends on the local
details of social change, but a common development is for women to leave the farm fastest,
andhence to lead in the shift from the vernacular. Susan Gal’s (1978, 1979)ethnographic study of
a Hungarian-speaking village in Austria docu-ments a case of this sort in detail. Gal found that
women’s linguisticchoices in the community of Oberwart were influenced not so muchby the
availability, but by the preferability, of occupations (and mar-riage opportunities) that required
the standard language, German. Inthis community, Hungarian was the language of a peasant life
thatdid not offer the same advantages to women that it did to men. Meninherited and controlled
households and land, while women did bothagricultural work and all of the domestic work on
tying women in farm households to long hours of hardphysical labor while their husbands’
burden in the field was lightenedby modern farm equipment. Gal found that women were leading
themen in the shift from Hungarian to German in this community, be-cause many of them were
attracted to the factory jobs available in thenearby German-speaking town. These jobs offered
greater autonomyand a town lifestyle that involved an easier domestic life than that ofthe
Hungarian woman in a farming household. Few of them were will-ing to marry the Hungarian
peasant men with whom they grew up,many opting instead for marriage to German-speaking
factory workersand the town life that these marriages entailed.Jonathan Holmquist (1985) found
a similar situation in the villageof Ucieda in the Cantabria mountain region of northern Spain.
Oncea very poor community, the modern cash economy that followed theFranco era has
improved the economic situation in Ucieda considerably,and brought about considerable social
change. The traditional moun-tain agricultural life consisting of shepherding native stock
(ponies,long-horned cattle, sheep, and goats) has given way over the past cou-ple of generations
to factory work in the nearby town, and to dairyfarming, which can be managed in addition to a
factory job. The lo-cal vernacular of this village was a Romance dialect closely related
toCastilian Spanish, but with significant differences. Holmquist focusedon word-final /u/, which
corresponds to Castilian /o/. Thus, for example,in a quote from an older person in the village
speaking of young menand their new way of life, Holmquist records (p. 197):el trabaju del
campu no lo saben ‘field work, they don’t know it’which in Castilian would be:el trabajo del
campo no lo sabenAs people in the village moved away from the traditional mountainagricultural
way of life, they began to lower this vowel to conform moreclosely to the standard Spanish [o]
economy. Andas in Oberwart, young women have found factory work preferable toagricultural
work, even dairy farming (as Holmquist says [p. 200], theydon’t want to be ‘‘stuck at home with
the cows”), while many youngmen continue to engage in dairy farming. The fate of word-final
/u/ isclosely linked to the fate of the agricultural way of life, as shown inTable 8.1, which shows
a numerical value for the height of this vowel.The numerical value is the average of all
pronunciations on a scaleof 0 to 4, where 0 is the lowest [o] pronunciation, and 4 is the high-est
(most conservative) pronunciation [u]. As this table shows, factoryworkers are the most likely to
lower this vowel and those who engagein the mountain animal economy are the least likely, with
dairy farm-ers in the middle. And in all agricultural groups, furthermore, womenare more likely
to use the lowered pronunciation than men, suggest-ing their attraction to town life. Only among
those working solely infactories -- those whose pronunciations are consistently low -- does
thegender difference disappear.In these two cases, one would not say that the jobs that were
draw-ing women into the standard language market were themselves partic-ularly gendered.
Many of the cases in which we see gendered shifts inlanguage use, though, do involve the
communityin South Carolina, Patricia Nichols (1983) found that in general thevariety spoken by
women was closer to standard English than that spo-ken by men. This difference corresponded to
the different employmentopportunities open to women and men on this island. The men wereable
to make good money as laborers on the mainland -- jobs that re-quired physical skill, but that did
not depend on the way they spoke.The women, on the other hand, found the best jobs as teachers
or asmaids in wealthy homes or hotels -- all settings in which they were expected to use more
standard language. In this case, it was the placeof gendered jobs in the language market that led
to gender differencesin speech.David Sankoff et al. (1989) have argued that gendered roles in
theworkplaces of western society tend to engage women more than menin the standard language
market. Women in their traditional workroles are often what Sankoff et al. have called
extensionof middle-class women’s domestic role into the workplace, allowing ed-ucated women
to make a living, while keeping them out of the publicsphere. This employment mixed raising
children with providing educa-tion. Some governesses taught academic subject matter, but all of
themtaught manners and refinement, of which linguistic propriety was animportant component.
While governesses provided private instructionto children of the elite, other women were playing
teachers, women were responsible for providing children whodid not have access to private
instruction with both ‘‘moral” and intel-lectual education, including access to ‘‘correct’’ or
standard grammar.As more women have moved into the workplace, many of the jobsaccessible
to them have been as front women, whose role is eitherexclusively or primarily to embody and
give voice to corporate stan-dards to outsiders: receptionists, hostesses, phone operators, flight
at-tendants, secretaries. These jobs require technicians of language notbecause language qualifies
a person as directly able to handle the prac-tical demands of the job, but because it serves as the
cultural capitalnecessary to be the ‘‘kind of” person who is qualified to occupy thatposition. In
order to be a receptionist in the front office of corporateheadquarters, one needs to be able not
only to communicate with one’semployer and those who come in the door, but to represent the
com-pany’s desired image. In this sense, the receptionist’s command of stan-dard language is not
only part of that individual’s cultural capital butalso part of the company’s cultural capital.In her
study of the language of corporate managers in Beijing, QingZhang (2001) examined a prime
example of women providing cul-tural capital for their company, as different work trajectories
shapedwomen’s and men’s use of Mandarin. Zhang compared the speech ofmanagers in two
enterprises. All of these people, maleand female, had entered their businesses with the same high
level ofeducation. In the state-owned businesses, women’s and men’s career tra-jectories were
the same. In the foreign-owned businesses, however, the men moved directly into sales positions
and quickly into management,whereas the women were initially given secretarial jobs from
whichthey only gradually moved into management positions. Hired primar-ily for their linguistic
skills, these women’s initial value to their com-panies was their ability to represent the company
for the company. These women, therefore, developed a style ofChinese that was more
‘‘cosmopolitan” -- less locally ‘‘Beijing” -- as be-fits a globally based enterprise. This variety
was notably different fromthat employed by the men in the same businesses, while the speechof
managers in the state-owned businesses showed very little genderdifference. We will discuss this
further in chapter nine.Employment is only one way in which the gendering of activity leadsto
gender differences in the development and use of linguistic varieties.For women who do not
work in the public marketplace, linguistic needsand preferences will depend on the nature of
individualmen as well as by institutions. The expression trophy wife refers to justsuch a practice,
with the emphasis on the woman’s physical propertiesand on her ability to consume with
refinement. A cultured and well-spoken wife, even one who lacks the physical accoutrements of
thetrophy wife, bespeaks a man of refinement and substance. Standardlanguage, therefore, has
often been emphasized as an important partof a woman’s capital on the marriage market.
Building on a traditionin which women do not compete in the financial market, a womanmay
well use standard language in the social market.Women’s and men’s social networks may also
lead to differential lin-guistic patterns. In a study of the English spoken in Belfast, LesleyMilroy
(1980) found that the use of vernacular language was rein-forced in close locally based social
networks. Density (the number ofconnections among the members of a network) and
closely related to class. In working-class communities, peopletend to live near, and spend leisure
time with, relatives -- and to makefriends through neighbors and relatives. They also tend to find
jobsthrough those same connections. As a result, there tend to be con-nections among more of
the people in a working-class network, andthese connections tend to cover more domains (e.g.
work, church, activ-ities). Milroy found that because of the poor employment situation
forwomen, though, women’s networks tended to be less dense and multi-plex than those of their
male peers. Correspondingly, their vernacularusage was lower than that of their male peers. In
one neighborhood Milroy studied, however, it was women rather than men who were em-ployed
and who were involved in more dense and multiplex networks.In that situation, the women’s
vernacular usage was on many measuresahead of that of their male peers.In a study of migrants
from rural communities to Brazlandia, a satel-lite city near Brazilia, Stella Maris Bortoni-
Ricardo (1985) found thatmen adapted their rural dialect more readily to the urban variety
thanwomen. The apparent reason was that men found work opportunitiesin Brazlandia that gave
them access to social networks in which the ur-ban variety was used. The women, on the other
hand, were restricted tothe neighborhoods where they lived. Since these neighborhoods
werepopulated by others who had migrated from the same rural areas, theyhad little access to the
urban variety. She did find four exceptions tothis pattern, who turned out to be four women who
belonged to largenuclear families, and who were exposed to many interactions withschool-aged
people in the home, providing considerable exposure tourban culture and language.
the quarter century since Lakoffs (1975) and Key's (1975) pioneering stud-ies, there has been a
mountain of research on gender and discourse - researchwell documented in the present volume.
In recent years, discourse analystshave also undertaken studies of language in the context of
family interaction.For the most part, however, the twain haven't met: few scholars writing in
thearea of language and gender have focused their analyses on family interaction,and few
researchers concerned with family discourse have focused their analysison gender and language.
This is a gap this chapter addresses.Drawing examples from an ongoing research project in
which dual-careercouples with children living at home recorded all their interaction for a week,as
well as videotaped excerpts of naturally occurring family interaction thatappeared on public
television documentaries, I examine (1) how gender-related patterns of interaction influence and
illuminate family interaction, and(2) what light this insight sheds on our ideology of language in
will argue - and, I hope, demonstrate - that poweris inseparable from connection. Therefore, in
exploring how family interactionis mediated by gender-related patterns of discourse, I will also
suggest thatgender identity is negotiated along the dual, paradoxically related dimensionsof
Researchers routinely interpret family interaction through the template of anideology of the
family as the locus of a struggle for power. In my view, thisideology needs to be reframed.
Power is inextricably intertwined with connec-tion. Discourse in the family can be seen as a
struggle for power, yes, but it isalso - and equally - a struggle for connection. Indeed, the family
is a primeexample - perhaps the prime example - of the nexus of needs for both powerand
means not only to understand more deeply gender andlanguage but also to reveal, contest, and
reframe the ideology of the familyand of power in discourse.Among recent research on discourse
in family interaction, three book-lengthstudies stand out. The earliest, Richard Watts' Power in
Family Discourse (1991),is unique in analyzing conversations among adult siblings and their
spousesrather than the nuclear family of parents and young children living in a singlehousehold.
For Watts, as his title suggests, power is the force defining familialrelations.Published a year
later, Herve Varenne's Ambiguous Harmony (1992) examinesa conversation that took place on a
single evening in the living room of ablended family: mother, father, and two children - a
teenage son from themother's previous marriage and a younger child born to this couple.
Varenne,too, sees power as a central force. He writes: "The power we are interested inhere is the
power of the catalyst who, with a minimal amount of its ownenergy, gets other entities to spend
large amounts of their own" (p. 76).Shoshana Blum-Kulka's Dinner Talk (1997) is unique in
which the parents were born and raised in the United States. Al-though Blum-Kulka does not
directly address the relationship between powerand connection, she discusses the parents' dual
and sometimes conflictingneeds both to socialize their children in the sense of teaching them
what theyneed to know, and at the same time to socialize with them in the sense ofenjoying their
company. This perspective indirectly addresses the interrela-tionship of power and connection in
the family.Psychologists Millar, Rogers, and Bavelas (1984) write of "control maneuvers"and
note that in family therapy, "Conflict takes place within the power dimen-sion of relationships." I
do not question or deny this assumption, but I wouldcomplexify it. I have emphasized, in a
number of essays (especially Tannen1994), the ambiguity and polysemy of power and solidarity,
which are inparadoxical and mutually constitutive relationship to each other. Thus
familyinteraction (including conflict) also takes place within the intimacy dimension. and we can
also speak of "connection maneuvers." My goal in this chapter is to explicate how what
researchers (and participants) would typically regard as control (or power) maneuvers can also
be seen as connection maneuvers, in part because connection and control are bought with the
solidarity - or, in different terms, of status and connection. HereI briefly recap the analysis
developed in that essay.In conventional wisdom, as well as in research tracing back to Brown
andOilman's (1960) classic study of power and solidarity, Americans have had atendency to
conceptualize the relationship between hierarchy (or power) andconnection (or solidarity) as
unidimensional and mutually exclusive (see fig-ure 8.1). Family relationships are at the heart of
this conception. For example,Americans frequently use the terms "sisters" and "brothers" to
indicate "closeand equal." So if someone says "We are like sisters" or "He is like a brother
tome," the implication is, "We are as close as siblings, and there are no statusgames, no one-
Thus, in work and military contexts, mostAmericans regard it as self-evident that friendships
across levels of rank areproblematic and to be discouraged, if not explicitly prohibited.I suggest
that in reality the relationship between power (or hierarchy) andsolidarity (or connection) is not a
single dimension but a multidimensionalgrid (see figure 8.2). This grid represents the dimensions
of power and ofconnection as two intersecting axes. One axis (I represent it as a vertical
one)stretches between hierarchy and equality, while the other (which I represent asa horizontal
relationships along an axisthat runs from the upper right to the lower left: from hierarchical and
distantto equal and close. Thus we would put business relations in the upper rightquadrant
(hierarchical and distant) and relationships between siblings andclose friends in the lower left
quadrant (egalitarian and close) (see figure 8.3).In contrast, members of many other cultures,
such as Japanese, Chinese, andJavanese, are inclined to conceptualize relationships along an axis
that runsfrom the upper left to the lower right: from hierarchical and close to equal and distant. In
this conception, the archetypal hierarchical relationship is the parent-child constellation:
extremely hierarchical but also extremely close. By thesame token, sibling relationships are seen
as inherently hierarchical. Indeed, inChinese (and in many other non-Western languages, such as
Sinhala), siblingsare addressed not by name but by designations identifying relative rank, suchas
"Third Eldest Brother," "Fifth Younger Sister," and so on (see figure 8.4).It is also instructive to
note that Americans are inclined to see power asinherent in an individual. Thus, Watts defines
power as "the ability of anindividual to achieve her/his desired goals" (1991: 145). Yet this, too,
alliances. Even in the most apparently hierarchical situation, such as aworkplace, an individual's
ability to achieve her/his goals is dependent on con-nections to others: the proverbial friends in
high places. In other words, poweris composed in part of connection, and connection entails a
kind of power.
2.6.3 A Paradigm Case of the Ambiguity and Polysemy of Power and Connection
The family is a key locus for understanding the complex and inextricablerelationship between
power (negotiations along the hierarchy-equality axis)and connection (negotiations along the
closeness-distance axis). And nowheredoes this relationship become clearer than in the role of a
key family member,mother. For example, Hildred Geertz (1989 [1961]: 20) writes that there are,
inJavanese, "two major levels of language, respect and familiarity." (I would pointout that, in
light of the grid presented above, these are two different dimen-sions: respect is situated on the
hierarchy-equality axis, whereas familiarity isa function of the closeness-distance axis.) Geertz
observes that children usethe familiar register when speaking with their parents and siblings until
aboutage ten or twelve, when they gradually shift to respect in adulthood. How-ever, she adds,
"Most people continue to speak to the mother in the same wayas they did as children; a few shift
to respect in adulthood" (p. 22). This leavesopen the question whether mothers are addressed in
this way because theyreceive less respect than fathers, or because their children feel closer to
them.I suspect it is both at once, and that trying to pick them apart may be futile.Although the
linguistic encoding of respect and familiar registers is a lin-guistic phenomenon not found in
English, nonetheless there are phenomenain English that parallel those described by Geertz.
Ervin-Tripp, O'Connor, andRosenberg (1984) looked at the forms of "control acts" in families in
order togauge power in that context. They found that "effective power and esteem wererelated to
age" (p. 134). Again, however, "the mothers in our sample were animportant exception to the
pattern . . ." (p. 135). "In their role as caregivers,"the authors note, mothers "received
nondeferent orders, suggesting that the children expected compliance and believed their desires
to be justification enough." As with Javanese, one could ask whether children use more bald
imperatives when speaking to their mothers because they have less respect for them, or because
A great deal of the research done on family discourse has focused on talkproduced in the context
of dinner-table conversation. The dinner table is afavorite site, no doubt, both because dinner is a
prime time that family mem-bers typically come together and exchange talk, and also because it
is a boundedevent for which speakers gather around a table and which is therefore relat-ively
easy to tape-record. Both Blum-Kulka and Elinor Ochs and her students(for example, Ochs and
Taylor 1992) identify a ritual that typifies Americandinner-table conversation in many families: a
ritual that Blum-Kulka dubs"Telling Your Day." When the family includes a mother and father
(as thefamilies recorded in both these studies did), mothers typically encouragechildren to tell
their fathers about events experienced during the day.Ochs and Taylor give the examples of a
mother who urges, "Tell Dad whatyou thought about gymnastics and what you did," and another
who prompts,"Chuck did you tell Daddy what happened at karate when you came in yournew
uniform? What did Daisy do for you?" (p. 310). Ochs and Taylor note thatin a majority of the
instances recorded in this study, fathers responded to theresultant stories by passing judgment,
assessing the rightness of their children'sactions and feelings, and thereby setting up a
constellation the researchers call"father knows best."In the families Ochs and her students
observed, mothers usually knew whatthe children had to say. This was true not only of mothers
who had been athome with the children during the day but also of mothers who worked full-time,
because generally they had arrived home from work earlier than thefather, and they had asked
the children about their day during the time theyhad with them before Daddy came home. At the
dinner table. Daddy couldhave asked "How was your day?" just as Mother did before dinner. But
inthese families, he usually didn't.Ochs and Taylor identify the roles in these narrative exchanges
account of an experience in a way that is critical of how the speakerhandled the situation. For
example, when an eight-year-old child. Josh, whohas been doing homework, announced, "I'm
done," his father asked in a "dis-believing tone," "Already Josh? Read me what you wrote." Thus
the fatherquestioned whether Josh really was finished or not (p. 313). In Ochs andTaylor's terms,
he "problematized" Josh's announcement "I'm done." The family power structure, Ochs and
Taylor observe, is established in these story telling dynamics. Just as Mother typically prompted
a child to tell Daddywhat happened, older siblings were much more likely to urge younger ones
totell about something that happened than the other way around. In this sense,older siblings were
treating their younger siblings more or less the way parentstreat children - something that, I
would note, younger siblings often perceiveand resent, especially if the older brother or sister is
not all that much older.Ochs and Taylor found that children were most often problematizees -
theones whose behavior was judged by others. Rarely were they problematizers -the ones who
questioned others' behavior as problematic. This puts childrenfirmly at the bottom of the
hierarchy. Fathers were the most frequent prob-lematizers and rarely were problematizees: rarely
was their behavior held upto the scrutiny and judgment of others. This puts them firmly at the top
ofthe hierarchy. In keeping with the findings of Ervin-Tripp, O'Connor, andRosenberg, mothers
were not up there, as parents, along with fathers. Mothersfound themselves in the position of
problematizee (the one whose behaviorwas held up for judgment) as often as they were
problematizer (the one whowas judging others). Thus fathers were in the position of judging
their wives'actions in addition to their children's, but mothers judged only their
children'sbehavior, not their husbands'. In other words, the storytelling dynamic placedmothers
in the middle of the family hierarchy - over the children, but underthe father.The authors also
observe that mothers often problematized their ownactions. For example, a woman named Marie
owns and runs a day care center.At dinner, she tells of a client who was taking her child out of
the center, andpaid her last bill. The client handed over more money than was needed tocover the
time her child had spent in day care, so Marie returned the excess.But she later wondered
whether she had made a mistake. After all, her policyrequired clients to give two weeks' notice
before withdrawing a child, and thismother had not given notice. So perhaps the client had
intended the overpay-ment to cover those two weeks, and Marie should have kept it, enforcing
herpolicy. The father made clear that he endorsed this view: "When I say somethingI stick to it
unless she brings it up. ... I do not change it" (p. 312). Marie wasthe "problematizee" because her
action was called into question. She had"problematized" herself by raising the issue of whether
she had handled thesituation in the best way; her husband then further problematized her
byletting her know that he thought she had not. Ochs and Taylor found that thispattern was
common: if mothers questioned their own actions, fathers often"dumped on" them by reinforcing
the conclusion that the mothers had notacted properly. In contrast, the authors found that in the
rare instances whenfathers problematized themselves, mothers did not further problematize
them.In this revealing study, Ochs and Taylor identify a crucial dynamic in middle-class
American families by which the family is a power structure with thefather at the top. They
further show that mothers play a crucial role in settingup this dynamic: "Father as
problematizer," they argue, is "facilitated... by the active role of mothers who sometimes
(perhaps inadvertently) set fathers up as potential problematizers - by introducing the stories and
reports ofchildren and mothers in the first place and orienting them towards fathers asprimary
recipients" (p. 329).For me, the most important word in this excerpt is "inadvertently." I
wouldargue that the father-knows-best dynamic results from gender differencesin assumptions
about the place of talk in a relationship, and that it reflectsthe inextricable relationship between
power and connection. When a motherasks her children what they did during the day, she is
creating closeness byexchanging details of daily life, a verbal ritual frequently observed to
charac-terize women's friendships (see, for example, Tannen 1990; Coates 1996). Inother words,
it is a connection maneuver. If the father does not ask on hisown, "How was your day?" it does
not mean that he is not interested in hisfamily, or does not feel - or wish to be - close to them. It
just means that hedoes not assume that closeness is created by the verbal ritual of telling
thedetails of one's day, and he probably does not regard closeness as the mostimportant
barometer of his relationship with his children.When Mother prods a child, "Tell Daddy what
you did in karate today," sheis, it is true, initiating a dynamic by which the father will assess the
child'sactions and thus be installed as the family judge. But I would bet that her goalwas to
involve the father in the family, bring him into the circle of intimacy shefeels is established by
such talk. From this point of view, the father-knows-bestdynamic is as much a misfire as is the
common source of frustration betweenwomen and men that I have described elsewhere (Tannen
1990): for example,a woman tells a man about a frustrating experience she had that day,
perform-ing a ritual common among women friends that Gail Jefferson (1988) dubs"troubles
talk." Since troubles talk is not a ritual common among men friends,he thinks he is being asked
to solve the problem, which he proceeds to do - toher frustration. She protests, which frustrates
him. Similarly, the mother whoprods her children to tell their father what they did that day, or
who talksabout her own day, is trying to create connection. But the father, not recogniz-ing the
ritual nature of her comment, thinks he is being asked to judge.In this view, it is not the mothers'
initiation of the "Telling Your Day"routine in itself that sets fathers up as family judge. Instead,
Fatherstake the role of judge of actions recounted in stories because they figure that'swhy they
are being told the stories. Fathers are less likely to talk about theirown work problems because
they don't want advice about how to solve prob-lems there, so they see no reason to talk about
them. Many men feel that re-hashing what upset them at work forces them to re-live it and get
upset allover again, when they'd rather put it out of their minds and enjoy the oasis ofhome. They
may also resist telling about problems precisely to avoid beingplaced in the one-down position of
receiving advice or of being told that theydid not handle the situation in the best way. On the few
thatmothers did not further dump on them - not necessarily because mothers felt they had no
right to judge, but more likely because they took these revelationsin the spirit of troubles talk
rather than as invitations to pass judgment. Theseclashing rituals result in mothers finding
themselves one-down in the familyhierarchy without knowing how they got there.I have
discussed this example from Ochs and Taylor at length to demon-strate how gender-related
and how what has been accur-ately identified as a matter of negotiating power is also
mycontentions that (1) power and connection are inextricably intertwined; (2) therelationship
between power and connection is fundamental to an under-standing of gender and language; and
interaction.
Conversational Ritual
The "How was your day?" ritual, for many women, is just one way that con-nection is created
and maintained through talk. Another way is exchanginginformation about personal relationships
and emotions. Here, too, conversationsthat take place in families reflect the divergent
expectations of family membersof different genders.For example, one way that many women
create and maintain closeness is bykeeping tabs on each other's lives, including (perhaps
especially) romanticrelationships. When male and female family members interact, gender
differ-ences in expectations regarding the use of talk to create closeness can lead tounbalanced
interchanges. The following example, which illustrates just such aconversation, comes from the
research project in which both members of dual-career couples carried tape-recorders with them
for at least a week, recordingall the conversations they felt comfortable recording. (The digital
recorders ranfor four hours per tape.) In this example, one of the project participants recorded a
conversation withher unmarried brother. The sister (a woman in her thirties) is asking her
brother(who is a few years younger) about his girlfriend, whom I'll call Kerry. Clearlythe sister is
Brother: Cool.
Brother: Yeah.
Sister: PRETTY much? When you say PRETTY much, what do you mean?
The conversation takes on an almost comic character, as the sister becomesmore and more
probing in reaction to her brother's minimal responses. Evid-ent in the example is a process I
call, adapting a term that Gregory Bateson(1972) applied to larger cultural processes,
complementary schismogenesis. Bythis process, each person's verbal behavior drives the other to
more and moreexaggerated forms of an opposing behavior. In this example, the sister
asksrepeated and increasingly probing questions because her brother's responsesare minimal, and
his responses may well become more guarded because herquestions become increasingly
insistent. Indeed, she starts to sound a bit likean inquisitor.Moreover, this conversation between
sister and brother sounds rather like amother talking to a teenage child. It is strikingly similar to
the conversationrepresented in the next example, which took place between a mother andher
Yeliseyeva in connection with a seminar I taught on familyinteraction. The excerpt comes from a
documentary made by filmmaker JenniferFox entitled "An American Love Story." The
documentary aired in five two-hour segments on the USA's Public Broadcasting System in
September 1999.In preparing the documentary. Fox followed the family of Karen Wilson,
BillSims, and their two daughters, in Queens, New York, over two years begin-ning in 1992. In
this episode, the younger daughter, Chaney, was anticipatingher first "date" - a daytime walk -
with a boy, despite her parents' misgivings.But the boy (who is thirteen) failed to appear on the
appointed day. After theentire family spent several hours waiting for him, Chaney got a
telephone callexplaining that his grandmother had refused permission for him to go. Karentries
Chaney: No.
Karen: Disappointed?
Karen: Relieved?
Through her questions and comments, Karen is showing her daughter thekind of conversation
she expects to have - a conversation about how Chaneyfelt about what happened to her. I doubt
that Chaney is unable to hold suchconversations; I would bet she has them frequently with her
best friend, Nelly.But, like many teenagers, she seems reluctant to divulge her feelings to
hermother.
The significance of these gender patterns in definitions of closeness, and the significance of
closeness in women's (but not men's) evaluations of family relationships, emerges in the
discourse videotaped in another public television documentary, "An American Family," which
aired in twelve weekly hour-long segments in 1973. For this series, filmmakers Alan and Susan
Raymond filmed the family of William and Pat Loud and their five children in Santa Barbara,
California, for seven months. My student Maureen Taylor examined conversations between the
parents regarding their children, and in particular their teenage daughter Delilah.Pat Loud had
taken Delilah on a trip to New Mexico. Delilah came home early - and Pat, on her own return,
tries to get her husband to tell her what Delilah said when she arrived home. A recurrent theme
in Pat's discourse is her assumption that her daughter should confide in her. Furthermore, Pat's
distress that Delilah left New Mexico without confiding her reasons for leaving to her mother is
associated with Pat's general distress at seeing her children leave home. Maureen Taylor, in a
seminar paper, pointed out that Pat and Bill have very different reactions to their teenage
children growing up and growing away.In talking to Pat, Bill explains that he is not concerned
because he believes the separation is inevitable: "You've got to learn, Patty," he says, "that
they're going to leave you." To back this up, he suggests that she think back to her own youth.
Taylor points out that the contrast between Pat's sense of desolation and Bill'ssense of liberation
at their daughter's - and all their children's - growing up reflects the gender-specific roles they
took in the family. Since Pat had devoted her married life to caring for her children, she
experiences their departures as abandonment. As she tells her brother and sister-in-law, "All my
kids a releaving me. And what have I got left? I haven't got anything left. And that scares the hell
out of me."In contrast, Taylor points out. Bill has spent his life traveling: first in the navy and
then in connection with his business. This reinforces the interpretation hegives to his children's
growing up: although Pat sees them as leaving her, he sees them gaining freedom and
independence for themselves. Furthermore, I would point out. Bill's description of what he won't
worry about makes it clear that the burden of family for him has been a financial one: the
responsibility for supporting everyone. His children's growing up liberates him from that
burden.Thus Bill's and Pat's different reactions can be explained not only by the different roles
they took in their family but also by differences in what women and men tend to focus on in
relationships in general and family relationships in particular. The example of Bill and Pat Loud,
then, demonstrates that family elation ships are a complex intertwining of connection and power,
that re-sponses to and interpretations of these forces pattern by gender, and that an understanding
In this final section, I examine several examples from the family discourse recorded by one of
the couples who participated in the research project Idescribed above by recording their own
conversations. (This is a different family from the one in which the sister/brother conversation
occurred.) In each of the following examples, the mother and father use complex verbal
strategies to balance the needs to negotiate both power and connection as they go about the tasks
required to maintain the daily life of their young family.In addition, as we will see, their
pseudonymously called Molly and Ben, have a two-year-old daughter, Katie. Both Molly and
Ben work outside the home: Ben full-time and Molly at a reduced schedule of thirty hours per
week. Each regularly takes off one day a week to spend with Katie, who consequently attends
daycare only three days a week. At one point in the taping, Molly and Ben, both at home,
become embroiled in an argument about making popcorn. Molly is int he kitchen by herself and
Molly responds, from the kitchen, "I'm making popcorn." And then she adds,
control maneuvers. Ben wants to switch roles with Molly, so that she will take
over child care and he will take over popcorn preparation. Molly resists this
switch. In a direct confrontation over power, Molly might simply refuse: "No,
I don't want to switch." Instead, by saying "You always burn it," she resists
relinquishing her task by appealing to the good of the family rather than her
own preference. Insofar as she resists doing what Ben wants her to do, her
statement is a control maneuver. But to the extent that she appeals to the
Because Molly has based her resistance on her husband's putative deficiency,
I make it perfect.
Although they continue to exchange attacks, self-defense, and counter attacks focused on
popcorn-making skills, Ben and Molly execute the switch: Ben takes over in the kitchen, and
Molly takes charge of Katie. But she continues totry to engineer her return to the kitchen. In this
Katie: Okay.
Katie: Okay.
Here, again, Molly's utterances are a blend of power and connection. To the extent that she is
trying to get her way - take back control of the popcorn preparation - Molly is engaged in control
maneuvers. But by proposing that Katie "help Mommy make popcorn," Molly is proposing to
satisfy both her-self and her husband: she would thereby return to the kitchen, yes, but she would
also fulfill Ben's request, "You take care of her." Moreover, by involv-ing Katie in the plan,
Molly is involving the child in the interaction. Further-more, her linguistic choices ("Let's not let
Daddy do it") align herself with her daughter: "Let's" merges mother and daughter; "not let"
includes the child in the mother's perspective as someone who has authority over Ben's actions,
amnd "Daddy" includes the mother in the child's point of view. All these are connection
maneuvers, though they create connection to Katie rather than Ben. From the kitchen, Ben
In all these examples, I have tried to show that whereas family interaction is, as researchers have
identities and roles. In ana- lyses of the interactions tape-recorded by this family, as well as
others in the study, Shari Kendall has shown that whereas both mother and father espouse an
ideology of equal co-parenting and wage-earning, in their ways of speak- ing, the mothers
position themselves as primary childcare providers and their husbands as breadwinners (see
Kendall, this volume). Alexandra Johnston, the research team member who spent time with
Molly and Ben and transcribed their conversations, observed that one way Molly positions
herself as primary caretaker is by frequently correcting Ben's parenting. In contrast, Ben rarely
corrects Molly's parenting. This, indeed, is what Molly is doing in the last example when she
tries to reframe Ben's interpretation of why Katie is being a pest, and to suggest how he might
"save [his] family a little stress" by getting a snack on the way home. In this way, the final
example, like all those preceding it, illustrates that we need to understand family interaction -
like all human interaction - not only as negotiations for power but also as negotiations for
connection. Linguistic strategies that can be identified as control maneuvers must also be
examined as connection maneuvers. Power and connection are the dimensions along which
human relationships are negotiated, and they are also the dimensions along which gender identity
The family is a microcosm of gender relations. In order to understand talk in the family, we have
to first understand gender patterns in talk. And the family is the training ground on which we
learn to inhabit, express, and manipulate the patterns of behavior, the ways of talking, that are
associated with gender. In this lecture, we’ll examine how gender patterns emerge as family
members negotiate the dynamics of power and connection. It brings together my work on
gender with my work on family interaction, as laid out in my book I Only Say This Because I
Love You. We’ll also introduce a concept that is crucial for all communication but which has
special power in the family—what Dr. Tannen calls “metamessages. ‘Metamessages,’ from
anthropologist Gregory Bateson, refers to two levels of meaning in what we say. What it says
about the relationship that we say these words in this way at this time—our emotions are usually
reacting to the metamessages. The meaning of metamessages comes from the way something is
said, the fact that it’s said, and meanings already on our mind from previous conversations. An
example of this is a woman who points out salmon that her husband might like on a menu after