Académique Documents
Professionnel Documents
Culture Documents
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=cup.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access to Language
in Society.
http://www.jstor.org
Language in Society 25, 1-26. Printed in the United States of America
ALICE GREENWOOD
AT&T Bell Laboratories
Murray Hill, NJ 07974
ABSTRACT
has become increasingly apparent that the results of early cross-sex studies,
and work that investigated only white speakers living in the West, cannot be
unproblematically transferred to generalized conclusions about the speech
styles of all women or all men in all contexts (see also Edelsky 1981, Holmes
1986). We believe that many assumptions and conclusions about female and
male speech styles, drawn from research carried out before 1990, need to be
reconsidered.
In particular, the conception of gender which underlies much of the work
in this field - where gender is unquestioningly treated as a dualistic, polar-
ized, and explicitly assigned category parallel to sex - seems simplistic and
naive (see also Eckert 1989, Cameron 1990, Butler 1990, Gal 1991, 1992,
Ochs 1992). We object to conclusions about speakers'general communicative
style (or specific phonological characteristics)which are based on the exam-
ination of single linguistic variables, isolated from their full conversational
and communicative contexts (Lakoff 1975, Zimmerman & West 1975, Maltz
& Borker 1982, Tannen 1990b, Labov 1991); and we reject, as have others,
claimed correspondences between single linguistic structures or pragmatic
expressions and invariant functional meanings associated with any group of
speakers (see also Hymes 1974, Goffman 1983, Holmes 1984a,b, 1986, 1993,
Cameron et al. 1989, Tannen 1992, Coates 1993, Greenwood 1996). The
value of sociolinguistic research that fails to situate the examination of lan-
guage structures in the arena of human communication, where the goals of
specific communicative acts can be considered, is highly suspect. We agree
with Cameron's admonition (1990:186) that "The social practice of language-
using is not defined simply by the act of speaking (or writing or signing) ...
What most crucially defines this social practice ... is the act of addressing
someone, in some context, for some purpose" (see also Eckert & McConnell-
Ginet 1992a,b, Gal 1991, 1992).
In this article we examine two discourse features which have been widely
studied and regularly associated with gendered speech styles. We investigate
the occurrence of you know, and the use of questions in casual conversations
between eight same-sex pairs of friends. These features were chosen precisely
because of the amount of attention they have received in language and gender
research. They can serve as exemplars of other linguistic forms and devices
which should be reanalyzed before reliable conclusions about the character-
istics of specifically situated same-sex verbal style can be made. In this dis-
cussion, we offer evidence of the following:
(c) Women and men of the same speech community, speaking in same-sex
pairs in the same conversational context, with equal access to the conversa-
tional floor, do not differ either in the frequency of the use of you know or
in the number of questions uttered.
(d) Women and men of the same speech community, speaking in same-
sex pairs in comparable settings, not only utter equivalent numbers of you
know and of questions, but use you know and questions to achieve compa-
rable discourse goals.
(e) It is more accurate to associate a style of speaking with a particularlin-
guistic task, or with a specific kind of speaking situation within a given socio-
cultural context, regardless of the sex of the individuals participating, than
to attribute a language style to any casually defined group of speakers.
The data for this study are from eight dyadic conversations, of approximately
35 minutes each, recorded in an experimental setting. Students from linguis-
tics and women's studies classes at a university in New Jersey were invited
to participate in a study of friendship. The participants were White, 18-28
years old, from middle- and working-class backgrounds. Each volunteer was
asked to bring a friend of the same sex (from inside or outside the univer-
sity community) to a specified location. The volunteers were informed in
advance that they would be audio- and video-recorded during the study. In
all, 30 conversations were recorded. We report here on conversationsbetween
four female and four male pairs of friends, roughly matched for age and
length of friendship.
To manipulate the talk situation, we divided the conversation into three
parts, each of which had distinct conversational requirements.When the par-
ticipants arrived, we gave the appearance of being overwhelmed with equip-
ment problems, and we appealed to them to be patient until we had
everything in working order. Microphones and a tape recorder were in full
view on the table at which they were asked to sit. The equipment was run-
ning and they were so informed, but we explained that the formal part of the
project was not yet underway. We then left them alone and said that we
would return as soon as we were organized, encouraging them to relax and
to enjoy the doughnuts and juice that we had provided. We label this por-
tion of the conversation the SPONTANEOUS talk portion because the speakers
themselves, not the researchers, dictated the interaction and controlled the
conversational material. They sat at the table, waiting for us to officially
begin, and chatted about matters unrelated to our project or to any other
assigned topic.
After 10 minutes we returned, apologized for the delay, and told the par-
ticipants that we were ready to proceed. We explained that, ratherthan inter-
Language in Society 25:1 (1996) 3
ALICE F. FREED AND ALICE GREENWOOD
view them separately, we wanted them to talk together about the nature of
friendship between women as compared to friendship between men. We
attempted to be as casual as possible in giving our instructions, hoping to
mitigate the speakers' self-consciousness. We call this the CONSIDERED talk
portion because the speakers were focused on a particulartopic, assigned by
the researchers.
After 15 more minutes we interrupted the conversations, thanked both
speakers, and asked each one to fill out an anonymous demographic ques-
tionnaire and a form granting us permission to use the taped conversations.
Because the documents had to be filled out individually, no talk was
required. However, there was usually a good deal of conversation. It was
apparent that the pairs of friends believed that their task was completed; they
proceeded to talk to each other in an informal and relaxed manner. They
made jokes about the questions, read them aloud to each other, sometimes
collaborated on their answers to the questions, and engaged in general com-
mentary about filling out the forms. This section, which lasted 6-10 minutes,
provided us with an opportunity to observe COLLABORATIVE talk.
This approach had several important methodological advantages for the
type of analysis that was undertaken. The research design allowed us to ana-
lyze the pragmatic expression you know and the use of questions in the lan-
guage of 16 individuals, in three different types of speech where both the
relationship between the interlocutorsand the setting remained constant. Fur-
thermore, this experimental design ensured that comparable speech samples
were analyzed for women and men. Holmes warns (1986:12) that "any com-
parison of the number of forms used by each sex clearly needs to control for
differential opportunities for producing such forms," a point frequently over-
looked in earlier research. Our approach guaranteed like access to speech
production for all female and male participants.
YOU KNOW
The expression you know has often been described as a female hedging
device, and interpreted as a marker of both insecurity and of powerlessness
(Lakoff 1975, Fishman 1978, 1980, O'Barr & Atkins 1980, Ostman 1981,
Coates 1986). For example, Fishman 1980 reported that, in the conversations
that she studied of three heterosexual couples, women's frequent and dispro-
portionate use of you know signaled both the subordination of the female
speakers in the conversations and the heavy conversational workload that
they carried. At the time that Coates' first edition of Women, men and lan-
guage was published (1986), researcherswere still making assumptions about
women's speech based in part on evidence from such mixed-sex conversa-
tional data. Thus you know was included in the list of gendered speech char-
acteristics typical of women and not of men. In her review of the literature,
4 Language in Society 25:1 (1996)
WOMEN, MEN, AND TYPE OF TALK
Coates pointed out that "A statement like It was, you know, really interest-
ing is considered less assertive than its unhedged version It was really inter-
esting" (1986:102). She also reported that the frequent occurrence of you
know revealed "malfunction in turn taking" (102). However, in the second
edition of the book (1993), these comments were revised to reflect the impor-
tance of examining linguistic forms in context. Our data show that impor-
tant information about the discourse function of you know in conversation
is masked if it is analyzed as a single pragmatic device, isolated from its full
conversational context.
Our corpus consists of 612 instances of you know,2 from approximately
4' hours of informal conversation between eight pairs of same-sex friends;
compare this to the mere 104 instances of you know in 11 hours of intimate
conversation between the three heterosexual couples reported by Fishman
1980. This large difference between our data and Fishman's corroborates the
observation made by Holmes (1986:14) of an increase in the frequency of
occurrence of you know in same-sex interaction as compared to mixed-sex
interaction. In addition, this contrast in the frequency of you know may
result from the type of discourse; i.e., casual conversation between friends
may elicit a greater use of this expression than intimate talk between couples.
A comprehensive analysis of the distribution of you know in our data
reveals the following:
(a) Every speaker in the sample, female and male, used this expression
eight or more times in 35 minutes of conversation.
(b) In every conversational dyad, one member of the pair used the expres-
sion more frequently than the other.
(c) The frequency of occurrence of you know varied widely with the type
of talk of the three differcnt segments of the conversations studied.
(d) All speakers in the sample, female and male, varied their usage of you
know in identical ways, in accordance with the three different types of talk.
Table 1 shows the number of occurrences of you know in the speech of
each participant, and compares it to the frequency of occurrence of this form
in the speech of that individual's conversational partner. The total number
of occurrences of you know in the eight female and male conversations out-
lined in Table 1 are almost identical. Out of the total 612 tokens of you
know, women used the form 310 times, and men 302 times.3 The totals
show that every member of these female and male pairs of friends used this
expression - and that, in each pair, one speaker used the form more often
than the other. For example, in female dyad 3, Speaker A used this expres-
sion 37 times, which represents 70% of the total occurrences of you know
in that conversation. The number was more than double that of her inter-
locutor, who used you know only 16 times. Similarly, in male pair 8, you
know occurred 20 times in the speech of Speaker A; again, this was more
Language in Society 25:1 (1996) 5
ALICE F. FREED AND ALICE GREENWOOD
Female Male
1. A: 36 B: 138 = 174 5. A: 11 B: 41 = 52
(21o) (79%) (21%Vo) (79%)
2. A: 8 B: 25 = 33 6. A: 38 B: 60 = 98
(24%) (76%) (39%) (61%)
3. A: 37 B: 16= 53 7. A: 71 B: 53 = 124
(70%) (30%) (57%) (43%)
4. A: 18 B: 32 = 50 8. A: 20 B: 8 = 28
(36%) (64%) (71%) (29%)
Total 310 Total 302
than double the eight occurrences of the form in his partner's speech. These
numbers show that, in informal dyadic conversation, it is not unusual for one
conversational partner to use you know more than the other.
When Fishman 1978, 1980 found an imbalance in the use of you know in
cross-sex conversations, she attributed it to the imbalance in power between
the speakers. Because her studies found that women used you know signifi-
cantly more than men, she interpretedyou know as a linguistic signal of the
interpersonal powerlessness of the women. However, in the conversations
under study here, where there is no overt (or societal) power differential be-
tween the participants, an imbalance is nonetheless found in the use of you
know. Given the limited knowledge that we have about the speakers' friend-
ship, it is unwise to speculate about the interpersonal function of the greater
number of you knows in the speech of one member of each pair. This imbal-
ance may reflect greater involvement or conversational effort on the part of
one speaker over the other - or more simply, may reflect a personal speak-
ing style of one of the participants. Yet it is clear that a specific linguistic phe-
nomenon or communicative feature, in this case the imbalance between
speakers in the use of you know, cannot be explained outside the full com-
municative context in which it occurs, and cannot be generalized from one
context to another.
In addition to analyzing the number of occurrences of you know by indi-
vidual speaker and by sex of speaker, we examined the distribution of you
know across the three distinct parts of the conversations. We found that a
striking 546 instances, or 89% of the total occurrences, occurred in Part II,
rising from 56 instances (or 907o)in Part I. Only 10 examples (2%) of you
know occur in Part III of the conversations. When adjusted for the time dif-
ferences between the three parts of the conversations, 13'07o of the instances
Legend
P kns
Pati:56youb
80- 546you
PathI: k&s
PaI
III : youb*s
pat
60-
PadlI:CIrdTalk (15mh.)
1:SptbneoTalk
Part min.)
(1O PadIII:
Collaborative
Talk min.)
(6-10
FIGURE 1: Distribution of you know by type of talk, adjusted for time.
of you know are seen to occur in Part I, 84%1o in Part II, and 2% in Part III;
see Figure 1.
Women and men were nearly identical in the number and distribution of
you know, with both groups showing an extraordinary increase in the usage
of you know in Part II; see Figure 2. Since the fluctuation in use is so dra-
matic for all eight dyads, regardlessof the sex or gender of speaker, it is clear
that associating this form with female speakers (or with males for that mat-
ter) is incorrect. Rather, these data lead us away from generalities which
focus on categories of sex or gender, and toward a conclusion grounded in
discourse requirements:women and men speaking in same-sex pairs respond
to the differences in the talk situation in exactly the same way. The only
change that takes place in the speaking situation in going from Part I to
Part II is the requirements of the talk itself. Thus it is the demands of this
particular talk situation, in which the participants respond to an assigned
topic - plus the constraints associated with this particular topic - that elicit
this verbal behavior. It is not the sex or gender of the speakers, or the nature
of their personal relationship outside this setting, that explains their increased
use of you know in the CONSIDERED section of the interaction.
For researcherswho have moved away from a reliance on gendered speech
styles as explanations for linguistic choices, there is now general agreement
that you know is a hearer-oriented expression, providing the speaker with a
100
Legend
Fern*E|
801
20 m-E
ParII:
Corsidred
Tall
(15min.)
Part Talk
l:Spontneous(10
min.) PadIlIl: Talk
Collaborative mi.)
(6.10
FIGURE 2: Distribution of you know by type of talk and by sex.
male pairs, you know can be seen as a device that effectively connects the
speaker to the hearer. This interactive function seemed to vary little from one
conversational segment to another; what changed was the frequency of
occurrence - brought about, we believe, by the participants' need to work
in concert to produce a discussion on an externally assigned topic of conver-
sation. This communicative behavior, manifested by both the women and
men in the sample, appears to reflect the speaker's desire for mutual orien-
tation, especially in the CONSIDERED talk segment.
These examples, which are representative of the sorts of utterances that
occurred in all the conversations, show that, after an utterance containing
you know by one speaker, the other speaker often follows with an explicit
yeah, just like ... , or a clear continuation of the topic initiated by the other
speaker in the immediately preceding utterance. (All these examples are taken
from Part II of the conversations.)
You know is thus a tool that works to reinforce mutual involvement in the
conversation, whether as an expression of speaker certainty or uncertainty,
and assists in the joint production of conversation. It is therefore understand-
able that 897o of the occurrences of you know in the data reported on here
take place in Part II of the conversations, where the speakers are most self-
consciously engaged in talking to one another. In this segment, speaker turns
are longer, phatic communication through question use increases, and a
greater amount of attention to a single topic is evident. It appears that both
female and male speakers are using you know to maintain close contact with
one another as they attempt to jointly construct a reply to an assigned topic
of conversation. It is the task at hand, the imposed face-to-face encounter -
and perhaps the nature of the subject itself, a discussion of friendship -
which best explains the frequent occurrence and the function of you know
in these conversations. Thus a discourse-based analysis, unlike previously
proposed explanations for the frequency of you know in the speech of
women (gender, sex, hesitancy, or conversational workload), provides an
adequate explanation for the findings reported here.
Question use is another linguistic device which has been stereotypically asso-
ciated with the conversational style of women. Lakoff 1975, among the first
to claim that women used more questions than men, declared that women
use tag questions as a hedging device, and that women have a greater ten-
dency to use rising intonation on declaratives, thereby turning their state-
ments into questions. As with you know, this usage was interpretedas a sign
of women's hesitancy and societal powerlessness. Lakoff's claims, which were
not based on empirical data, inspired a significant amount of research that
produced conflicting conclusions about tag and other question usage (Dubois
& Crouch 1975, Brouwer et al. 1979, O'Barr & Atkins 1980, Holmes 1984a,
Preisler 1986, Cameron et al. 1989, Coates 1989).
In other important research that subsequently formed the basis for widely
accepted beliefs about women's question-asking behavior, Fishman 1978,
1980 found that, in naturally occurring conversations between three hetero-
sexual couples, women asked many more questions than men. As with you
know, she interpreted this usage as evidence of the disproportionate conver-
sational workload carried by women when interacting with men. The differ-
ential use of questions by women and men has continued to be a much
discussed topic, fueled most recently by Tannen's (1990b) assertions about
women and men's different communicative styles, where question-asking is
seen as part of women's cooperative speaking style and as a device for shar-
ing the floor.
Despite the limited data on question use by women and men and the
paucity of definitive empirical evidence to support one of these three inter-
pretations over the others - hesitancy, conversational workload, or cooper-
ativeness - many linguists still accept the claim that, as a part of a female
speaking style, women ask more questions than men. Our treatment of ques-
tions demonstrates, to the contrary, "the absolute necessity of considering
forms in their linguistic and social context, not in general, and suggest that
[researchers]should regard multifunctionality as the unmarked case" (Cam-
eron et al. 1989:77). We believe that questions constitute another discourse
phenomenon that has been incorrectly associated with gender, and has been
inaccurately assumed to serve one invariant communicative function for a
particular group of speakers.
Our corpus consists of 787 questions uttered by the 16 different speakers
in the eight 35-minute conversations described earlier. We used syntactic and
intonational criteriato identify utterances as questions. Six types of questions
were found in the corpus:
(a) Yes/no questions characterized by simple subject-auxiliary inversion;
these included reduced yes/no questions where the auxiliary is deleted and
alternative questions.
(b) WH-questions.
(c) Full declaratives and other syntactic phrases with a final phrase rise.
(d) Tag questions, including both canonical (or auxiliary) tags, e.g. They
didn't hit you, did they? and invariant (or lexical) tags, e.g. That's whereyou
lived, right?
(e) WH-questions, followed by a phrase with a final rise in tag position,
e.g. What's today's date? the 25th?, sometimes called "WH-questions plus
guess" constructions (Norrick 1992).
(f) Questionsof the form How/ What about ..., e.g. What about when
women get older, like when they get married and stuff?
Female Male
1. A: 26 B: 45 = 71 5. A: 54 B: 65 = 119
(37Wo) (63Wo) (45%0) (55%)
2. A: 84 B: 41 = 125 6. A: 43 B: 38 = 81
(67%) (33%) (53%) (47%)
3. A: 58 B: 64 = 122 7. A: 27 B: 52 = 79
(48%) (52%) (34%) (66%)
4. A: 27 B: 59= 86 8. A: 65 B: 39= 104
(31%) (69%) (63%) (37%)
Total 404 Total 383
1X0
Legend
* Parti:328Ques0f
80- Partl:259Questions
[II PatIII:200Questions
60
20-
PatII:CafdlTalk(15niin.)
1:Sptneous
Part Tak(Omin.) Par
III:
Mo akTa k(6-1l0min.)
FIGURE 3: Distributions of questions by type of talk, adjusted for time.
number of you knows by that same speaker, based on the assumption that
the speaker was displaying a particularly female style, a powerless style
(O'Barr & Atkins 1980), or a cooperative style. However, this generalization
is contradicted by the data presented here. We find that, in only four of the
conversations (2 female and 2 male) does the speaker who uses more you
knows also ask a greater number of questions. In the other four dyads, the
speaker who uses the greater number of you knows either asks approximately
the same number of questions as the other speaker, or fewer.
Examining the distribution of questions across the three parts of the con-
versation (spontaneous, considered, and collaborative), we found within each
part, just as with you know, that male and female speakers all followed the
same pattern. Our results show a non-random distribution, with 4207oof all
the questions occurring in Part I, 3307oin Part II, and 25% in Part III.5
That is, the first, spontaneous talk portion of the interaction produced many
more questions than either of the other two parts. The distribution, adjusted
for time, is 45%oin Part I, 2401oin Part II, and 3107oin Part III; see Figure 3.
Our data again establish that neither sex nor gender is a salient variable in
these conversations, and that discourse requirementsunderlie the use of par-
ticular linguistic forms. Speakers adapt the number and functional type of
questions they ask to the demands of the particular conversational situation.
The same general patterns are followed by all eight pairs, with women and
men adjusting their question use in strikingly similar ways. The rate of ques-
Language in Society 25:1 (1996) 13
ALICE F. FREED AND ALICE GREENWOOD
1X--
Legend
Feniae
80-g Male
40-
-
20
PitII: Tak
Cosidered
(15
mn.)
Pat1:
Sp uUSTaIk
n.)
(1On P III:
CbtoeTaIk
(6Omh)
FIGURE 4: Distribution of questions by type of talk and by sex.
tioning, adjusted for time, is greater for both women and men in Part I of
the conversation, and the lowest in Part II. Consistent with the pattern found
for you know, these data show that women and men speaking in same-sex
pairs respond to the differences in the talk situation in exactly the same way;
see Figure 4. Once again, the only change in the speaking situation, between
one part of the conversation and the others, is the requirements of the talk
itself. It is these requirements that govern the frequency of occurrence of
questions. We conclude that, unless conversational participants are observed
in several different talk situations, with other variables controlled, assertions
about characteristics of their speaking style are suspect.
Not only did the rate of questioning change in the same way for all speak-
ers in the three parts of the conversation, but the functional type of ques-
tions being asked differed as well. Each section of the conversation had a
distinct distribution of question categories. Our earlier work (Freed & Green-
wood 1992, Greenwood & Freed 1992, Freed 1994) describes in detail how
each question in our corpus was separately identified, and then categorized
along an information continuum. Questions were found to fall into four
broad categories - depending on their functional use within the conversation,
and on the type of information sought or conveyed by the speaker. Our anal-
ysis makes it apparent that previous research which casually groups all ques-
14 Language in Society 25:1 (1996)
WOMEN, MEN, AND TYPE OF TALK
Legend
80 E
RdtiQuen
40 - _
0__~
ber of EXTERNAL questions rises to a high of 54% - more than half of all the
questions asked in this section of the conversation - and the RELATIONAL ques-
tions drop to 9'70. Again, EXPRESSIVE STYLE questions make up 26% of the
questions used. In fact, the use of such questions, through which speakers
convey rather than seek information (see Freed 1994), remains the most con-
stant throughout the three parts of the conversation.
The following are examples of the sorts of questions that most frequently
occurred in each of the three parts. We present one exchange from a female
pair and one exchange from a male pair to illustrate each of the major func-
tional categories of questions for each part of the conversation.
TALK questions (underlined), which provided the speakers with the oppor-
tunity to confirm or verify the preceding utterance of the other speaker or
clarify what the other speaker meant, occurred most frequently in Part I.
(5) Male pair
A: I heard she's a flight attendant.
B: Oh, is that what she is? Okay. I know it was, had something to do with the airlines.
A: That's what Schultz said. He said he probably never sees her.
B: She speaks French and he speaks German.
A: Fred speaks German?
10
Leqend
Qe
80 mTalk
ReWWQuesSom
0-
I Cmied Tak(15w.
Padll:
1:Span Tak
Pad (10*i.) III:
PartCoab Tak
(6lOninL)
FIGURE 6: Distribution of question types by type of talk, female speakers.
mation by sex. These figures show that women and men use questions in
almost identical ways. The distribution of functional categories used by both
women and men generally conforms to the configuration seen in Figure 5,
where no such breakdown is made.
The women and men in this sample do essentiallythe same kind of question-
asking, e.g. more EXTERNALquestions in Parts I and III than in Part II, more
TALK questions in Part I than in Part II, more RELATIONALquestions in Part
II than in either of the other two sections, and slightly more EXPRESSIVESTYLE
questions in Part II than in Parts I and III. However, some subtle differences
emerge. In all three parts, women ask a higher percentage of RELATIONAL
questions than men, and men use more EXPRESSIVE STYLE questions than
women. Of the total number of questions women asked, 29% were RELA-
TIONALquestions. For men, relational questions represented only 16% of the
total questions used. By contrast, men use questions for EXPRESSIVE STYLE
36% of the time, as compared to women's 24%. This finding is suggestive
of the research which reports that different socialization practices within the
same speech community may result in different language behavior for women
Language in Society 25:1 (1996) 19
ALICE F. FREED AND ALICE GREENWOOD
Legend
Exrnal
Questons
E S1leQuesbons
PatII:
Considered
Talk
(15
min.)
1:Spontneous
Pad Talk
(10
min.) Pad CotIabrative
IlI: Tak
(610
min.)
FIGURE 7: Distribution of question types by type of talk, male speakers.
and men (Goodwin 1980, Maltz & Borker 1982, Tannen 1990b). Women, it
is thought, are socialized to actively engage their conversational partners,
whereas men have been taught to display their verbal uniqueness. The slight
differences between question-asking style that we observe may well result
from the fact that, in our gender-differentiated society, men typically find
themselves in the sort of talk situations which elicit RELATIONALquestions less
often than many women; instead, they are in settings which encourage them
to express their individuality.6 We wish to emphasize, however, that the
comparable use of questions by this group of women and men when speak-
ing in identical talk situations is of prime importance, and that the similari-
ties between the four pairs of women and the four pairs of men are more
striking than any differences.
CONCLUSION
and to be involved in distinct practices; as a result, women and men may not
routinely participatein equivalent types of discourse. That is, we have no rea-
son to doubt that some differences in the everyday speech of women and men
may result from distinct socialization practices for girls and boys, and from
various gender-assigned activities; women may engage in cooperative talk in
a wider range of settings than men. Yet we believe that, just as the commu-
nicative style of women has been overly stereotyped as cooperative, so too
the verbal style of men has been overgeneralized as competitive and lacking
in cooperativeness. We are not convinced that the sort of verbal behavior
found in these conversations is absent from men's natural speech, and we
continue to object to the characterization of women and/or men as having
distinct conversational styles, when these differences are generalized to all
women and all men in all contexts and situations. As Eckert has argued
(1989:253):
gender does not have a uniform effect on linguistic behavior for the com-
munity as a whole, across variables, or for that matter for any individual.
Gender, like ethnicity and class and indeed age, is a social construction and
may enter into any of a variety of interactions with other social phenom-
ena. And although sociolinguists have had some success in perceiving the
social practice that constitutes class, they have yet to think of gender in
terms of social practice.
It is our hope that the conclusions from this study will motivate researchers
in sociolinguistics, particularlyin language and gender studies, to direct their
work away from simple correlations between linguistic form and communi-
cative function, away from dependence on global binary categorizations of
human beings, and toward a more nuanced analysis of gender and its inter-
action with other linguistic and social phenomena.
NOTES
* Earlier versions of this article were presented at the American Association of Applied Lin-
guistics in Seattle (1992), in Atlanta (1993), and at the 1992 BerkeleyWomen and Language Con-
ference. We thank the director of the Psychoeducational Center at Montclair State University
for allowing us to use their facilities for taping these conversations. We also thank Marc Freed-
Finnegan for technical computer assistance.
l In this article we have not attempted to resolve the important conceptual distinction
between "sex"and "gender."As the term gender has become the commonly accepted one in lan-
guage and gender research, we have used gender when referring to speech styles thought to be
associated with women and men. We use the term sex to refer to the biological categories of
women and men; and we use both terms in cases where the distinction is blurred. This usage
is intended to alert our readers to the fact that the issue remains unresolved.
2 For the purposes of this analysis, we considered only those instances of you know which
occurredas separateand intonationally discreteunits. We did not count as instances of you know
the initial part of such reduced yes/no questions as You know where I'm going? (from Do you
know whereI'm going?) or You know what? (from Do you know what?). In this we differ from
the analysis proposed by Schiffrin 1987, who considers both sorts of occurrences of you know
together. In our view, these forms are syntactically and functionally dissimilar.
3 In female pair 1, Speaker B uses more cases of you know than any other woman or man.
In fact, this one speaker, who utters you know 138 times in 35 minutes, accounts for 45% of
all instances of you know uttered by the eight female speakerscombined. The next highest occur-
rence of you know is the speech of a man, speaker A from pair 7, who uses the form 71 times.
Consider the strikingly different conclusions that would be drawn if a different female pair,
matched for age and length of friendship, were substitutedfor pair 1. In another such pair, which
we designate as la, one woman uses 10 instances of you know, and the other uses 61. If we sub-
stituted this female pair for pair 1, the combined number of occurrences of you know for all
eight pairs would come to 510 instances of you know, instead of 612; women would be shown
to use the form 207 (not 310) times, and men 302 times. Based on this other set of numbers,
we concluded in an earlier study (Freed & Greenwood 1993) that men used you know one-third
more often than women. At that time, we stated that our findings were consistent with the num-
bers reported by Holmes 1986, who found that, in same-sex samples, men used you know twice
as often as women.
4 When we first reported on this analysis of questions in 1992 (Freed & Greenwood 1992,
Greenwood & Freed 1992), we found that the total number of questions asked by women was
slightly higher than those asked by men. Our numbers were then based on inclusion of female
pair la described in note 3. It was our impression that one woman in this pair skewed the results,
because she used 133 questions, which was 49 more questions that the next highest questioner.
(The range of questions for all other speakers was 27 to 84 per individual.) As we have stated,
when we examined and compared the speech of this pair with that of female pair 1, matched
for age and length of friendship, we found that the occurrence of questions in 1 was consistent
with that of all the other dyads. However, one of the women in 1 used an unusually high num-
ber of you knows. We were obliged to choose between these two female pairs, each of which
included one individual whose speech was markedlyidiosyncratic. We chose to report on 1 rather
than la, because la would have forced conclusions that were not clearly substantiated by the
other speakers.
5 If the number of questions asked in these conversations occurred in a random distribution
(adjusted for the different length of time of each segment), we would expect 30%0to occur in
Part I, 4407 in Part II, and 26%oin Part III.
6 There is a considerable literature (Aries 1976, Edelsky 1981, Tannen 1990a, Bischoping
1993, Johnstone 1993) that explores sex differences in conversational topic choice. The discus-
sion goes beyond the scope of this article, but may be related to the issue raised here.
REFERENCES
Aries, Elizabeth (1976). Interactionpatternsand themes of male, female and mixed groups. Small
Group Behavior 7:7-18.
Bergvall, Victoria L.; Bing, Janet M.; & Freed, Alice F. (1996), eds. Language and gender
research: Rethinking theory and practice. London: Longman, to appear.
Bischoping, Katherine(1993). Gender differences in conversational topics: 1922-1990. Sex Roles
28: 1-18.
Brouwer, Dede; Gerritsen, Marinel; & deHaan, Dorian (1979). Speech differences between
women and men: On the wrong track? Language in Society 8:33-50.
Butler, Judith (1990). Gender trouble. London: Routledge.
Cameron, Deborah (1990). Feminism and linguistic theory. London: Macmillan.
; McAlinden, Fiona; & O'Leary, Kathy (1989). Lakoff in context: The social and linguis-
tic functions of tag questions. In Coates & Cameron (eds.), 74-93.
Chafe, Wallace, & Danielewicz, Jane (1987). Properties of spoken and written language. In
Rosalind Horowitz & S. Jay Samuels (eds.), Comprehending oral and written language, 83-
113. New York: Academic.
Coates, Jennifer (1986). Women, men and language. London: Longman.
(1989). Gossip revisited: Language in all-female groups. In Coates and Cameron (eds.),
94-121.
(1991). Women's cooperative talk: A new kind of conversational duet? In Claus Uhlig
& Rudiger Zimmerman (eds.), Proceedings of the Anglistentag 1990 Marburg, 296-311. Tu-
bingen: Niemeyer.
(1993). Women, men and language, 2nd ed. London: Longman.
(1994). Discourse, gender and subjectivity: The talk of teenage girls. Paper presented
at the Third Berkeley Women and Language Conference: Communication in, through, and
across Cultures, Berkeley.
(1996). Women talking to women. Oxford: Blackwell, to appear.
, & Cameron, Deborah (1989), eds. Womenin their speech communities. London & New
York: Longman.
DuBois, Betty Lou, & Crouch, Isabel (1975). The question of tag questions in women's speech:
They don't really use more of them, do they? Language in Society 4:289-94.
Eckert, Penelope (1989). The whole woman: Sex and gender differences in variation. Language
Variation and Change 1:245-67.
, & McConnell-Ginet, Sally (1992a). Communities of practice: Where language, gender
and power all live. In Hall et al. (eds.), 1:89-99.
; ~~ (1992b). Think practicallyand look locally: Languageand gender as community-
based practice. Annual Review of Anthropology 21:461-90.
Edelsky, Carole (1981). Who's got the floor? Language in Society 10:383-421.
Fishman, Pamela (1978). Interaction: The work women do. Social Problems 25:397-406.
(1980). Conversational insecurity. In Howard Giles et al. (eds.), Language: Social psy-
chological perspectives, 127-32. New York: Pergamon.
Freed, Alice F. (1994). The form and function of questions in informal dyadic conversation.
Journal of Pragmatics 21:621-44.
, & Greenwood, Alice (1992). Why do you ask?: An analysis of questions between friends.
Paper presented at the Annual Meeting of the American Association of Applied Linguistics,
Seattle.
; ~~ (1993). An examination of you know in same-sex conversation. Paper presented
at the Annual Meeting of the American Association of Applied Linguistics, Atlanta.
Fromkin, Victoria, & Rodman, Robert (1993). An introduction to language, 5th ed. Fort Worth:
Harcourt Brace Jovanovich.
Gal, Susan (1991). Between speech and silence: The problematics of research on language and
gender. In Micaela di Leonardo (ed.), Genderat the crossroadsof knowledge: Feminist anthro-
pology in the postmodern era, 175-203. Berkeley: University of California Press.
(1992). Language, gender and power: An anthropological view. In Hall et al. (eds.),
1:153-61.
Goffman, Erving (1983). Forms of talk. Philadelphia: University of Pennsylvania Press.
Goodwin, Marjorie Harness (1980). Directive-response speech sequences in girls' and boys' task
activities. In McConnell-Ginet et al. (eds.), 157-73.
(1990). He-said-she-said: Talk as social organization among Black children. Blooming-
ton: Indiana University Press.
(1994). Co-construction in girls' hop scotch. Paper presented at the Third Berkeley
Women and Language Conference: Communication in, through, and across Cultures,
Berkeley.
Greenwood, Alice (1996). Floor management and power strategies in adolescent conversation.
In Bergvall et al. (eds.), to appear.
, & Freed, Alice F. (1992). Women talking to women: The function of questions in con-
versations. In Hall et al. (eds.), 1:197-206.
Hall, Kira; Bucholtz, Mary; & Moonwomon, Birch (1992), eds. Locating power: Proceedings
of the Second Berkeley Women and Language Conference. Berkeley: Berkeley Women and
Language Group.
Hall, Kira, & O'Donovan, Veronica (1996). Shifting gender positions among Hindi-speaking
Hijdas. In Bergvall et al. (eds.), to appear.
Hirschman, Lynette (1973). Female-male differences in conversational interaction. Paper pre-
sented at the annual meeting of the Linguistic Society of America, San Diego.
(1994). Female-male differences in conversational interaction. Language in Society
23:427-42.
Holmes, Janet (1984a). Hedging your bets and sitting on the fence: Some evidence for hedges
as support structures. Te Reo 27:47-62.
(1984b). Women's language: A functional approach. General Linguistics 24:149-78.
(1986). Functions of you know in women's and men's speech. Language in Society
15:1-22.
(1993). New Zealand women are good to talk to: An analysis of politeness strategies in
interaction. Journal of Pragmatics 20:91-116.
Hymes, Dell (1974). Foundations in sociolinguistics: An ethnographic approach. Philadelphia:
University of Pennsylvania Press.
Johnstone, Barbara (1993). Community and contest: Midwestern men and women creating their
worlds in conversational storytelling. In Deborah Tannen (ed.), Gender and conversational
interaction, 62-82. Oxford & New York: Oxford University Press.
Labov, William (1972). Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press.
(1991). The intersection of sex and social class in the course of linguistic change. Lan-
guage Variation and Change 2:205-54.
Lakoff, Robin (1975). Language and women's place. New York: Harper & Row.
Maltz, Daniel & Borker, Ruth (1982). A cultural approach to male-female miscommunication.
In John Gumperz (ed.), Language and social identity, 195-216. Cambridge& New York: Cam-
bridge University Press.
Macaulay, Ronald K. S. (1978). Variationand consistency in Glaswegian English. In Peter Trud-
gill (ed.), Sociolinguistic patterns in British English, 132-43. London: Arnold.
McConnell-Ginet, Sally; Borker, Ruth; & Furman, Nelly (1980), eds. Women and language in
literature and society. New York: Praeger.
McLemore, Cynthia (1991). The interpretation of L*H in English. Linguistic Forum (Univer-
sity of Texas) 32:175-96.
Morgan, Marcyliena (1991). Indirectness and interpretation in African-American women's dis-
course. Pragmatics 1:421-52.
Norrick, Neal R. (1992). WH-questions with guesses in tag-position. Journal of Pragmatics
18:85-89.
O'Barr, William, & Atkins, Bowman K. (1980). "Women's"language or "powerless" language?
In McConnell-Ginet et al. (eds.), 93-1 10.
Ochs, Elinor (1992). Indexing gender. In Alessandro Duranti & Charles Goodwin (eds.), Rethink-
ing context: Language as an interactivephenomenon, 335-59. Cambridge & New York: Cam-
bridge University Press.
Ostman, Jan-Ola (1981). You know: A discoursefunctional approach. Amersterdam:Benjamins.
Preisler, Bent (1986). Linguistic sex roles in conversation. Berlin: Mouton de Gruyter.
Romaine, Suzanne (1978). Postvocalic /r/ in Scottish English: Sound change in progress? In
Trudgill (ed.), 144-57.
Schegloff, Emmanuel A.; Jefferson, Gail; & Sacks, Harvey (1977). The preference for self-
correction in the organization of repair in conversation. Language 53:361-82.
Schiffrin, Deborah (1987). Discourse markers. Cambridge & New York: Cambridge University
Press.
Sheldon, Amy (1990). Pickle fights: Gendered talk in preschool disputes. Discourse Processes
13:5-31.
Tannen, Deborah (1990a). Gender differences in conversational coherence: Physical alignment
and topical cohesion. In Bruce Dorval (ed.), Conversationalorganization and its development,
167-206. Norwood, NJ: Ablex.
(I 990b) Youjust don't understand: Women and men in conversation. New York: Wil-
liam Morrow.
(1992). Rethinking power and solidarity in gender and dominance. In Claire Kramsch
& Sally McConnell-Ginet (eds.), Text and context: Cross-disciplinaryperspectives on language
study, 135-47. Lexington, MA: Heath.
Thorne, Barrie, Kramarae, Cheris; & Henly, Nancy (1983), eds. Language, gender and society.
Rowley, MA: Newbury House.
Trudgill, Peter (1972). Sex, covert prestige and linguistic change in the urban British English
of Norwich. Language in Society 1:179-95.
(1978), ed. Sociolinguistic patterns in British English. London: Arnold.