A rallying cry of the early women's movement was "The personal is political." This was not only a slogan, but a directive for analysis that would be a means for understanding and reinterpreting our private lives. I was a first year graduate student in sociology when I joined the women's movement. My specialty ethnomethodology, had already shown me the importance of everyday interactions, that they are not simply a reflection of reality but the means by which people construct and maintain their understanding of the world and of themselves. Through interactions, we constantly define reality, as well as the form and substance of our social relationships.
However, this approach did not deal with power, which, as a feminist, I found was a central concern in everyday interactions. I wanted to know the ways in which interactions between men and women express power relations; I wanted to see exactly how our personal interactions are political.
We know much more now about the social sources of our oppression than we did a few years ago. The relative positions of the sexes in the economy have been well documented. Analysis has shown the ways in which women's unpaid domestic labor contributes to our position of powerlessness. Still, we feel this powerlessness individually, and it is maintained concretely on an everyday basis.
I began to research the ways in which the general social organization of male-female hierarchy exists in our daily activities. Specifically, I have analyzed male-female conversations between intimates in their homes.
Initially, I hoped to tape record arguments, since these would be conversational power struggles. But as it turned out, I didn't get any arguments. Instead, I discovered that such overt power struggles were not necessary to illuminate women's oppression in everyday conversation. Rather, supposedly trivial daily talk reveals a division of labor between men and women in conversation which supports our more general positions of power and powerlessness.
It is not simply two people talking that produces conversation, but their active
agreement to mutually cooperate in talk. Interactions are always potentially
problematic and are sustained only by the participants' continuing efforts, turn
by turn. They must be begun, developed and ended jointly.
How, then, is the work of interaction accomplished in conversations between male and female intimates? Who does what kind of work? To find out, I listened to 52 hours of tape-recorded conversation between three couples who had agreed to have a Uher 4000 tape recorder set up in their apartments. The apartments were small enough for the recorder to pick up all conversation from the kitchen and living room as well as louder talk from the bedroom and bath. The tape recorder was in place from four to 14 days. The tapes could run for four hours without interruption. Though automatic timers could switch the tapes on and off, all three couples insisted on doing the switching manually. Uninterrupted recording varies from one to four hours. The talk did not seem self-conscious because of the recorder's presence and conversations seemed natural.
The couples had lived together for varying amounts of time—from three months to two years. All were white and professionally oriented, and all but one woman (a social worker) were in graduate school They were between the ages of 25 and 35. Two women were declared feminists and all three men as well as the third woman described themselves as "sympathetic" to the women's movement.
I made some interesting discoveries even before I began my analysis, through casual conversations with the people involved in the taping. First, in all three cases, the men set up the tape recorders and nearly always turned them on and off, sometimes without letting the women know. The reverse never occurred. Now, controlling a conversation is more than controlling the topic; it is controlling the situation in general—not only what wil be talked about, but whether there will be conversation, and under what terms it will occur. The tape recorder introduced a new element into the routine home situation. Controlling the tape recorder meant controlling part of the situation in which conversation took place.
Second, there was the issue of normally private interactions being available to a third party, the researcher. In addition to the men doing most of the censoring of the tapes, they made other attempts to control what I heard. For instance, the clicks that are recorded when the machine is turned off (and my own sense of the conversations) helped me to separate time segments. One man carefully erased all clicks on his and his partner's tapes, confusing my attempts to identify different time segments.
The second case was worse. I had made the mistake of asking one couple to help me transcribe a particularly difficult tape of theirs—a conversation about a book club selection. In fact, they couldn't hear themselves any better than I could, and the man wanted to know why I was interested in the conversation. He kept guessing what I was looking for, was annoyed that I wanted a literal transcription, kept saying that only the gist of the conversation (which he could tell me) was important, and repeatedly tried to tell me the meaning of the conversation, explain ing his motivations for saying things.
These preliminaries already suggested that men are more likely than women to control conversation. The men made sure that they always knew when their interactions were available to me by their control of the recorder, but they were unconcerned, if not sneaky, about letting the women know this. When they had the chance, they attempted to control my interpretations of the conversations.
I listened to all 52 hours of tape and transcribed five hours of it for
close analysis. Given the small amount of conversation analyzed, my findings
from the transcripts are provisional. But the implications from such a small
amount of material suggest this is only the tip of the iceberg. From
listening to all the tapes, it appeared to me that the men controlled the
subject much more than the women did and that the men were consistently
When I analyzed specific transcripts, it appeared that the men did very little to gain such control, and by doing very little, they displayed power without specifically exercising it. A turn-by-turn analysis revealed the women's constant conversational work. They struggled to get responses to their own remarks or were busy responding to the men's remarks. In my analysis, I identified a number of strategies used in conversational work. These strategies reveal the differential work men and women do.
Questions are a powerful kind of utterance. They are half of a paired relation, with answers the second half. That is, Q-A is a unit, not two separate things. Questions demand answers, while statements do not demand other statements. If no answer is given, its absence is noticeable and can be complained about. A question in conversation works by beginning a two-part sequence; it ensures an interaction of at least one utterance by each of two participants.
Counting the questions in about seven hours of tape, I found that the women asked two and a half times more questions than the men did—150 to 59. At times, it seemed that all the women did was ask questions. Then I began noticing my own speech and discovered the same pattern. I often made questions out of remarks that could have been statements: "Isn't it a lovely day?", "Shouldn't we go grocery shopping?" I then tried to break myself of the "habit" and found it was difficult. Many remarks came out as questions before I noticed. When I did succeed in making a statement, I usually did not get a response. It became clear that I wasn't asking so many questions out of habit, but because unless I did, my attempts at interaction would fail.
Attention-getting devices are remarks that, whether they give information or not, attempt to get the attention necessary to begin conversation. The women used these devices consistently while the men did not, indicating the women's difficulty in assuming that what they said would be treated seriously and would generate a response. l've identified three types of attention-getting devices.
The first was originally discussed by Harvey Sacks when he explored
children's restricted rights to speak in the presence of adults.
Second, there were many instances of y know" in the transcripts. This phrase is an attempt to see if the other person is paying attention. It was used 34 times by the women and three times by the men. The use of the phrase increased as the men's responses decreased.
Third, "This is interesting," or a variation, was used throughout the tapes.
By engaging in conversation people indicate that there is a mutual interest
in what is being said. When one uses "This is interesting" as an
introduction, it signals that the remark itself may not be seen as worthy of
attention. The speaker is working to establish the interest of the remark.
In the five hours of transcribed material, the women used this device ten
times, the men seven.
A minimal response refers to a speaker saying "yeah," "umm," "huh," and
little more. Both men and women used this but often in different ways. The
men used the monosyllabic response as one step short of not responding at
all. For example, the woman made a long remark, and the man responded with
"yeah," which neither encouraged her to continue nor elaborated the topic.
Minimal responses
The women also used this kind of minimal response at times, but mostly in a way that is best termed "support work. While the man talks, the woman skillfully inserts "mmm, "yeah, oh," throughout the stream of talk. By doing this, the woman signals that she is paying attention to what is being said, that she is fully participating, that she has a continuing interest in both interaction and interactant. This is done with agility during the speaker's pauses, seldom causing even slight overlaps. Nothing in either tone or structure suggests that the woman is trying to take over the talk.
Finally, there are statements that do no specific conversational work. Of course, by its mere existence, a statement provides for a response, and thereby does some minimal work. However, it also shows the speaker's assumption that any attempt will be successful and interesting, that he or she will be responded to. It is as if the speaker assumes there are no problems, that success is naturally his or hers. In the transcribed material, the men made this kind of statement more than twice as many times as the women. The substance of their remarks was not more interesting than the women's, but took on that character because they generated interaction.
One segment of conversation from my transcripts is the beginning of an interaction. The woman is reading a book in her academic specialty and the man is making a salad. I originally transcribed it for analysis because conversation seemed difficult for the woman, and she used strategies to ensure some response. (I suggest reading the transcript before continuing with the text.)
The woman (F) starts off in Set I with two "d'ya know" question sequences, signaling that she is not sure she can get the man's (M's) attention. It seems to be the right assumption, since M's response in Set 3 is minimal and both fall silent. F's next attempt, in Sets 5-6, uses more attention-getting devices. This time, F prefaces her remarks both with a "That's very interesting" and "Did you know ...? The double-barreled attempt is more successful and generates two exchanges, Sets 7-8. Fs success is short-lived, however; M fails to respond to her last contribution and thus ends the exchange.
F's third attempt begins in Set 10 with "That's really interesting." As M makes no attempt to follow up on any of her statements, F continues talking for 30 seconds, using two "y'knows" and a question intonation. M finally responds with a "Yeah" after F repeats herself and says "y'know" a third time. Though there are further attempts made by F later in the interaction, a conversation on this topic never develops.
The transcript shows some ways the strategies described earlier are used in actual conversation. It shows the woman working at interaction and the man exercising his power by refusing to participate fully. As the interaction develops and F becomes more aware of her difficulties, F brings more pressure on M by increasing her use of strategies. Even so, she only ensures immediate, localized responses, not a full conversational exchange.
Immediately following this transcribed material, M begins and continues a conversation about a soft drink and Richard Nixon being a former lawyer for Pepsi Cola. F gives elaborating responses, with a series of exchanges between them that end when M opts not to continue. This Pepsi-Cola/Nixon exchange shows that the man was willing to engage in discussion, but apparently only on his own terms.
There is a division of labor in conversation. Though the women generally do more work, the men usually control the conversations couples have. Since the men's remarks develop into conversation more often than the women's, men end up defining what will be talked about and which aspects of reality are the most important.
Women are required to do specific kinds of interactional work in conversations with men. More than that, we are generally required to be available. The conversational work expected of women differs according to the situation; some times we are supposed to be an audience, "good listeners," because we are not otherwise needed. We may have to fill silences and keep conversations moving. Sometimes we are supposed to develop other people's topics and other times to present and develop topics of our own. There are subtle demands on women to be available for interaction. If women do not answer these demands—if we are not "naturally" available, we get in trouble. Women who sit silently while a conversation flounders are seen as hostile or inept. Women who consistently and successfully control interactions are criticized, particularly by men who question the female's status as a woman. They are likely to be called "bitchy," "domineering," or "aggressive." When women attempt even temporarily to control conversation with men, it often starts an argument.
Women who do not behave interactionally are punished; often it is implied that they are not "real women." One's sexual identity is crucial. It is the most "natural" differentiating characteristic there is. But it is not simply our bodies that define gender. We must constantly behave as male or female in order for our gender to be taken for granted in interaction. We must prove our gender continually.
The active maintenance of gender requires women to be available to do what ever needs to be done in interactions. Since interactional work is tied up with female identity, with what a woman is, the fact that it is work is obscured. It is not seen as something we do, but as part of what we are, similar to caring for the house and children, which have also been seen as part and parcel of what women are. In both cases, the value and necessity of the work are hidden behind the screen of the sexual division of labor.
I am grateful to Harvey Molotch, with whose guidance I began this research, and to Myrtha Chabran, Mark Fishman, Drew Humphries, Linda Marks, Morgan Sanders and Susan Wolf for their help or ideas and criticism on this and earlier drafts.
Pam Fishman is 32, grew up in Arizona and lives in Brooklyn. Her Ph.D. thesis, from the University of California at Santa Barbara, is on power in everyday conversation. She and Linda Marks are writing up what they and their friends often talk about—using their daily experience to do feminist social analysis.