That's Not What I Meant!: How Conversational Style Makes or Breaks Relationships

by Deborah Tannen

A perfectly tuned conversation is a vision of sanity—a ratification of one’s way of being human and one’s place in the world. And nothing is more deeply disquieting than a conversation gone awry.

To say something and see it taken to mean something else; to try to be helpful and be thought pushy; to try to be considerate and be called cold; to try to establish a rhythm

glide effortlessly about the room, only to end up feeling like a conversational clod who can’t pick up the beat—such failure at talk undermines one’s sense of competence and of being a right sort of person. If it happens continually, it can unde...

“We didn’t go to the party because you didn’t want to.” “I wanted to go. You didn’t want to.” It would turn out that he had taken something I’d said as a hint about what I wanted, and I mistook his agreement with what he thought I wanted for being what he really wanted. He kept acting on hints I hadn’t thrown out, and I kept missing hints he had. With both of us loaded with good will, we kept doing what neither wanted. And instead of thanks, we both got recriminations. We were driving each other crazy.

What makes misunderstandings like these so hard to straighten out is that our ways of communicating seem self-evidently natural to us. He didn’t feel he was hinting; he felt he was communicating. He didn’t feel he was picking up hints from me; he felt he was hearing me communicate.

That’s why the frequently heard advice to “be honest” doesn’t help much. We were being honest. But our ways of being honest were different—and mutually unintelligible. When I missed his hint,

The only way we knew of treating the disease was precisely what was causing it—talking.

She saw for the first time that what she had thought of as being polite was actually indirect and possibly not clear communication.

Indirectness, ways of using questions or refusing politely, are aspects of conversational style. We also send out signals by how fast we talk, how loudly, by our intonation and choice of words, as well as by what we actually say and when. These linguistic gears

are always turning, driving our conversations, but we don’t see them because we think in terms of intentions (rude, polite, interested) and character (she’s nice, he’s not).

Many instances of rudeness, stubbornness, inconsiderateness, or refusal to cooperate are really caused by differences in conversational style.

Often the most effective repair is to change the frame—the definition or the tone of what’s going on—not by talking about it directly but by speaking in a different way, exhibiting different assumptions, and hence triggering different responses in the person we’re talking to.

But the most important thing is to be aware that misunderstandings can arise, and with them tempers, when no one is crazy and no one is mean and no one is intentionally dishonest. We can learn to stop and remind ourselves that others may not mean what we heard them say.

Many Americans, especially (but not only) American men, place more emphasis on their need for independence and less on their need for social involvement. This often entails paying less attention to the metamessage level of talk—the level that comments on relationships—focusing instead on the information level. The attitude may go as far as the conviction that only the information level really counts—or is really there. It is then a logical conclusion that talk not rich in information should be dispensed with. Thus, many daughters and sons of all ages, calling their parents, find that their fathers want to exchange whatever information is needed and then hang up, but their mothers want to chat, to “keep in touch.”

American men’s information-focused approach to talk has shaped the American way of doing business. Most

Women are negatively stereotyped as frivolously talking at length without conveying significant information. Yet their ability to keep talking to each other makes it possible for them to maintain close friendships.

Our needs for involvement and independence^—to be connected and to be separate—are not sequential but simultaneous. We must serve both needs at once in all we say.

Linguist Robin Lakoff devised another set of rules that describe the motivations behind politeness—that is, how we adjust what we say to take into account its effects on others. Here they are as Lakoff presents them: 1. Don’t impose; keep your distance. 2. Give options; let the other person have a say. 3. Be friendly; maintain camaraderie.

Negotiating the offer of a drink is a fairly trivial matter, though the importance of such fleeting conversations should not be underestimated. The way we talk in countless such daily encounters is part of what constitutes our image of ourselves, and it is on the basis of such encounters that we form our impressions of each other. They have a powerful cumulative effect on our personal and interactive lives.

You can be nice to someone either by showing your involvement or by not imposing. And you can be mean by refusing to show involvement—cutting her off—of by imposing—being “inconsiderate.” You can show someone you’re angry by shouting at her—imposing—or refusing to talk to her at all: the silent activity called snubbing.

Partners in intimate relationships often differ about how they balance involvement and independence. There are those who show love by making sure the other eats right, dresses warmly, or doesn’t drive alone at night. There are others who feel this is imposing and treating them like children. And there are those who feel that their partners don’t care about them because they aren’t concerned with what they eat, wear, or do. What may be meant as a show of respect for their independence is taken as lack of involvement— which it also might be. Maxwell

But what he intended was not what she understood—which was what she would have meant if she had said what he said in the way he said it. These paradoxical metamessages are recursive and potentially confusing in all conversations. In a series of conversations between the same people, each encounter bears the burdens as well as the fruits of earlier ones.

The American expected a show of involvement; they were being polite by not imposing.

But because we can’t step out of the situation—the human situation—we keep trying to balance these needs. We do it by not saying exactly what we mean in our messages, while at the same time negotiating what we mean in metamessages. Metamessages depend for their meaning on subtle linguistic signals and devices.

Conversational style isn’t something extra, added on like frosting on a cake. It’s the very stuff of which the communication cake is made. Aspects of conversational style are the basic tools of talk—the way we show what we mean when we say (or don’t say) something. The main signals are pacing and pausing, loudness, and

pitch, all of which make up what is commonly thought of as intonation.

These signals are used in linguistic devices that do the work of conversation, complex work that includes, always and simultaneously: creating conversation by taking turns talking; showing how ideas are related to each other; showing what we think we are doing when we talk (for example, we’re listening, interested, appreciative, frien...

Such differences are not a matter of some people expecting long pauses and others expecting short ones. Long and short are relative; they have meaning only in comparison to something—what’s expected, or someone else’s pause. Someone who expects a shorter pause than the person she’s speaking to will often start talking before the other has a chance to finish or to start. Someone who is waiting for a longer pause than the person she’s speaking to won’t be able to get a word in edgewise.

That’s why slight differences in conversational style—tiny little things like microseconds of pause—can have enormous impact on your life. These little signals make up the mechanics of conversation, and when they’re even slightly off, conversation is thrown off—or even cut off.

When you hear others talking more loudly than you expect, they seem to be shouting—and seem angry or brash. When you hear others talking more softly than you expect, they seem to be whispering—and withholding or unassertive.

If they use loudness at unexpected points in their talk, you can get confused about what’s important, or even what the point is. If you expect extra loudness to express emotion—for example, anger—and you don’t hear it, you may not notice when those with different styles are angry. If you discover they are, you may think there’s something wrong with them for not expressing it in what seems to you a normal way.

If you expect extreme shifts in pitch and don’t hear them, what you hear sounds monotonous. You get the impression that the speaker is a bland sort of person, or doesn’t care much about this conversation, or even is emotionally disturbed, suffering from “flattened affect.” If you don’t expect such extreme pitch shifts and you hear them, you may conclude that the person is overdramatizing or overemotional.

Since signals such as pitch shifts (as well as loudness and pacing) are also signs of emotional expression, it is probably no coincidence that women tend to use greater shifts in pitch than men, and that women are often perceived as overemotional. The same goes for members of certain cultural groups, including Greeks.

Bearing this in mind, psychiatrists, psychologists, and social workers, whose jobs entail assessing the appropriateness of levels of emotional expression, must make efforts not to take ...

The disbelief was aimed not at David but at the person about whom David was complaining, so the metamessage to David was intended to be “I agree that this other guy is ridiculous; this story is really worth telling, and I’m on your side.”

The daughter in this example, like David in the preceding one, was getting more of a reaction than expected. The flip side of such differences is getting less reaction than expected and hence the impression that the toned-down listener isn’t listening, isn’t following, or isn’t interested. When this happens on the telephone, you may actually ask, “Are you still there?”

Another way of showing interest and appreciation is asking questions. But questions can also seem nosy, overbearing, or hinting at something else. Questions, like everything we say, work on two levels at once: the message and the metamessage.

Some people show interest by asking questions, and others expect people to volunteer what they want to say. Some people encourage others to talk by getting the ball rolling themselves. Others wait to be asked. If Mary is waiting to be asked, and John is waiting for her to volunteer, she will never talk—and each will blame the other for the resulting imbalance.

Self-revelation, asking questions, and complaining can all be used according to the conversational principle “Do as I do.” The expectation that others will follow suit explains what otherwise seems like irrational or even hypocritical conversational behavior.

These conversational signals and devices are normally invisible, the silent and hidden gears that drive conversations. We don’t pay attention to the gears unless something seems to have gone wrong. Then we may ask “What do you mean by that?” And even then we don’t think in terms of

the signals—“Why did your pitch go up?”—but in terms of intentions—“Why are you angry?”

Many of these signals and devices can be changed if we’re aware of them, either across the board or with certain others. And minor changes can have major results. For example, when conversations just don’t seem to be going well, we can try making little adjustments in our volume, pacing, or pitch—speeding up or slowing down, leaving longer pauses or shorter ones—in an attempt to get closer to a shared rhythm. And realizing that ritual complaining and apologizing do not have the same meaning for every...

What we say is also an important clue to what we mean, but we don’t always say what we mean in so many words. We balance the conflicting needs for involvement and independence by hinting and picking up hints, by refraining from saying some things and surmising what other people mean from what they

refrain from saying. Linguists refer to the way people mean what they don’t exactly say as indirectness.

Why don’t we just say what we mean—directly? We’ve seen that it’s more satisfying to communicate indirectly; it would be boring simply to say what we mean, and we’d lose the metamessage of rapport. It’s useful to cover ourselves by not going on record with what we think. But even if we wanted to be direct, we couldn’t, for the following reasons: First, deciding to tell the truth leaves open the question, which of the infinite aspects of the truth to tell. Second, being direct isn’t enough because countless assumptions underlie anything we say or hear. We don’t think of stating them precisely because they are assumptions. Third, stating just what we mean would often be hurtful to others. And finally, differing styles make honesty opaque. Saying what we mean in our natural style conveys something different to those whose styles differ. Attempts to get others to communicate in a way that seems natural to us will seem manipulative to them—and won’t work. Let’s look at examples of why we can’t say what we mean.

Thus, one reason we can’t solve the problems of indirectness by being direct is that there are always unstated assumptions—both the speaker’s and the hearer’s—that may not match. We don’t state them precisely because they are assumptions—by definition, ideas that are not stated because they are taken for granted. We don’t become aware of assumptions until there is unmistakable evidence that they are not shared.

As the conversation proceeded, each one kept hearing the other say things that were surprising and odd because Claire assumed that Ross was calling from his own home, and he assumed that he was calling from John’s. He forgot to state where he was, and she didn’t think of asking because she assumed she knew. Neither one came out and said, “WHAT ARE YOU TALKING ABOUT?” They kept dismissing the oddness or devising interpretations for it until Claire heard something she couldn’t interpret at all: Twenty-two twenty-two Regent Street. Since we are all walking through life on an individual path, there are many times when information assumed by one is unimagined by the other. If the assumptions turn out not to be shared, we may later be blamed—and blame ourselves—for not having stated them. The right to remain silent is of no use in conversation. But it is neither natural nor possible to state all the assumptions underlying everything we say. And when problems arise, we often can’t trace them back to the specific conversations, let alone the assumptions underlying them, that led us astray.

Those who do not expect or like directness are not so much unwilling as unable to use it.

Each feels manipulated by the other, but they’re both just trying to get comfortable—and to do things right. This is analogous to what happens when two people who are standing and talking have different ideas about how close to stand when they talk. Both try instinctively to adjust the space between them to what is natural and comfortable, with the result that one keeps backing up and the other keeps advancing. They end up edging their way down the hall. Each one feels maneuvered by the other—and is. But neither consciously intends to force the other into anything. They’re both just trying to make the situation feel right. The danger—and inaccuracy—of a term like “manipulative” is that it blames others for the way we feel in response to them. The Uses of Indirectness Why can’t we just say what we mean? Why is so much communication indirect, hinted at in metamessages, picked up in tones of voice and glimpsed in facial expressions instead of confronted head on and clearly stated in words? First, there is a payoff in rapport. It is far better to get what we want, to be understood, without saying what we mean. It makes us feel the fine pleasure of being on the same wave length. This is the pleasure of those magical conversations when we say just a few words—or no words at all— and feel completely understood. It’s the communication jackpot, the pursuit of which makes us play the birthday-present and related do-you-love-me games. Second, there is a payoff in self-defense. If wha...

The payoffs of indirectness in rapport and self-defense correspond to the two basic dynamics that motivate communication: the coexisting and conflicting human needs for involvement and independence. Since any show of involvement is a threat to independence, and any show of independence is a threat to involvement, indirectness is the life raft of communication, a way to float on top of a situation instead of plunging in with nose pinched and coming up blinking. Through indirectness, we give