Zulawski observes that blundering itself, or a certain number of errors, or certain types of errors, doesn’t point directly at deceptive intent. What’s more significant, he believes, is the frequency of errors in a stressful situation relative to a speaker’s normal speaking style. Individuals vary widely in how much they blunder verbally, a fact that police officers and skilled interviewers recognize. The next time you’re pulled over, the cop will often begin with a question that sounds like idle chitchat: what a sunny day we’re having; where are you going? But he or she is actually checking how you respond to something that presumably should be easy to answer. They start with the mundane in order to gauge how you give answers you don’t have to think about. Theoretically, a person could confound an interrogator from the outset with liberal doses of “uh” and “um,” like a jet spreading chaff to throw off a pursuing missile.
From here, how verbal blunders vary among individuals becomes stickier. It can be tempting to attribute a speaker’s verbal blunders to his or her intelligence (or lack of it); only the cruelest misanthrope would blame all verbal blunders on a failure of smarts. The sociologist Erving Goffman offered a better distinction. He called some of them “knows better” errors and others “doesn’t know better” errors. Both occur in abundance in our behavior, including our verbal lives.
The “doesn’t know better” errors are caused by stupidity or lack of experience. These “boners” and “gaffes” (as Goffman dubs them) occur when someone violates the rules ignorantly. Boners provide “evidence of some failing in the intellectual grasp and achievement required within official or otherwise cultivated circles,” Goffman writes. When a government official mispronounces a word outside his subject area, that’s a boner. In 2002, then secretary of Health and Human Services, Tommy Thompson—not a health specialist—said, “We haven’t had any attacks as of anybody receiving West Nile virus or encephalopoulus.” A diplomat who addresses a prime minister as a president makes a boner, too.
A gaffe is a “breach in manners” that is “unintended and unknowing.” In some circles, wearing white shoes after Labor Day is a gaffe; in others, a nongang member who wears gang colors as he walks down the wrong city street is making a gaffe that’s potentially dangerous. Nonnative speakers of a language like Spanish or French commit gaffes when they presume an intimacy that doesn’t exist by using a familiar pronoun instead of a formal one such as, in Spanish, tú instead of usted.
The “knows better” errors represent a momentary, involuntary loss of control. The broadcast announcer who said, “We have a new Miss America, a lovely Georgia piece” committed a “knows better” error. The telecommunications expert who said “phoine” instead of “phone” made one, too, as does the baker who oversalts or the surgeon who undersews. Verbal blunders are the “knows better” errors, in the sense that they’re accidents, inadvertent and unintentional, and outside of our control. To Goffman, slips and disfluencies are both “hitches in the smooth flow of syntactically connected words.” That is, someone expects you to get something in order in time. When you don’t, that’s a blunder. One mark of the “knows better” fault is that the speaker knows, often immediately, when she’s slipped. As a result, she corrects herself.
The naïveté or ignorance represented by the “doesn’t know better” errors is easy to explain: someone didn’t know the norm. On the other hand, the person who makes the “knows better” error does know the norm, yet couldn’t prove or perform it. How many people have realized their shoes are on but their socks are still in their hands, or that they’ve put the child outside and poured a bowl of cereal for the dog? Competence doesn’t always produce perfect performances—just ask Olympic athletes, who are highly practiced and coached athletes. Rather, you judge competence from other measures, such as repeated success. (On most mornings the dog goes outside and the child gets cereal.)
This doesn’t answer the question of why perfectly competent, skilled, trained, intelligent people who can speak English, walk, open doors, play piano, or fly airplanes suddenly screw up. I’m fascinated by “knows better” errors of all types, because they’re microinsurrections of our biochemical vassals. Against our control—and the pride of our control—our neurons rise up, invincible. We often treat our bodies and our selves as if they were nations under the control of a central government. From the perspective of the distant capital, the regions seem to be full of rebellion, all of it immune to surveillance, curfews, and punishment. But if you traveled to those areas, the tumult would seem less violent. You’d see that the uprisings don’t threaten the entire nation but are more a response to local conditions. How should the capital respond? Not by becoming less tolerant of uprisings but putting into perspective their actual threat.
Curiously, “knows better” faults are often treated as “doesn’t know better” ones. In the right circumstances, an honest slip of the tongue can appear to be (or made to be) a moral failing. This often happens in public life—indeed, it’s an occupational hazard of living in the public eye. In 1865, before leaving to celebrate the retaking of Fort Sumter by the Union, the preacher and popular orator Henry Ward Beecher assured his listeners that the Union flag would fly over the fort “on Good Friday, on which occurred the resurrection of the Saviour” the newspaper, The World, promptly criticized Beecher for the error.*23 Conversely, a person who makes a gaffe often tries to mitigate his or her stupidity by insisting after uttering the blunder that it was an inadvertent loss of control, an accident that anyone could have made. Tommy Thompson might claim that he of course knows the word “encephalitis” it simply exited his mouth wrong. Claiming a Freudian slip can be convenient, too, because it blurs the line between the “knows better” and “doesn’t know better” errors. If the Freudian slip represents the intention of the unconscious self, it takes the conscious self, and what it could or could not know, off the hook.
What further complicates the way we manage our speech and live with verbal blunders is that we don’t speak in a vacuum. Everything we say is part of a broader social context. Talking happens in a range of places, from the pulpit to the cereal aisle of the supermarket, and occurs within many contexts, from pillow talk to boardroom talk to baby talk. Different moods and settings shape what’s said and how it’s heard, determining the scope of our attention to verbal blunders as well as our tolerance for them. Some verbal blunders don’t invariably register as errors: depending on who’s talking, we might be willing to overlook them. The reverse also holds: many people correct their own speaking even when there’s nothing audibly wrong. In Japanese culture, it’s considered deferential to restart your sentences even if you know what you want to say.
Goffman noted that verbal blunders are as much sociological events as they are linguistic ones. Joseph Williams, an English professor at the University of Chicago, once illustrated this phenomenon for his fellow English teachers, who were always looking for the most productive (and efficient) way to respond to the errors that students make when they write. For his part, Williams wondered why some errors “excite this seeming fury while others, not obviously different in kind, seem to excite only moderate disapproval.” Why, for instance, does mistaking “which” for “that” generate so much more attention from teachers, editors, and grammar mavens than “that” for “who”? Why does writing “should of,” not “should have,” get treated as a massive misfiring of grammatical pistons, when the error is simply a problem with spelling? In his essay “The Phenomenology of Error,” Williams argued that errors don’t exist only in texts, in writer’s heads, or in the grammar books. They mainly exist in readers’ experiences—hence the term “phenomenology,” in reference to the subjectivity of error perception. Some readers “resonate” with language errors more than others.
He sweetened his point playfully. At the end of his essay, he revealed that he’d embedded about one hundred grammatical errors in the text. Many of them were the types of mistakes that student writers often make: split infinitives, misplaced a
postrophes, dangling modifiers. As an experiment in expectations, he challenged readers to recall which errors they’d noted while reading without going back to search for the rest, and mail in the list to the editor of the journal. At one level, this was ivory tower fun and games. At another level, Williams had a serious point to make: the professor reading a student essay is cued to find errors. If she were reading the text in a professional journal, she would be less primed to notice errors.
Parents will recognize the parallax involved: when your kid trips, you’re concerned. When the obnoxious neighbor kid trips, it’s a hilarious pratfall. Goffman made the same observation: all grammatical faults may be something you hold a writer or speaker accountable for—a faultable. However, not all faultables are actual faults. That is, not everything that a teacher, editor, or transcriber directs a writer or speaker to change is, in fact, incorrect.*24 Moreover, quite often there are faults that never attract attention in the first place. Speakers themselves may want to fix something—which Goffman calls a “repairable”—that isn’t necessarily technically wrong.
For a variety of reasons, we don’t notice deviant speaking. But the converse can also be true: competent speaking isn’t always beyond criticism. To put it another way, sometimes it’s impossible to say something that won’t be perceived as a problem. Take, for instance, the name of one of my favorite bands, Sigur Rós, which is from Iceland. I was first told that the bands’ name was pronounced “Sigger Ross,” which I said for years until someone gently pointed out that the Icelandic pronunciation sounds more like “Siyer Rouss” (rhymes with “gross”). No matter which version I choose to say, I always run the risk of being corrected. With “Siyer Rouss,” someone can dislike the pretension of the “proper” Icelandic pronunciation. With “Sigger Ross,” someone else might hate the American banalization. Or take the example of the young man asking the young woman to marry him. Should he ask smoothly and risk sounding rehearsed? Or should he stumble and blurt his question like the romantic fool that he really is, honest and passionate? As Goffman pointed out, a person can aim for the ideal delivery, but in some moments the best delivery is a blathering mess. As for me, I chose to propose as a blathering mess, and that’s made all the difference.
4
What We, uh, Talk About When We Talk About “uh”
Once people become aware of them, speech disfluencies come to represent weeds in the garden. Nails that hang us up. Bumps in the road. I like them because they’re signs of the wild—like viruses and sexual attraction, they’ll always slip out of our grasp, evading our thickest armor. But such wild things make a lot of people uncomfortable. Actors train to deliver lines clean of them. Teachers criticize students who utter them. At the Federal News Service, the transcribers complain about the powerful men and women who litter their speech with them. Rose Mary Woods, Richard Nixon’s secretary, cleaned all the “uhs” and “ums” out of the transcripts of his taped conversations, along with the obscenities that littered the sentences. Likewise, the Presidential Recordings Program at the University of Virginia omits the “verbal debris” from transcriptions of presidential recordings. That is, unless the transcriber decides that a pause filler or repair “conveys any information about what the speaker was trying to say,” as the introduction to the published edition of some of Lyndon Johnson’s recordings puts it. Along with oral historians, an array of modern institutions—schools, the media, the legal system, religious institutions—work to stamp out verbal blunders.
Verbal blunders can also be seen (or heard) as yet another form of human behavior—not misbehavior. This view isn’t incompatible with the purpose of producing clean transcripts, but people whose criticisms about deviant language have a moral charge don’t accept the neutral view. Consider the facts, however: as a result of research over the last fifty years, we now know that anxious talkers have no monopoly on disfluencies, which pile up in people’s unrehearsed speaking an average of two to twenty-six times every one hundred words. We know that longer utterances attract disfluencies. So do the beginnings of utterances, over their middles or ends. Men, young children, and older adults make them more often. So do people on the phone,*25 nonnative speakers of a language, and even people with their hands in their pockets. (Apparently gesturing reduces disfluency.) Speakers also say they feel more comfortable saying them around certain people: friends, family. In general, anyone who thinks or acts at the same time as they speak, especially under pressure, will blunder. Put a tape recorder in front of any of these folks, and you’ll trap disfluencies like lint in a clothes dryer.
Verbal blunders may be universal and profuse, but how people in a particular era conceive of speech, language, and communication also determines, in part, what faults are noticed, what faults acquire the status of “faultables,” and what speakers are expected to do about them. Reverend Spooner had his name applied to a rare verbal blunder at a time when engineers and scientists were beginning to wrestle with how much of their thoughts and actions humans could control. And it’s not a coincidence that Sigmund Freud’s and Rudolf Meringer’s study of speech errors tapped into an undercurrent of Vienna’s fascination with social performances. In the same way, the conception of disfluencies as an aspect of normal communication arose after World War II out of an attempt to quantify human interactions so that they could be dissected, engineered, optimized.
The first scientist to count adult disfluencies was a Yale psychologist named George Mahl, who, in the 1950s, became interested in studying fear and anxiety in psychiatric patients. It was during the Cold War, and there was a climate of suspicion about undercover Communists and who might have been brainwashed by the enemy. One method for getting at a person’s secret intentions or their cloaked meanings was a technique called “content analysis,” which looked at patterns in word use to uncover what a person truly intended. The technique became integral to Cold War Kremlin watching, propaganda reading, and spy work and remains a key part of interrogations and analyses of evidence by law enforcement. For instance, if a new leader suddenly showed up in a foreign country that your country didn’t have diplomatic relationships with, you might analyze the leader’s speeches for clues about his priorities, his moral character, or his decision-making abilities. Though Mahl received money from the U.S. government, his research was confined to making psychiatry sessions more effective, as was that of his colleagues at Yale, who analyzed tape recordings of interviews to find out how their patients actually felt.
“Getting these tape recordings of therapy interviews was an eye opener for me,” Mahl says. “We would have this material transcribed, and I noticed it being loaded with disfluencies.” Because content analysis required clean, unbroken sentences, the hesitations, repetitions, pause fillers, and other fragments of speaking were scrubbed out of the transcripts. Though his work was little noted at the time, Mahl recognized the potential value of the real, verbatim transcripts and he determined to study how people actually talked. Their dross became Mahl’s treasure.
When we met, George Mahl was an active, spry eighty-seven-year-old who had just finished a singing lesson when he picked me up at the New Haven train station. He lived in nearby North Haven, Connecticut, and intrepidly drove his station wagon several times a week to the Yale campus to work out at the gym and attend lectures. With his red suspenders, a white beard, and white hair, he looked like a cross between Burl Ives and Moses. During World War II, he worked as a psychologist in an army hospital where psychosomatic medicine, the notion that some physical ailments had psychological causes, was in fashion. Back at Yale after the war, Mahl wrote a dissertation about peptic ulcers caused by anxiety and in 1947 joined the Yale faculty and studied stomach-acid secretions in dogs and macaques. Some of his psychologist colleagues would record interviews with patients, have the interviews transcribed, and use the transcripts for analysis.
“I looked at the transcripts that we were getting, and I saw all this, quote, garbage,” Mahl said. The analysts cleaned the transcripts s
o that counting words was easier, but Mahl was fascinated by what he came to call “speech disturbances”—everything that was cleaned out. “I thought—that’s where the gem is,” Mahl told me. “To study fear and anxiety, I’m not going to start counting sentences where the patient talks about being frightened or being anxious. I’m going to study the speech process itself, because I think speech disruption is a much better indicator of anxiety.”
Content analysis didn’t sit well with Mahl—he didn’t like how it assumed a one-to-one relationship between a word and some intention. Simply because a patient says “fear” doesn’t mean he’s afraid, just as someone who doesn’t say “fear” isn’t afraid. To Mahl’s way of thinking, some language use doesn’t “mean” anything at all; in some situations any word could provoke the same response. If a piano were tumbling out of a window, a person could shout “Piano!” or “Look out!” or “Geronimo!” to get people’s attention. This was what he called the “instrumental” aspect of language. Content analysis overlooked this aspect in lieu of counting words and noting their literal meanings. To Mahl, this wasn’t sufficiently probing. Furthermore, in the case of anxiety—a negative emotion—a patient was unlikely to address it so directly that an analyst couldn’t mistake it for something else. So rather than relying on a therapist’s perceptions or the patient’s own (often unreliable) self-reports, Mahl turned to speech disturbances as a reliable, unobtrusive, and objective way to check and measure the moment-to-moment fluctuations of a patient’s emotional state.
For instance, when a Russian immigrant told the therapist about a pogrom he’d witnessed as a child:
Um-- Slips, Stumbles, and Verbal Blunders, and What They Mean Page 7