Um-- Slips, Stumbles, and Verbal Blunders, and What They Mean

by Erard, Michael.

  Here is a related experiment that you can do yourself that shows how Baars’s method worked. It also underlies another theory about the cause of slips. Find a person and ask her to repeat the word “poke” a few times. (Don’t tell her what you’re doing.) Seven is a good number of repetitions. Then ask what she calls the white of an egg. Almost inevitably the person will reply “yolk.” (Likewise, she’ll slap her forehead when she realizes what she’s said.) What happens here is that you’ve “primed” her perceptual system in a particular way, overstimulating the neural connections that map onto the “oke” sound. Her brain reached for the overstimulated version because it was the most immediately available one.

  This priming is fundamental to a theory about what causes slips of the tongue, called “spreading activation theory.” It was developed by Gary Dell, a cognitive scientist at the University of Illinois Urbana-Champaign. In his model, associations between words, sounds, and meanings are triggered electrically in the brain when the units are selected to be spoken. Because the units are arranged in a weblike fashion, these electrical activations sometimes overlap, stimulating the wrong word or sequence of sounds as much as the correct ones. This creates momentary confusion about which is the strongest activation. This occurs because activation “spreads.” It goes top down, from sentences to words to sounds, as well as bottom up, from sounds to words to sentences. The theory of spreading activation predicts the following errors, which Rudolf Meringer would have called “contextual” errors, because interfering sounds appear around the error:

  Sim swimmers sink

  Swum simmers sink

  Some simmers ink

  Some swummer sink

  Some swinkers sink

  Some sinkers swim

  Some simmers swim

  Some swimmers drown

  Dell’s model has more facets than I can present here, but it does explain why faster talkers produce more slips. It’s not because more words equal more opportunities to err. Rather, more talking keeps certain electrical activations from fading. This is the interference apt to spread and get in the way. Dell’s model also predicts, not surprisingly, that slower talkers make fewer errors. When they do err, their errors will be the future-looking anticipations. This model is powerful in other ways, too. It accounts for noncontextual errors, those where the interference is not visible in the utterance. If you were to ask someone to close the door, your brain will have registered (or “activated”) that the door is open. This activation might lead to the error, “Could you open, I mean, close the door.” The spreading activation model also accounts for a wide range of speech errors as well as their patterns: the bias for spoonerisms to produce real words, their tendency to involve the first sounds and syllables of words, their tendency to come from the same part of speech. This is hardly the last word on explaining slips of the tongue, and Dell is apparently now extending his model beyond normal speakers to people with language disorders.

  Ultimately, however, language is only one sort of behavior that needs to be performed in a sequence, though errors in language are unique because they are less expensive than errors that occur in other realms. For example, according to the Institute of Medicine, between 44,000 and 98,000 people a year die in the United States from medical errors. More than seven thousand of these die because someone gave them the wrong drug or dosage, or because two drugs had an adverse interaction. In the 1970s, NASA research showed that crews were responsible for two-thirds of air crashes. During the Gulf War, in which highly technologized battles were waged with long-range weapons, “friendly fire” caused 23 percent of American casualties: we were dropping our bombs on ourselves.*52 From warfare to nuclear power plants to robotic surgery, minimizing the cost of errors by redesigning systems to be redundant and flexible when errors occur has sparked research in the psychology of error.

  One basic idea is that “control” consists of both voluntary and involuntary elements. In the classic example by James Reason, an expert on the psychology of error, a person comes out of a subway turnstile and takes out a wallet to pay, though no money is required. While speech errors reveal the invisible seams of language, slips of action show units of action—how they’re stored in the brain and how the brain uses them. Taking out a wallet when payment is required forms one unit. That it’s done unconsciously is a sign that such a unit automatically exists in your head. These units can also be temporarily deleted, as in the example of the person who goes into the room to fetch a book, takes off her rings, then comes out of the room without the book. There are also exchanges of behavior, such as offering the cat a can of pudding instead of a can of cat food.

  Most human action can be broken down into sequences of deliberate action and automatic action. Lifting a foot, for instance, is more conscious than putting it down; walking through the turnstile may be more conscious than paying the fee. That we shift back and forth between conscious control and unconscious instinct or habit in many activities sheds light on the psychology of action. We’re more familiar with the intentional part of the cycle, because the habitual is taken for granted and off the map. As Reason notes, whenever conscious cognitive plans go awry, they’re usually taken over by automatic plans. Both slips of action or slips of the tongue result in something that is familiar and predictable and takes the shape of what we already know.

  What we call “intentional activity” is, itself, made up of actions that have been learned and can be repeated—they are not new behaviors that come from nowhere. It also means that what we see as an “error” can merely be the appearance of a competent behavior in the wrong context. This could provide insight into a range of everyday situations, from dialing the wrong number to traffic accidents to picking the wrong item off the supermarket shelves. It also explains why slips of the tongue are constrained by a language’s grammar and sound system. These make up a sort of safety net to which your cognitive system automatically reverts in times of trouble.

  Admittedly, this doesn’t leave much room for slips as creative acts—after all, no one would call walking out of the room without the keys an unintentional innovation in behavior. In the early days of the slip renaissance, slips of the tongue were celebrated, sometimes even fetishized—if you were a linguist or around linguists enough, you didn’t feel embarrassed about saying them, because you knew that your slip of the tongue, a singular, irreproducible instance, represented a transmission from the place where language dwelled and a connection to its creative forces. (You also knew that it would be written down and kept forever.) How disappointing it must have been for linguists to see slips of the tongue lumped in with errors of the mundane and dangerous sort.

  As a result, linguists tend not to discuss nonlanguage errors. Nevertheless, they are studying slips of the tongue in more sophisticated ways that are sensitive to factors that Meringer or Fromkin hadn’t considered. One such factor is the individual’s experiences with language. This is not Freud’s sense of experience—linguists still don’t believe that repressed feelings, traumatic experiences, or neuroses are the central or universal cause of slips. They do, however, recognize that a person takes idiosyncratic paths through life, paths that can shape how they blunder and who hears them do it.

  9

  Fun with Slips

  Arnold Zwicky, a linguist at Stanford University, has spent the last thirty years researching spelling errors, double entendres, spoonerisms, “schoolboy howlers of the malaprop variety,” mixed metaphors (such as “the Internal Revenue Service appears to be totally impaled in the quicksands of absolute inertia”), and a whole menagerie of linguistic deviations and innovations. Zwicky is such a slip geek that he has expanded the accepted scientific labels far beyond the primitive categories of “knows better” and “doesn’t know better,” the intentional and the voluntary, the true and the false. Others use mistakes as props in their intellectual vaudeville acts. Zwicky turns them into performances of connoisseurship more akin to wine tasting than linguistics: he appears to be able to sn
iff the bouquet of a verbal blunder and tell you whether it was spoken in the morning or the evening. To him there’s as much difference between the “Blue Bonnet plague” (or bubonic plague) and “the pastor cut the shermon sort” as between a merlot and a chardonnay.

  Zwicky is a genial, white-haired man in his midsixties who, on the day I meet him in his office, is wearing a T-shirt with a pocket out of which peeks the small, inevitable notebook of the inveterate slip collector. Somewhere in his office is a box of similar notebooks, the precious reward of thirty years of listening like, as he calls it, “a Martian.” (By which he means, you assume nothing about what you hear.) People who work with him know that if Zwicky grabs for his notebook midconversation, you’ve said something interesting.

  His fascination with slips of the tongue began in the early 1970s, when he sat in on a course taught by Vicki Fromkin. (He appears seven times in her slip collection, six times making a slip and once when someone observes that “Zwicky has gotten skwinny.”) It occurred to him that slips might be useful for teaching linguistics, to illustrate the abstruse technical concepts of linguistics (imagine trying to enliven phonological features without “glear plue sky” or illustrate the abstract past tense without “Rosa always date shranks”)—and to invest serious analysis with a bit of fun. Fromkin taught him how to take dual notes at lectures: on one pad you write the slips; on the other you write notes about the lecture. Zwicky doesn’t have a reputation for making many speech errors, but after listening to him give a talk once, Fromkin told him she’d recorded him making one slip a minute. (The average may be closer to one every seven minutes.) “The other people in those audiences heard maybe one an hour,” Zwicky told me. “A damn good thing, too.”

  In 1980, he published Mistakes, a slim but exhaustive workbook (intended for students) that listed and discussed the gamut of things called “errors.” On one end were the true accidents, like “a bright flire,” “loosely flew” or the “tropic of cancercorn” at the other, the inventions and creations of writers like Lewis Carroll (“gyre and gimble”) and James Joyce (“museyroom,” “bisexcycle”); in between was the jargon of subcultures and the deliberate misspelling of words from dialects (“I kin git yew”). Such distinctions can seem unnecessarily detailed. After all, what do intentions matter if the end product is annoying, offensive, or deviant? For Zwicky, it’s to account for what our brains contain: words from speaking and reading; social conventions; formal education; unpleasant or happy experiences speaking or writing; and personal preference and talent. Too often, he thinks, people fault those who make errors for not having enough discipline, rules, or knowledge. But suppose that we don’t assume deficits; instead, assume that someone knows rules and words but applies the inappropriate ones. The question becomes, what is the speaker doing in that instance? And why? How much does linguistic performance depend upon intelligence and memory—and on community and experience, too? After all, language takes brains communing, not isolated brains.

  Zwicky has also written about the language of menus, the language of police blotters, and the names of gay male porn stars, the ordinary material of everyday language that is important to him as a working-class kid who grew up in eastern Pennsylvania, then went to Princeton and MIT. His intention is “to celebrate the speech and behavior of ordinary people, including an assortment of otherwise marginal groups, such as the rural working class,” he says.

  The slips that vary among individuals perhaps the most are word errors—or, as word slips have come to be called, “malapropisms.” The word “malaprop” was inspired by the wildly word-misusing character, Mrs. Malaprop, in Richard Brinsley Sheridan’s 1775 romantic comedy, The Rivals.*53 Sheridan crowned his character with the French term, mal à propos, or “inappropriate.” And the play’s comedy is not just that Mrs. Malaprop uses words incorrectly as she blocks a love affair between the hero and heroine; it’s that she claims to be a verbal expert. Her name is the play’s signal that her verbal pretentiousness is going to be skewered.

  If you go to a university library for a copy of The Rivals, you’re likely to find that generations of students have already underlined some of her butcheries, which takes away your pleasure in finding them yourself. “Oh!” Mrs. Malaprop exclaims at one point. “It gives me the hydrostatics to such a degree. I thought she had persisted from corresponding with him.” At one point she intercepts—and has read aloud—a letter in which Captain Jack Absolute, the play’s hero, mocks her use of language. As he does not have the word “malapropism” to use, he charges her instead with “deck[ing] her dull chat with hard words which she don’t understand.”

  Understandably, Mrs. Malaprop is incensed—and sent into a fit of mal à propos speech: “There, sir! an attack upon my language! What do you think of that?—an aspersion upon my parts of speech! Was ever such a brute! Sure, if I reprehend anything in this world, it is the use of my oracular tongue, and a nice derangement of epitaphs!”†54

  A couple of years ago, I saw a production of The Rivals in New York City, where the role of Mrs. Malaprop was played by Dana Ivey, an accomplished actress who specializes in playing power matrons on stage and screen. Just as college student underlining had removed some of the surprise of the malapropisms, so did the live performance. Each time Ivey malapropped, the audience laughed, and each time I recognized something odd about the laughter but couldn’t identify it. Then I noticed that the laughter followed on Mrs. Malaprop’s line far too soon for the audience to have known that the word was ill-chosen. Imagine telling a joke and having your listener begin laughing before the punch line’s out of your mouth, as if they knew already where the joke would be. Listening more closely, I realized that Ivey was slightly overenunciating each malapropism—effectively signaling to the audience that a humorous moment was coming. The actress’s portrayal of the character was appropriately dragonlike, yet vulnerable. But the difficulty of deliberately saying an incorrect word adds another layer to an understanding of verbal blundering.

  Sheridan’s play is an artifact of a historical period when people thought differently about verbal blunders than they do now. In Sheridan’s day, Mrs. Malaprop was more than simply a figure of linguistic derision; she also embodied anxieties about manners, language, and social class. Such anxieties materialized on American and British stages in characters like Jemima Blockhead, in a play titled The Blockheads, or Mrs. Sententious, who appeared in The Better Sort.*55 All three of these characters flaunted their high social status yet couldn’t fulfill its linguistic expectations. Not that they noticed, however—if performing at a high level outstripped their abilities, they were incapable of monitoring their own performances. In his history of American eloquence, historian Kenneth Cmiel writes that Blockhead, Malaprop, and Sententious were humorous (not tragic) characters because they thought they expressed themselves as others in their community did, but could not. As Cmiel puts it, “They did not strike the pose of restraint necessary in a republican world.” By comparison, Spooner and Freud questioned the relevance of restraint in an increasingly complex world.

  As the study of verbal blunders became more sophisticated over the twentieth century, researchers wanted to know how intention and intelligence figured into a malapropism, whether it was a one-time occurrence or a repeated performance. Patterns in malapropisms also held the power to tell about language’s hidden structures, as two psychology graduate students at the University of Texas, Anne Cutler and David Fay, found in the early 1970s. Fay, who had been inspired by Franklin to begin collecting slips, discovered Meringer and Mayer’s book, Misspeaking and Misreading. Fay couldn’t read German, so Cutler, who could, translated for him. She not only became a leading expert on Meringer, she and Fay began to do what other scientists were doing: collecting slips.

  They listened for instances where someone said “equivocal” when they meant to say “equivalent,” or “tambourines” when they meant to say “trampolines.” Fay (now at Verizon Laboratories) and Cutler (now director of the Max Planck I
nstitute for Psycholinguistics) found that 99 percent of the time, the words a speaker meant to say (the target) belonged to the same grammatical category as the inadvertent mistake. If you meant to say a noun, the mistake was almost always a noun, and likewise for other parts of speech. Lawn, line. Week, work. Accurate, adequate. Map, make. Gaudy, gory. Among other things, this means that a sentence in which a malapropism occurs doesn’t suddenly become ungrammatical. (Nonsensical, certainly: “That’s a gory piece of jewelry.”) Fay and Cutler also found that 87 percent of target-mistake pairs had the same number of syllables and 98 percent shared a stress pattern, so that the stress would fall on the second syllable of both words, for example.

  For the sake of simplicity, Fay and Cutler ignored the pairs that were related only in meaning. For instance, you might say “black” when you mean “white,” or “write” when you mean “read.” Good, bad. Nearly, barely. Specific, general. Saying “she got hot under the belt” when you meant “she got hot under the collar.” These types of errors are fairly obvious when they occur, so one might be tempted to think they are the more frequent ones. In fact, just as there is a bias for sound exchanges to produce actual words, there is a tendency for malapropisms to be related in meaning to the target word.*56 Vicki Fromkin wondered if there was a semantic reason for saying “art of the flute” instead of “art of the fugue” and not, say, “art of the flue.” (“Flute” and “fugue” coming from the world of music, and “flue” from the world of chimneys.)