Word by Word

by Kory Stamper

  This 1893 citation is excellent, but “it suggests material prosperity without the crucial element of egalitarianism,” Joanne explains. It’s not enough that the citation almost covers the current meaning: it must unequivocally do so. Could you read egalitarianism into that citation? Maybe, but it would be reading something into the citation, not basing your judgment on what’s actually there.

  The citation Joanne found as the first written use of this particular sense of “American dream” ironically tolls the bell for the American dream itself:

  Every republic runs its greatest risk not so much from discontented soldiers as from discontented multi-millionaires. They are very rarely, if ever, content with a position of equality, and the larger the population which is said to be equal with them, the less content they are. Their natural desire is to be a class apart, and if they cannot have titles at home, they wish to be received as equals by titled people abroad. That is exactly our present position, and would be the end of the American dream. All past republics have been overthrown by rich men, or nobles, and we have plenty of Sons of the Revolution ready for the job, and plenty of successful soldiers deriding the Constitution, unrebuked by the Executive or by public opinion.

  This citation sounds as if it were written last year. It’s from 1900.

  This is one of the joys of dating: everything is older than you think it is. The linguist Arnold Zwicky has coined the term “the recency illusion” to refer to the misbegotten assumption that anything that strikes you as new in language is a recent innovation, when, in fact, it’s not. Those dates in the dictionary prove that many modern linguistic bête noires go back quite a ways. Clipped words like “wevs” for “whatever” and “obvi” for “obviously” often show up in the mouths (and in the texts) of youth and are therefore derided as lazy. The articles bewailing the death of English at the hands of young people with cellphones are too numerous to cite, but few authors of these jeremiads against technology realize that clipping slang predates the cellphone by centuries: John Gower used “hap,” a clipping of “happening,” in his Confessio Amantis (“A wonder hap which me befell”), and that was written in the late fourteenth century. Ah, but those horrid initialisms, naysayers cry—“LOL” and “OMG”—surely a mark of modern laziness, moral decline, and the end of Good English as we know it! Never mind that these naysayers use plenty of initialisms themselves—please RSVP ASAP and BYOB. Or that “OMG” goes back to 1917, when it was first used in a letter to Winston Churchill. What now? Shall we blame the decline of English on typewriters?

  People rarely think of English as a cumulative thing: they might be aware of new coinages that they don’t like, but they view those as recent incursions into the fixed territory they think of as “English,” which was, is, and shall be evermore. The dates put the lie to that assumption. “The point at which [a word] entered the language is something that never occurs to people unless they see that date, and it just gets them to thinking a little bit about the history of the word,” says Joanne. She’s right: you realize that much of our vocabulary sits in a current of English that you don’t see on the surface of your patch of river. The rise of Indian cuisine in America over the last thirty years means that more of us are familiar with the word “korma,” but it first appeared in English in 1832. “Child support” sounds like a coinage born of the late-twentieth-century rise in divorce; in fact, the word dates back to 1901.

  —

  People also want to have some deeper connection with the word, to take responsibility for it, or to know some secret part of it. It’s not enough to know how to spell a word or what a word means; you have to know it.

  The most heartfelt people who write in are the ones who have come across an entry in our dictionary and write to tell us that they have a surefire antedating for us: they coined the word. This is almost always untrue, but the phenomenon is endlessly interesting. They always come with a personal story attached, with brilliant clarity of detail: I coined the term “wuss” in my dorm room at Princeton University in 1969, long before the date you give; you say that “noogie” appeared in 1968, but I grew up with kids getting and giving noogies in grade school, and I was already in graduate school by the time you say the word was created; I was born in Staten Island in 1926, and by 1932 I was ordering ice-cream cones with jimmies, and by 1942 I was adding jimmies to ice-cream cones and sundaes as a soda jerk, which proves that your date of 1947 for the word “jimmies” is wrong. People are unswayed, even when we turn up hard evidence of the word in print that antedates their own date: “When I come back and give them earlier evidence, they continue, some of them, to assert that, no, no,” says Joanne. They are claiming their stake in this language: the word “cyberstalk” or “vlog” or “ginormous” belongs to them first and foremost, regardless of the evidence to the contrary.

  Though hunting down dates for words can sometimes feel like gearing up for a lexicographical pissing contest—can we beat the OED?—there is something inspiring and educational about tracing a word back to a particular point in time. You learn things about little lexical whims that ended up sticking: the fad for contractions in the seventeenth century that never went away, the mania for Latin hypercorrection through the nineteenth century. You realize that it’s not just coincidence that we ended up with a bunch of food words from Yiddish and Chinese and Polish in the mid- to late nineteenth century: they’re there as a direct result of the waves of immigrants who came to English-speaking countries in the nineteenth century. Chasing dates is a journey through history; sometimes you get to your destination quickly, and other times you get a long scenic drive past Winston Churchill’s private correspondence.

  * * *

  *1 Even though we explain what those dates represent in the front matter of the dictionary, which we evidently write and proofread entirely for our own amusement.

  *2 “Shoe-rose” is used in Pride and Prejudice (1813), but a quick search turns up evidence of it in print back to 1801; “shaving glass” appears in Persuasion (1817, give or take), but I can trace it back to 1751. The OED entries for “shoe-rose” and “shaving glass” haven’t been updated since 1914, but I’m sure that when the editors at Oxford get to S, they’ll find my quick-and-dirty antedatings—and then some.

  Nuclear

  On Pronunciation

  It was Emily Vezina’s first week of work, and she was still getting used to the atmosphere of the editorial floor. She had gleaned one office rule very quickly: no talking. So imagine her surprise when her mandatory reading of the Third’s front matter was interrupted by a voice, speaking at a normal volume and in a calm, measured tone: “Pedophile.”

  She shook it off—ooookay—and tried to return to reading front matter when she heard another voice chime in, equally dispassionate: “Pedophile.”

  Now it was getting creepy. The whole office was taking part, like some horrible incantation: person after person, down the row of cubicles. “Pedophile.” “Pedophile.” “Pedophile.”

  Suddenly at her arm was Josh Guenter, our pronunciation editor. He held a blue index card out to her, upon which was written the word “pedophile.” “How do you say this?” he asked, and she obliged. “Pedophile,” she said. He scratched something out on a pad of paper, nodded, and moved to the next cubicle, where another editor looked up, blinked, and announced “pedophile.”

  It was not some bizarre hazing ritual. He was polling the office to figure out how to order the pronunciations for the Collegiate Dictionary.

  —

  Anyone who visits the office can feel the busy quiet settle on them like a blanket. On a warm afternoon, it can be very hard to stay awake in the cosseted hum of computers and people breathing. So it was very odd, about a year into my tenure, to be standing in one of the more remote hallways on the floor looking for a German dictionary and hear what sounded unequivocally like conversation coming from a nearby office. Then I heard a laugh track. A pause, and then the interlude repeated: same murmured conversation, same laugh trac
k. A light went on: I was standing outside the pronunciation editor’s office. He was taking citations.

  All lexicographers are collectors and collators of information, but the pronunciation editor is the only lexicographer who doesn’t pay any attention to the Reading and Marking list or worry about finding space in their office for all the source materials they’re sifting through. They collect pronunciations from instances of speech. That means, in short, that the pronunciation editor gets to spend all day watching TV. (Josh hesitates. “I spend more time on YouTube,” he says.) There is a small television in his office, along with a radio, a tape recorder, and the requisite books spilling from shelves.

  It may shock you to hear that the pronunciation editor doesn’t just make up the pronunciations you find in your dictionary. “I’m flattered that you think I have a vocabulary of 100,000 words,” says Josh, “but this is not me.” The pronunciations found in dictionaries end up there through a process of collection, arranging, and analysis that is similar to the process used for definitions.

  Taking citations for pronunciations (or prons, as the lingo goes) is a different critter, however. Context doesn’t matter at all for pronunciations; what does matter is a clear articulation of the word in question. Pron citations come from three main sources: broadcast media (which includes radio, TV, movies, and cable), audio or video on the Internet (YouTube and podcasts being the big generators), and human contact (phone calls, the aforementioned office polls, and face-to-face conversations). Steve Kleinedler, who handles the pronunciations for The American Heritage Dictionary, has occasionally sent me links to promotional videos pharmaceutical companies have put together for various products. Both Josh and Steve regularly call up companies, town halls, and famous people (mostly Nobel Prize winners and presidents) to ask them how they say the name of their product, their hometown, or themselves. If there’s more than one pronunciation and Josh isn’t sure which one to list first, he might go around with a clipboard and an index card for an office poll to see if he can get a better sense of which pronunciation is more common after making us all say something in our raspy, unused voices. When watching or listening to a source, Josh will transcribe what that source says onto a white citation slip, give the bibliographic evidence, and then file it away.

  How Josh writes it down is the real trick. English is not a phonetic language—that is, there’s not a one-to-one correspondence between a letter and a sound that it represents. So when it appears at the beginning of a word, the g in English has two sounds—the sound at the beginning of “girl” and the sound at the beginning of “giraffe”; c has three-ish sounds, depending on where it is in the word and what’s around it—the sound at the beginning of “cat,” the sound at the beginning of “citrus,” and the sound in the middle of “politician”; and so on. That means that when writing down a pronunciation, you can’t rely on the twenty-six letters in our alphabet to give an accurate representation of what was said. “SPRAHCH-geh-FYOOL”—huh? Is that CH as in “chat” or as in the first raspy sound in “Chanukah”? Is that AH as in “ax,” or the first a of “again”? All that uncertainty means you might end up pronouncing “sprachgefühl” like “SPRATCH-gay-FULL” or “SPRUTCH-geh-FOOL,” and you would be very wrong.*1

  That’s why dictionaries use a proprietary alphabet*2 full of odd letters like ā and ǝ and ƞ. They help us accurately portray the pronunciation of “sprachgefühl” as ˈshpräḵ-gə-ˌfᵫl,*3 even if they do drive people crazy because you need a key to decipher them. These alphabets are phonemic, not phonetic. The letters in a phonetic system represent one sound per letter; the letters in a phonemic system represent a group of sounds per letter, because an individual phoneme (the smallest unit of sound in an utterance, and the thing that our pronunciation alphabets represent) can vary depending on your accent and dialect. Take the symbol that we use to represent the i in “pin”: i. In a phonemic system, we tell you that i is pronounced like the i in “pin” regardless of how that vowel actually sounds when it leaves your face. If, in your native dialect of English, you say “pin” or “pen” or “payin” or “pehhn” when you read the word “pin,” that symbol represents all the sounds represented by that vowel. This means that we can cover a variety of accents and dialects and not have to privilege one over the other. This means that if you say PIN for “pin” and “pen” (a phonological feature in some parts of the country called pin-pen merger), but I say PIN for “pin” and PEN for “pen,” the dictionary doesn’t, through its pronunciations, say that one of us is right and the other wrong. Those weird alphabets allow for different phonological quirks.*4

  Pronunciation editors need training not just in phonemic and phonetic alphabets but in listening. They need to know about the phonological characteristics of most of the dialects in at least the country they’re writing in, if not a good sketch of the phonological traits of English around the world. Josh asked where I was from at one point, and when I told him, he said, “Oh, General Western. So you have ‘cot-caught’ merger, and ‘Mary-merry-marry’ merger?” I evidently did, though I didn’t know what those were until he explained that some dialects use different vowels for “cot” and “caught” and for each of the “Marys.” I do not; when I talk phonology with people, I talk of the “cot-cot” and “Mary-Mary-Mary” mergers.

  Because pron editors can hear these fine distinctions, they can also articulate these distinctions, so they’re also often called upon to be the ones to record the audio pronunciations in the electronic dictionaries. When Josh was in grad school, before he worked for Merriam-Webster, he had been hired by The American Heritage Dictionary to record some pronunciations for its website; now he records the new pronunciations for ours. It’s a relatively manageable job now, only a few thousand new words to research and say at most, but back in the mid-1990s we had the entire Collegiate Dictionary’s prons to record for the new website—close to 150,000 pronunciations. It was too many for one person. Four actors—two men and two women—were hired to record the vast majority of the pronunciations. Only two groups of words were held back and done in-house by our pron editor: words that had pronunciations with sounds that were difficult for native English speakers to make (like “sprachgefühl”) and profanity. Peter Sokolowski remembers being in the hallway outside the pronunciation editor’s office one day and hearing from within the office a very measured voice say, as blandly as possible, “Motherfucker. Motherfucker. Motherfucker.” It was one of our old pronunciation editors, trying to get the intonation right for the audio file. He left a few years later to become a priest.

  —

  Audio prons are one of the great advantages of an online dictionary. You no longer need to figure out that ridiculous key; now you just click on the little speaker icon, and you can hear the word pronounced for you. Pronunciations in dictionaries have not always been so accessible.

  Many early dictionaries, written as they were for those learning English as a foreign language, provide some basic guidance in pronunciation, though it’s not where we moderns expect it to be. Often, the pronunciation help was given in a long section of prose at the beginning of the dictionary, sometimes disguised as grammar or orthography or prosody. Mulcaster’s 1582 Elementarie (“Vvhich Entreateth Chefelie of the Right Writing of Our English Tung”) is a good early example. It gives lengthy explanations of the writing and sound of most of the letters in the alphabet before getting to his dictionary, though he is not exhaustive: he goes on about the sounds of e for several pages, but has two sentences devoted to the sound of a (“A Besides this generall note for the time and tune, hath no particular thing worth the obseruation in this place, as a letter, but it hath afterward in proportion, as a syllab. All the other vowells haue manie pretie notes”).

  Johnson handled pronunciation in the grammar that he appended to his 1755 Dictionary, and he set up his aim with a short excerpt from one Dr. Wilson, written in 1553. “Pronunciation,” it begins, “is an apte orderinge bothe of the voyce, countenaunce,
and all the whole bodye, accordynge to the worthines of such woordes and mater as by speache are declared. The vse hereof is such for any one that liketh to haue prayse for tellynge his tale in open assemblie, that hauing a good tongue, and a compelye countenaunce, he shal be thought to passe all other that haue the like vtteraunce: thoughe they hauve much better learning.” In short, good words require good pronunciation, and good pronunciation will make you look smarter than everyone else in the room. Johnson claims that English has two pronunciation conventions: one that’s cursory and colloquial, and one that’s regular and solemn. Given Johnson’s irascibility and use of the word “vitiate”*5 throughout the preface to his dictionary, you can guess which convention he feels has won out: the writers of English grammars have “established the jargon of the lowest of people as the model of speech.”

  Dictionaries, as Emily Brewster says, are aspirational texts, but the way we talk has always been a problem for the person aspiring to just the right social level. One must steer a steady path through a narrow strait: vulgarity, low, and classless speech on one side; prissy, overreaching affectation on the other. Is it low-class to pronounce “cadre” as ˈka-ˌdrā (KA-dray), or is it preening and pretentious to pronounce it ˈkä-dər (KAH-dur)? Which one will transform me from a lumpen buffoon into a lithe and elegant viscountess? Help me, O Dictionaries!

  But by modern standards, Johnson does diddly-squat to help those who want to help themselves. Johnson’s discussion of pronunciation is nestled tightly into a treatise on orthography, not handled as a discipline all its own. In the grammar, he occasionally mentions the pronunciation of certain letters and how they have historically sounded, but he notes that “by writing in English I suppose my reader already acquainted with the English language, and consequently able to pronounce the letters, of which I teach the pronunciation; and because of sounds in general it may be observed, that words are unable to describe them.” At individual entries in his Dictionary, he includes marks to tell the reader which syllable is accented, but there is no attempt to render any sort of transcription of what Johnson thinks a word’s pronunciation should be.