by Kory Stamper
But the fact remained: How were pupils of English going to learn how to say the words correctly? Pronouncing dictionaries were created to step into the breach. The first dictionary devoted primarily to pronunciation was James Buchanan’s 1757 Linguae Britannicae Vera Pronunciatio, which he wrote specifically for educating youth. His dictionary is the first to include diacritical marks to help distinguish between long vowels and short ones, as well as accent marks to tell the reader which syllable was stressed.
Half a dozen more pronunciation dictionaries were published in the last half of the eighteenth century. Most were both prescriptive and aspirational; some intended to fix the defects of speech that had crept even into the upper echelons of society. So what was their model of elegance and propriety? Each claimed to represent the speech of the London gentry, and yet they offer vastly different pronunciations for the same word. One book’s pronunciation of “fear” gives the digraph “ea” a long-a sound, ā, like in the English “day”; another’s gives it a long-e sound, ē, like “meet” and “deceit”; another lexicographer notes it has the sound in “beer” and “field”—though “beer” and “field” use different vowel sounds (the short “ih” i and the long-e ē, respectively). So whose standard is the standard?
There’s a technical problem with these prescriptive pronouncing dictionaries. Each of their systems for rendering pronunciation relies on the reader’s somehow intuiting what the vowel quality of their reference word is supposed to sound like, as opposed to what it actually sounds like in their dialect. To a Londoner, that ā in “fear” sounds pretty Scottish; to a Scotsman, that i or ē sounds pretty southern. This really isn’t that surprising. “It’s not that there’s written language and spoken language,” says Josh. “There’s language, and there’s writing. You start with the phonetics, and you design a writing system based on the phonetics and phonology of the language. English just has a poorly matched system, but that’s anomalous.”
There have been attempts to remedy this. Buchanan, Benjamin Franklin, and Noah Webster all proposed, with varying degrees of success, alternate spellings and alternate spelling systems that would bring the orthography of English more in line with its pronunciation. Only Webster succeeded, and his success was very limited: while Americans do use some of his spelling reforms, like “plow” for “plough” and “honor” in place of “honour,” his more extreme suggestions (“wimen” for “women” and “tung” for “tongue,” both of which show up in his 1828 dictionary) never caught on.*6 Josh thinks these reformations are bound to fail. “Regularizing spelling to match phonology—you could, in theory, do that. In theory. It would be a hell of an undertaking, and you couldn’t do it in practice. You’d have to get everyone to agree to the new system.” He searches for a good analogy. “It’s not herding cats. It’s herding 500 million cats.”
Mismatch between English’s pronunciation and its orthography is something that everyone, native speaker and learner alike, harps on. It feels like a bait and switch: after all, we learned as children that if words have the same cluster of letters at the end, they rhyme: hop on pop, cat in the hat. And then we encounter “through,” “though,” “rough,” “cough,” and “bough”—five words that all end with “-ough” and not only don’t rhyme but don’t even have similar pronunciations. But “won” and “done” and “shun” rhyme? Are you telling me Dr. Seuss lied to me about English?
The two biggest complaints we tend to get at Merriam-Webster about English pronunciations are about that desire for regularity. The first is always a request to change the way we say something based on how it’s spelled. This isn’t a new concept; Johnson posited in the preface to his Dictionary that “for pronunciation the best general rule is, to consider those of the most elegant speakers who deviate least from the written word.” But because English isn’t phonetic, that’s a ridiculous proposition. Josh gives an example: “What if we said, okay, from now on, we’re going to count ‘own, twoh, three, fowar, five, six, seeven.’ Those are the most common numbers, and there’s a big mismatch between spelling and pronunciation. You think we can just say, no, it’s ‘own,’ not ‘won’?”
The second complaint is that the pronunciation of a foreign borrowed term is too Englishy. The usual suspects are almost always French: “croissant,” “chaise longue.” This is one area of pedantry that merely serves to show off the peever’s educational bona fides—by recognizing that a word is a French borrowing and knowing what its French pronunciation is. Unfortunately, when foreign words are snatched into English, they are often given pronunciations based on English, not the origin language.
Josh offers three reasons why French gets singled out as the subject: English has a lot of French borrowings; French has a set of sounds that we don’t have in English, so any Anglicization will be off; and there is, for English speakers, a prestige association with French, which means we see French words as high-class. But he points out that most French words have been so Anglicized that we don’t even think of them as French anymore: “clairvoyant,” “bonbon,” “champagne.” Anglicization is the norm; in fact, it was prescribed in early twentieth-century usage guides like Fowler’s Dictionary of Modern English Usage.
But sometimes the desire to sound smart gets the better of us. Look at “lingerie.” When it first came into English, it was Anglicized to lan-zhə-(ˌ)rē or lan-zhrē, which was as close to the proper French pronunciation as we could get. Three things conspired against keeping lan-zhrē: the way we think French should sound, the way that English works, and our desire to be fancy.
French has five nasal vowels (that is, vowels whose quality changes when followed by an n), but for whatever reason English speakers really have only latched onto one of those vowels as sounding authentically French—the “an” vowel, which we render as ӓn. We throw it into a whole bunch of French words, even when it doesn’t belong, because it makes us look like we know French and are therefore smart: ˈän-və-ˌlōp is the textbook example, and so is län-zhrē. We also tend to think of French words as ending in ā—“café,” “résumé”—so we subbed the long ā in for the final ē, giving us län-zhrā.
Next, we had to assign stress to “lingerie.” French is an unstressed language, by which I mean that the language itself doesn’t use stressed syllables. English, however, is highly stressed; stress plays a part in distinguishing homographs and meaning, like “PRO-duce” and “pro-DUCE.” Where do we put the stress in a word that’s unstressed? Pretty much all over the place: English speakers have stressed the first and last syllables of “lingerie” in turns.
And then there is our desire to be fancy. The closest Francophone pronunciation that English provides, lan-zhrē, sounds…ordinary. “It sounds like an everyday word for something that is, after all, considered to be an exotic, fantasy thing,” says Josh. We cannot fathom that something so full of allure and mystery, something so French, could have a pronunciation that rhymes with “can tree.” It’s an exotic thing, so it must have an exotic pronunciation.
This general confusion about “lingerie” means that we collected so many pronunciation variants of it that we listed thirty-six of them in the Third and sixteen in the Collegiate.*7
—
We get particularly snippy when words have one pronunciation that seems to match the spelling of the word and one that doesn’t, especially when one of those pronunciations is considered nonstandard. Nonstandard pronunciations are like nonstandard words: they are used naturally but attract the wrong kind of attention for being “uneducated” or otherwise stigmatized. Many of these are, like nonstandard words, dialect pronunciations: “LIE-berry” for “library,” common in some dialects of American English, is one of the best known (and excoriated) examples. Nonstandard pronunciations, like nonstandard words, get entered into some dictionaries, like ours, when they’re remarkably common. But trying to figure out what’s a nonstandard pronunciation and what’s an acceptable variant is not easy. “Judging whethe
r something is nonstandard is not something you can do objectively from the data in a recording,” Josh says. “That’s just something you have to feel.”
Why bother entering the nonstandard pronunciations? Don’t we want people to say things correctly? That question presumes that there are tons of nonstandard pronunciations out there, when, in fact, there aren’t. “Take a look at the vast majority of the 100,000 words in the dictionary,” Josh says, “and you’ll find precious few of them.” That’s not to say there aren’t lots of variant pronunciations, but those are different from nonstandard pronunciations. You can say də-ˈle-mə or dī-ˈle-mə for “dilemma,” and both are considered standard and correct. “The great majority of words have no particular prescription to them for the pronunciations,” says Josh. “There is nothing to decide.”
That’s not to say that people don’t attempt to prescribe. An alternate pronunciation of “nuclear”—ˈnü-kyə-lər, the pronunciation that often gets spelled phonetically as “nucular”—has had a very hard run of it. It is, in modern times, universally reviled; usage commentators have called it a “spectacular blunder,” an “aberration,” and “beastly,” and our correspondence files are full of outrage over its very existence. In fact, we received so much e-mail about the pronunciation ˈnü-kyə-lər that we devoted an entire page of our online FAQ to the issue so we wouldn’t spend most of our time answering e-mail about it.
Much of the Sturm und Drang surrounding ˈnü-kyə-lər is oxymoronic: ˈnü-kyə-lər is at once lazy and uneducated and yet still spoken by some of the most prominent (and highly educated) people in our public lives. The former president George W. Bush is often blamed for somehow promulgating this pronunciation of “nuclear,” and the people who write to us all in a froth about it often accuse us of political pandering by including this pronunciation in our dictionaries.
And that, ultimately, is why people write to us: we have included that pronunciation in our Collegiate Dictionary. We give four possible pronunciations of “nuclear” in the Collegiate: ˈnü-klē-ər (NEW-klee-ur), ˈnyü-klē-ər (NYOO-klee-ur), ˈnü-kyə-lər (NOO-cue-lur), and ˈnyü-kyə-lər (NYOO-cue-lur). Those last two—the shudder-inducing -kyə-lər pronunciations—are preceded by an obelus (÷), which is our shorthand way of marking nonstandard but widely used pronunciations in our dictionaries. In the event that you don’t know what the obelus is for,*8 we have a short usage paragraph at “nuclear” that tells you right away that the pronunciations that end in -kyǝ-lǝr are disapproved of by many. Unfortunately, we don’t stop there, which is why we receive so much hate mail regarding “nuclear”:
Though disapproved of by many, pronunciations ending in -kyə-lər have been found in widespread use among educated speakers including scientists, lawyers, professors, congressmen, United States cabinet members, and at least two United States presidents and one vice president. While most common in the United States, these pronunciations have also been heard from British and Canadian speakers.
The paragraph is actually a little out of date as of this writing. The pronunciation ˈnü-kyə-lər has now been used by four-ish U.S. presidents—Dwight Eisenhower, Gerald Ford, Jimmy Carter (who also used the pronunciation ˈnü-kyir),*9 George W. Bush, and sometimes Bill Clinton and George H. W. Bush. It has also been used by a number of international leaders, U.S. congresspeople and cabinet members, governors, weapons specialists, military personnel, a Nobel Prize–winning theoretical physicist—people who are in a position to use the word almost daily and yet still say ˈnü-kyə-lər. And as the tip of the hat to Eisenhower shows, it’s also not a recent invention: we have evidence of this use back to at least the 1940s.
The linguistic process by which “nuclear” became ˈnü-kyə-lər is called “metathesis,” where two phonemes within a word switch positions.*10 This is the process that gave us the standard pronunciations of “iron” (“EYE-urn” instead of “EYE-run”) and “comfortable” (“KUMF-ter-bul” instead of “KUM-fert-uh-bul”) and other nonstandard pronunciations like “PURR-tee” for “pretty.” Some lexicographers and linguists posit that “nuclear” underwent metathesis because there are not any other common English words that end in -klē-ǝr (just “cochlear”),*11 but there are a good number of words like “molecular” and “vascular” that end in -kyǝ-lǝr, and the gravitational pull of these words and this more common speech pattern dragged “nuclear” into -kyǝ-lǝr orbit. But the “how” is not as interesting to people as the “why,” and we don’t have a compelling “why.” Why do people who have gone through some of the most rigorous graduate programs in the world use a pronunciation that’s just not right?
If it can be believed, linguists haven’t undertaken a study of it. We know it’s not a regionalism: though Carter, Bush, and Clinton all hail from the South, Eisenhower and Ford did not, nor did Walter Mondale, the vice president who said ˈnü-kyə-lər. Rose Kennedy corrected Ted Kennedy’s pronunciation of it in a letter—“I wish you would check your pronunciation of the word nuclear”—and Ted Kennedy was most certainly not southern. Evidence of ˈnü-kyə-lər appears nearly all over the English-speaking world: in Australia, Canada, England, California, Iowa, Utah, Indiana, Pennsylvania, Texas. It’s not just that we have recordings of people from all over saying it. We also have unironic use of the spelling “nucular” in printed obituaries, restaurant reviews, and press releases from biomedical companies, to name a handful of published and edited sources. In other words, that verbal tic is natural enough to some people that they automatically spell “nuclear” as “nucular.”
The linguists Geoff Nunberg and Allan Metcalf suggest it might be military jargon. Nunberg has talked to some of these people, who all manage to say “ˈnu-klē-ǝr family” and “ˈnu-klē-ǝr medicine,” but anything having to do with weaponry is ˈnü-kyə-lər. “I once asked a weapons specialist at a federal agency about this, and he told me, ‘Oh, I only say “nucular” when I’m talking about nukes.’ ” Metcalf notes the same general pattern, though he gives no anecdata*12 to support his contention. Steven Pinker, another linguist who got in something of a mild slap fight with Nunberg over ˈnü-kyə-lər, drops that in 2008 he spoke to the Strategic Studies Group at the Naval War College in Newport, Rhode Island, and heard ˈnü-kyə-lər from two senior analysts there. Interestingly, the earliest print instances we have of “nucular,” the spelled-out variation of the pronunciation that everyone hates, are in stories about and bulletins from people in the military, government, or nuclear sciences.
Some linguists, including Josh, think that because this variant occurs primarily in contexts like “nuclear weapons” and “nuclear power,” it “suggests that ‘nucular’ may not be an alteration of ‘nuclear’ but formed from adding the ‘-ular’ suffix to the root ‘nuke.’ ” If “nuke” started as military jargon, that accounts for the appearance of this pronunciation mostly among people who have spent a fair amount of time in the military or the federal government. But then we’d have to account for “nuke” or “nuc” showing up well after the spelled-out “nucular”; our first attested print use of “nuke,” in the word “thermonukes,” dates to 1955, while recorded spellings of “nucular” go back to 1943.
But, as Josh says, “this is all conjecture.” We don’t have much information about the original ˈnü-kyə-lər users, and we don’t have much information about current ˈnü-kyə-lər users—nothing on their native dialects, languages, ages, nada. That, of course, makes it much easier to jump to a few conclusions about the people who use ˈnü-kyə-lər:
But which of these stories explains why [George W.] Bush says “nucular”? Most people seem to assume he’s just one of those bubbas who don’t know any better. But that’s hard to credit. After all, Bush didn’t have to learn the word nuclear in middle age, the way Eisenhower did. He must have heard it said correctly thousands of times when he was growing up—not just at Andover, Yale, and Harvard, but from his own father, who never seems to have had any trouble
with the word. [Ed. note: Not true. We have evidence in our files of Bush senior using ˈnü-kyə-lər.] But if Bush’s “nucular” is a deliberate choice, is it something he picked up from the Pentagon wise guys? Or is it a faux-bubba pronunciation, the sort of thing he might have started doing at Yale by way of playing the Texas yahoo to all those earnest Eastern dweebs?
Love him or hate him, that’s not very fair to Bush. It’s true that people do sometimes intentionally take on an accent to appeal to a group somehow or to distance themselves from a stigmatized group. But more often it’s the case that people unconsciously change the way they pronounce something through sustained contact with other speakers. This has happened to me: after years of living in Greater Philadelphia, I find that sly Philly o, which sounds like all the vowels smeared together, in alphabetical order, popping up occasionally in words like “hoagie” and names like “Sophie.” I used to say “FOUN-tin” and “fill-uh-DELL-fee-uh,” the standard pronunciations for “fountain” and “Philadelphia”; now, when I am in a hurry, I say “FAN(t)-in” and “fill-LELL-fee-uh,” just as the old-timers here do. I can no longer remember if I am supposed to name my home state as “cah-loh-RA-doh,” making “RA” sound like the beginning of the word “rad,” or “cah-loh-RAH-do,” making “RAH” rhyme with “ma” and “pa.” I listen intently to my ma and pa say “Colorado” to see if this jogs any lingual muscle memory, and because I am listening intently, I find that they say it inconsistently as well. “This has happened to me,” says Josh. “I’ll call the Springfield Public Library, say. They answer, ‘Sprungfeld Public Library.’ I’ll say, ‘I’m trying to verify a pronunciation. The name of your town, how is it pronounced?’ They say, ‘Springfield.’ ‘Never “Sprungfeld”?’ ‘Nope, never “Sprungfeld.” ’ ‘Okay, thanks.’ And then I wait a while and call back just to see if I caught them on an off day, and—‘Sprungfeld Public Library.’ ”*13