Righting the Mother Tongue: From Olde English to Email, the Tangled Story of English Spelling

by David Wolman

  Although there are other entities in cyberspace named Penny Lane, a search for Penny Lane is, not surprisingly, going to lead first to Web sites about the Beatles song because the most abundant stuff (technically, the most linked-to stuff) regarding Penny Lane relates to the Beatles, not the Wisconsin farm called Penny Lane that makes and sells a southeast Asian sweet and sour dipping sauce.4 If you search for peni lane, you’ll still get steered toward the Beatles, but you can also make your way to a page about a New Zealand-based fishing vessel, Peni Lane. It all comes down to probabilities, which Norvig and company can build with unprecedented precision because they have an unprecedented amount of data about the language. This year, Google’s search engine will perform four hundred billion searches worldwide.5

  When I enter rubarb, Google search redirects me toward the information I want. That may sound scary if you’re a technophobe. Be afraid. Google knows what you’re thinking. But really what we’re talking about is one helluva tool. I want information about a certain red and green leafy vegetable. I know it. Google knows it. And there’s no point in letting an h—a silent one at that—slow my effort to get it.

  The last thing Google people want is to be perceived as setting rules or boundaries around what users do. A company as big as Google already has enough trouble dispelling fears of Big Brotheresque practices. “The question, ‘Do you mean?’ is deliberately ambiguous,” said Norvig. “What we’re not saying is, ‘Here’s how you spell.’” In this way, Google can be authoritative without being authoritarian, providing a snapshot of what’s out there in cyberspace without presuming to correct your English.

  A few years back, Norvig spent a cross-country flight writing what he calls a “toy” spell-checker, to help illustrate what’s going on under the hood of Google’s suggested spelling function. Outlining one of the program’s parameters, he writes: “It would be bad form to say the probability of a word is zero just because we haven’t seen it yet.” From a mathematical standpoint, this matters tremendously for making the spell-checker as accurate as possible; probabilities of zero have a habit of crashing calculations, so it’s easiest to go with a tiny number instead. Yet this computational detail also reflects the linguistic reality of language innovation on the Web. “You never know when things will change,” said Norvig. The words iPhone, mashup, Skype, and kiteboarding didn’t exist until recently. If the code is too doctrinaire in its operation, it will erroneously flag those types of words. Phone isn’t spelled i-p-h-o-n-e. But because of the possibility of something like iPhone appearing on the Web, Google’s spelling checker has to be inclusive. Or, as Norvig put it: “We let the data decide.”

  That data is created by the public. When searching for information about a species of dog well known for black-and-white coloring, the majority of humans conducting Google searches type dalmation. Google’s algorithms don’t flag this “misspelling.” As one employee put it: “Nobody can spell dalmatian, so neither can we,” or more accurately, neither can Google’s code. A search for dalmation leads directly to the same list of results delivered by a search for dalmatian. In one sense, this widening acceptability harkens back to the pre-printing, Haight-Ashbury–style free spelling of Old English manuscripts. Then again, maybe it’s not as much a matter of acceptability as it is a shift in how we think about correct and incorrect spelling. As individuals, we may not be conducting our lives with spelling rebellion in mind. Yet cumulatively, we, the wired populace of English users, are asserting control over the orthography of our beloved language, one Internet search at a time.

  Launch a search for airplayn and Google’s program, after calculating that you probably meant airplane, will bring up a suggested spelling of airplane and a gazillion links to sites with the word spelled correctly. The same goes for renaisance, kachitori, milennium, accomodation, entreprenur, cematery, floresent, and countless other common misspellings. Where things start to get interesting is with words that would be considered misspellings to some, but not to others, and not to Google search—words like dalmation, altho, thru, nite, supercede, int’l, attn, gonna, lemme, wanna, lite, legit, straight-laced, and journo (that’s Aussie for journalist).

  One question that’s been nagging me for a while now is whether the members of the Simplified Spelling Society, if armed with the money and technical know-how to put their enuf, lojic, frend, speling, rime, laking, and rubarb all over the Internet on thousands of pages, might be able to hack the lexicon by forcing their preferred spellings onto so many screens across the planet. “No,” said Norvig. “They can’t take over common words that are already everywhere.” With accepted spellings populating most of interstellar cyberspace, usurping these commonly recognized forms would require a mountain-moving effort that makes spelling reform at the turn of the twentieth century look easy. The Society members might be able to influence some more obscure words, said Norvig, but little beyond that. Even through the Internet, far-reaching language reform by a few, aimed at the many, appears to be impossible.

  That doesn’t mean, however, that creative spellings sprouting up organically won’t take hold and challenge the hegemony of at least a portion of the current standard. The Internet, Crystal had told me, is the one area where a genuine bottom-up alteration of spelling stands a reasonable chance of success in the long term. “You have to ask: Why is a word tough to spell? Because someone once made it difficult. There shouldn’t be a b in debt or an h in rhubarb. In the past, with print, you could never get away with leaving it out because of copy editors and publishers,” he said. “But on the Internet, you can! We’re seeing a return to natural, instinctual spelling. This is people voting [for spellings] with their fingertips.” Crystal isn’t suggesting that these grassroots changes take effect right away, but thirty or fifty years down the road, who knows what might happen? In that sense, the Simplified Spelling Society members should be hopeful. It’s not the reform they envisioned—that is, a panel of experts steering the language. But a shake-up is happening, fueled by what looks like an expanding attitude of inclusiveness about words and spelling that has not been seen for centuries.

  Blogs, Web sites, chatrooms, email, and text messaging—out on the Web or in our speedy electronic communications, language curmudgeons look more anachronistic by the day. Yet they persevere. In Ireland and the United Kingdom, there have been some recent spats between those who think text message code should be allowed on school papers and tests, and those who vehemently oppose the idea.6 Online commentary gives the impression that the court of public opinion sides with the view that texting has a poisonous influence on language and literacy.7 In 2007, Irish education officials concluded that texting erodes students’ writing skills. “Text messaging, with its use of phonetic spelling and little or no punctuation, seems to pose a threat to traditional conventions in writing,” their report declared. Teens are “unduly reliant on short sentences, simple tenses, and a limited vocabulary.”8 Take that, Hemingway.

  Text message code is still too new for long-term rigorous studies of its effect, although Crystal, who recently completed a book about texting, says they are starting to appear and that they suggest texting betters literacy. Negative reports about its impact are generally geared toward a specific—which is to say, stylistic—view of English language skills, not literacy in the developing-brain sense. But since when has sound evidence been a prerequisite for a strong opinion about anything, least of all language? A few years ago, a British Broadcasting Corporation bigwig wrote the following:

  There are so many threats to the survival of good, plain English that it is not easy to be optimistic. Email has a great deal to answer for. Punctuation is no longer required and verbs are abandoned with the speed of a striptease artiste late for her next performance. Text messaging is worse—much worse. Yet I have seen it suggested that students be allowed to use “texting” abbreviations in examinations. Ultimately, no doubt, we shall communicate with a series of grunts—and the evolutionary wheel will have turned full circle.9

>   Really? Consider for a moment this celebrated construction: OK. OK is a blend of both spelling reform and abbreviation, and took shape long before there was any such thing as the Internet. Although there’s some dispute about its exact origin, the leading theory is that it stands for “oll correct” or “ole kurreck,” and traces back at least as far as the 1830s, when trendy misspellings and abbreviations were part of the vernacular, such as NG for no go, SP for small potatoes, and GT for Gone to Texas.

  Abbreviating and truncating are time-honored traditions not just in English but in languages, period. This trend goes beyond taxi from taxicab, bus from omnibus, and OK from oll kurrect. The likes of BBQ, R & R, R & D, ASAP, TLC, PR, and FYI are already well established. And if you’re not yet acquainted with GHGs, I might as well introduce you to greenhouse gases. There’s also the slew of corporate-or military-world acronyms and rejiggerings like MREs, for meals ready to eat, and reorg and recon for reorganize and reconnaissance.

  Texting is just an extension of truncating and acronymming, whether it’s TBH (to be honest), FWIW (for what it’s worth), PLOS (parents looking over shoulder), AIMP (always in my prayers), ADN (any day now), AAR8 (at any rate), or 4COL (for crying out loud). Ever seen pwn for the word “own”? At some point in the recent history of online gaming, someone, somewhere, possessing an ample amount of clout, entered pwn instead of own (p and o are keyboard neighbors) in a trash-talking context along the lines of: “I own you this round.” The slipped keystroke became a form of insider speak among this networked community of gamers, and then it somehow spread, all the way to the desk of Merriam-Webster dictionary editors.10 On the Internet, nobody’s word concoctions are isolated, and every last one is a candidate for entry into the lexicon, or at least the more welcoming versions of it.

  Spend a little time thinking about OK, as well as ASAP, FYI, AWOL, BBQ, and SCUBA (now just scuba), not to mention the tried-and-true expression that history repeats itself, and you may find yourself agreeing with Crystal’s assessment that the forecast for digital English isn’t dire at all. The most devastating blow to traditionalists, aside from the inevitable narrowing of the mind that comes from strict prescriptivism, would be further evidence suggesting that people become more sophisticated in their understanding and usage of language, not less, as a result of increasingly digital lives. After all, you have to know what the code stands for. The representation AAR8 doesn’t mean anything to sender or recipient if both don’t already know the words and the expression, at any rate. Arms crossed in dissent, pundits, if they’re not careful, may end up knowing only one thing where others know two—the words and the coded shorthand. Perhaps science will even reveal that the brains of people texting are more active when composing written communications than those who reject it because of some vague notion of correct treatment of English.11

  Texting is different from spelling, of course, but opinions about it inform the discussion of the future of English orthography. A fundamental question is whether people in cyberspace typing rubarb, altho, nite, dalmation, int’l, and even seperate are categorically wrong. Maybe you feel they are, or that they’re so far adrift from convention that they’re as good as wrong. If that’s the case, you may be pleased to know that while new words and alternative spellings—or maybe it’s new “words” and “alternative spellings”—are piling onto the Web, that doesn’t mean they’re piling unfiltered into some of the more revered records of the lexicon.

  At the offices of Merriam-Webster, I asked John Morse about his company’s relationship with the forces of the Internet. Although challenging, cyber-catalyzed linguistic evolution doesn’t get under Morse’s skin. “Google is constantly on any information provider’s radar,” said Morse. “It’s like a giant star coming into your solar system, bending and changing space time.” Merriam-Webster staff tries to keep up with the times, introducing features like an online open dictionary, which is essentially Wiktionary without calling it that, and without giving users the keys to the editing kingdom. One recently added word variant: redonculous, also redonkulous, for ridiculous.12

  From the Merriam-Webster perspective, there are limits to the Internet’s usefulness as source material for compiling the contents of a dictionary. When talking about the word whack, for instance, Morse said: “We’re searching edited prose. That’s a lot different than a blog’s mention of a rock band called Whack.” In Morse’s judgment, just because words are on the Web doesn’t make them published. A lot of chatroom content, he said, could be likened to the notes kids pass to each other in school, which doesn’t pass muster, he said, for dictionary source material.

  The general rule at Merriam-Webster is to stay in “protected spaces” online. Yahoo News, Salon, Slate, Boing Boing, and scientific journals on the Web are all safe terrain. But the latest unpaid blogger’s posts about country music, a Listserv about caribou hunting or a chatroom devoted to the latest Brazilian pop music? That type of Web content, said Morse, has to be passed over, otherwise we’d end up with “mob rule.” Take that, Wikipedia.

  I find this view of the lexicon a little redonkulous. First, the flimsiness of the criteria, “edited prose.” Surely thousands, if not millions, of well-written blogs and zines that fly under Morse’s radar are legitimate sources for new words and meanings, just as an unedited Joyce Carol Oates diary entry, or a back-of-a-napkin essay outline scribbled by Tom Wolfe, are also legit sources. It seems like a huge omission, waving aside this massive volume of words that may be, and most likely is, contributing to language change. That Merriam-Webster embraces mob rule a little with its partially open dictionary, yet applies a different standard to its flagship dictionaries, calls into question the strength—and utility—of modern-day levees upholding language authority.

  A more wide-angle view of language in the digital age suggests that the days of the professional word arbiter are numbered, and that some spellings may loosen because of it. Noah Webster’s first dictionary was an engine of spelling reform, but there hasn’t really been one since. The next revolutionizing force may be the global citizens of English, using what Crystal calls “instinctual spelling.” Milennium, tendancy, reverant, enuf, brocoli, comraderie, privledge, miselanious—who knows how far it may go, or how far we will let it go? The idea may sound radical, and all together those altered spellings certainly look jarring. But the real changes won’t be nearly so disruptive—orthographic rebellion that leads to less intelligibility doesn’t make much sense. More likely, the process will be natural, almost invisible, like the Great Vowel Shift, or how goode became good, glycerine became glycerin, sette became set, omnibus became bus, catalogue became catalog, and hiccough became hiccup. But whereas once this process took centuries, it could go much faster on the Web.

  Throughout the story of English orthography, attempted reforms (with the exception of Webster’s Americanized spellings) have collapsed, in large part because there was no governing body, no Académie L’Anglaise. The language, as Benjamin Ide Wheeler told Stanford’s Calamity Class of 1906, is the “precious possession of the English-speaking world”—all of us, and most of us share a resistance to anything that smacks of the few telling the many how to behave. It’s this populist tradition that adds elements of both hypocrisy and tragedy, I think, to the failed efforts of people like Dewey, today’s Simplified Spellers, and other reformers. Many of them held, and hold, power-to-the-people ideals. They want spelling to be easier because they want reading and writing to be easier for everyone. Yet the way they tried to get there was through some sort of orthography oligarchy. No dice.

  The criterion of “published material” and “edited prose” used by some present-day lexicographers in their attempts to proscribe the entire language, echo this problem of top-down control. Things aren’t nearly as narrow as they once were, back when Samuel Johnson almost single-handedly laid down the laws of English orthography, or when Webster was making himself dizzy with etymologies as he turned circles within his doughnut-shaped desk. Today, Morse’s
staff can turn to a corpus of text comprised of some one hundred million words. But there’re trillions of words in cyberspace, and millions more next week and the week after that. They can’t be ignored.

  Yet Morse has no option but to limit what his editors read, short of quitting and devoting his energies (without pay) to the growth of Wiktionary. “The largeness of the Web,” he said, “makes it impossible. So you sort of have to raise the bar. It’s a question of scale. There’s just so much electronic text available.” It’s easy enough for a curious outsider to pester Morse about how lexicographers draw arbitrary lines in the sandscape of the lexicon, but the brave souls on the Merriam-Webster crew still have to put out bound dictionaries. In contrast, the editors of the OED have the luxury of just saying yes to everything because the bulk of OED editorial energy now goes into OED Online, which can grow without limitation. There are still plenty of judgment calls, like which spellings are most common, versus what’s slang and what’s informal, and how long a word must survive in the vernacular before it earns an official place in the lexicon. But the virtual dictionary has no binding.

  Morse, master definer Jim Lowe, and the rest of the Merriam-Webster staff have to stop somewhere—for the collegiate or smaller-sized dictionaries especially, but also for the unabridged edition, which is ironic when you think about the definition of unabridged: “1. not abridged or shortened, as a book. 2. a dictionary that has not been reduced in size by omission of terms or definitions….” Morse tells me that of the tens of thousands of h-less rhubarbs on the Internet, “if they were all found on message boards written by people who aren’t experienced writers, and not appearing anywhere else, then no, I doubt rubarb would get in.” (On the other hand, all it takes is one US president to use a construction like nucular to slingshot it into the realm of officialdom.)