Word by Word

by Kory Stamper

  That hasn’t stopped people from telling us where the word comes from. In the days of steamer travel between England and India, wealthy patrons traveling with the Peninsular and Oriental Company reserved the choicest cabins on the ship, which were the ones that got the morning sun but were shaded in the afternoon—no air-conditioning in the nineteenth century. These cabins were on the left side of the ship on the way out, and the right side on the way home, and so were stamped “P.O.S.H.” to indicate that the ticket holder had a cabin that was port side out, starboard side home. The “posh” ticket, then, was for the moneyed, elegant folk, and it was this association with wealth that gave us the “elegant” and “fashionable” sense of “posh” we know today.

  This is a fabulous story: it conjures images of women in bustles swooning on deck, canapés, servants in white linen shuffling deck chairs about for the quality. It’s also one of those great historical tidbits that sinks into the language and presents a finely sculpted detail to the modern reader. It’s beautiful—and total bullshit.

  Etymology requires evidence, and in spite of all of the e-mails and letters we’ve gotten over the years (all the way back to the 1930s, in fact), there is bugger-all written evidence for this origin story for “posh.” It’s not that there’s no one writing about these trips: there’s lots of nineteenth-century literature out there about the British interactions with and trips to India. In fact, more and more of that writing is being made public, and we’ve turned up nary a contemporaneous hint that the “port out, starboard home” theory is true.

  The first evidence we have for this theory goes back to a 1935 letter to The Times Literary Supplement of London:

  Sir,—In the Oxford New English Dictionary, the supplementary volume, the word “posh” is said to be of “obscure origin.” There is reason to believe that it is made up of the initial letters of “Port Out, Starboard Home,” an American shipping term describing the best cabins.

  Right away, alarm bells sound. The current hagiography of “posh” is that it is British in origin, and yet here’s a Brit claiming that it’s an American shipping term. Further, the story always includes the tidbit about cabins and sun—it’s an integral part of the story—and yet it’s missing here. Again, this is the earliest citation we have for this theory, and already the details of the story have been confused.

  There’s another red flag: the acronymic etymology. In spite of the foot stomping over newer acronyms like “OMG” and “LOL,”*2 we really do love acronyms, and especially acronymic explanations for words. “Constable on patrol,” “to insure promptness,” “gentlemen only, ladies forbidden,” “without passport,” “fornication under consent of the king” (or “for unlawful carnal knowledge,” or “forbidden under charter of the king,” or “file under carnal knowledge”)*3: all of them are the punch line for excellent stories about the supposed origins of those words, and they are all complete “ship high in transit.” As a word group, acronyms are not very productive in English: they don’t give us a lot of new words that stick around. The words that do stick around tend to be technical—“radar” from “radio detecting and ranging,” “laser” from “light amplification by stimulated emission of radiation,” “scuba” from “self-contained underwater breathing apparatus,” “CAT scan” from “computerized axial tomography.” But these are transparent, and so uninteresting. We like our false acronymic etymologies to be sexier, like “good old raisins and peanuts” or “north, east, west, south,” and it’s saying something about the field when “gorp” and “news” are considered sexy.

  Acronyms also weren’t terribly common until World War II, where they were deployed with aplomb. Unsurprisingly, most of the general words we have in English today that have true acronymic etymologies had their origins in the military: the aforementioned “radar,” “GI” (originally “galvanized iron,” if you can believe it, but misconstrued by soldiers and others as “government issue”), “snafu” and “fubar” (“situation normal: all fucked up” and “fucked up beyond all recognition,” brought to you by government bureaucracy). It’s true that a few of them snuck into English before the early twentieth century, but very few: “RSVP” (répondez s’il vous plaît) and “AWOL” (absent without leave) are the only two that could be considered general vocabulary, and some people will complain that one of those is not a proper acronym but an initialism and so doesn’t count.*4

  The more elaborate, the more detailed a story about a word’s origin is, the more skeptical the etymologist becomes. Look at the general vocabulary words that have come into English through interesting stories: look at “sandwich.” The name was taken from the title of John Montagu, the fourth Earl of Sandwich and quite the gambler. He loved gambling so much that he once sat at the gaming tables for a twenty-four-hour stretch, so absorbed that he didn’t stop to take a proper meal but ate cold beef between toast while playing. This bread-and-filling concoction became very much in vogue and came to be called a sandwich.

  This is as good a verifiable story as you get in etymology: It’s about wayward aristocracy and gambling and snacks. And it has some evidence to back it up—this story was taken from an account of the earl that was written and published during his lifetime, so if the earl wanted to refute it, he could. The account also mentions that the word “sandwich” for the bread-and-filling snack was in popular use in London during the late 1760s, which was contemporaneous with Montagu. The only thing etymologists are missing is a signed affidavit from Montagu himself and a sketch of the sandwich in question.

  But you’ll notice that the story’s details are painted in broad strokes. There’s no iconic cry from Lord Montagu asking for beef between bread; no triumphant win and subsequent mythology about the sandwich that made it a lucky charm of players; no shock from the onlookers that he refused to leave the gaming tables. The story we have is a good one, but it lacks a compelling narrative arc. Humans are inveterate storytellers. If history is lacking, we are happy to embellish it. And so, if a story behind the creation of a word has too many good details, etymologists grow suspicious.

  And with “posh,” the details change with remarkable frequency. The tickets were issued for journeys between England and Europe, for the shady side of the ship; or for any unspecified journey leaving Portsmouth, England, to ensure a sunny trip; or for journeys leaving Southampton, England, with the “posh” cabins getting the nice, scenic view and not the crappy one; or for journeys between England and America, so the afternoon sun would warm the passengers (or the morning sun—people are not clear on the position of the sun with regard to the assumed path of the voyage); or for journeys between America and India, direction of travel unspecified, though it would have mattered, because those “posh” tickets would have gotten the heat of the South Pacific sun traveling west from San Francisco. There are so many variations on the story that you need a map and a sextant to navigate them.

  It’s not as though lexicographers are actively working against this theory for the origin of “posh.” We have begged and pleaded for evidence—stamped tickets, earlier uses that tie this somehow to ship travel, journal entries, pamphlets. Anything that is contemporaneous for the era and written down would be the clincher that substantiates this. The answer to our cries: crickets. One of our etymologists wrote, on the back of a citation propounding the stamped-ticket theory, “Attractive but undocumented. We should live to see a ticket so marked.”

  The “port out, starboard home” theory isn’t the only one that’s been proposed. Some people hold that the word is Urdu, an adaptation of safed-pōś, meaning “white clothed” and then “affluent” or “well dressed.” Others have told us tales of a screen in some courts in the British Raj that was called a “posh” and that screened the unwashed masses from the toffs. Some insist that “posh” comes from “pasha,” a word of Turkish origin that referred to a high-ranking individual in the Ottoman Empire. Great theories all, and completely unsubstantiated. Because there was a lot of published material about th
e British Raj, you’d expect the Urdu word pōś to show up somewhere in the writing of the early to mid-nineteenth century and be explained as the Urdu word for “affluent” if this was the origin, but the record is silent. Ditto on the screens, and the luxury association with the P&O. “Pasha” came into English about three hundred years before “posh” did, which means that if “posh” were an adaptation of “pasha,” we’d likely have some written evidence linking the two together.

  Linguists have a couple of theories that are slightly more probable than the ones proposed by our correspondents—that the adjective comes from the earlier noun that means “a dandy” or the earlier slang noun for “money,” or that it’s somehow related to a one-off use of “push” in a P. G. Wodehouse story that means “fashionable,” but even the best guesses of etymologists aren’t good enough. We truck in hard facts.

  This upsets people, because it bucks against their assumptions of what etymology is: the logical, fairy-tale story of English. But it would be sloppy scholarship to elevate an unsubstantiated theory to the level of researched etymology simply because that theory is au courant. Further, the way that English grows doesn’t make sense. The history of English is full of messiness and illogic because the English language is a true democracy, built entirely by the people who use and have used it, and people, generally speaking, are messy and illogical. What genius, for instance, looked at the ragged edge of their sweater, laddering and unknitting itself with energy, and thought, “This is so bad that it’s not just raveling; it’s super-raveling. No: über-raveling. No, no, I got it: it’s frickin’ unraveling! Like, unreal amounts of raveling. Yeah, I’m going to call this ‘unraveling’ from now on.” Who thought that “pumpernickel” was a good name for a dark rye bread? Because when you trace the word back to its German origins, you find it means “fart goblin,”*5 and now you cannot help but blench and giggle whenever you see pumpernickel.

  Some people can’t deal with this wending and wandering. If English will not bend to their logic, then their logic will bend to English.

  —

  Devotees of what linguists call “etymological fallacy” ardently believe that the best uses of words—by which they mean the purest and most correct uses of words—are the meanings of that word’s grandparent, or etymon. “Decimate” is the soldier most often called up for duty by fallacists. How dare, they trumpet, you use “decimate” to mean “to cause great harm to”? Its real meaning is “to destroy one-tenth of,” as is clear from its etymology (from the Latin decimare, “to select by lot and kill every tenth man”). So if you care about language, then you shouldn’t use the hyperbolic extended meaning of “decimate.” You should only use it to refer to destroying a tenth of something.

  It is true that the earliest English use of “decimate” meant “to select and destroy one-tenth of.” That’s because when “decimate” was first used in English, it was primarily in contexts that described the harsh military discipline of the Romans. “Decimate” in the “one-tenth” meaning came into English in the late sixteenth century, and by the mid-seventeenth century its use had been expanded to refer to causing great harm. For about two hundred years, these two senses lived side by side without any peevery touching them.

  In the late nineteenth century, however, Richard Grant White took it into his head that the extended sense of “decimate” was wrong because of its etymology—“to use decimation as a general phrase for a great slaughter is simply ridiculous”—and though it took a while for White’s objection to be established, once it was, a long peeving tradition began. A handful of usage commentators in the twentieth century took up White’s charge: the “greatly harm” sense of “decimate” was ignorant nonsense. Starting around this time, we have evidence of people’s talking about “literally decimating” something—meaning they have selected and killed or destroyed exactly one-tenth of whatever the unlucky target is.

  But even usage commentators can’t ignore that “decimate” rarely means “to destroy one-tenth of” in English; it’s most often used with that extended “greatly harm” sense that’s been in frequent use since the late seventeenth century. This is truly logical: as Ammon Shea put it in his book Bad English, “How often does one really have the need to say, in a single word or so, that something has had exactly one-tenth taken from it?” Evidently, not very often. So the fallacists changed tack: “decimate” really means “to destroy one-tenth of,” and by extension, then, it can mean “to greatly harm,” but any uses of “decimate” that imply total annihilation are beyond the pale. This view shows up from the mid-1960s onward, and it is the objection raised in the usage literature nowadays.

  Etymological fallacy is the worst sort of pedantry: a meaningless personal opinion trying to dress itself up as concern for preserving historical principles. It misses that language change itself is a historical principle: a language that doesn’t change is a dead language, and as much as etymological fallacists seem to love the purity of Latin,*6 you’ll notice that none of them have abandoned that whore English for it.

  For people who love logic and straight lines, fallacists get squiggly if you press them on where, exactly, we draw the cutoff line for etymologically supported use and arrant anything-goes semantic nonsense. For instance, no fallacist suggests that we need to reorder the months of the year because the names for a bunch of them—September through December—don’t match up etymologically with their placement in the calendar. September (seven) is the ninth month of the year; October (eight), the tenth; November (nine), the eleventh; and December (ten), the twelfth. No one complains that “redact” is now used to excise writing from text when its Latin root means “to put in writing.” No fallacist objects now to words that came about through mistakes or misreadings, like “apron” (which was a fifteenth-century misapprehension of “a napron” as “an apron”) or “cherry” (taken to be the singular of the Old French cherise, which Middle English speakers wrongly assumed was plural because of the final -s), because these mistakes are so very old and well established. If we were, however, to excise these from the language, how would we do it? Would we pick a watershed date for “right” use, and if so, which date? The death of Shakespeare? Dryden? Pope? The zenith of etymological fallacy would be to refuse to admit any words after the Norman Conquest or perhaps the Danelaw. You don’t see any agitation for that, however; fallacists know, in their hearts, that the language is going to keep changing, and no amount of tantrums or threats on their part will stop that.

  I speak of etymological fallacy as if it were a new thing, but it’s truly not. People have been fomenting for an English academy, a ruling body that shapes the language and decides what’s official and what’s not, since the early eighteenth century. Daniel Defoe and Jonathan Swift were the key movers: both felt that English needed to be cleansed of impurities, polished, and subsequently policed. Nothing new would get in unless the academy gave the new coinage its imprimatur. Defoe wrote, in his 1697 proposal, that “twou’d be as Criminal then to Coin Words, as Money.” Swift, writing later, noted snarkily that English was in such a terrible state that it has “not arrived to such a Degree of Perfection, as to make us apprehend any Thoughts of its Decay.”

  French and Continental Spanish both have academies, founded in the seventeenth and eighteenth centuries, respectively, and it is true that they issue edicts on which constructions and words are “allowed,” and they have lexicographical arms that issue dictionaries with the “correct” words in them. Be that as it may, that doesn’t stop the French and Spanish from using whatever words they damned well want.

  The call for an academy of English fizzled, though there are still stragglers to the game, either crying out for its creation or volunteering to be that academy. In its absence, though, people who yearn for definitive guidance have created something similar: The American Heritage Dictionary’s Usage Panel.

  When Webster’s Third was released in 1961, it represented a marked departure from previous Merriam-Webster dictionaries
in many ways, but the thing that gained the most notoriety was a new Websterian adherence to describing as much of the language in use as possible, including contested uses (like the extended sense of “decimate”). There was a swift backlash from the literati of the day who deplored what they perceived as laxity in Webster’s defense of the language. There was extensive bewailing of the death of English, sped along by the “passel of double-domes at the G. & C. Merriam Company joint,” as one New York Times editorial put it, and along with the pearl clutching came several high-profile requests to pulp the entire run of the Third and go back to the good old days of Webster’s Second, which didn’t shy away from using labels like “illiterate” and “uneducated.” But Merriam sailed onward, heedless of the calls for reform (no doubt because the idea of junking the Third, a dictionary that took more than a hundred editors twenty-seven years to complete, sent the then president of the company into apoplexy). The Third was one of its flagship products, and the G. & C. Merriam Company would abandon it when hell froze over.

  One James Parton, the publisher of American Heritage magazine, decided to do all he could to make it snow. In 1962, a few months after the release of the Third, he began buying up shares of Merriam stock with the view of buying out the company. His reason: the company “badly needs new guidance,” and the Third was “an affront.” His plans were to “take the Third out of print! We go back to the Second and speed ahead on the Fourth.” When the hostile takeover failed, he did the next best thing: he created his own dictionary to right Webster’s wrongs and give America the authority it clamored for.

  The authority upon which The American Heritage Dictionary rested was its Usage Panel, a group originally of 105 writers, editors, and professors who were convened as a cabal of English experts, people who would decide which particular words and phrases were acceptable and which were not. Once a year, the lexicographers at the AHD send ballots to the panelists asking if they would find certain things acceptable in speech or writing. The first panel was reactionary in its founding—among the experts were a number of very vocal critics of Webster’s Third—but some of their opinions are surprisingly moderate. The December 1964 ballot for “ain’t,” a word that was singled out in criticisms of Webster’s Third, gives a slightly more nuanced approach than you would expect for a group of people who had it in for the Third. When asked about the acceptability of “I’m right about that, ain’t I?” 16 percent of the panel allowed it in speech, and 23 percent regarded it as more acceptable in speech than “It ain’t likely” (which was roundly dissed; only 1 percent of the panel found that acceptable in speech or writing). Malcolm Cowley, author and literary critic, finished off his 1966 ballot for the first edition of The American Heritage Dictionary with a cautionary epitaph, “There is always the danger that we, the so-called authorities, should become too damned pedantic,” and Isaac Asimov pleaded, “My opinions are strong, but not necessarily authoritative. Please realize that.”*7