by John Simpson
And so every three months from March 2000 we had to hand over a substantial quantity of text (sometimes the size of a large novel) for online publication—every word in the alphabetical sequence we were working on. Nothing was left for later—I didn’t believe we would have any more time “later” than we did now. The book version was becoming—to us—largely an irrelevancy, and was slipping into the past. We thought that the dictionary might be published later as a series of book volumes, but not until the full cycle of update was over and not unless the book-buying market wanted it and made the publication economically viable. The dictionary’s “centre of gravity,” as we proudly explained to people, had shifted to the online version, which was where any change would in future take place. M had moved into N and O, and then we spent ages in P and R, which are both massive letters, dwarfed only by S—which the original dictionary had had to divide into two enormous volumes in order to prevent the spine from splitting. But each quarter the work was ready, and we never missed a publication deadline. There were no volume ends to worry about—we just closed the publication range after the end of whatever range of words surrounded our numerical cut-off point.
We were hampered, though, by our old system software, which was gradually sinking into obsolescence. We knew this because whenever something went wrong with the computer system, we were told that it couldn’t be fixed properly, but that we would have to use a workaround. Eventually I think there were more workarounds than there were simple bits of code. As each programmer left, another piece of code became worthless, as the fiction was maintained that no one else knew how to fix it.
When we started working with computer scientists, we heard for the first time quite a few words (such as the noun kludge: a makeshift software fix) to describe the process or result of encouraging a faulty computer program to work even though the computer scientists didn’t really know why it was broken in the first place. When talking to real people, the technicians politely called these informal fixes workarounds. Even workaround was quite a new word in the 1990s, though the concept is age-old. This time the word does not derive from the world of ice cream, but from the aeronautical and the astronautical industry, where safety mattered in triplicate. By 1961, if a system wasn’t working and the engineers couldn’t go back to basics to fix it, they euphemistically employed what they blandly denominated “workarounds,” so as not to frighten anyone—which doesn’t really seem particularly good news for the customer, or the astronaut. As with the OED’s computer system, if you couldn’t fix a problem, you employed a workaround. After a few years—and at least by the late 1970s—people realised that it was a useful word to get you out of any tricky situation.
Eventually all the pieces of string involved in holding together our software were stretched to the breaking point, and in 2005 we took delivery of a brand-new system in which most aspects of the editorial and publication process were integrated, and which finally allowed us to edit wherever we wanted in the database. In addition, we used the opportunity to make further changes in the look of the dictionary entries, making them easier to read and understand by breaking up huge columns of quotation material and implementing thousands of backroom standardisations to the text.
When the software was ready for us to install, we had a small competition to give it a name. Old OED software suites had traditionally been allocated names based on any word that included the letter sequence O-E-D. Predictably, our first set of programs had been called “OEDipus.” Unlike Welsh, English has only a few words containing the OED’s letters, especially if you leave out unlikely program names like “three-toed.” The winner of the naming competition this time (radically) did not contain the letters O-E-D. It was “Pasadena,” wryly named by my colleague Jeremy Marshall as the “Perfect All-Singing All-Dancing Editorial and Notation Application.”
We argued amongst ourselves (maybe for the first time) and with the University Press’s senior managers on whether it was right to revise entries out of alphabetical order. In general I was in the conservative camp, keen to edit everything in detail as we encountered it on our sequential pass through the alphabet, but in this I was apparently one of a dwindling band of druids. And indeed there was an argument to be made that it was worth making exceptions to our alphabetical progress in order to address the most relevant entries, those that our audience might be most eager to see updated—typically the big, core-vocabulary words, the sensitive words, words that “really mattered.” For the editors, though, these were the tough ones to edit, and a constant diet of these would be crushing. I felt that editors also needed the small, surprising, intriguing entries that the alphabet throws up to counteract the unremitting gigantism of the “important” entries (black, earth, field, ground, nation, self-, work—all of which go on for pages). What followed was a sort of tug-of-love between the two ideologies—slow and careful editing in alpha order and slow and careful editing wherever in the alphabet we felt it was most needed.
In the end, we shifted our editorial priorities so that in one quarter we’d progress along the old alphabetical sequence that we knew well—further into P (the new computer system hit us around panache). And then in alternate quarters we devised ranges of “big” words, to which we’d give the full OED treatment.
I have to confess that, despite my inclinations to the contrary, it was rather exciting to put together that first range of non-alphabetical words. These were published—once we’d become used to the new system, in March 2008. We had selected some meaty, high-profile terms to update in this ground-breaking batch: affect/effect, air, American, cancer, gene, heaven, hell, language, love, sad, thing. These words all involved some heavy-duty editing—most had been in the language for centuries and had developed a multiple subsense-structure and numerous compounds.
At the top of the list was the word gay, a word we were at the time continually asked about by journalists, lecture audiences, radio DJs, and just about anyone else we happened to come into contact with. We set about updating this entry straightaway—most of the text still dated from the original instalment published in 1898. Let me take you through some of the issues that gay raised for us.
From at least the 1960s there had been extensive social advocacy for gay rights, and social change typically influences language. At first there was often a deep-rooted worry in conservative quarters that “we could no longer use a traditional old word,” gay, because it had been hijacked by its new meaning. On the other side, there was a powerful cohort of language users who actively promoted the use of gay as a symbol of a new-found personal liberation. The opposing pulls on gay in the modern era, coupled with its extensive history in the past, made it a challenging and yet intriguing word on which to work.
We were updating a composite Second Edition entry consisting of the old OED1 entry (first published in 1898) augmented by additions in the twentieth century—so we already had hints of the “homosexual” meaning from the days of the Supplement I had worked on in the less-enlightened 1970s. When the OED had first defined gay at the very end of the nineteenth century, it was a merry, lively word, and one that in its dotage had become slightly naughty. Its first use in English dated, according to the nineteenth-century editors, from around 1300, and had found its way into English from French gai (itself recorded from the twelfth century). The old OED was puzzled about how gai had got into French, so it offered a number of suggestions, only to distance itself from them all—as usual—in as polite a variety of English as it knew how.
Once gay had found its way into English, the First Edition of the OED described it in a number of meanings, mostly circling round the idea of “full of joy or mirth”: people could be mirthful, and women (typically in poetry) were apparently particularly prone to being joyful or gay. Gay horses pranced, poetry was known as the “gay science” (apparently after an idiom from Provence). But by the mid-seventeenth century, gay veered off in a rather more dissipated direction: we start finding “gay lotharios” (and “gay dogs”), and the
subtle vocabulary of laxness and incipient immorality. By the early nineteenth century things had become even worse, and it was possible to regard prostitutes as “gay women,” leading colourful lives described by a word which allowed the sanctimonious middle-classes to discuss but not condone them. By the nineteenth century, gay had two faces: on the one hand, happy, carefree, and joyous, and on the other, abandoned to pleasure and morally lax.
These two strands blended into the new meaning “homosexual,” which the Supplement to the OED had identified only as recently as 1935, in an underground glossary published in Upland, Indiana, and this use has blossomed explosively since then. Our revised entry would need to tease out the process of how this new meaning, apparently promoted by the people to whom it applied, had come into being.
Dictionary work is not all about definitions: we also needed to clarify aspects of the etymology. For example, we found a bridging word gai recorded in Anglo-French (the variety of French used in the upper echelons of British society after the Norman Conquest). This discovery allowed us to document the progression of gay from French into English via an interim dialect. And that’s important to us. We didn’t get the word willy-nilly from Paris or central France—it came as part of the Norman hang-over package. Rather than limiting ourselves to around 1300, we could now plot an earlier use in about 1225, which is much more satisfactory for a working Anglo-French word. (You must forgive me if these niceties seem unnecessary to you.)
The First Edition of the dictionary had associated gay with immorality and louche behaviour, but too late. New evidence from online sources and elsewhere helped to focus the picture. With the help of the Middle English Dictionary, we found that the medievals also had given the word gay a lascivious spin (missed by the original OED), although only for fifty years around the early fifteenth century. Geoffrey Chaucer wrote of “som gay gerl” in the late fourteenth century: OED1 thought this use of “gay girl” was—in its context—simply an expression of praise for a lady; more recent scholarship preferred the reading “lascivious girl.” The usage now seems to lay the seed for the general sense of promiscuousness or immorality (the gay lotharios, etc.), which now dates not from 1637 and the time of playwright James Shirley (The Lady of Pleasure), but from William Shakespeare’s time, with a first example weighing in from 1597 (in John Payne’s Royall Exchange). At that time, you might go gay, if you wished to lead a life of licentious immorality. Even the prostitution sense creeps backwards in time from the early Victorian era into the late eighteenth century, in Richard King’s long-forgotten New Cheats of London Exposed (around 1795). After research, meanings often turn out to be older than we had imagined.
But new forces were at work around gay as we moved into the twentieth century. The old dictionary firmly stated that gay, in the meaning “homosexual,” dated from Noel Ersine’s underworld glossary of 1935. That says explicitly, after all, that a geycat (or gaycat) is a “homosexual boy.” But geycat is recorded, as we discovered, much earlier—from 1890s America, without explicit homosexual connotation, and meaning a “young or inexperienced tramp.” We took the view that this 1935 use apparently only implied homosexual conduct contextually, and was not prima facie evidence for the equivalence of gay and homosexual.
One reason why gay was a tricky entry to revise was that readers could—incorrectly—read later meanings into earlier evidence. I don’t think they had a political agenda here: it’s just a natural tendency to wish that what you’ve found is what you were looking for. I’ve done it myself—everybody has. But editors are trained to reject everything unless it demands inclusion with a cast-iron certainty. If they start bending the rules, or letting in maybes and possibilities, then the dictionary’s credibility is blown. It’s better and safer to err on the side of caution and conservatism than to run with every idea that is sent along to distract you.
Because the “homosexual” meaning of gay had a complex history, and clearly attracted so much interest and misjudgement, we took the highly irregular course of printing a series of false quotations before the “real” ones. These false—or near-miss—quotations illustrated early halfway usages which might be interpreted in the “homosexual” sense, but which in fact only exemplified the older meaning “merry” or “happy,” in contexts suggestive of homosexuality or lesbianism. Here is the first of these, from Gertrude Stein’s “Miss Furr and Miss Skeene” of 1922:
Helen Furr and Georgine Skeene lived together then. . . . They were together then and travelled to another place and stayed there and were gay there . . . not very gay there, just gay there. They were both gay there.
It’s not hard for someone to think that this is a true example of the later meaning gay = “homosexual” (and specifically “lesbian” here), but it’s probably just the older mirthful and colourful meaning, used in the context of homosexuality or close companionship. There are many examples like this, where the context suggests more than the word itself conveys. Here is another, from Noël Coward’s song “I Went to a Marvellous Party” (1939):
Everyone’s here and frightfully gay,
Nobody cares what people say,
Though the Riviera
Seems really much queerer
Than Rome at its height.
You can see how this might be mistaken by the over-zealous for “homosexual.” But we need to look at every aspect of the word and wait until the evidence is overwhelming. In this case, we waited until Gershon Legman’s heavy-duty Language of Homosexuality, published in New York in 1941 as part of George W. Henry’s Sex Variants. We made this the new first example of the “homosexual” sense of gay, bravely five years later than the Supplement to the OED’s first use (the geycat)—which we now considered to be a misreading. You couldn’t argue with Legman, and his explanation suggests that the term arose as a positive use amongst homosexuals (which fits its first appearance in a text which would not have had a general readership):
Gay, an adjective used almost exclusively by homosexuals to denote homosexuality, sexual attractiveness, promiscuity . . . or lack of restraint, in a person, place, or party. Often given the French spelling, gai or gaie by (or in burlesque of) cultured homosexuals of both sexes.
One final observation on gay: we found that, as usual, the productivity of gay as a word in the second half of the twentieth century was marked not by new basic meanings (we added only two), but by the proliferation of compounds, and especially compounds involving the “homosexual” meaning. Whereas our starting-point entry had twelve compounds, we easily expanded this to thirty-three, and doubtless could have gone much further. Here are a few, though I dislike lists in language books: gay bar, gay boy, gay day, gay gene, gay icon, and gay lib (not to mention gaydar). Sorry—that’s enough of them for now.
Sorry makes its appearance in English in the Anglo-Saxon or Old English period, before the Norman Conquest of 1066. Early manuscripts containing English are pretty thin on the ground, and so we have little surviving documentary evidence for that period of the language (about 3 million words of text, and even then often of a religious or a legal nature). From today’s perspective it’s remarkable that sorry is directly related not to sorrow, as you might naturally think, but to sore (the noun: bodily pain, a sore place, etc., etc.). So if you’re sorry, you are not primarily displaying your sorrow, but betraying how much you hurt. But there is a secondary link with sorrow, as similar words often influence each other’s development: it is probably from sorrow that sorry gets its short
To the Anglo-Saxons, sorry meant “expressing pain or sorrow,” but even then it could also mean “penitent” or “apologetic.” So you could say that you were sorry about something from the early days, but the emergence of the single word sorry, used on its own as an apology for something you have done, occurred much later than the Old English period. We don’t have any records of Anglo-Saxon merchants saying “sorry” for overcharging an Anglo-Saxon peasant. In fact, we don’t have a
ny evidence for this particular lexical usage until 1843, in Sargent’s New Monthly Magazine, an American periodical published by Epes Sargent, a Boston journalist and litterateur and a member of the American East Coast literary “Knickerbocker Group,” which included the more famous Washington Irving and James Fenimore Cooper. Fascinating though Sargent doubtless is as a minor Boston literary figure, he’s most important to us because, in March 1843, his short-lived magazine New Monthly included the short sentence “Miss Marion, good morning, see you tomorrow—sorry I’m in such haste,” which (as far as the OED has so far been able to ascertain) is the first known occurrence in print of “sorry” used apologetically, and shortened from “I am sorry.” It’s a small thing, but it kept us occupied, and it’s from these small tiles that the grand mosaic of the language develops.
Although the media profile of the dictionary was now consistently high, it was highest of all in the United States. By now the OED had a small editorial office within the main Oxford University Press operation on Madison Avenue in Manhattan, New York City. This was led by Jesse Sheidlower, the first port of call for American journalists contacting the OED. Jesse had been involved with the OED for about five years by then: a lean, super-smart, super-slick New Yorker with a passion for slang, words, and cuisine, both in the oven and on the plate. To his closest friends, he was the Jessemeister, ruling the word-waves of American English. He had previously been working with Jonathan Lighter on the Historical Dictionary of American Slang, and seemed to remember almost every fact he had ever read about language. Unlike most of his Oxford counterparts, Jesse would get up in the morning hoping that the press would ring him or email him, offering a “publicity moment.” Maybe he even sent letters to himself at work to provoke publicity opportunities. The press was mesmerised by the fact that he’d had the temerity to publish a treatise on the word fuck called The F Word. That wasn’t what got him the job on the OED by a long chalk, but it made our dictionary publicists sleep well at night.