The Word Detective

by John Simpson

  In order to develop a project plan, the Admiral needed facts (not yet available in any quantity or with much reliability). We couldn’t generate facts without a detailed editorial policy, and so the business of hammering one out—which would guide us in transforming a nineteenth-century dinosaur into a twentieth-century thoroughbred—became our No. 1 priority.

  Finally, there was a chink of light. After three years in limbo, there was once again a palpable sense of excitement and optimism amongst the small group who constituted the project—Ed Weiner, Yvonne Warburton, and I, and several others, had at last been invited by the University Press to consider the options for forging ahead with a comprehensive update of the big dictionary. Starting from now, we could plan the changes we wanted to introduce to the dictionary. With those changes, we could open up the content, increase access, and free it from the dusty air of the nineteenth century.

  We busily divided the number of words defined in the existing dictionary (414,800) by the number of years until 2000 (7), and came to the conclusion that (a) it couldn’t be done, and (b) we’d better see if we could do it. Maybe we thought that if we more or less achieved this target, then that might be okay. It turns out that isn’t how businesses work.

  It being a universal truth acknowledged by anyone managing a dictionary project that lexicographers should not be permitted to control the flow of facts themselves, it was agreed that a number of the University Press’s Delegates, or their representatives, would meet regularly with senior members of the dictionary staff (the Admiral, Ed, and myself) as the OED Advisory Committee. This group would establish a comprehensive editorial policy acceptable to all parties. It made sense: you can’t start revising a dictionary at A and just wander hopefully through until you reach Z. You have to plan the whole structure before you start. We were updating an antiquated mammoth of a dictionary, but we needed to have a vision of what our new version was meant to look like. We certainly had our own vision—but did it square with that of the University Press’s Delegates? The Advisory Committee also had another, unwritten function. Back in those days, there were people who were extremely resistant to any change being made to the text of the dictionary: they regarded it almost as biblical writ. So the committee would, by its approval of our plan, confer upon us academic licence to revise the Victorian tablets of stone.

  And so it happened that on regular occasions over the next two years the Admiral, Ed, and I participated in a series of meetings with some of the sharpest, no-nonsense professors that the University could throw at us: their specialisms included the history of English, English literature generally, comparative philology with particular reference to Linear B (the syllabic and ideographic script used to write early Mycenean Greek), and (almost the easy option) sociolinguistics. Several were reputed—even in the rarefied world of Oxford scholarship—to have extra toppings of academic rigour. We certainly did not enjoy the prospect of the first of these meetings: they were such a powerful group of academics—some of whom we only knew previously by name and reputation—that if they took a position in opposition to our vision of ourselves, then we would be in deep trouble.

  First we had to decide whether the dictionary as it stood offered a satisfactory general model for the future. Were the current components of the entry the right ones—definitions, etymologies, illustrative quotations, all set within a historical structure? Or did some other structure suggest itself? Should we throw away the mould and start from scratch?

  There were alternative options, one of which was that we should completely reassess all available sources and start afresh. Fortunately, it was generally agreed that the current dictionary, if properly updated, would still provide an excellent model for its users, and that it would be madness to discard what we already had and begin with a clean sheet. Scholars still wanted the precise, detailed information that the OED had always contained, but they wanted that information to be up to date, not (as in many cases) over one hundred years old. Even to the academics of Oxford one hundred years sounded a long time. And could we start to imagine what budget would be necessary for restarting from the beginning all over again? That, to all of us, was a no-brainer. It would be enormous, and well beyond the resources of the University Press.

  Another big question we needed to discuss: Should the OED continue to order its subsenses historically, or would it be best—as some people argued—if the most common meanings were to appear first, with the rest presented in order of frequency further down the tail of the donkey? Well, frequency order helps you find the common meanings, but isn’t much help (in a big entry) if you want to find anything else. How can you come up with a ranking order of word meanings that is relevant for all varieties of English, where frequency order is different, and constantly changing over time? Anyway, if you want the most common meaning first, go to a dictionary where this is all that matters. It’s not the OED. With the OED, historical order leads on from the etymology, and should demonstrate how meanings change over time. Consider magazine (or nice, or table, or watch). Commonest first? Sometimes I just get cross.

  Here’s a good example of how meanings change over time, along a logical but unpredictable route: the word magazine. What is the commonest meaning today? A periodical publication with feature articles and glossy pictures. But go back to the seventeenth century: What was the commonest meaning then? The same? I don’t think so. The modern “publication” meaning didn’t even exist until 1731. In the seventeenth century magazine meant “a storehouse for goods,” especially provisions or explosives (preferably not at the same time).

  The purpose of the OED is to show that magazine came into English in the sixteenth century (in the “storehouse” sense), and that we got the word from Middle French (magasin), which derived it from Italian (magazzino), and that it comes ultimately from an Arabic word (makzan) for—guess what—a place where things are stored, i.e., “a storehouse.” Arabic influences (architecture, literature, etc.) along the southern Mediterranean are an indication of contact between Arabic people and Europeans (especially in medieval Spain). And where you have contact between peoples, you have language interaction and borrowings (also from Arabic: algebra, giraffe, mohair, etc.). The later senses of magazine in English are tendrils from the original “storehouse” meaning, leading up to the present day.

  If you want to follow the evolution of magazine from Arabic to now, you don’t start with the glossy periodical publication. Maybe not a no-brainer, but a brainer involving only a modicum of reflection.

  The committee also considered another issue, involving one of those wild suggestions which sound ludicrous but from which you learn (in the Oxford way) simply by discussing and rejecting it. The idea was proposed by one of our senior academics on the committee that we would get the whole job finished earlier and with a considerable slashing of the budget if we disregarded everything in English before, say, 1500, and refocused the OED for the future as a dictionary dealing only with Modern English. There was no doubt that we would finish sooner that way. It was a bit like saying you could beat the Olympic hundred-metre sprint champion if they gave you a ninety-metre start. When the idea was proposed, we looked at each other as if someone had asked if we could squeeze Esperanto, Vulcan, or the language of fishes into the OED. But this was a deadly serious suggestion. And certainly, nothing would have made our financial colleagues happier than our truncating the project in this way.

  The idea was not completely madcap, however: elsewhere in the Western world (Toronto and Ann Arbor, Michigan) there were already large dictionary projects whose objective was to capture—in much more detail than the OED could expect to manage—the entire vocabulary of English in the Old and Middle English periods, respectively. Shouldn’t we just cede anything before 1500 to these dictionary projects, and simply concentrate on the later period? The idea was that we could save ourselves from dipping even a toe into the murky waters of Anglo-Saxon and Middle English vocabulary, and the equally murky culture of those early days. We could s
tart with fresh pencils at 1500.

  But that’s not how English works. You can’t start right in the middle and forget the distant origins. If we fudged everything before 1500, we would not be able, authoritatively, to comment on deeper etymology or even earlier meanings that influenced the senses of post-1500 words. We’d break in on important meanings halfway through their existence, unable to speak from proper knowledge about their early life. What we would be doing, in fact, would be representing English as if it were mainly a Renaissance creation—Latinate, or with words from modern Dutch, or French, or German, and full of English-only compounds, but without the organic complexity of its early Germanic and Romance past, where the real questions about English are set and have to be answered.

  After a few minutes’ discussion we had one of our rare votes, and all—including the wily proposer—agreed nem con against the stripped-back and curtailed OED. At a stroke, we had committed ourselves to a further ten or fifteen years work.

  All of our discussions with the Advisory Committee were premised on the assumption that our principal objective was to illustrate—through the dictionary—the complex emergence and flowering of English as a major world language. This objective had been present in the days of the Supplement to the OED, but had not been right at the forefront of our minds back then. Moving back in history to the earliest days of English ensured that we all had a new focus on our work.

  It was easy to be blinded by the superficial attractiveness of later French borrowings in English and not to pay proper attention to the Germanic bedrock of English. Our committee was largely peopled by Anglo-Saxonists and Germanists, so there was little chance we would be allowed to forget this, even if we had been inclined to. English is called a Germanic language because it was brought to the British Isles by Germanic invaders and settlers (northern European tribes) speaking Germanic dialects—from an area much larger than Germany today. After the departure of the Romans from Britain in the early fifth century AD, Britain was once again peopled for the most part by Celts, and in the far north by grumpy Picts.

  There are very few Celtic words in English nowadays. The OED offers just five from the earliest period of the language, Old English from before 1150 (brat, “a coarse or makeshift over-garment”; mind, “an Irish neck ornament”; bannock, “a northern round or oval home-made bread”; pen, “a hill, a height”; and conveth, “a Scottish land tax”). The better-known ones, such as clan, pet (a tame animal), and slogan, for instance, date from well after the first Germanic incursions. We used to think that this was because the German invaders obliterated all trace of the older Celtic language as they stormed through Britain. But in the absence of written records, it can be hard to interpret how the original Germanic invaders and the invaded Celts jostled along. Modern scholars tend to think that after the initial incursions, the Germans probably spent decades intermarrying with the Celts and learning from them how to tame the land. At the same time they effectively and probably almost unwittingly eradicated the Celtic language from large parts of its original domain, and Germanic English became the “prestige” language. Ironically, after the Viking invasions from the eighth century this new language underwent a comparable change itself—at least in the north and east of England (the area of the Danelaw)—as Scandinavians eventually settled and brought their own variety of Germanic to bear on the new Anglo-Saxons. And then again, after the Norman Conquest of 1066, the new Germanic English suffered a trough in fortunes and visibility similar to that which the Germanic invaders had inflicted upon the Celts. The Germanic Anglo-Saxon language almost disappeared from official records, subsisting principally in the mouths of the “ordinary” people outside the court and the aristocratic, legal, and religious elite—by whom, of course, the documentation was preserved.

  By 1500—as we enter the earliest period of what is now known as Modern English—the records show English reasserting itself as the symbol of an expanding, more self-confident nation. Even today, words of Germanic origin dominate large areas of our basic vocabulary: the number system; most of the short prepositions and conjunctions we use; the basic verbs be, can, do, have, may, must; the “strong” verbs which change their stem vowel in the past tense (swim/swam, ride/rode)—in fact, out of a list of the one hundred most frequently used words in English today, not far from 100 percent are of Germanic origin, and only one or two (such as nation, or according to some lists, people) are of post-Conquest French origin.

  The incursion of French after the Conquest brought English into immediate contact with the classical heritage of medieval Europe; French is often regarded as a more beautiful, sonorous influence than the guttural Germanic. The OED tells us that we have two and a half words of Romance origin in English today for every word of Germanic, but that doesn’t reveal the whole picture—those Germanic words are still often the big-hitting, high-frequency ones.

  The expansion of English has meant that it has gradually become a major language in many parts of the world which had, at some point in their history, come into extensive contact with English speakers, normally at the expense of the indigenous peoples (North America, Australia, New Zealand, South Africa, and numerous other places). But those original Germanic/Romance percentages have still generally held good into the modern, international era of English. In the new world in which such a large number of non-first-language speakers of English may be found, it is possible that speakers of Romance languages (French, Spanish, Italian) will naturally choose to use a higher proportion of English words of Romance origin when they speak English (speaking words with which they feel more comfortable). In the longer term, perhaps that will unsettle the old percentages.

  Some of the policy decisions made by the OED’s Advisory Committee confirmed areas where the OED had been doing the right thing all along: its historical principles, the necessity for evidence rather than subjective impressions, the grand scope and detailed implementation. It was in one of these meetings that we agreed that the OED should not omit any of its entries during its update, that it should not jettison entries which some people might consider no longer significant, perhaps because they were only attested once or twice, or in a very specialist discipline, or—worse still—only in poetry. We decided that once you started peeling layers away from the dictionary, you would never know where you were going to stop. So we stopped right at the beginning, and decided that nothing should be jettisoned if it was already on board our lexicographical Noah’s Ark—though we should continue to vet words very carefully before allowing them on board in the first place.

  There aren’t many things we like doing as a language group more than shortening words, to hurry us along in our fast-paced lives. We don’t have the time for all those syllables and, in the case of SMS messaging, even for most of those letters. The OED knows of about one hundred new words coined by “shortening” in the first decade of the twentieth century, including Aussie, auto, benny, demo, home ec, pneu, and post-op. That’s about the same number as during the previous decade. It seems that the number of new words created by shortening had progressed at roughly the same rate from Shakespeare’s time (arith. for “arithmetic”; cit for “citizen”) up to around 1800 (advert, mag for “magnitude” and “magpie”; van for “caravan”); then the rate started to increase in the early nineteenth century (perhaps as the pace of life hotted up), before hitting and maintaining a peak from around 1900 (ad for “advantage” or “addict”; path for “pathology”). The late nineteenth century was when this linguistic radicalism took a strong hold, and the mass of evidence for these shortenings gave them a place in the dictionary. Nowadays there are many forces producing abbreviations and initialisms: text-messaging, email, and websites (FAQ, LOL, etc.), technology (DVD, SIM, SMS), medicine, politics—and more or less any area which wants to present jargon as short and approachable.

  The original editors of the OED wouldn’t have known this meaning of this verb to vet—at least not before 1904. That’s when the first reference to the word occurs, in
the works of Rudyard Kipling (“These are our crowd. . . . They’ve been vetted, an’ we’re putting ’em through their paces”). It’s a good example of a word that rapidly developed numerous usages.

  The verb derives, as the historical documentation shows us, from the noun vet. And the noun vet is a shortening of veterinarian or veterinary (which are long words, and so of seventeenth- and eighteenth-century origin, respectively). The verb to vet originally had what you would imagine (correctly) was its literal meaning—to examine or treat a sick animal. 1891. It was a throwaway contraction, but the humorous British decided as early as 1898 that you might want to “vet” a person to see if there was anything wrong with them. And from there it was only a short hop to vetting things—examining them carefully for problems.

  These vet words are not related to veteran. The former derive from Latin veterina, “cattle,” and the latter from Latin vetus, “old.”

  There was one area where we did want to innovate extensively and to open wide the ground base upon which the dictionary was raised, and that was in the choice of texts from which we would find information about the language. We knew of old that the dictionary had favoured classic authors (at top of its list came William Shakespeare, Alexander Pope, Jonathan Swift, Walter Scott, and John Milton—not to the exclusion of all others, but enough to make some people think that the dictionary was based principally on “good” sources). This practice had skewed the analysis of the language, since not everything was written by literary giants. We planned to rewrite the policy and rebalance the typical sources from which we would seek evidence in future. We weren’t out of line with the spirit of the age. Since the nineteenth century there had been something of a proliferation in the publishing of more informal sources such as diaries, journals, and private letters. We were also well aware that there were many unpublished texts residing in archives which could yield a slightly different picture of the English language if they were studied properly from a linguistic perspective.