The Word Detective

by John Simpson

  Everything needed to be looked at again in the light of modern knowledge: definitions needed to be rethought; quotation sections needed to be updated to the present day; bibliographical details needed to be checked and standardised; entry structures needed to be reconsidered; references to outmoded systems of currency, geographical names, and so on needed to be fixed; some people whose birth years were given in old nineteenth-century OED etymologies had now died, and we needed to give them closure in their entries; label names needed to be reconsidered; more new—and especially old—words needed to be added. So it needed not just a tune-up, but a complete service. I don’t think we thought that many people would reach page 55 of the introduction, as introductions are normally completely overlooked by anyone who opens a dictionary. In fact, dictionaries must be one of the few types of books that are more or less opened at random—or where the user thinks, for example, that particular words might be found—rather than at page 1. But you miss things that way.

  By now we had a clear sense of where we wanted the dictionary to move, but we had no budget to fund this, and certainly not enough editorial staff to help bring it about. At the same time we were relieved to find that people in the real world were starting to take a greater interest in what we were doing. Scholars were fascinated by the possibilities of the OED on CD-ROM and the idea of an enormous online dictionary, constantly updatable, incorporating all of the documentary evidence on language that had been missed while the main OED remained in its mothballed cocoon throughout most of the twentieth century. The Second Edition of the OED on CD-ROM had become a reality in 1992 (superseding the beta-test version of only the First Edition, produced in 1989), after months in which editors were fully stretched, pushing and shoving the data into new tags to make searching more efficient. Researchers started to knock on our door, at first just in dribs and drabs, but then in numbers, which helped to feed our sense that the dictionary had a bright future.

  Dribs and drabs is an odd expression. Nowadays we don’t really use the two principal terms in other contexts. Do we say, “I’ll just put a drib of butter on the toast,” or “Clean up that drab of mud from the floor”? I don’t think so. These are effectively fossilised words, only used in this expression (like hue in hue and cry and beck in beck and call earlier). Still, there’s no harm in finding out what they used to mean.

  As far as we know, drib and drab were not drafted into English from abroad, but arose amongst the tussocks and streams of Britain. We can start with drib. The OED first passed its investigative eye over the word in 1897, when it found that its first occurrence in English was actually in Scots, in around 1730, where it meant a drop, or a small quantity. Casting far and wide for a possible etymology, the editors noticed that the verb to drib predated it—right back to 1523, but not characteristically from Scottish sources. The verb is said to derive ultimately from drip or drop, and here the OED comes up with another of its masterfully cryptic statements: “The modified consonant express[es] a modification of the notion.” So a drib is not quite the same as a drop, it is more of a drip or drop forming part of a dribble. (Unfortunately the dictionary does not contain illustrations.) To cover itself, it sums up by saying that drib itself may come straight from dribble. These words aren’t invented on paper, so you have to piece together the early history from later records, and you can’t expect exactness every time.

  Now we’ve tied up drib, we need to unpick drab. Obviously it means much the same as drib. The OED tells us that it found our drab from dialect records in 1828, in the sense of “a small sum of money.” That meaning seems to persist. It’s not a one-off contextual example.

  What we don’t know is how these two words drib and drab collided. The First Edition of the dictionary didn’t even include dribs and drabs. We added it very much later, in 1993—when we swept up lots of our interim unpublished material into two occasional OED Additions volumes. And then we found the expression in 1809 (so we’ll have to review that first dialect example for drab in 1828). What is starting to look likely is that dribs and drabs is one of those nonsense pairs where both elements sound similar but one or both make no real sense independently: hurdy-gurdy, hubble-bubble, higgledy-piggledy, riff-raff, bits and bobs, etc.

  One of the first of this wave of new academic researchers to knock on our door, an educator from Canada, John Willinsky, had obtained some seed money to start a pilot study on the OED—not by looking at page after page of tightly printed definitions, but by analysing the machine-readable data on which the dictionary was based. He was able to make use of the database held at the University of Waterloo in Canada since our work there in the 1980s, and with our assistance had come up with some interesting early stats and tables about the language and the dictionary, published in his book Empire of Words: The Reign of the OED (1994). He used this data to indicate which parts of the dictionary needed further specialist care and attention.

  For example, he examined the most-cited authors in Burchfield’s Supplement. Top of the league tables of authors cited here (as opposed to elsewhere in the dictionary) were George Bernard Shaw, Rudyard Kipling, James Joyce, and P. G. Wodehouse. Given the world of text available to the editors, this is generally (apart from Joyce) rather a conservative, conventional list. The first American author pops up at No. 6, in the form of Mark Twain. How about the top twenty women authors in the full (unrevised) Second Edition? Right at the top comes George Eliot, with 3,310 quotations (her top-cited work was Daniel Deronda, not the equally long but more popular Middlemarch); then Harriet Martineau, Elizabeth Barrett Browning, popular Victorian novelist Mary Braddon, OED contributor Charlotte Yonge, Fanny Burney, Jane Austen, and Mary Mitford. There was scope here, I think, for the new editors to search out other kinds of female discourse. In case you think these stats are all helpful, bear in mind another table, of titles given to authors in OED quotations: the league table omits “Mr.” (because—in the manner of the day in the nineteenth century—that was assumed), and runs from the most popular—“Sir,” down through “Mrs.,” “Bishop,” “Miss,” “Lord,” and “Lady.” But even this table is of interest, in that it highlights the assumptions and deference of nineteenth-century bibliographers and lexicographers (and of the reading public generally). This old-fashioned classificatory system has now been consigned to the dictionary’s editorial dustbin.

  This sort of information hadn’t been easily discoverable until now, though we had heard of a legendary Anglo-French scholar who spent much of his adult life working slowly through the print dictionary isolating all of the words in English which derived from Anglo-French. Nowadays that work can be done almost in an instant.

  The OED on CD opened up the dictionary to questions that people had not been able to ask before—simply because there had been no hope of receiving an answer: not just “What does this word mean?” but larger ones, like “How did the set of words first used in English in the early fifteenth century differ from those first encountered in the late eighteenth century, and what does this show about how language and society differed in these two periods?”

  When you consider this question, you see extensive differences between the concerns of language and society in the two time-periods. Back in the earlier period, late Middle English, we are still heavily dominated by post-Conquest French. The concerns of language users at that time (at least in the documentation that has survived) were with the details of everyday life in the Middle Ages: the day-to-day issues of feudal life, the concerns of the lord of the manor, activity in the law courts, religion, farming, food (especially amongst the upper layers of society), and, increasingly, shifting thought patterns towards modern analytical discourse. It’s arguable whether that last one was something everyone was aware of, and it’s perhaps an odd addition to the list—but the vocabulary of French (and its source, Latin) brought with it an intellectual culture that had largely not troubled the Germanic Anglo-Saxons up to this date, with words that we still use: abstraction, debater, negotiation, sentiment. Man
uscript texts—you have to remember—typically recorded the language of the successful classes of society.

  Here is a snapshot of some of those words first introduced into English in the early fifteenth century, mostly from post-Conquest French: some are religious (abbatial), some legal (abetting, abjure, above-written), some characteristic of the lives of feudal villeins (abasement, abject). If we move further into the alphabet, we just find more religion and conflict, pain, publishing, and pseudo-science: pacification, pact, paganism, palliative, palmistry, pamphlet. There are over 11,000 words I could choose from in this category. Don’t forget papelote = “porridge,” which is probably another welcome donation from Anglo-French. You might think that we didn’t necessarily need another word for porridge, as we already had a perfectly good one (i.e., porridge), but in 1450 we didn’t know about porridge, which only came on the scene in the sixteenth century, and even then only in the meaning “soup” or “broth.” Our oatmeal concoction was so named, according to the records, around 1640.

  That gives you a mini-picture or cameo of life on the ground in the late Middle Ages. If you now switch to the second set of words, from the late eighteenth century, you find an entirely different reality. English is under the influence of classical models; it has an international role, evidenced by vocabulary from across the world; its interests are far more “modern” than they were in the late medieval era. If we look at words starting with A entering the language in the early eighteenth century, we find long, Latinate terms—ways of describing things in a fairly complicated fashion: aberrating, abevacuation, abruptedly (rather than the simpler and earlier abruptly), absenteeship, accountability. The French influence is still there, but it’s overlaid with the Latin of the Renaissance and by a patina of learning. Music shows Italianate influence, as in a cappella (literally “in the style of the chapel,” i.e., unaccompanied).

  And if we head off again into the letter P, it’s not pacification, but the abrupt and perhaps redundant pacificity; not above-written, but the elegant and technical pagination. The arts of the Continent influence British fashions: paintbrush and palette knife. Instead of inventing the pamphlet, as we did in the Middle Ages, we now find pamphleteer as a new verb and pamphleteering as a new adjective. But those eighteenth-century types still enjoyed their simple entertainments, as attested by the staunchly non-classical able whackets, nicely described by the OED as “blows given on the palm of the hand with a twisted handkerchief.”

  Researchers were fascinated by the patterns and statistics they could find. Sometimes these patterns told you structural information about the English language; at others they just told you things about how the dictionary was compiled. As John Milton was one of the top-cited authors throughout the dictionary’s twenty volumes, did this translate into a towering role for him as a shaper of the language? There are dangers in ascribing lexical creativity to writers on the basis of their position in OED league tables, because their appearance there may not in fact be related exclusively to their influence on the language. This is something that I only worked through to slowly, and it surprised me, because I had been guilty of accepting stereotypes. Take Milton, for example. At the moment he ranks as the fourth most-cited author in the OED, with 12,400 quotations (Shakespeare is way out front with 33,000; close behind are other legends of the literary page of the likes of Chaucer, Dickens, and Tennyson). How did writers achieve such a high position on the dictionary’s leader board? Largely through their popularity as writers to the Victorian readers of the dictionary, since the vast majority of those references were in place in the First Edition of the dictionary. And how did they achieve this popularity? In different ways, but not just for their use of words: it was also because they had written outstanding plays, epic poetry, heartrending novels, etc. Arguably, the writers at the top of the OED’s league tables are there simply because they were the most-read authors, and therefore quotations from their works were avidly carded, and therefore readily available to OED editors when they came to compile entries, and so by default they float to the top of the dictionary’s statistics mill.

  Are authors who are cited first for specific words and meanings in the OED inherently more creative? This seems a promising line. Milton is shown, on this reasoning, to be the earliest reference for just over 2,000 terms (remember Shakespeare had around 8,000). But again there is a strong caveat. The original OED tended to “read” authors who had entered the literary canon at the expense of other authors who may have been more experimental but less finished and less popular. There is statistical evidence that canonical authors were over-invested with lexical creativity. James Joyce is cited as the first user of around 575 terms in the Second Edition of the OED, but the revision of the dictionary currently in progress has found earlier references for over 40 percent of these. The original data was flawed, or at least it was not intended to be used to make value judgements of this sort. It is not, of course, that the information to be derived from the dictionary is wrong; it is just that it needs to be interpreted sensitively, in context, and on its own terms—and not brashly used to create new, incorrect stereotypes. Milton is a towering figure of English literature, but not necessarily for his language. The case isn’t yet proven.

  As well as asking questions about authors, we could—as a matter of course—ask questions about words. Which words had the most subsenses, and did they develop rapidly, or over a long period of time? It’s often the shortest words that have the most complex history (can, do, set, make, pull, put, red, to, out). There are thirty-eight words which have forty or more main meanings in English: 66 percent of these entered English in the Anglo-Saxon period, and all of them are monosyllables in their basic form; of the remaining 34 percent, only the noun quarter is not monosyllabic. These Germanic words from the earliest period of English have had many years to shift meaning, enter colloquial phrases, and generally act as the glue for the language.

  How many words entered English from Russian, and why? As of today, the OED says 404, peaking in the years following the Russian Revolution and going into the Cold War. (Again, this is a good question, and one which can tell us much about cultural contacts and mutual respect—or even distrust—between two language groups over history.)

  Are there more adjectives than nouns in English? The dictionary thinks there are twice as many nouns as adjectives. This means that there are more things named (nouns)—whether abstract or concrete—than there are words to describe them. Some adjectives are based on earlier nouns anyway, and so often depend for their existence upon pre-existent nouns. But figures suggest that overall there are about two nouns for every adjective, and yet in the seventeenth century it seems that the figure is reduced to 1.5 new nouns to every new adjective. Why? In the nineteenth century, the ratio is higher again: 1.8 nouns to each adjective. These questions are as yet unanswered. Statistics are always right, but we may sometimes be asking the wrong questions of them. Was the data faulty for the seventeenth century, or was this a period when derivative adjectives were particularly likely to be formed, or was there another reason for the fall-off in the introduction of new nouns?

  As we entered the early 1990s, we embarked upon what I regarded at the time—and also in retrospect—as a dark phase of the project, which lasted perhaps for the first half of the decade. The Second Edition of the dictionary had been published to acclaim, but we had been working with our sights set so closely on this goal that we had completely overlooked the need to plan for—rather than just to expect—a future involving the comprehensive update to which computerisation was only the prelude. And my dark mood paralleled something of a dark period at home, as Hilary and I gradually realised that our younger daughter, Ellie, had severe developmental problems of the sort that no amount of funds, effort, care, support, or love would overcome. Not everything can have a happy ending, but at this point we were deep in a tunnel with no sign of light ahead.

  NINE

  Gxddbov Xxkxzt Pg Ifmk

  By 1993, the dictio
nary had outgrown its palatial surroundings in St Giles’. We moved back on to the main University Press site into two long and uninviting corridors in a characterless modern building which only existed as a passageway between two of the more historic parts of the Press. These gloomy corridors had been abandoned by another group that had gone on to more satisfactory premises. As usual, the office mirrored the sort of work we were doing, which was now linking, preparatory, in transit.

  By now, our old friend and supporter the Shark had left the University Press to become a director of a major London publisher, so we needed to activate our remaining Oxford friends to convince the assembled dons of the University and the executives of the University Press that updating the dictionary was the right and only thing to do. They would then, we assumed, loosen the institutional purse strings for us. Our primary editorial objective was to develop our new policy and then to test-edit a range of dictionary entries, updating them according to the tenets of our new vision. The University Press would then, we assumed, enter some sort of internal conclave, and with luck decide to give the enormous update project the go-ahead.

  With prompting from two or three of the dictionary’s friends—senior professors in the University—the Delegates of the University Press, that body of professors ultimately responsible for approving the Press’s academic publishing policy, kindly enquired of us how we were getting along, and in due course our business director, the Admiral, was formally asked to develop a project plan. The University Press was bending our way, mindful of the expectations it had raised that we would be able to work on the big dictionary once the Shorter had been wrapped up. After the Delegates met one long day in 1993, we knew that a project plan could be sketched out. It was politely intimated to us that a satisfactory project would involve a completely revised and updated OED by the year 2000, and that a budget to show that this was feasible would be even more satisfactory.