The Word Detective

by John Simpson

  Phrases live and die through the amount of use they enjoy. In order to survive, they often need to leap from a small world into the big one. But they don’t always jump continents. By a long chalk (“by far,” “by a long way”) is an expression first recorded in the 1840s, just into the Victorian era. It is commonest in British English, with less evidence—for example—from Australia and New Zealand, and less still from America.

  The phrase comes originally from the small world of bar-room games. If you were engaged in a long drinking session in a public house in Britain in the sixteenth century, the landlord might chalk up on a slate just how much you owed. In the seventeenth century, people found it useful to use chalk to keep the score in games (often also enjoyed in alehouses)—and each point you scored would be represented by a chalk mark. If you ran rings round your opponent in the game, then you’d win by a larger margin, or a “long chalk.” So they say.

  The lexicographical arm of the OUP in the United States was entrusted with updating the North American component of the OED. Jesse would broker contacts with many of America’s most successful word-hunters, such as Fred Shapiro of the Yale Law Library, and we’d suddenly find early uses for jazz terms, film terms, and (as I had come to expect) baseball terms flooding into our electronic files and hence online for all the world to view. When we revised the entry for jazz we suddenly found that the network of American jazz historians had been encouraged by Jesse to let the OED into their secrets. The 1910s were a crucial time for the development of the word jazz (1912—energy or excitement; 1913—misleading talk, nonsense; 1915—the new popular music style; 1919—dance styles associated with this music; 1918—sexual intercourse). We discovered that the Supplement had made another mistake, claiming 1909 as the first recorded use of the word jazz to describe a type of music; the source should have been dated 1919 (a recut version of Cal Stewart’s Uncle Josh in Society, which had originally been recorded without the word jazz in 1909).

  Dictionaries and their editors are often not considered prime-time television material, so we were in for a surprise in 2005 when an upbeat London TV production company, Takeaway Media, was commissioned by the B.B.C. to create a television programme based around a national word hunt. The idea was to help the OED record the language better by tapping into the everyday experiences and reminiscences of the television-watching public.

  The plan was for us to come up with a set of words—maybe fifty would do—for which the public would be asked to hunt out earlier evidence from the piles of special-interest magazines generally considered to be piled up in their houses. So we dutifully created such a list. We suggested a group of fairly serious words for which we thought a public appeal might help—biodiversity, mega-rich, playlisting, that sort of thing. Then we were told that the list had to contain the sort of words people might be interested in and might be able to discover hidden away on their bookshelves. So we went back to square one.

  The sort of expression people might be interested in, apparently, was the full monty. It turned out that they would also be interested in cocktail and codswallop and boffin. In general, they would be interested in the history of curious, but everyday, informal words and phrases—nothing too serious, and only terms where there was a fighting chance that the public might turn up gold. That was fine by us—any new information would be useful to the dictionary, whatever sort of word it related to. So we pulled together a search list and launched it on the production company, which then launched it on the national media. People saw the list for the new TV programme (called—rather dismissively and against my lexical intuition—Balderdash & Piffle) in their newspapers, or on the radio, or on the programme’s website, and they rushed off in their droves to try to help, and then began emailing or posting their discoveries to the programme’s producers.

  In principle I object to “quaint” words, but balderdash intrigues because of what it used to mean, its possible correspondences with Scandinavian dialect, and the leap from its original meanings to the only one we know today. It is a word that defies etymology, at least in the eyes of the cautious OED. But the dictionary does draw our attention to the dialect word balder, which means “to use coarse language,” as well as to various similarly aggressive Scandinavian words.

  The problem in English is that the earliest documentary evidence insists that balderdash was a drink. In the sixteenth century it apparently meant “froth, or foamy liquid” and, as things hotted up in the seventeenth century, “a jumbled mix of liquors,” which might mean milk and beer, beer and wine, or brandy and mineral water. Only by the late seventeenth century do we start to find evidence for the modern sense of “a senseless jumble of words,” or “nonsense, trash, spoken or written.” The feeling of the nineteenth-century editors was that really this sense might turn out to be the original one, but they didn’t quite have the nerve to assert this. In cases such as this, we need to await the OED’s update, to see if the tectonic environment of balderdash has changed.

  The B.B.C.’s word hunt started to take shape. All of the programmes for the series were recorded on the same day, which gave us no chance to develop in the role. And we were given no advance notice of what relevant new information the public had turned up. The show worked like this: There were three OED lexicographers (Peter Gilliver, Tania Styles, and me) sitting behind the inquisition table, caped in immense learning. The programme’s presenter and the leader of the word hunt, Victoria Coren Mitchell (daughter of the languid, quick-witted humourist and writer Alan Coren), would sweep down upon us from her eyrie to ask us whether we would accept her word-hunt searchers’ latest findings, and if we would promise to put them straight into the OED as the earliest evidence ever discovered for the word we were reviewing.

  She was so forceful and decisive that we really should have collapsed and just put anything she asked into the dictionary, but deep down we had our principles. So when she told us yet another shaggy-dog story about the origin of shaggy-dog story, we were all (after learned discussion of its merits) prepared to throw it out the window, until we realised that she did have a point: the evidence she had in her hand was indeed earlier than anything we had in the dictionary, and someone had trumped us (or at least trumped the editors who last worked on that entry twenty-five years earlier). On some occasions we would dispute the evidence hotly (and we were pretty conservative—we couldn’t just accept anything), but at other times we grudgingly accepted it, and doubtless made someone very happy.

  As far as we knew before the programme, the expression shaggy-dog story, a long-winded rigmarole tale, amusingly pointless at best, dated from 1946 and an anthology enticingly called A Collection of Shaggy Dog Stories, published by a small company based in Hunstanton, on the Norfolk coast. It is inherently unlikely (though not entirely impossible) that the expression first saw the light of day in a seaside town in East Anglia with no apparent links to shaggy dogs, but the Supplement to the OED had no way of improving this data. No one really knows what the first-ever shaggy-dog story was. People make up stories about what it might have been, through some impulse to complete the picture. They say there was an archetypal rambling shaggy-dog story in which the dog’s coat grows longer and shaggier, until at a pitch of excitement someone observes that the dog isn’t shaggy at all. Fortunately the OED isn’t too concerned with speculation of this order.

  After the programme, we had better information, thanks to the avid watchers and word-hunters. Someone had access, at home or online, to a copy of Esquire magazine from May 1937. Esquire here offers helpful advice in techniques for separating sheep from goats—as you might say, or the sharp- from the dull-witted: “One of the more sporting ways of finding out which ones are not [sane] is to try shaggy-dog stories on them.” When presented with this information, and a verified photocopy of the original, we had to concede, and after the filming we rushed back to our offices to add the new information to the dictionary.

  It turned out that a few of the word-hunters had been cunning (or you might ca
ll it using their common sense) and had used big text databases to find better evidence than we had previously discovered. But that was fine. In the end, the dictionary benefited from the series, and the additional national publicity fed further interest in the dictionary. Once it was finished, the B.B.C. commissioned another series, and so more new findings trickled into the dictionary. Funding for entertaining cultural programmes then suddenly took one of its cyclical dips (or the producers realised that television wasn’t the dictionary’s natural milieu), so the second series turned out to be our last.

  It was in fact a refreshing change to find people who were not simply interested in the new words that the OED was adding. Most of the Balderdash & Piffle words came from the twentieth century, and they were predominantly informal words or expressions, but they weren’t froth. They were words, for the most part, that had seeped into everyone’s vocabulary without being noticed, but they had left little unresolved issues for lexicographers which needed to be cleared up. Two of my favourite finds from the series were nit nurse and pass the parcel. I’d known one of these when I was growing up in the 1950s, but not the other.

  Fortunately the one I knew was the game, pass the parcel, which in its original Spartan form was a staple of every birthday party I had attended when at primary school. For those too young to remember, a prize is wrapped in multiple layers of gift-wrap, the music starts, the parcel is passed round the circle of partygoers, the music stops, and the lucky holder opens a layer of wrapping. Mostly there is no prize there, but eventually the music stops with only one layer of wrapping left, and the lucky holder gains the prize. In its later form it involved unwrapping a prize from the parcel in every round of the game, but in my day you just sat through five minutes of passing the dreary parcel around and around the circle of your fellow partygoers, with occasional halts when the music stopped—until eventually someone else won the prize.

  But the OED had clearly not been playing enough party games: it had only discovered evidence for the expression from 1980. That was very late, so we put it on our appeal list for Balderdash & Piffle. Sure enough, as the weeks passed, new evidence arrived at mission control, and when we came to film the series, we had a description of the game (which talked about “passing the parcel”) from the aptly named Foulsham’s Fun Book of 1932, plus a real, concrete example for 1953. The history of the expression now extended back at least as far as my memory allowed.

  My other favourite find was nit nurse, an informal expression for the school nurse who would comb through the itchy heads of wartime children looking for head lice. We had often found this sort of informal children’s term hard to pin down. It was as elusive as the lice themselves, and we had only discovered written examples from 1985. That was slightly worrying, as the expression had already drifted into childhood mythology by then: our evidence from 1985 actually read, “Whatever happened to the nit nurse?” So we threw it open to the word-hunters, and this one attracted the research community.

  Someone with online access to the top UK medical journal, The Lancet, instantly took it right back to 1942 for us. All at once it had been demonstrated that OED research doesn’t find everything, that words arise from the social environment that demands them, that there are always people out there who know more about particular terms than the OED editors, and that computers can save a lot of legwork (though you shouldn’t necessarily believe everything they say). It was also a reminder that The Lancet database was still too expensive for simple publishing companies such as Oxford University Press to subscribe to.

  This increased publicity and interaction with the media brought another issue to the fore for me: the popular misconception that the chief editor of the OED doesn’t edit, but spends the day organising things so that other editors can get on and edit. That would be easy to do. There were always meetings I could have attended to discuss organisational or publishing matters. But I wondered whether there was a point in being editor of the OED if you didn’t spend the day editing. By continuing to edit, I showed the seventy or so other editors on the project that I thought it was the most important thing for me—and hence for them—to be doing. Editing is also (fortunately) what I enjoy doing most. Maybe I could have become a successful commercial publisher (well, you know me—I sort of doubt it), but I’ve always prided myself on the fact that I can edit, and edit pretty fast. It is the OED, after all. It doesn’t deserve to have people using it to further their own ends. There are always enough people around who want to do that. But someone has to stand up for the dictionary. The editors don’t want to be run by ciphers who creep from meeting to meeting discussing budgets and office moves. I always felt it was too important for that. Call me soft, but I did what I thought the chief editor should do, and that was lead the editing.

  Even though the dictionary was now living happily online, there was still one general problem that I had with it: it still looked like a book (or at least a series of books) on computer. There was nothing wrong with that, and many people just wanted to look up a word and then move on with their lives. However, there are many ways of looking at data, and the book view is only one. Those of you with a predilection for classifying things, and I’m including all lexicographers in this catch-all group, will know that putting index cards in piles according to word meaning is only the bottom rung of knowledge. With all of the complex software behind the digital OED, we knew that there were new ways of visualising the language on-screen that could make the dictionary (and the language) come alive to a far wider set of people.

  From the days, back in the early 1990s, when the OED had first appeared in its entirety on CD, we had been encouraging scholars not just to look up the meaning and history of individual words, but also to start exploring the language on a broader scale. We did this by suggesting that they take a look—as they then could—at comparative data, to search it for patterns and differences. This was a brave attempt to draw people away from using dictionaries simply to look up definitions, and to encourage them to glimpse all of that extra information hidden in the big, historical OED. I wanted people to be able to see charts, diagrams, animations, and videos that would give them further clues about language history, and also to be able to collaborate much more closely with editors, perhaps through the medium of wikis.

  Clue is unusual in that it is a word that nicely knits together the Germanic origins of English and the fascination we have had—even from the days before the Renaissance—with classical mythology. The older spelling of clue is clew. The OED reminds us of other similar pairs in which the -ew spelling is the older one: blew and then blue; glew and glue, rew and rue, and trew and true.

  We first encounter clew in Anglo-Saxon times, when it was spelled kliwen or klewen (with a final -n) and had widespread equivalents around Germanic Europe. It meant “a ball” (“a round ball”), and often a “ball of thread,” so it was a domestic, homely sort of a word. As Geoffrey Chaucer and his contemporaries knew, various figures from classical mythology had used balls of thread to get them out of scrapes, and the most famous was probably Theseus, when on his Minotaur-slaying quest. Ariadne had given him a ball of thread (a clew, to the medieval mind), which he fed out gradually when entering the Minotaur’s Labyrinth, so that he could follow it back when he returned—at speed if need be, with the Minotaur at his heels. So strong was the connection between these myths and the purpose of the balls of thread that we formed the idea that anything that gave you a hint or a clue as to how to extricate yourself from a mazelike or perplexing situation was itself a clew, or—as had become standard by the seventeenth century—a clue. Although I don’t like crosswords, I like the origin of clue and its labyrinthine complexity.

  Visualisations and animations had been on our to-do list for a number of years as we approached the year 2010, but we had little hope that we would find the resources to experiment in this area. With my realistic hat on, I can see that most people will always want to search for basic definitions, pronunciations, and etymologies. But we st
ill hoped that we could increase the percentage of users who would be interested in carrying out more creative, global searches of the data. We had become quite familiar with seeing trends of, say, economic data predigested by way of graphs and other visualisations in the newspapers or on the TV news. Why couldn’t the OED offer similarly inviting digests of language data?

  We sensed that there was a latent appetite for this new perspective. We also knew that the way the OED appeared online when it was first published back in the year 2000 was not helping. However much we stressed that the text was a brand-new computer-searchable animal, it still just returned to you lists of words. But all around us people in their everyday lives were positioning themselves within a world of fluid, dynamic, visual data.

  Everyone, not just the lexicographers, could see that the online dictionary needed a new look and a new feel: one that empowered the user more actively. Those with more of an idea of these things had cottoned on to the concept of filtering data. You looked up one thing, and you could then throw a second search at the data which refined and filtered your results until you had drilled down, in the technical jargon, to what you were looking for. Suppose you searched for all of the words in the category Visual Arts (2,768). Then, having found those, you decide you want to filter out all of the ones that didn’t originate in the eighteenth century: you’d be left with 209. Then you might want to drill down further to discover how many of these words were of French origin. That would leave you with 29 of your original 2,768 Visual Arts terms, including costume (first appearing later than you might expect, in 1715); fine arts, from 1767 (and from French, inasmuch as it is a translation of the French term beaux-arts); and also statuette (1738). This methodology gives you the ability to decide what you want next as you go along, rather than having to predetermine where you want your searches to take you.