by John Simpson
One or two people may ask about the word’s origin. In case you are one of them, I should say that the verb to hone derives directly, without deviation, from the noun hone. The noun hone—back in the Anglo-Saxon, Old English period—meant “a pointed or overhanging rock,” and especially one that marked a boundary. They rather liked stones which acted as boundaries in those distant early days of the language, as one of the main activities in Anglo-Saxon villages was checking where your boundaries were and keeping other Anglo-Saxons (and sometimes Celts or Picts or Vikings) on the other side. But for us, the relevant meaning arose in the fifteenth century (so just back into the Middle Ages), when a hone started to mean a “whetstone,” on which you might sharpen your razor of a morning. It’s surprising that we had to wait so long, from the fifteenth to the late eighteenth century, for someone to find a use for hone as a verb. But nowadays, as I believe I’ve said several times already, we use it too much.
Once we had read through all the application letters and scanned through all the handwritten and printed CVs, and once we had filtered out all the honers, holders of clean driving licences, lonely cinematophiles, and ramblers, then we had to boil the remaining applicants down to around thirty by considering hard whether their experience, personality, and life skills warranted our moving them through to the next round. Those who satisfied most of these requirements were sent whichever version of the aptitude test we were currently employing.
The whole point of the aptitude test wasn’t to ensure that the candidates failed, but to give them some idea about what work on the dictionary might be like. We started from the assumption that, whatever they thought, they really had no idea what working on a historical dictionary meant. We also started from the assumption that about a third of the people to whom we sent the test would disappear off our radar into the ether, laughing raucously to themselves at the stupidity of anyone wanting to do that sort of thing for a living. That’s okay. They’ll find a niche somewhere.
It can be instructive to watch the struggle for precedency of one pronunciation over another through the years. When we first adopted the word balcony from Italian in the early seventeenth century, we didn’t immediately pronounce it BAL-co-ny; we adopted the Italian style of pronunciation, with the stress on the second syllable: bal-CO-ny. The OED says that the modern pronunciation was winning through by 1825, but it adds that the English poet Samuel Rogers (1763–1855) said that this newfangled way of saying the word “makes me sick.”
Do you pronounce niche as /nitch/ or /neesh/? If you say /nitch/, then you are probably old-fashioned and proud of it. Most people pronounce it roughly /neesh/ these days. I suppose I do now, too, but I wouldn’t have fifty years ago. The word derives from French, but developed its old-fashioned pronunciation in English quite early on. English speakers have used the word, according to the OED, since around 1610—so it’s a late French borrowing influenced by Renaissance architecture and design. With such a strong cultural force behind it, it found its way into most of the European languages—even German (Nische) and Russian (niša). The First Edition of the OED (the letter N was published in 1907) gave /nitch/ as the only pronunciation option. Even up to 1977, most of the big pronouncing dictionaries gave /nitch/ as the standard and /neesh/ as a newcomer. Now, they have mostly decided that /neesh/ is the dominant form, and that /nitch/ (sadly) has had its day.
It is clearly pertinent to ask why these pronunciations change. And there are clearly different reasons. We began by using an approximation to the Italian pronunciation of balcony, but clearly came to feel that it was not consistent with English pronunciation norms. The derived form balconied (stressed as an English word on the first syllable) may have helped shift the accentuation of balcony, as might the more familiar word falcon. In general, there has been a tendency in English for stress sometimes to move forwards towards the front of words over time.
It is different with niche. The older pronunciation may have been influenced in the first place by late Latin and Italian use, or users of the word might have wished to make it sound more Italianate and rather snobbishly adopted this /nitch/ variety. But there are other possibilities, too. As usual, there is no one answer to questions such as these.
Niche wasn’t a word you needed very often in the old days, unless you lived in a stately pile or spent your summers in Italy with friends who owned villas. This eventuality may lend weight to the snobbishness argument. A niche is a hollow recess in a wall where you might place a small statue. In English it has been a busy word. New meanings have been recorded in these years (divided by century): 1662 / 1725, 1733, 1749, 1756 / 1822 / 1911, 1913, 1963—so it’s been good to lexicographers. I’ve separated the meaning shifts by century so that it is easy to see that the period of greatest activity was in the eighteenth century; and if you divide the centuries further, by quarters, you can see that the busiest period for the word was the second quarter of that century. That makes sense. That was a period of great classical influence amongst the style-blazers in Britain, so a word for an alcove in which to display your classical statue probably would have had a reasonable profile. But people were also interested in how individuals fitted into a society that was rapidly developing a class system—where their own position (or niche) in life was along the wall of society. That’s the 1733 date, filled by Jonathan Swift (Epistle to a Lady: “If I can but fill my Nitch, I attempt no higher Pitch”). Also in the eighteenth century it came to mean a hole or lair for an animal (1725), or a natural recess in a rock or a hill (remember the celebrated faux-naturalistic landscape gardener, Lancelot “Capability” Brown?) (1756). We had to wait for 1913 for niche to develop into the place of an organism within an ecosystem, and as late as 1963 for consumerism to bring us the niche market.
Whatever home that one-third found who didn’t attempt our aptitude test, it was the two-thirds who did have a go at it in whom we were now interested. First of all, they will have been surprised at how low-level and data-oriented the test was. They didn’t need to know complex linguistic formulas, but how to analyse data sensibly. So we gave them a set of, say, thirty sentences, all containing one keyword. And we gave them a copy of the OED’s entry for that word. The task was to tell us which subsense of the OED’s long entry each quotation fitted into. On one occasion we used the verb to list. The verb has had about ten different meanings in English.
As with most things, it’s easier to understand if you have a go at it yourself. So here is a quotation from Samuel Pepys’s diary of 1668: “The persons therein concerned to be listed of this or that Church.” You are not given any extra context, though you might check online to see if that helps, or go to your local library to extract a copy of Pepys’s diaries. Now, you have to read that quotation carefully, and decide under which of the existing dictionary definitions it fits. Here are three possible ones (all abbreviated, and you would normally have around ten to select from):
(1) to set down together in a list;
(2) to include or enrol in the number or membership of;
(3) to enter on the list of a military body.
Read the quotation and the definitions carefully, and trust your instincts (just this once), or you will never finish. Samuel Pepys is not talking about enlisting in the army, so you can discount No. 3. Nos. 1 and 2 are more tricky. What do you think? Pepys means that his “persons” are put in a list, but that could be either No. 1 or No. 2. So what distinguishes these two definitions? Pepys is talking about people listed in separate lists (“of this or that Church”), so they are not being “set down in a list together” (No. 1), which the dictionary exemplifies (amongst other quotations) with: “About one hundred species of butterflies have been listed.” Here the butterflies are all written out in the same list. But Pepys is talking about people being listed in the membership rolls of different churches, so it’s No. 2. I hope you followed that. This sort of puzzle is one of the reasons I enjoy reading sixteenth- and seventeenth-century text.
Mostly the candidates got thes
e questions right. Sometimes we would throw in an example of another verb to list (such as “to listen,” or “to please or delight,” or “to enlist”). Sometimes we’d throw in a noun, to see if anyone was concentrating. Sometimes we’d include quotations that might be attached to two or three different subsenses, depending on how you read them. There were other tricks, too.
Once they were starting to feel that everything was simple, we’d chuck them in the deep end with question two: a copy of a full dictionary entry, as compiled in, say, 1904, and not yet updated. And we’d just ask them the simple question: What would you do to update this in line with modern-day usage and scholarship? Their hearts should have shot through their mouths then. This was the big one (at least as far as I was concerned). We were giving them just enough rope to hang themselves, and we’d sit back and watch. Obviously some people had no idea what to do—despite having honed their skills for this precise day in their lives—and then wrote a short paragraph of waffle, before meandering on to the next question. Others set about it with gusto. It was all too easy to unseat yourself by sticking to generalities. What we needed were concrete ideas on updating the vocabulary of definitions, ideas for how to find new documentary evidence (quotations), and most of all some sense of the overall structure that the entry should have in its revised form.
Lexicography is pretty sharp-edged. There’s no place for wobbly or brittle thinking. You see a problem and leap in to solve it; you don’t wallow in it, indulging yourself in the beauties of the language. It’s necessary to compare the usage you are addressing with hundreds of other examples from the same semantic area, to see what is special about your use. Or if you are trying to write or update a definition, you assume all of your source material is wrong until it proves itself not to be. You need a scientist’s sense of distrust and a writer’s sense of elegance.
People often think lexicography is easy: it’s not. It involves qualities most people don’t have. Stamina, for one. There are times when I think that’s the most important quality for a lexicographer. Mental, and also physical. Mental is obvious; physical—well, it’s not factory work, but it wasn’t static desk work. All the time you were up and down stairs, carting huge books from shelves on to tables and back again, lugging open heavy filing cabinets that should have been power-assisted but weren’t—relentlessly, because you had to hunt out that last piece of evidence. There is less of that to do now that the basic work is on computer, so maybe the profile of the lexicographer will change as time passes. Lexicography is a long haul, and you need to stay with it. It’s not something you can jump in and out of, or you will never build up the necessary expertise.
And other people think that lexicography is hard: it’s not. It’s easy—if you remember to keep things simple, that normally things happen as you might expect them to, and that you just need to follow the data and not try to impose grand schemes on it yourself.
From the test we normally settled on about five or six applicants to interview. On the big day, two or three of us would be gathered together in one of the dictionary offices awaiting the arrival of the future generation of OED editors. We knew how the candidates had done in the aptitude test. But that wasn’t everything—it was essential that people could fit in (fairly) easily with the other editorial staff that we already had. We wanted people to be very ordinary and quite extraordinary at the same time, and sometimes we got it right.
In general the applicants were just leaving university—either after a first degree or a higher degree. We didn’t require doctorates, as we’d heard was the case in American dictionary houses. What we wanted was promise and competence. We hardly ever appointed people trained in linguistics, as we found they tended to come to us with theories of their own, and wanted to work from the top down—whereas we lived in a world of data, trying to piece the jigsaw together from within, or to dig ourselves up towards the light.
Once someone had reached the interview stage, they had a pretty good chance of being appointed. They were the cream of the crop. Obviously we couldn’t use all of them, but they were all on short odds.
What sort of people were we looking for—what sort of temperaments and personalities? We wanted people who could listen, observe, and not necessarily just sound off—as some of them appeared to have been encouraged to do at university. The job is to hear language, to absorb it, think about it, and analyse and classify it. That’s not really the sort of thing you can do if you are talking all the time. Let me mark myself on these criteria, to see how I’d fit in. Listening, not talking (nine out of ten). Sometimes I don’t think I expend more than two hundred words a day, but I’m listening hard and monitoring what I hear.
We need to ensure that the candidate will fit into a team. There’s a place for loners, but they make things more difficult for everyone else. It goes without saying that you can’t tell everything about someone’s social adeptness at an interview, but you have a chance to gauge it. I think I scored really low here when I started—maybe a three. You’d have had to see past that to employ me. But I just needed a year or so to integrate. I’m up with the eights by now.
Are you a rambler or a finisher? Do you prefer to research for research’s sake, or as a means to an end? You’d be surprised how many of the ramblers there are. It is rather fun sniffing them out. You ask a leading question and see how long it takes them to bring their answer to a definite conclusion. And sometimes you add to that how long they have been working on an unfinished doctorate. The higher the aggregate climbs, the less likely they are to land the job. I would have done reasonably well here—maybe six or seven. I wouldn’t have rambled, but I may have dried up prematurely if I didn’t feel I knew what I was talking about.
I’ve already mentioned stamina. Lexicography is slow and involved; the excitement comes as the fuse burns slowly towards the answer. But it’s not always climactic. So can the candidate demonstrate any achievement that would not qualify as gnat-like in its longevity? Concentration. Think about that one, and then think about it again. But don’t waste time thinking; it has to be productive. We need slow-burning but explosive bursts of concentration. It may not be possible, but perhaps that gives you the idea. I get a nine for that.
What about the aptitude test they have all done? Well, we need to bring them back to that too. We need to see how they take criticism, because sure enough (if my experience was anything to go by), you’ll come in for plenty of that before you’re done. Do they sulk? People don’t usually sulk at interview, but they can go into themselves and seem happier avoiding the issues. Can they dig themselves out of any holes they jumped into during the test? Are they ready to change their mind? If you spot a weakness, you have to try to explore it—bring it out into the open. Eventually even your fellow interviewers will see the point.
From time to time we wouldn’t find any suitable candidates after two days of interviewing. If that was the case, it was best not to appoint second-best, as that would only cause problems down the line. What was exciting was when you could see glimpses of the strange potential a candidate needed to become a lexicographer. It could come when you didn’t expect it, but it always left us on an up when the interview ended. Sometimes we thought we had a fantastic candidate, based on the aptitude test and the CV, but with some deft questioning things started to fall apart. At others, borderline interviewees pulled it round from a low base. You couldn’t second-guess who would finally make it. They might come from any discipline, any level (bachelor’s to postdoc), any university or college. And any of them might be the next chief editor in fifteen or twenty years.
There are good things about dictionary work and there are bad things about dictionary work, and the higher up you go the more bad things there are. The best thing was the work—detective, creative, important, as we liked to think. The bad things involved spreadsheets and budgets and frowns from our senior colleagues, who thought we were overcomplicating a problem that a couple of sparrows could sort out while chatting at a bird-table.
/> One day in 1993, Ed and I had a short conversation. Our little discussion went along the lines of:
ED: This will surprise you, but I’m more interested in writing the dictionary than managing its composition.
JOHN: Okay.
The end result of this brief but considered exchange of views was that rather than remaining co-editors of the dictionary, I bravely took over as chief editor, with Ed as my deputy. I heard that some people thought I was now at the top of the crazy tree; the expression “poisoned chalice” was heard in the dictionary halls, as we were still battling against a completion date of 2000. But I fondly thought I could fight the dictionary’s corner whenever the necessity presented itself, and Ed intended—sensibly—to nestle down to the excitement of editorial discovery and creation. As it happened, I planned to be there, too, and to spend only a little while on other matters. We did, after all, have the Admiral, whose job it was to steer the ship round any rocks that reared their ugly masses.