The Oxford English Dictionary and its chief word detective
Oxford English Dictionary Chief Editor John Simpson is to retire after 37 years at the famous reference work. Here he writes of a life hunting for the evidence behind the birth of words.
Historical dictionaries are not just about definitions.
Every word or phrase has a story, and the historical lexicographer has to tease this story out from whatever documentation can be found. That is one of the pleasures of working on the Oxford English Dictionary.
An enduring myth is that the word pom (as in whinging pom and other more colourful expressions) is an acronym from either "Prisoner of His Majesty" or even "Permit of Migration", for the original convicts or settlers who sailed from Britain to Australia.
The first recorded use of pom comes from 1912, which is quite - but not unnaturally - early for an acronym.
There is no historical documentation to support these myths (rather like the disproved theory that posh derives from tickets for the upmarket cabins on the old P&O liners - port out, starboard home). Instead the etymology is apparently more circuitous.
We start with the word immigrant, well-established by the mid 19th Century as a settler. In a joking way people would play with immigrant from around 1850 or so, turning it into a proper name (Jimmy Grant), to give the strange immigrants a pseudo-personality.
Equally playfully, a Jimmy Grant morphed around 1912 into pomegranate and immediately into pom, which it has stuck as till today.
Etaoin shrdlu is an expression well known to newspaper compositors and little-known to readers.
It comes from the same stable as Anthony Burgess's Homage to Qwert Yuiop. Qwertyuiop is what you find on a computer screen (or, in the old days, on a typewriter) if you run your fingers along the top row of letters on a keyboard.
Etaoin shrdlu is the equivalent sequence of letters that an old-style Linotype printing machine operator would have put out by running his (or her) finger down the first two (leftmost) columns of Linotype keys.
But etaoin shrdlu had a purpose. The Linotype operator would hit these keys intentionally to signal that an error had been made and the preceding line should be removed from the type before it was printed.
Sometimes the type-setters and proofreaders were asleep and missed this alarm bell.
The OED finds its first reference to this practice in 1931, but Linotype machines were used for many years before this.
When we come to revise the entry, we will be able to make use of the enormous online databases of newspaper text, not for the purpose they were principally established, but to hunt out evidence of compositors hitting the etaoin keys.
At present we can easily find examples of this practice back to the late 19th Century. On 17 April 1899, the Dublin Freeman's Journal inadvertently published this segment:
Mr McCall was called on to preside pending
The election of chairman.
'muwG R, * etaoin shrdlu cmfwypvbgkqjcmff
Garbled type (and not a line which Microsoft Word's spellcheck function feels particularly easy with), because the compositor went wrong. And they missed the line when they printed the paper that day.
Not English? The dictionary helps you to understand text, and if you encounter etaoin in a document you need to know why it is there.
New Model Army
Everyone knows that Cromwell's New Model Army defeated the Royalist forces in the English Civil War. The army was established in 1645 and disbanded with the restoration of Charles II, to whom doubtless it provided unwelcome memories.
So the historical lexicographer will of course be able to find evidence for the expression New Model Army all the way back to 1645 or thereabouts. But if you type it into any database of 17th Century text (or 18th Century text for that matter) you will be wasting your time.
You will not find any reference until much later. The OED dates the phrase New Model Army to 1845, exactly two hundred years after the army was set up, in the works of the historian Thomas Carlyle. So it's a term we apply historically to Cromwell's force.
Now Cromwell and his fellows must have called it something, and it turns out (from a trawl of the available records) that they called it the New Modelled Army - from at least February 1646. They talked about an army established on a "new model", but they did not (as we would nowadays) simply add "army" to "new model", but preferred in the idiom of the day to say it was an army "new-modelled".
Everyone is familiar with the word pal in the sense "friend, mate". It's one of the few common English words said to derive from the old Romani (or gypsy) language.
The previous edition of the OED dated the word from 1682, in a manuscript document held in the Hereford Diocesan archives. There was a troubling gap in the evidence then until towards the end of the following century. But maybe that didn't matter. Maybe it was just a hard term to find in the records at that period.
When we came to revise the entry for pal for the current online version of the OED I was slightly concerned that we were offering as our first evidence for the word a sentence copied down by one of the dictionary's readers from an unpublished document which had not been reverified since then.
Here was the sentence:
Wheare have you been all this day, pall?.. Why, pall, what would you have mee to doe?
It feels that there may be something wrong with the sentence. In some ways it has too modern a ring.
I decided to do some field work, so one day I drove over to the Herefordshire County Record Office, which is where the diocesan records are now housed.
With the help of an ever-courteous archivist I had soon identified document containing the use of "pal", a sworn statement made in a court case of 1682 involving Mary Ashmoore and her lover Ed Broughton.
After an adulterous affair Ashmoore had found herself pregnant. They had passed the child off as her husband's until a letter from Ed to Mary had come to light. Then they were in deep trouble.
We already felt that the syntax of the sentence felt odd - now we had a handle on the context. It seemed inherently unlikely that Ed would call his lover pal.
Was this really the same word? The context revealed everything. The couple are using familiar names: Ed in the man's case, and Poll (of which Pall is a variant) in the case of the woman.
Mary becomes Moll which becomes Poll and from there to Pall.
After a day's careful research we actually lost a first reference to pal, but secured a more accurate origin at the end of the 18th Century instead.
You can't necessarily be certain about an etymology, though.
Several years ago one of my colleagues was busy researching the word nacho, the tortilla chip.
She combed the books and newspapers, and roamed from eating house to eating house, restaurant to restaurant, in search of the true origin.
Along the way she found many references to a Mexican chef, Ignacio (= "Nacho") Anaya, who worked on the north-east border of Mexico in the 1940s.
The first reference we found for nacho came from a Texan newspaper of 1948. It might all fit. But there is still that seed of doubt.
At the last moment we had to pull back from claiming that Nacho Anaya was the chef behind the tortilla. Close, but no cigar - as yet.
One last research curiosity - what were the words for the new "horseless carriage" in the closing years of the 19th Century? This could go either way.
New findings could change the picture, though we've hunted high and low for evidence.
At the moment here is the leader board:
1)automobile, squeezing ahead of its competitors by virtue of its appearance in print on 22 August 1895
2)motor car, just rounding the final bend behind automobile - 10 September 1895
3)two months later we tried out autocar - 15 November 1895.
All of these were well ahead of car (in this meaning), which limped over the finishing line the next year, in 1896.
But the fascinating thing about these discoveries is that they represent the best information available at the moment. If you can trump them, then just let us know.