On seeing the second line of the title, some experts in Shakespeare’s diction may have jumped to the conclusion that they are in for another essay on a scurrilous topic. Not quite, unless the subject English spelling is considered obscene by definition. How is it possible for a single vowel letter to have so many values? “Elementary, my dear Watson,” as Sherlock Holmes did not say in any tale told by Conan Doyle. (Supposedly, the phrase was first used in 1915 by P.G. Wodenhouse in his novel Psmith Journalist. In Conan Doyle, the exchange between Watson and Holmes runs as follows: “’Excellent!’” I cried. “’Elementary’,” said he”. Those famous familiar quotations that everybody knows! They are like the proverbs of Alfred and the sayings of King Solomon. Dozens of works on word history open with Voltaire’s witticism that in etymology vowels count for nothing and consonants for very little. Yet it does not turn up in any of his written works.)
Unless counterbalanced by drastic reforms, long tradition usually makes spelling appear at best antiquated and at worst irrational. “This happens in all languages. For example, let us take English,” to quote a linguist of my acquaintance. We will follow his advice and “take” the letter o. Consider the following list:
bosom, Boleyn, woman;
love, dove, above, come, done;
move, prove;
on, gone;
one, none;
so, toe, nose.
On and nose (the short and the long of it) are taken for granted (so and toe are, in partly like nose), but the others?
I’ll begin with woman. The Old English for woman was wifman. Its long i designated a sound comparable with Modern Engl. ee in wee. Later that vowel underwent shortening, so that the word’s pronunciation began to resemble Modern Engl. wifman, rather than weefman, whereupon f was assimilated to its neighbor and wifman first turned into wimman (with regard to assimilation, compare lem’me go from let me go and leman “lover” from leofman) and then into wiman, for, as time went on, English lost long consonants. Contrary to professors of elocution, “common people” mispronounce words, slur as much as they can, and in general do not care about their delivery. Otherwise they would not have allowed wifman to degenerate into wimman. But they did not stop there. To articulate w, speakers protrude their lips and are not always in a hurry to spread them again. The result of this laziness was that Old Engl. widu “wood,” for instance, yielded wudu. Likewise, wiman became wuman. In the Middle English period, scribes disliked the sequences wu, um, mu, un, nu, and uv (because of too many vertical strokes the letters were hard to separate in reading, the more so as the usual signs for v and w were u and uu respectively) and substituted o for u. This is how uuuman became uuoman, that is, woman. Present day English has no words spelled with initial wu-. The few exceptions are dialectal forms recorded by linguists centuries after the phonetic processes mentioned here had been completed, and the only one most of us know is wuther, thanks to Emily Bronte’s title Wuthering Heights. In the early modern period, short u, except in the north of England, changed to the vowel of Standard English one now hears in shut up. Hence love, dove, above, come, and others. The story of done is more complicated: the change from long o (as in the modern paw or pore) to long u (as in the modern school), the shortening of that u, and the last step to the vocalic value of u in shut up. Womb and woman, which also have o contiguous to w, are still pronounced with the vowel of wuther. The original sound remained intact under the influence of w-.
The lips are active not only in the production of w but also in the production of p and b, and this is why pull and bull are pronounced the way they are. However, sometimes p- and b- could not save the following vowel from change, and alongside put, pull, and bull we have putty, pulp, and bulb. Unfortunately, the pernicious habit of designating the vowel in words like womb with the letter o resulted in the modern spelling bosom. The long stressed vowel of Old Engl. bosom (again as in Modern Engl. paw, pore) changed to long u (the equivalent of Modern Engl. oo), underwent shortening, and has been preserved. Boozom, boozam, or buzom would have made sense. Bosom reminds us of the word’s image that has not existed for at least half a millennium, and this is its only virtue. Anne Boleyn’s name was also spelled Bullen, but the unnatural variant has triumphed. When a word of Modern English is spelled with oo, we may assume that in the past it had a long vowel, regardless of whether its today’s reflex is long (as in food, mood) or short (as in good, hood). But the vowel of wood hardly ever was long. It is often said that conservative English spelling comes students in good stead, for it provides a window to the history of the language. It does, but those who look out of that window should be warned that the glass distorts the picture more than once.
It is now clear why prove and proof are spelled differently. The digraph oo in proof causes no surprise. Prove joined the words with v after o. The difference between prove, move and love, dove is that in the first group the vowel has remained long. Had love and dove withstood shortening, the four words would have rhymed, as they probably did in Shakespeare’s days. Today love/move is a so-called rhyme to the eye—a fact of no importance, since rhyming poetry is all but dead.
Old Engl. an “one” (with long a, as in Modern Engl. father) should have developed like stan, which is now stone, and it did, judging by the pronunciation of only (from anlic) and alone (a fusion of two words). In Middle English, an became on (on as in today’s awning). The rest is less clear. At that time, long vowels and diphthongs behaved similarly in that they could be pronounced with stress on the beginning and on the end, and this is why leosan, for instance, existed in two variants: leosan and leosan. As a consequence of this alternation, Standard Engl. lose, the reflex of leosan, has a dialectal variant lease, which continues leosan. This is also the reason show has a competing spelling shew (among the greats G.B. Shaw used only shew). If choose had sheared the fate of lose, today we would be asked “to cheese/chease our cheese.” Apparently, Middle Engl. on, that is, oon could be oon or oon, depending on the rhythm of the sentence. The variant oon was pronounced uon and won, rhyming with on. Several other dialectal variants of the same type have also been attested. Won became wun and later won, indistinguishable from the past tense of win. The pronunciation wonly was already known in 1570. As is usual with phonetic novelties, educated people first rejected the “vulgar” pronunciation of one with initial w– but were overwhelmed. The result is that today one is not a homophone of own. Most language historians trace the novelty described here (from oon to wun and won) to the British southwest, but it is hard to understand why the local pronunciation of such an important word should have been adopted by the Standard. Perhaps the forms with w– developed in the London area in the “allegro speech” of the capital (a great melting pot at all times) or under the influence of the “lower classes.” Once and none have aligned themselves with one. Spelling passed this tempest by.
The conclusion is obvious: the letter o has so many values because spelling has not caught up with the history of English sounds. Language retaliates sluggishness by producing spelling pronunciations. The fairly recent innovations often and fore-head are not the only examples of this type. Those who know about Coventry only from books sometimes pronounce Cov- as in cover. And indeed, who won’t be lost among Coventry ~ cover ~ over? Other people think that the name of the poet Donne, a homophone of done and dun, should be pronounced with the vowel of on. We can pity the naïve foreigner who missed the difference between worsted, the past tense of the verb worst, and worsted, the fabric, but sad is the lot of a native speaker who so often feels like a foreigner at home.
Anatoly Liberman is the author of Word Origins…And How We Know Them as well as An Analytic Dictionary of English Etymology: An Introduction. His column on word origins, The Oxford Etymologist, appears here each Wednesday. Send your etymology question to [email protected]; he’ll do his best to avoid responding with “origin unknown.”
How _do_ you pronounce Coventry?
“Elementary, my dear Watkins?” Conan Doyle would have been very unlikely to get Watson’s name wrong, surely? ;) Now I can go back and read the rest of the article.
OED lists the pronunciation of Coventry as in cover, so it must be common. Although I know of no dictionaries that show them, I myself use a somewhat rounded vowel in pulp, pulpit, bulb.
“Womb and woman, which also have o contiguous to w, are still pronounced with the vowel of wuther.” Merriam-Webster gives three different vowels in these: respectively those of GOOSE, FOOT and STRUT. I personally have the STRUT-vowel in ‘woman’ and the FOOT-vowel in ‘wuthering’.
[…] at the Oxford University Press OUP Blog, Anatoly Liberman wrote a 1,437-word “blog” entry, complaining that common people ruin pronunciation […]
I suspect that the spelling “wuther” is an illiteracy of Bronte’s, or her publisher’s, anyhow; the OED is clear that the initial consonant is /W/, from a Norse etymon in kv-, and it uses the spelling “whither” for the headword (the oldest uses are Scots, and have the expected “quh-” representing /xw/).
We might well expect /E/ and /U/ as variants of the normal /I/ vowel, and indeed we find them in the OED. But I suspect that the STRUT vowel is a modern spelling pronunciation; the Earnshaws surely spoke Yorkshire dialect with no FOOT/STRUT split, and would have used a vowel close to FOOT.
Coventry shares the same first-syllable vowel as dozen, as does Romsey.
Tell me how do you pronounce “coven” ??