Monday 28 February 2011

comparable niches

Comments on Friday’s blog somehow managed to segue into the pronunciation of the word niche and the matter of whether BrE is becoming Americanized (and whether therefore AmE is ‘destroying the language’).
Here you see the LPD entry for niche. My poll shows that the French-style pronunciation is now overwhelmingly preferred in BrE, while (as far as I know) remaining virtually unknown in AmE.

The OED comments
N.E.D. (1907) gives only the pronunciation (nitʃ) /nɪtʃ/ and the pronunciation /niːʃ/ is apparently not recorded before this date. H. Michaelis & D. Jones Phonetic Dict. Eng. Lang. (1913), and all editions of D. Jones Eng. Pronouncing Dict. up to and including the fourteenth edition (1977) give /nɪtʃ/ as the typical pronunciation and /niːʃ/ as an alternative pronunciation. The fifteenth edition (1991) gives /niːʃ/ in British English and /nɪtʃ/ in U.S. English.
So it’s BrE that’s done the innovating here.

In one of my talks about pronunciation preference polls I show three slides to illustrate the point that the evidence is that — contrary to what many people suppose — in some cases change spreads from AmE to BrE,in other cases from BrE to AmE,while occasionally we find that both are moving in tandem.
(Sorry about the Comic Sans. This is a very old talk. I’ve moved on typographically since then.)

Friday 25 February 2011


On Sky News yesterday Adam Boulton reported David Cameron as (metaphorically) paying ˈhɒmɑːʒ to various Middle Eastern rulers (ones still in power, that is). I wonder if this was just his fancy way of pronouncing the word homage. This word was borrowed from Old French into Middle English some eight hundred years ago and is usually pronounced ˈhɒmɪdʒ. Or was he consciously reborrowing the modern French equivalent, hommage? If the latter, then you would expect him to know that French h is silent. In French the word is pronounced ɔmaʒ.

As experts on French will tell you, words spelt with initial h- fall into two classes phonetically. Some have the so-called “h aspiré” (aspirated h), which is unpronounced as such, or might sometimes be realized as a glottal stop; but in any case it blocks liaison that would otherwise occur. Others have “h muet” (silent h), which is equally unpronounced but allows liaison. For what it’s worth, the one in hommage is h muet.

As is well known, English has restored an h-sound in the pronunciation of many French loanwords (e.g. habit, heritage, hospital), though not in all (not in heir, honest, hono(u)r, hour and their derivatives). For herb the h has been restored in BrE but not in AmE. In BrE hotel there are a few laggards who have not yet restored the h, though most speakers pronounce it (except of course those who drop h anyway). You’ll also know about humble and (h)ostler. Let’s not get into h in unstressed syllables (historical, heretical, hysterics).

An awkward case is the visibly French loanword hauteur. In French it’s otœʀ. In BrE it may be pronounced as in French, or more likely in some such halfway anglicized form as əʊˈtɜː. That’s what you’ll find in LPD and CPD, without any restoration of h. But I notice that ODP gives an alternative hɔːˈtɜː — I can’t say I’ve ever heard this myself. In AmE I think the h tends to be restored insofar as the word is used at all.

The problem with having no h-sound in hauteur is that it makes it a homophone of auteur. I suppose there might just about be a misunderstanding if one referred to the (h)auteur in a discussion of some cinéaste’s oeuvre. But then of course they’re homophones in French anyway.

If we wanted to anglicize hauteur really thoroughly, perhaps we ought to spell it haughteur (or perhaps haughture, attested in the sixteenth century), like its cognate haughty, where the h-sound has been firmly restored. The spelling with -gh- is nevertheless etymologically entirely unjustified, being attributed by the OED to the influence of caught, taught and perhaps high, height.

The OED’s take on the pronunciation of hauteur is a strange mixture of French and English (o is meant to be “as in French eau”, while œ is “as in boeuf”).

Thursday 24 February 2011

cavalla, cavally

There’s a village on the island of Montserrat called Cavalla Hill. It can be reached only along a particularly steep and narrow road, but is home to the largest remaining Methodist church on the island since the loss of Bethel to the volcano.

I have been struck by the fact that there is considerable variability among Montserratians concerning the pronunciation of the first word in this place name. Some call it kəˈvalə, which I suppose is what you would expect from its spelling (the double ll signalling stress on the preceding vowel, as in umbrella, vanilla).

But many others make the final vowel i, and I have the impression that this is the general popular form, used by those who know it mainly as a spoken rather than as a written name. Yet even this group are divided in two. Some people stress it on the penultimate, but others on the initial syllable. The majority, it seems to me, say ˈkavəli. Since I too heard the name spoken before I saw it written, I too tend to give it initial stress and final i.

I haven’t been able to find out where the name comes from. Most placenames in Montserrat are geographical descriptors (Little Bay, Lookout, Woodlands), religious names (St Peter’s, St John’s, Salem), transplanted British or Irish names (Plymouth, Kinsale, Olveston), or based on former estate owners’ names (Tuitts, Brades, Nixons). But Cavalla Hill doesn’t seem to fit in any of those categories.

Dictionaries and Wikipedia tell us that cavalla is the popular name of several species of fish, including Caranx hippos and Scomber scombrus, aka the horse-mackerel. Not being a fisherman, I have never heard this word in use and do not know how fishermen pronounce it.

Webster’s Collegiate, however, gives not only cavalla kəˈvælə for the name of the fish, but also a variant form cavally kəˈvæli. (It also reports that the etymology is Spanish caballa, from a Latin word for ‘mare’.)

The OED, which does not record the spelling cavalla at all, gives the word only as cavally kəˈvæli.

There is also a river in West Africa called the Cavalla or Cavally.

It is not clear why a hill in Montserrat should be called after a fish or after a West African river.

The main point of phonetic interest here is the alternation between final schwa and final i. It puts me in mind of the Grand Ole Opry (= opera). According to Wikipedia, the term opry is quite recent, though, having been coined in 1928 by a Texan broadcaster, George D. Hay.

Are there other English words in which final ə alternates with i? I suspect there are, but I can’t actually think of any.

Wednesday 23 February 2011

mysteries of existence

Like other teachers of elementary phonetics to NS students, when teaching I would regularly set exercises in “doing transcription”. I would give the students a passage of English in ordinary spelling. Their task would be to convert it into phonetic transcription.

This is a valuable exercise for NSs as much as for NNSs. It familiarizes them with the phonetic symbols. It makes them conscious of the difference between spelling and pronunciation. It alerts them to the characteristics of connected speech as opposed to individual words in isolation.

Good students sail through this task. The weaker ones often find it remarkably difficult. I would frequently have to point out that write does not actually begin with a w-sound, nor looked end with a d-sound (still less a syllabic ). The first vowel in particular is not normally pronounced ɑː, and the second vowel in information is not ɔː. Orthography has a distressingly dazzling effect on the phonetically unsophisticated.

But setting and marking (AmE ‘grading’) transcription can also be valuable for the teacher. Early in my career I noticed some students transcribing exist as ɪkˈzɪst (and similarly with example, exhausted, exams etc). Since the usual pronunciation is ɪɡˈzɪst I would mark this wrong. (Initial e-or ə- rather than ɪ- is OK, of course, though you have to check that that’s what they genuinely say, not just spelling-driven.)

However some students protested: they really do pronounce the word as ɪkˈzɪst. I checked it out, and they appeared to be correct. I came to realize that some speakers in the southeast of England, at least, have an unexpected dissimilation of voicing in these words. Their kz in exist seems to be different both from the gz of eggs or big zits and from the ks of exceed.

Being now a fortis plosive, it is also a candidate for glottal reinforcement: ɪkʔˈzɪst.

Convinced now of their reality, I decided to include these variants in LPD. But no other dictionary seems to recognize their ɪkˈzɪstəns.

Tuesday 22 February 2011


Ulrike Stange, who is doing some research on English interjections, wrote
When analysing the functions of wow! I came across sentences like
Wow. Calm down.
Wow. Take it easy.

I'm not a native speaker of English but I'm fairly sure that wow! is pronounced differently in these cases (rhyming with American go); it does not sound like a British English diphthong.
The dictionaries I checked only provide the usual pronunciation (rhyming with "cow"). I was wondering if they might be homographs, the "normal" wow! expressing admiration, etc. and the other one apprehension or something along these lines. However, no wow! with this function is listed in any of the dictionaries I consulted.
I replied
The interjection meaning “slow down!” and rhyming with (British or American) go is normally spelt “whoa!”.
Perhaps these are just spelling mistakes.

I wonder if there is anything more to it than that.

I’m not aware that the slow-down interjection whoa! is pronounced with something that “does not sound like a BrE diphthong”. As far as I know, it is straightforwardly a homophone of woe, pronounced with the local GOAT vowel whatever it may be.

Other languages seem to use quite different interjections in this sense: French holà!, German brr!, Spanish ¡so!

As a child I learnt a nursery rhyme
If I had a donkey that wouldn’t go
would I beat him? Oh dear, no.
I’d say “Gee up, Neddy! Whoa!”
(That’s the version I know. The internet offers a rival nursery rhyme that goes “…I'd put him in the barn and give him some corn, The best little donkey that ever was born!”)

The OED gives spellings and definitions that seem exactly right.
whoa wəʊ ‘a word of command to a horse or other draught-animal to stop or stand still’ (from 1843)
wow waʊ ‘now chiefly expressing astonishment or admiration’ (from 1892)

On the other hand the on-line Urban Dictionary offers further, alternative, definitions of whoa that make it a partial synonym of wow:
1. To express surprise (interj)
2. To express astonishment (interj)

So perhaps there are two possible ways of interpreting this cartoon.

Monday 21 February 2011

deaccented apposition

When you have two noun phrases in apposition, the usual treatment in English intonation is to accent both. If, as often, they are placed in separate intonation phrases, there is usually tone concord between the two (English Intonation, p. 85).
ˈLet me introˈduce you to my \colleague, | \Charles.
Oh, and ˈthis is \Mary, | a\nother colleague.
I’ve got ˈMr /Bun, | the /baker, | and ˈMrs \Plod, | the po\liceman’s wife.

In another kind of apposition (or something similar), one of the elements is more like a title. In this case the two parts are usually placed within a single IP.
the ˈnovelist ˈMartin \Amis
the ˈsculptor ˈHenry \Moore
my ˈman \Jeeves (= my servant J.)

In yet a third type the second element is more or less redundant as far as information content is concerned. Unsurprisingly, in view of this, it is typically deaccented.
I ˈcan’t ˈstand that \Thatcher woman.
ˈWhat’s the ˈname of that \writer chappie? (| The ˈone you were \talking about.)
ˈWho's this \Smith fellow?

I was reminded of this type of apposition (or whatever) when reading the story “The Aunt and the Sluggard” in P.G. Wodehouse’s, My Man Jeeves, 1919 (free download on my Kindle).
Jeeves is my man, you know. Officially he pulls in his weekly wages for pressing my clothes and all that sort of thing; but actually he’s more like what the poet Johnnie called some bird of his acquaintance who was apt to rally round him in times of need — a guide, don’t you know; philosopher, if I remember rightly, and — I rather fancy — friend. I rely on him at every turn.

Here, the construction the poet Johnnie is not like the poet Milton or the poet Keats. Johnnie is not the poet’s name. It means no more than ‘fellow, chap’, implying that the hearer must be familiar with the poet although the speaker can’t for the moment recall his name. And Johnnie is not accented: the ˈpoet Johnnie.

In this phrase I can see that NNSs might be misled both into accenting Johnnie and into imagining that Johnnie was the name of a poet.
But then Wodehouse is full of traps for the NNS. He writes in a particularly idiomatic way — and in a colloquial idiom that is no longer to be heard in contemporary English.

Friday 18 February 2011


From a universal typological perspective, we sort of expect vowel systems to be more or less symmetrical. There is a feeling of orderliness about the vowels of Latin, Spanish, Modern Greek, Japanese, and Swahili, with their 5-vowel system nicely balanced between front and back, high and low.
i   u
e   o
Orderly 7-vowel systems like Italian add a further distinction of height, with ɛ and ɔ.

Adding a backness distinction between low vowels, æ vs ɑ, as in Farsi, and/or adding front rounded vowels y, ø, Finnish-style, does not disrupt the equilibrium. Nor does adding back unrounded vowels ɯ, ʌ, Korean-style.
Nor does a tense-lax distinction, pairing i with ɪ and u with ʊ as in English or German.

From this typological perspective English is however rather messy — particularly the types of English where historical monophthongs have developed into assorted diphthongs. (Yes, it’s GOAT that I’m thinking of particularly.) The Anglo-American NURSE vowel, ɜ or ɝ, doesn’t actually disrupt the balance, but from an international perspective is an extremely unusual type of strong vowel.

The prize for unbalanced vowel systems seems to go to Pacaás Novos, aka Wari’, an Amazonian language spoken near the Bolivia/Brazil border. It has four mid/high front vowels, but only one back vowel. More unusual still in Wari’ is one of the items to be found in the consonant system, namely the ‘dental affricate’ t͡ʙ̥, which consists of a dental stop plus a bilabial trill, together forming a single unit.
Dan Everett, co-author of a descriptive grammar of the language, says that the Wari’
are a helpful, kind people who speak a groovy language.

(The above photo of a Wari’ man relaxing was taken by the anthropologist Beth Conklin.)

Sadly for us, though, younger speakers are reportedly simplifying their consonant system. Rather than articulate t͡ʙ̥ they prefer to use just the first part of the affricate, so reducing it to t.

Thursday 17 February 2011

Latin double velars

Consider the everyday words accelerate, accept, accident, success, vaccine. They demonstrate the English spelling-to-sound principle that where double -cc- is followed by e or i the pronunciation is ks. Thus we have əkˈseləreɪt, əkˈsept, ˈæksɪdənt, səkˈses, ˈvæksiːn.

There are two rather rarer words where not all speakers follow this principle. One is flaccid, where ˈflæsɪd competes with the expected ˈflæksɪd. The other is succinct, where I have just heard an American narrator on the BBC pronounce səˈsɪŋkt rather than the səkˈsɪŋkt that I would say myself. I can think of no reason why these two words should be exceptions to the general rule. Like the others, they are of Latin origin. Their pronunciation goes back to a Latin double -cc-, which was classically kk but subsequently had the second velar ‘softened’ when it developed into or s in late Latin and the successor Romance languages.

As you would expect, this being English spelling, there are a handful of other words that violate the rule. Soccer ˈsɒkə, ˈsɑkɚ obviously “ought” to be spelt socker. And lovers of classical music will be familiar with Italian names such as Puccini puˈtʃiːni and Pagliacci.

Strangely, the corresponding rule doesn’t work in the case of the voiced etymological equivalent. In exaggerate ɪɡˈzædʒəreɪt we have simple , not *ɡdʒ. In suggest Brits have simple , thus səˈdʒest, but most Americans have ɡdʒ, thus səɡˈdʒest. Otherwise the velar remains unsoftened; there don’t seem to be any other Latin-derived words with -gge- or -ggi-. Double gg stands for simple g in occasional non-Latin-derived words such as druggist and digging. And then there’s ciggy ˈsɪɡi, colloquial abbreviation of cigarette, which demonstrates the use of -gg- in informal spelling.

If you’d like to wallow further in the irregularities and inconsistencies of English spelling, allow me to recommend Masha Bell’s blog.

Wednesday 16 February 2011


There are many words in English that have ju, jʊ, jə or some derivative of this in a weak syllable: words like regulate, educate, executive, ambulance, particular. For some of them there are variants with plain ə or even zero instead. These are stigmatized. It is not considered admirable to pronounce particular pəˈtɪk(ə)lə (”partickler”) rather than pəˈtɪkjʊlə.
Biddy nodded her head thoughtfully at the fire as she took up her work again, and said she would be very particular; and Joe, still detaining his knees, said, "Ay, ay, I'll be ekervally partickler, Pip;" and then they congratulated me again, and went on to express so much wonder at the notion of my being a gentleman, that I didn't half like it.
—Dickens, Great Expectations

Where there’s stigmatization there will usually be hypercorrection. This is the origin of forms with -j- in words such as escalator and percolate. The standard ˈeskəleɪtə, ˈpɜːkəleɪt seem too plain, so people feel they would be nicer as ˈeskjəleɪtə, ˈpɜːkjəleɪt. I have marked these in LPD with a warning triangle.

Nevertheless, there are one or two words in which I have the impression that the hypercorrect form is somewhat better established.

In the hymn The Church’s One Foundation there is a verse running
’Mid toil and tribulation,
And tumult of her war,
She waits the consummation
Of peace forevermore;
Till, with the vision glorious,
Her longing eyes are blest,
And the great Church victorious
Shall be the Church at rest.
The spelling of consummation (double -mm-) tells us that -j- is inappropriate in this word. So does the etymology (cognate with sum and summation, nothing to do with consume). Yet I think congregations rather often sing ˌkɒnsjuˈmeɪʃən rather than ˌkɒnsəˈmeɪʃən or ˌkɒnsʌˈmeɪʃən. Perhaps the adjacent tribulation pushes one in that direction, too.

In LPD, accordingly, I have placed no warning triangle for -j- in consummate, consummation. Ought I to have?

Tuesday 15 February 2011

worrying about Wyre

Anyone who watches BBC TV or listens to BBC radio will have heard the name of Wyre Davies, currently the BBC’s Middle Eastern correspondent. Everyone pronounces his first name as ˈwɪrə.

And therein lies something of a mystery. I wonder if anyone can shed light on it.

Wyre is Welsh, and his first name is presumably Welsh too. After all, it can’t be English: if it were the English name so spelt it would be pronounced ˈwaɪə, as in the Lancashire placename.

No, it must be Welsh, and presumably a hypocoristic form of some longer name. But — (i) I cannot discover what this longer name might be, and (ii) there is a phonotactic rule in Welsh to the effect that the vowel ə cannot occur in final syllables (as it does here).

As a Welsh name you would expect Wyre to be the pet form of some name beginning Wyr- or Gwyr-. (Welsh g soft-mutates to zero.) But I know of no such name. Perhaps the nearest is Gwilym, the Welsh for ‘William’.

Given the Welsh spelling -e you would expect the pronunciation to be e, as in bore ‘morning’ ˈboɾe, which you may know from the greeting bore da! ‘good morning!’. Since English doesn’t allow the DRESS vowel in final position, that would map onto English . But all the other BBC presenters are consistent in calling Wyre not ˈwɪreɪ but ˈwɪrə.

In Welsh the schwa vowel is always spelt y. It can be stressed or unstressed. But it cannot be in a word-final syllable. So for example the word for ‘mountains’ is mynyddoedd məˈnəðoið; but if we strip off the plural suffix we are left with the singular mynydd, which is pronounced ˈmənið.

Correspondingly, given the Welsh spelling y (not counting the digraph wy) the reading rules tell you that the pronunciation will be ə, except in a final syllable, where it will be i. (The sole syllable of a monosyllable is inevitably final. There are footnotes here, which we can ignore, relating to north Welsh ɨ and to proclitic function words.) One of the words for ‘is’, ydy, is ədi; one of the words for ‘valley’ is dyffryn ˈdəfrin. One of the words for ‘man’ is dyn diːn, plural dynion ˈdənjon. All of these exemplify the reading rule for y.

OK, the spelling wy can be an exception to this rule, so we needn’t worry about the first syllable of Wyre ˈwɪrə. But the final vowel remains very unexpected, and from the standpoint of Welsh phonetics seems to be impossible.

Monday 14 February 2011

rhymes for “love”

There’s a Project Gutenberg ebook compilation called A Wodehouse Miscellany. It comprises some twenty short articles and stories by the humorist P.G. Wodehouse (1881-1975). The publication dates of the individual components are not always stated, but the compilation itself dates only from 2003.

Anyhow, one of the articles is entitled ‘On the writing of lyrics’. Here’s an excerpt.
“Love” rhymes with “dove”, “glove”, “above” and “shove”. It is true that poets who print their stuff instead of having it sung take a mean advantage by ringing in words like “prove” and “move”; but the lyricist is not allowed to do that. This is the wretched unfairness of the lyricist’s lot. The language gets him both ways. It won’t let him rhyme “love” with “move”, and it won’t let him rhyme “maternal” with “colonel”. If he tries the first course, he is told that the rhyme, though all right for the eye, is wrong for the ear. If he tries the second course, they say that the rhyme, though more or less ninety-nine percent pure for the ear, falls short when tested by eye. And, when he is driven back on one of the regular, guaranteed rhymes, he is taunted with triteness of phrase.

No lyricist wants to keep linking “love” with “skies above” and “turtle dove”, but what can he do? You can’t do a thing with “shove”; and “glove” is one of those aloof words which are not good mixers. And — mark the brutality of the thing — there is no word you can substitute for “love”. It is just as if they did it on purpose.

If only phoneticians (such as yours truly) could write about ear rhymes and eye rhymes in such an entertaining way!

Wodehouse was of course English, and speakers of BrE will nod their heads in agreement with what he says here. But Americans may be a little surprised. In AmE there is another word, a very frequent one, that lyricists can use as a rhyme for love: the strong form of the preposition of. In BrE this is pronounced ɒv (with the LOT vowel), so is not a suitable rhyme, but in AmE it is predominantly ʌv (with the STRUT vowel; some prefer to write it əv). Voilà!

Presumably this STRUT-vowel pronunciation arose as a restressing of the weak form əv.

Despite living for many years in the US, Wodehouse seems not to have noticed how (most) Americans say of.

As far as I know, ʌv (STRUT) is unknown in BrE. Conversely, though, there seem to be some Americans who say ɑv (LOT), but how many they are, and whether they have some distinguishing social or regional characteristic, I do not know.

Friday 11 February 2011

Chuka Umunna

When I first noticed the name of the up-and-coming British politician Chuka Umunna I fired off an email to him to ask him how he preferred his name to be pronounced. I got a brief acknowledgement promising a reply in a few days — but then nothing more.

Now, through the good graces of the BBC Pronunciation Unit (thanks, Martha Figueroa-Clark), I can reveal that he is ˈtʃʊkə ʊˈmʊnə . He has confirmed that this is the correct pronunciation of his name.

Chuka Umunna grew up in Streatham, south London, and is of mixed English, Nigerian and Irish descent. He is also well-connected, being the grandson of a High Court judge — Sir Helenus Milmo, who played a leading role in the Nuremberg trials — and the nephew of a leading libel lawyer, Patrick Milmo. He was elected as the MP for Streatham in the 2010 general election.

Martha tells me that the chap who handles Mr Umunna’s media enquiries says that a lot of people (unsurprisingly) do not know how to pronounce his name, and may attempt to give it a non-English-sounding pronunciation.

(I think the name Umunna is of Igbo origin. In that language it is presumably pronounced [umũna], with tones that I cannot guess at. If I am right, the first of the two letters n indicates the nasalization of the preceding vowel in Igbo, not gemination of a nasal consonant.)

[Further content removed because of privacy issues.]

Thursday 10 February 2011


I watched the first programme of David Attenborough’s new series Madagascar on the television last night.

As well as following the narration and admiring the stunning pictures of wildlife and scenery, I was able to check that Attenborough pronounced sifaka, indri, tenrec and other Madagascan animals in the way I would expect. It turned out that he said these names just as shown in LPD — but there was one which he pronounced in a way that I had not allowed for. That was lemur.

Most people pronounce this word ˈliːmə(r). That is the pronunciation I prioritize in LPD, and it is the only form given in the Cambridge EPD and the Oxford ODP, not to mention Webster’s Collegiate and the OED. In LPD I do also include the rare and rather prissy variant ˈliːmjʊə which I must have noted somewhere. But what David Attenborough said on TV last night, repeatedly, was ˈliːmʊə. I don’t think I have ever heard that before.

Etymologically, the word is ‘modern Latin’, a reconstructed singular form of the Latin plural lĕmŭrēs ‘spirits of the dead’, supposedly from the lemur’s spectre-like face (COD). The OED’s first citation in this zoological meaning dates from 1795.

With such coinages there is no choice really but spelling pronunciation, hence ˈliːmə(r). In the Latin of nineteenth-century England, though, lemures would presumably have been pronounced ˈlemjʊəriːz, and in the twentieth century ˈlemʊreɪz. So this was evidently not the determining factor in pronouncing the reconstructed singular as the name of the animal.

En français on dit lémur lemyʀ; auf Deutsch sagt man Lemur leˈmuːɐ̯.

Wednesday 9 February 2011


One of the books I read while away on holiday was Why Does E=mc2 (And Why Should We Care?), by Brian Cox and Jeff Forshaw — hence yesterday’s posting.

I wouldn’t say that I now have a thorough understanding of general relativity and special relativity, let alone of quantum mechanics, but at least I am clearer about what I don’t understand. I came away feeling that in a few months I’d better reread the book.

In the part of the book dealing with quantum mechanics there is naturally extensive discussion of elementary particles, among them the quark.
How do you pronounce quark? Does it rhyme with park or with fork?
Despite the fact that other words spelt with quar- are pronounced kwɔː(r)-quart, quarter, quarto — I have the impression that most people who know the word, whether physicists or not, pronounce it kwɑː(r)k. In LPD I included both possibilities, but prioritized kwɑː(r)k over kwɔː(r)k.

The man who invented the word, Murray Gell-Mann, wouldn’t agree. He found it in James Joyce’s Finnegan’s Wake, the point being that quarks always occur in threes.
Three quarks for Muster Mark!
Sure he has not got much of a bark
And sure any he has it's all beside the mark.

Gell-Mann explains it as follows, in his book The Quark and the Jaguar.
In 1963, when I assigned the name "quark" to the fundamental constituents of the nucleon, I had the sound first, without the spelling, which could have been "kwork". Then, in one of my occasional perusals of Finnegans Wake, by James Joyce, I came across the word "quark" in the phrase "Three quarks for Muster Mark". Since "quark" (meaning, for one thing, the cry of the gull) was clearly intended to rhyme with "Mark", as well as "bark" and other such words, I had to find an excuse to pronounce it as "kwork". But the book represents the dream of a publican named Humphrey Chimpden Earwicker. Words in the text are typically drawn from several sources at once, like the "portmanteau" words in "Through the Looking-Glass". From time to time, phrases occur in the book that are partially determined by calls for drinks at the bar. I argued, therefore, that perhaps one of the multiple sources of the cry "Three quarks for Muster Mark" might be "Three quarts for Mister Mark", in which case the pronunciation "kwork" would not be totally unjustified. In any case, the number three fitted perfectly the way quarks occur in nature.

Cox and Forshaw take this further:
Gell-Mann has since written that he originally intended the word to be pronounced “qwork,” and in fact had the sound in his mind before he came across the Finnegan’s Wake quotation. Since “quark” in this rhyme is clearly intended to rhyme with “Mark” and “bark,” this proved somewhat problematic. Gell-Mann therefore decided to argue that the word may mean “quart,” as in a measure of drink, rather than the more usual “cry of a gull,” thereby allowing him to keep his original pronunciation. Perhaps we will never really know how to pronounce it.

I think the moral of this story is that once you coin a successful technical term you have to relinquish ownership of it. As I see it, the speech community has collectively decided that it prefers to make the word rhyme with park, and Gell-Mann’s views are irrelevant.

The OBGP disagrees. Perhaps I (or someone) ought to do a pronunciation preference survey on the word.
Is the pronunciation influenced by another word with the same spelling, the quark that is a kind of curd cheese popular in Germany? Some British supermarkets sell this. I imagine that most people who buy it pronounce it to rhyme with park, though I have no evidence. That’s certainly how it is pronounced in German. But I do wonder whether German physicists follow the Duden Aussprachewörterbuch’s advice when referring not to soft cheese (Käse) but to the Elementarteilchen. (And notice how Duden hesitates interestingly between the German equivalents of the FORCE and NORTH vowels, presumably undifferentiated in Gell-Mann’s English.)
Lastly, what about QuarkXPress, the software package used by designers and publishers? Am I right in thinking that it too usually has a first syllable rhyming with park?

Tuesday 8 February 2011

E = mc2

There’s an entry in the Oxford BBC Guide to Pronunciation that got me thinking. It’s true — the default stress pattern in an algebraic expression involving a simple raising to the power of two is to deaccent squared.

The theorem of Pythagoras, which you will remember from school trigonometry classes, states (in the wording I was taught) that in a right-angled triangle the square on the hypotenuse is equal to the sum of the squares on the other two sides. Here, square and squares are regularly accented.
The ˈsquare on the hyˈpotenuse | is equal to the ˈsum of the ˈsquares on the ˈother two ˈsides.
Expressed as an algebraic equation, this takes the form
a2 + b2 = c2
which we say aloud as
ˈA squared plus ˈB squared | equals ˈC squared.

The word squared is not accented (unless of course it is contrastive). You might suppose that this is because we avoid accenting any item that is repeated or about to be repeated — a well-established principle in English accentuation.

However, this carries over, rather mysteriously, into cases where no repetition is involved.
ax2 + bx + c = 0
is said aloud as
ˈA ˈX squared | plus ˈ B X plus ˈC | equals ˈnought

Hence the constant in Einstein’s equation, the square of the speed of light, c2, has the spoken (recommended) default pattern ˈC squared.

So far, so good. But as soon as the item to be raised to the power of two is complex, involving more than one symbol, the word squared defaults to being accented. The expression (a + b)2 is said aloud as ˈA plus ˈB ˈsquared The expression (ax)2 is read as ˈA ˈX ˈsquared. I’m not sure why.

Hence the default interpretation of spoken ˈA ˈX squared is ax2, while the default interpretation of spoken ˈA X ˈsquared is (ax)2.

But a further way of avoiding the possible ambiguity in the latter is to say A X all squared. And in that wording you place the main accent on all — I think. Similarly, for (a + b)2 we can say ˈA plus ˈB all ˈsquared. Do you agree? It’s a long time since I did algebra.

None of this applies to cubed or higher powers. We say x3 as ˈX ˈcubed and x4 as ˈX to the ˈfourth.

Monday 7 February 2011


The sociolinguists have demonstrated that most native speakers of English fluctuate between two forms of the -ing ending: the ‘high’ variant ɪŋ, with a velar nasal, and the ‘low’ variant ɪn, ən, with an alveolar nasal.

The difference is stylistic, with the H variant being used in formal situations and the L in informal/colloquial situations. Just where the line is drawn between the two possibilities varies, depending on social class and other factors.

Using the L variant is popularly known as “dropping one’s g’s”, although in surface phonetics it is a matter of place of articulation rather than of omitting something. It is of course shown in writing by the use of an apostrophe, putting -in’ in place of -ing. (The tattoo in my picture omits the apostrophe.)

The L variant also has the subvariant of a syllabic , used particularly after t (→ ʔ) and d. Both ɪn and are obviously also subject to possible dealveolar assimilation, producing forms with m and (!) ŋ.

So for running we can have H ˈrʌnɪŋ or L ˈrʌnɪn, ˈrʌnən. For putting we have H ˈpʊtɪŋ and L ˈpʊtɪn, ˈpʊtn̩, ˈpʊʔn̩. Because of possible assimilation, if we are faced with ˈrʌnɪŋ ˈkwɪkli or ˈteɪkŋ̩ ˈpleɪs we cannot tell which of H and L is involved.

However at the top of the social scale there is a smallish group of speakers who use H (velar) under virtually all circumstances. At the bottom there is a rather larger group who virtually always use L (alveolar).

Personally I have to admit membership of the first group. I don’t believe I ever use the L variant except for jocular or other special effect.

The lyrics of popular music often mandate the L variant. I find it difficult to remember to do this as required when singing.

I remember when BBC English assigned me a producer (many years ago) who used the L variant extremely frequently. I was shocked at my own gut reaction, which was that he must be ignorant and uneducated to the extent that I found it difficult to take his opinions seriously.

Because it is so socially sensitive, this variable also generates hypercorrections. In Accents of English (p. 263) I mentioned as examples of this a braz[ɪŋ] (brazen) hussy and Badmi[ŋ]ton. The other day I heard a nice one from a railway station announcer: Harpenden pronounced as ˈhɑːpɪŋdən.

Friday 4 February 2011


Steve Doerr writes asking me to comment on
the name ElBaradei, prominent in the news just now. The location of the stress and the pronunciation of the -ei are the bones of contention for me. (I'm not an Arabist though.)

In LPD I give əl ˈbær ə daɪ, with alternatives el-, -deɪ. As far as the main pronunciation is concerned, in this I follow the recommendation of the Oxford BBC Guide to Pronunciation, which draws on the expertise of the BBC Pronunciation Unit and (presumably) the advice of Arabic speakers working for the BBC World Service (now so shamefully to be drastically cut back). The OBGP notes further that “this is his preferred anglicization”.

For some reason I neglected in LPD to supply the Arabic pronunciation, which to the best of my knowledge is [ʔelbæˈɾædʕi].

In Arabic the name is normally written البرادعي ʾlbrādʿī (see for example here), with short vowels as usual not shown. The final consonant is an ʿayn (pharyngealized glottal stop).

Classical Arabic has only three phonemic vowel qualities, /i a u/ (short and long), and I think the same is true of Egyptian Colloquial Arabic. The consequence is that there is considerable allophonic and other variability in the actual qualities of the vowels. So /a/ is typically æ in non-pharyngeal contexts, but in pharyngeal contexts ɑ, while in the definite article it can even range over e and ə. So I do not think there is any call to be dogmatic about whether we make the final sound in English or .

Word stress in Arabic can fall on any of the last three syllables of the word, and is predictable from the syllable structure. The rules are somewhat different in Egyptian from the rules for other varieties. If I am right in thinking that in this name it falls on the penultimate, that would be because of the consonant cluster following the penultimate vowel. But when we anglicize it we can’t cope with ʕ, which we omit, so that there is no cluster and the stress reverts to antepenultimate. (Or am I fantasizing? Is the vowel long, as written, and therefore stressed?) What Steve Doerr presumably doesn’t like is the final stress heard from some newsreaders.

Anyhow, you can see from these rather vague and tentative remarks that I am not an Arabist either. I would welcome comments from those who know more than I do about Egyptian Arabic phonetics.

Thursday 3 February 2011

being mischievious

Quite a few people found it rather shocking that in my 2008 LPD pronunciation preference poll as many as 15% of the BrE respondents said that for mischievous they preferred the pronunciation mɪsˈtʃiːviəs, with the figure rising to 29% for those born since 1981. An older poll had already reported 11% of Americans preferring it, too.My problem for today is that most of those who say mɪsˈtʃiːviəs would probably want to spell it correspondingly, as “mischievious”. If they do, should we simply condemn this as a spelling mistake, or is it time to consider recognizing it as an alternative spelling? It seems reasonable to suppose that in a hundred years’ time it may have more or less totally displaced “mischievous”. Or at the very least that the OED’s characterization of it as “regional, colloq. and humorous” may need to be revised.

There are several other similar cases. I have just seen the veteran broadcaster Esther Rantzen on television talking about something being prəʊˈtruːbərənt, i.e. protuberant. I do give this -ˈtruː- variant in LPD under protuberant, though with a warning triangle. Ought I to remove the triangle from “protruberant”, as I have from “mischievious”?

Then what about Sarah Palin’s “refudiate” for “repudiate”?

Or our old friend “pronounciation” for “pronunciation”? (You wouldn’t believe the number of journalists and radio and TV presenters who want to talk to me about “pronounciation”. I almost feel churlish when I pointedly reply with reference to “pronunciation”. But not quite.)

These are not exactly malapropisms, because a malapropism involves confusing the meanings of two distinct dictionary words. They’re not exactly mispronunciations, either. They’re more a recasting or regularization of what is felt to be irregular, in writing as much as in speech.

The same problem, in a milder form, arises with “re-occurrence” for what I call “recurrence”, and with “to disassociate” in place of “to dissociate”. I suspect the younger generation doesn’t even feel uneasy, as I do, at these forms. (I mustn’t fall foul of the recency illusion. The OED reveals that they’re not exactly new. No matter.)

Wednesday 2 February 2011

Kindle woes

The reason I have recently been reading Edward Lear (blog, yesterday) and various other authors of the nineteenth and early twentieth centuries is that I have acquired a Kindle, Amazon’s e-book reader. There are plenty of out-of-copyright books that can be downloaded for it free of charge, and I have been reading (or re-reading) not only Lear but also Lewis Carroll, Conan Doyle and P.G. Wodehouse.

I have to say that in general I am delighted with my Kindle. Rather than lug kilos of printed books to the West Indies for my visit last month, I was able to carry a sizeable library electronically in this one lightweight device.

For extended reading, I find the Kindle much more comfortable than reading a computer screen. Its screen is not backlit, and does not flicker. You can’t read it in the dark, but you can read it in direct sunlight. The battery lasts for several weeks without recharging.

However… judging by my experience, there are still certain problems with fonts and with non-ASCII characters. Despite the claim that the device, in this third generation, “supports additional fonts and international Unicode characters” (Wikipedia), it seems that ordinary downloaded books can only be displayed in three typefaces: “regular”, which means Rockwell, “condensed”, which is the same thing condensed, or “sans serif”, which is Arial or something similar. No other fonts are available. You can, though, adjust the size, the line spacing and the number of words per line.

(I should say that it is also possible to display pdf files on the Kindle, and they can of course contain an unrestricted range of fonts, characters, and graphics.)

Anyhow, as I read Sherlock Holmes I soon came across things like this:As you can see, the non-ASCII characters é and à are replaced by garbage. (It is not easy to photograph a Kindle page with the digital camera I have available. Flash off, macro mode.)

OK, so that book was a freebie, and we can tolerate the odd glitch in something that is free of charge.

But I also paid good money for other books. One was Guy Deutscher’s The Unfolding of Language (which I recommend).

As you would expect in a book about linguistics, Deutscher makes use from time to time of phonetic symbols. Disaster! This is what happens to a humble schwa:and as for exotica such as š and ḫ…The special characters take the form of low-res graphics, of an inappropriate size and font. I resent having such typographical incompetence in a book I have paid good money for, but I suspect the fault lies with the publishers (Random House) rather than with the makers of the Kindle.

It can be done. Here is a fragment of a screenshot from another e-book (a freebie!), coping routinely with the characters ĉ and ŝ.
So wake up, other publishers venturing into e-books!

Tuesday 1 February 2011

obsolete exonyms

In Edward Lear’s Book of Nonsense, first published in 1845, there are two limericks that give interesting indications of how two familiar European placenames were pronounced in English at the time.
One concerns a seaport in Spain:
There was an Old Person of Cadiz,
Who was always polite to all ladies;
But in handing his daughter,
he fell into the water,
Which drowned that Old Person of Cadiz.
In Spanish this place is spelt Cádiz and correspondingly pronounced ˈkaðiθ, or locally ˈkaðis. As shown by his rhyme, Lear must have called it ˈkeɪdɪz. In contemporary English, though, it is usually called kəˈdɪz, which is scarcely closer to the Spanish than Lear’s version. Why have we modernized the pronunciation by switching the stress to the wrong syllable?

Places of this name in the USA may retain the earlier pronunciation.

The other relates to the city that is now the capital of the Czech Republic.
There was an Old Person of Prague,
Who was suddenly seized with the plague;
But they gave him some butter,
which caused him to mutter,
And cured that Old Person of Prague.
The Czech name is Praha, pronounced ˈpɾaɦa. Lear evidently said preɪɡ. This variant is mentioned by Daniel Jones in EPD as late as 1963: although giving only the pronunciation prɑːɡ, he adds the note “There existed until recently a pronunciation preiɡ which is now probably obsolete”.

In the case of Copenhagen, interestingly enough, the innovating form with -ˈhɑːɡ-, despite being popularized by Danny Kaye, has not succeeded in displacing traditional -ˈheɪɡ-. And we still call Den Haag / ’s-Gravenhage by the name The Hague heɪɡ.

I expect that Lear, like other early Victorians, would also have rhymed Rome with loom and tomb, Milan with Dillon, and Calais with Alice. Again, the latter persists right up to the 1963 EPD, where it is characterized as no more than “old-fashioned”.