Tuesday 17 May 2011


Those of you who read Jack Windsor Lewis’s PhonetiBlog will already know that the English Phonetic Society of Japan has just published a festschrift in my honour. It comprises a double issue of their journal English Phonetics. Thank you, EPSJ!

Jack has already given you some idea of the contents of the volume, and I refer you to his posting.

He also tells you that if you wish to purchase a copy (post free) you will need to remit ¥4500 to Mitsubishi Tokyo UFJ Bank, Fujigaoka Branch # 252, SWIFT code: BOTKJPJT, account number 1698177, account holder’s name Eigo Onsei Gakkai. You’d need to tell them your name and postal (mailing) address, too, and what you are purchasing.

One of the many interesting articles to be found there is by my colleague Patricia Ashby, and is entitled ‘The l-vocalization trend in young London English speech — growing or declining?’. Her evidence is drawn from a fairly small sample (nine speakers, all from inner London, plus three non-London ‘controls’) reading a structured set of sentences exemplifying word-final l in various phonetic environments (pre-consonantally, pre-pausally, pre-vocalically). Nevertheless it demonstrates some important points.
• All the London subjects sometimes vocalize final l even before a following vowel, as in I’d like to ask that girl out […ɡɛo ˈæoʔ]; and
• in all phonetic environments male speakers were at least twice as likely to vocalize dark-l as female speakers.

In discussing the background history of earlier l-vocalizations in English, Patricia mentions the name of the Piccadilly Line tube station Arnos Grove. The etymology of the first part of this name is believed to be Arnold’s, via an intermediate stage Arnol’s.

What she does not discuss is the question of how the Arnos of Arnos Grove is pronounced today. Given its etymology, we might expect ˈɑːnəʊz. The omission of the possessive apostrophe need not surprise us in the name of a London tube station, given other apostrophe-free Underground names such as Canons Park, Barons Court, Carpenders Park, Golders Green and Colliers Wood.

But I confess that I have always said the name to myself as ˈɑːnɒs, making it parallel to chaos, ethos and non-possessive names such as Amos and Carlos. (Note to Americans: I know that for you these names are ˈeɪməs and ˈkɑːrloʊs respectively. But in BrE they’re ˈeɪmɒs, ˈkɑːlɒs.)

I wonder what other Londoners do. Does anyone call it ˈɑːnəʊz Grove? I don’t think I’ve ever heard that, but then it’s not a part of London I often go to.

I checked on Forvo and find that the speaker there clearly says ˈɑːnɒs. So I’m not alone.
_ _ _

I shall now be away until the end of the month. Next posting: 1 June.

Monday 16 May 2011


(Image by xerones.)

David Deterding expressed some surprise at the LPD entry for comfortable.
The LPD entry for 'comfortable' seems to suggest that the preferred US pronunciation is: [kʌmftərbl]. Is that correct? Does the /r/ get moved so that it is next to the /b/? Actually, the schwa is raised, which means it is optional; and then I am confused. Does the second syllable potentially have a syllabic /r/?

I answered
The r-metathesized pronunciation for AmE is shown in plain type, not bold. That means that the preferred form is the same as BrE, with no r. Its existence is confirmed by Webster's Collegiate etc.

Yes, my convention in LPD is to represent 'syllabic r' (lettER) as raised schwa plus r, bringing it into line with the conventions for syllabic n and syllabic l. If the schwa is "optionally deleted" then the syllable spacing guarantees that the sonorant becomes syllabic. (Unless of course there's a compression mark, in which case it can revert to being nonsyllabic.)

David came back with
Golly, I had never appreciated the distinction between bold and not bold. And I had to look for quite a while to find an explanation of this (p. xvii). I wonder if it might be included in the "typographical conventions" on p. 922, at the end of the book? I guess it makes sense, but hmmm ...... I've been using the dictionary for years, and I never picked up that distinction between bold and not.

I’m sure he’s not alone. My only defence is that it was the publishers who insisted on reducing explanations to the absolute minimum. Since no one reads them anyhow, they said, don’t use up space on them. And they are not to be found anywhere on the CD-ROM.

So let me set out again what it says on p. xvii of LPD.
Many English words have a number of different possible pronunciations. Some of the users of LPD will be teachers and learners of EFL/ESL […], and will look for advice on how to pronounce a given word. For them one main pronunciation, printed in bold, is given at each entry. This is the form recommended for EFL purposes. […] If the BrE and AmE recommended forms are different from one another, then both are given in bold. Other users of LPD, especially those who are native speakers of English, will be interested not only to see what form is recommended but also what variants are recognized. Where pronunciations other than the main one are in common educated use, they too are included, but as secondary pronunciations, printed in ordinary black type. […]
(In earlier editions, and on the screen display of the CD-ROM of the current edition, the main pron is in colour, secondary pron(s) in black.)

Back to comfortable. The entry reads

ˈkʌmpft əb |əl ˈkʌmpf ət əb |əl || -ərb-, ˈkʌmpf ət̬ əb |əl, ˈ•ərt̬-

As usual, a raised symbol stands for an optionally inserted segment. Ignoring the optional epenthetic p and the choices between ə plus a sonorant and a syllabic sonorant, we can unpack the remaining abbreviatory conventions as follows.

main pron, BrE and AmE ˈkʌmftəbl̩
secondary pron, BrE ˈkʌmfətəbl̩
secondary prons, AmE ˈkʌmftɚbl̩, ˈkʌmfət̬əbl̩, ˈkʌmfɚt̬əbl̩

I wonder whether north American readers, in particular, are happy to see a recommended pronunciation with omission of the vowel corresponding to the -or- of the spelling. And how do you evaluate the transfer of the r into the penultimate syllable? And would you voice the t?

Friday 13 May 2011


Sorry today’s posting is so late. The Blogger site has been unavailable (except as read-only) for well over 24 hours, and has only just become available again.
_ _ _

The manufacturers of Scrabble™ have decided to allow players to use various new words when playing the game. Among them are pharm, wiki, badware, Facebook and also, bizarrely, catflap, heatwave and pushbike, which for some reason weren’t already in the Collins Official Scrabble Words dictionary.

As far as the British popular press is concerned, the most sensational addition was innit, our Multicultural London English universal tag question (as in “He’s right, innit?”).

I’m relieved to find that I’ve already got innit in LPD. Likewise keema, pushbike, wiki and thang, not to mention Myspace and Facebook (including its non-standard -buːk variant). I've also got pharma (as in "big pharma"), but not pharm.

For the next edition, do I really need to add catflap and heatwave? I've got pushbike. I suppose I’d better get round to adding tik, gak and non-proper-name tina. But first I’ll have to find out how people actually pronounce them. One assumes they’re just tɪk, ɡæk, ˈtiːnə.

Can anyone help me with aloo gobi? It’s an Indian dish which is widely on sale in our UK supermarkets, so I ought to know. Wikipedia tells us that it’s Hindi आलू गोभी, Urdu آلو گوبی ɑːluː goːbh, so my guess is ˌæluː ˈɡəʊbi.

Then there’s grrl: is this just a spelling variant of girl, or do people give it a special pronunciation? If so, what?

I feel like a dad (or perhaps rather a granddad) at the youth club disco.

Thursday 12 May 2011

endangered, exotic

Thanks to my stay in hospital last weekend I’ve finally got round to reading Mark Abley’s Spoken Here: Travels among Threatened Languages.

The author succeeds admirably in keeping us entertained with fascinating facts as he visits speakers of assorted endangered languages: Australian and Amerindian indigenous languages with only a handful of speakers, Manx and Provençal, Yiddish and Welsh, vividly bringing home to us the difficulty of being forced to live your life through a language that isn’t your own.

Despite all his energetic research (which includes mentions of Whorf, Chomsky, and Trubetzkoy), Abley is not a linguist, and is conscious of not being one. I must admit I had never realized how forbidding some of our familiar linguistic terminology can be to outsiders.

One of the things he presents to our admiring eyes as part of “an elaborate array of phonetic devices”, as used in the description of Yuchi (an endangered language of Oklahoma), is that Yuchi distinguishes between “a dorsal-palatal x (as in German ‘ich’) and a velar (as in German ‘ach’)”. Well yes, German has this ç - x distinction too, and so for that matter do many other not terribly exotic languages, including Modern Greek.

Continuing with the “elaborate array of phonetic devices”, you’ll be amazed to hear that Yuchi vowels can be “open, closed, lengthened, or nasalized”. Just imagine!

If he knew more phonetics, Abley might reasonably have mentioned that Yuchi has not only an aspiration contrast but also ejectives, so that we have p vs. ph vs etc. And it has a voiceless lateral fricative ɬ. Now that’s a bit more interesting.

Yiddish … has a linguistic feature with the menacing name “devoicing assimilation”: voiced consonants like b and d are “devoiced” by other consonants that follow on, so that the vocal cords no longer vibrate.

In no time at all Abley is running with a metaphor, of a speech community “devoiced” (prevented from making their voice heard) by their gradual assimilation into the majority language surrounding them.

The take-home message is one that I agree with whole-heartedly.
One irascible, bloody-minded, language-speaking son of a bitch is worth a few dozen well-meaning hobbyists.

It’s no use just talking about endangered or minority languages, in the way that Abley found enthusiasts for Provençal doing among themselves (but in French). You’ve got to speak them, sing in them, play in them, use them. That’s the only way to keep them from disappearing.

Wednesday 11 May 2011

double-stressed◀ adjectives

Petr Rösel asked, gnomically,
What's the stress 'patten' in long-legged? Like the one in long-haired or long-headed?

The question is odd because as far as I am aware the two compound adjectives long-haired and long-headed have identical stress patterns. (And was he making some subtle point about the nonrhotic pronunciation of pattern, homophonous for me with Patton, patten and paten?)

So long-legged is like both of the other long- compounds he mentions, namely ‘double-stressed’. Indeed this is the usual pattern for all adjectives of this structure, in which an adjective is joined to a noun, the noun usually being turned formally into the participle of a denominative verb by the addition of -ed or -ing: big-boned, broad-chested, cold-hearted, fine-sounding, good-natured, half-hearted, old-fashioned. It’s a very productive process, and because of this dictionaries typically list only a small selection of the compound adjectives that are available. Linguists happily write and say long-vowelled (as in a long-vowelled verb stem), though it’s not in any dictionary known to me.

These adjectives are most typically used not predicatively but attributively, before a noun which is likely to be accented. Therefore ‘stress shift’ kicks in, and we get an accent on the first element of the compound adjective, but probably not on the second. That’s why the dictionary entry in LPD and LDOCE includes the stress-shift mark ◀.

• a ˈlong-haired ˈlayabout
• a ˈfine-sounding ˈslogan
• a ˈhalf-hearted atˈtempt
• ˈold-fashioned ˈclothes

If the conditions for stress shift are not met, we get the basic pattern.

• ˈThis ˈPeter, | is he ˈlong-ˈhaired?
• Their ˈslogan | was adˈmittedly ˈfine-ˈsounding.

The problem (if it is a problem) is that some adjectives of this type are virtually never used predicatively. Consider long-haul. We speak of ˈlong-haul ˈflights and ˈlong-haul ˈpassengers. Do we also say ˈthis ˈflight | is ˈlong-ˈhaul? Yes, I suppose, but rarely.

That is the reason for the discrepancy that some have remarked on between different dictionaries. LDOCE shows long-haul as having initial stress, while LPD and CEPD (I think correctly) show it as double-stressed.

In German it’s different. Adjectives with this type of structure are stressed on the first element under all circumstances (I think): ˈkurzfristig, ˈlangweilig, ˈschöngeistig. And the newly triple-elled ˈschnelllebig — don’t you love the reformed spelling?

The other interesting question with long-legged is how many syllables there are in the second element. But that’s another issue entirely. (See LPD s.v. legged.)

Tuesday 10 May 2011

what's this l like?

Felix Chan had a question about clear and dark l. You’ll recall that the standard rule taught to NNSs is that l should be clear before a following vowel, dark elsewhere. This rule is fine for words in isolation, but what about connected speech? Felix said,
I don't think Debbie will agree.

I am wondering whether the 'l' in the word 'will' should be pronounced
as a dark l (since the preceding sound is a vowel?) or a clear l
(since the following sound is a vowel?).

Let me say right away that the nature of the preceding sound is entirely irrelevant. (You get clear l in black, but dark l in tables. The preceding consonant is b in both cases.)

Felix continues

I am even more puzzled when I listen to the examples in Longman
Dictionary of Contemporary English for the word 'capital':

'Washington D.C., the capital of the United States' I heard a dark l
in the word 'capital'.

'Hollywood is the capital of the movie industry' Here, I heard a clear
l in the same word 'capital'.

So how can this be? In reply I said
You can get either clear or dark l in this position (word-final before a vowel in the next word). Partly it depends on syntax (how closely are the words linked together?), partly on speech rate, partly on personal differences. Generally speaking I would use a clear l unless there was a major syntactic boundary between the words, or unless I paused at that point.

I must confess that that reply is not based on any evidence beyond tradition and my own introspection. The account in the current (7th) edition of Gimson’s Pronunciation of English, edited by Alan Cruttenden, has this to say (p. 216):
When an affix beginning with a vowel is added or the next word begins with a vowel (fiddling, fiddle it, finally, parcel of books), the lateral may remain as dark and may remain syllabic or become non-syllabic; alternatively the lateral may become clear , in which cases it is usually non-syllabic. The lateral is less likely to become clear in those cases where the following word begins with an accented syllable, where a [ʔ] may intervene, as in real ale [riːɫ ˋʔeɪɫ], cf. real estate [ˋriːl esteɪt].

I think it would have been better to deal separately with the question of loss of syllabicity (which I call ‘compression’), because this is sensitive to the strength of the following vowel — it is an option available only if the following vowel is weak. The possible insertion of ʔ (‘hard attack’) is also a separate issue.

And personally I wouldn’t be caught dead saying riːl instead of rɪəl for real — but now I’m showing my age, and I’m well aware that hardly anyone nowadays distinguishes real from reel in the way that I do.

I expect someone somewhere has carried out experimental measurements of clear vs. dark l in a variety of contexts, including unscripted connected speech, but I can’t refer you to any such research. Nor can I lay my finger on Abercrombie’s interesting observation, made half a century or more ago, that we use a clear l in feel in the sentence I feel ill, but in the sentence I may not look ill, but I do feel ill we use a dark one.

I also suspect that Abercrombie used, and I use, a clear lateral in some contexts where Cruttenden has a dark one. (Although not sounding in general like northerners, Abercrombie and I grew up in the northwest of England, whereas Cruttenden grew up in London.)

Felix was delighted with my answer, but still came up with a further poser.
In the phrase 'annual leave', I should pronounce a dark l then a clear l at the boundary between the two words?

We might also consider such examples as tell lots, full length, table lamp, feel lonely. I said
Theoretically, yes. In practice, in fast speech at least, you get assimilation making both parts of the lateral clear.

You may not all agree with this, and there are obviously social and regional differences. Again, has anyone ever made systematic observations relating to this point?
_ _ _

I’m sorry there was no posting to this blog yesterday. This was for reasons beyond my control.

Friday 6 May 2011


A correspondent from Hong Kong, in an email with the subject line “English Pronunication” [sic], asked a number of straightforward questions about articulatory details of English phonetics. But the last paragraph of his message had a sting in the tail.
Finally, thank you for writing … such a good book on English
intonation. I think it will be perfect if rise-fall and level tones
are included in the book as well.

What? They ARE included in the book!

I could not forbear to reply,
The rise-fall tone is covered in my intonation book (p. 217). In my view it makes sense to treat it as a subvariety of the Fall. Likewise, I treat the mid-level nuclear tone as a subvariety of the (low) Rise (p. 224). Did you read the book to the end?

Crestfallen, the correspondent thanked me.
I'm glad that your book (English Intonation - An Introduction) covers
the 2 tones. I've been studying the book for 2 years but I haven't
finished it yet. Today, I'm studying p.152.

Moral: before writing to an author to criticize his work for its perceived shortcomings, make sure you have finished reading it.

I don’t think I’ve written here or anywhere else about my decision to treat the English rise-fall nuclear tone as a subvariety of the fall tone rather than as an independent tone. In this my approach is different from that of Halliday, for example, who treats it as separate, while uniting high and low falls as the same basic tone. For me, the fall tone covers O’Connor and Arnold’s High Fall, Low Fall and Rise-fall. All have in common the physical characteristic of a nuclear pitch movement that either falls throughout or ends in a fall, ending always on a lowish pitch, with the tail (if present) having a low level terminal tendency. In terms of pragmatic meaning, all share the general characteristic of being what Brazil calls “proclaiming”, typically implying definitiveness / finality / assertion. Any sentence type may have a Fall, but with exclamations it is obligatory: they are always said with a Fall of some kind (including the possibility of a rise-fall).

• What a reˈmarkable \hat! (excited, enthusiastic)
• What a reˈmarkable \hat! (deadpan, unimpressed)
• What a reˈmarkable ^hat! (impressed, perhaps gossipy)

• * What a reˈmarkable /hat! (ungrammatical except as an echo question)

As anyone who has taught intonation analysis to NS students knows, the inevitable response to demonstrating the rise-fall is that the class breaks out into giggles.

Pedagogically speaking, I have no evidence that treating the rise-fall as a variety of fall helps NNSs understand what is going on. But it seems a reasonable thing to do. Learners should only tackle it after the basic tones (rise, fall, fall-rise) have been thoroughly learnt and understood. Otherwise there’s a serious danger of students confusing the rise-fall with the fall-rise. (For NSs who might confuse them, just ask whether or not it makes you giggle.)

That’s why I relegated discussion of the rise-fall to the back of the book, in the section devoted to Beyond the Three Ts: Finer Distinctions of Tone.

Thursday 5 May 2011


There’s an interesting piece on the BBC news website about stenography.

Actually, I think there’s some confusion in the article’s headline. The topic it reports on is actually, as I understand it, not stenography as such but more precisely stenotypy.

“Stenography” is a general term that covers all forms of shorthand — handwritten shorthand systems such as Pitman and Gregg (the systems that are, or rather were, predominant in the UK and the US respectively), as well as machine shorthand. “Stenotypy”, on the other hand, refers specifically to the machine shorthand produced with the aid of a “stenotype”, which the COD defines as
n. 1 a machine like a typewriter for recording speech in syllables or phonemes. 2 a symbol or the symbols used in this process.

As explained in the article, and in greater detail in Wikipedia, the stenotype keyboard has a mere 22 keys, yet encodes the spoken material a full syllable at a time. This requires “chording”, i.e. the simultaneous depression of several keys. The left hand records the onset consonant(s), the right hand the coda consonant(s), and the thumbs the vowel. Many sounds are recorded by arbitrary key combinations, e.g. initial d as TK, initial m as P plus H but final m as P plus L, “long E” () as AO plus E, and so on.
All English shorthand systems, as far as I am aware, including stenotypy, encode the pronunciation of the words direct, rather than via the traditional orthography. This means that all shorthand writers have to be experts in instant on-the-fly phonetic transcription — though of course they do not have to pay attention to the way the given speaker says the word on the given occasion but rather to the citation form.

Both Pitman and Gregg, and also I assume stenotypy, require nonrhotic speakers to record the symbol representing r in words in which rhotic speakers use r. In Gregg, which I learnt to 120 wpm as a teenager, stork is written differently from stalk (though they are homophones for me), but the latter is written identically to stock and for that matter stoke (though they are not homophones for me).

However there are also numerous shortened forms (logograms, “briefs”) which make the relationship between sound and shorthand spelling less direct.

Oh dear. The Wikipedia article goes on to tell us
Some court reporters use scopists to translate and edit their work. A scopist is a person who is trained in the phonetic language, English punctuation, and usually in legal formatting.
How often do we have to say it? A writing system is not a language. English is still English, no matter whether you record it in traditional orthography, in morse code, in IPA, in shorthand, or as a .wav or .mp3 file. Converting ordinary spelling into one of the other forms is not translating it into a different language.

Wednesday 4 May 2011

Aussie dogs

In a newspaper story of the kind you’d expect to find in the doldrums of August or on April Fool’s Day rather than in a news-filled month such as we have just experienced, the Adelaide Sunday Mail tells us that
a restaurant that refused a blind man entry because it thought his guide dog was "gay" has been ordered by the Equal Opportunity Tribunal to pay him $1500.

What was strikingly missing from this story, as I read it, was a discussion of why and how the misunderstanding could have arisen. Pretty obviously, I would have thought, it must have been because the expression guide dog, pronounced in a mainstream English way as ɡaɪddɒɡ — with dd standing, as usual, for a geminated , with no release of the word-final consonant and no onset of the following word-initial one — was interpreted as the Australian sequence ɡaɪdɒɡ, where is the quality associated with the FACE vowel rather than with the PRICE vowel. So gui(de) was misheard as gay.

As is well known, Australian English, ‘broad’ and ‘general’, has undergone a diphthong shift like that of Cockney, with the onset vowel qualities of FACE, PRICE and CHOICE shifted anti-clockwise. This makes Australian FACE sound like the PRICE of most other accents.
Most readers will have heard the joke about an Australian patient who said “I came into hospital just [təˈdaɪ]” (today), only to be treated as a suicide risk.

Interestingly, the waiter who misheard guide dog may well, like the restaurant owners, have been Vietnamese. We are not told the origins of the customer’s partner who uttered the words. Perhaps she was a posh Australian who had an unshifted PRICE diphthong.

In other relevant and equally world-shattering news, Australians are apparently now complaining about their prime minister’s voice. (To an outsider such as me she just sounds typically Australian.)
What is really grating on a lot of Australians is the heavy ''Gillard twang''. She regards this as central to her distinctly Australian speech manner and she's proud of it. Yet it is weighing down her communication efforts and dominating the impression she leaves with the public. […] What Gillard regards as identifiably ''Australian'', the public largely regards as annoying.

What I regard as annoying is not Ms Gillard’s voice but the sort of advice offered by her vocal coach, one Dean Frenkel, who is “an overtone singer and author of the forthcoming book Evolution of Speech”, as well as being the holder of the world record for singing a continuous note.

I just don’t know what Mr Frenkel means by such things as
Vowel articulation - ''e'', ''i'' and ''o'' should be exercised in a far more understated way. No over-articulating of vowels.
Is he really advocating a reduction in clarity? Or is he asking for more vowel reduction, more weak forms?
More lightness - there's too much gravity in her voice. Add some occasional lightness that taps into a greater range of melody and more frequent higher melody. This would raise her energy and sound more natural.
Whatever his qualifications as a vocal coach, he doesn’t seem to know much physics.

Tuesday 3 May 2011

the evidence of the vows

Things change. Or perhaps I was just wrong. In 1982 I wrote that for the GOAT vowel
RP characteristically lacks the pre-/l/ allophone [ɒʊ] of many other {British} accents. (AofE p. 237)

Twenty-five years later I was able to report that Bente Hannisdal, in her admirable PhD thesis (blog, 25 April 2007), had found that in her investigation of the pronunciation used by British television newsreaders,

A correspondent wrote to me ten days ago asking for advice on suitable sound clips to demonstrate this change. I replied as best I could at the time.

Now, though, last week’s royal wedding has furnished us with an iconic example of before and after. Watch it here (at 1:25 into the clip), here (4:05 in), or below.

In the Abbey, as the bridegroom makes his vows, first the archbishop of Canterbury and then Prince William utter the words
to have and to hold from this day forward, for better for worse, for richer for poorer, in sickness and in health, to love and to cherish, till death us do part …

The archbishop, as is appropriate for an RP speaker of his age (born 1950), pronounces hold as həʊɫd. William repeats the word, but as hɒʊɫd, which is what is now arguably appropriate for an RP speaker of his age (born 21 June 1982). Here they are.

Hannisdal’s newsreaders constitute one kind of evidence for current RP — RP seen as the model of widely acceptable non-regional English-of-England pronunciation. The royal family provide evidence of a different kind — of RP seen as the de facto pronunciation of those at the very top of the traditional social hierarchy. Either way, it’s clear that the ɒʊ allophone before dark l is now a regular part of proper, pukka, echt, RP.

At least for those born, like the Duke of Cambridge, since those far-off days when I typed the manuscript of Accents of English.