Monday, 30 April 2012

did I hear you aright?

If you hear a word spoken, but then want to write it down, you may be in some uncertainty how to spell it.

At least in English, that is. If we had a thoroughly transparent spelling system, you might think, it wouldn’t be such a problem.

I remember some years ago when I was working as a pronunciation consultant on a dictionary project there was a young lexicographer who suggested that we ought to add the missing word apeth, since it was in general use in expressions such as ‘a daft apeth’. (She may have thought that it literally meant some kind of baby monkey.) We had to tell her, gently, that the established spelling of this word is ha’p’orth, or in full halfpennyworth, with the pronunciation ˈheɪpəθ, and that it refers to the British predecimal coinage, superseded before she was born. (Embarrassingly for her, her misapprehension also made it clear that she came from a social background that was not confident about which words are said with h and which not. We assured her it didn’t make an ‘apeth’ of difference.)

But it’s also clear that if you hear an unfamiliar word spoken you may not hear it in exactly the same way as the person who said it intended. That could lead to spelling it wrong (or differently) even in the most perfectly phonemic orthography.

The feck in feckless is, according to the OED, “apparently aphetic < effeck, variant of EFFECT n.”. Rather than ɪˈfekt, someone must have heard effect just as fek. They then spelt it correspondingly as feck, which subsequently was adopted by everyone in feckless.

I suspect there was a similar process in the development of flimsy from film(s) + -y (which the OED strangely labels an ‘onomatopoeic’ formation). In fact film is regularly pronounced flɪm in Jamaican Creole (as well as ˈfɪləm in Irish English, which is a different matter).

There is a tour guide in Montserrat who leads hikes into the rain forest. He is very knowledgeable about the local plants and animals, and particularly the birds. He has a Facebook page on which he regularly posts pictures of interesting things seen along the hiking trails.

Here’s something he posted a few days ago.

I’ve been on several hikes that he led, so I can confirm his skill and ability. He certainly knows the scientific terms fungus, fungi, and could probably tell you the full Latin name of the species here illustrated. So it’s interesting that he spells mushroom as mashoom. (“Reserve” is obviously meant to be “reverse”.)

It’s possible that this is just a joke, consciously reflecting the local pronunciation of the word. Or it might be that he really thinks that the word is ˈmæʃʊm, and spells it accordingly.

Actually, Montserratians don’t eat fungi. No restaurant will offer you a mushroom omelette, and you can’t buy mushrooms in the supermarket. Their word for a mushroom or toadstool is jumbie parasol, literally ‘ghost umbrella’. So mushroom, for a Montserratian, is presumably a word heard only from expats and visitors to the island. This could account for our guide’s uncertainty how to spell it.

The local STRUT vowel is typically back and rounded: we could represent it narrowly as ɔ̈. You can see that the AmE, and particularly the BrE, STRUT vowel might well be heard as the local TRAP vowel (= something rather more central than cardinal a). And you might well overlook the r in ʃr.

Mashoom soup, anyone? Sorry, that’s crème de champignons.

Friday, 27 April 2012


For my sins (‘especially BrE’, according to LDOCE) I sit on the Steering Committee of the Government of Montserrat’s UK office in London.

At yesterday’s meeting we had a presentation from someone introducing a discussion on the future of the island’s magnificent Cultural Centre.

Sir George Martin of international music fame built it in 2006 as a gift to Montserratians. This is the same George Martin who established a state-of-the-art recording studio, Montserrat Air Studios, at the historic Waterworks estate in the 1970s; it attracted international stars of the calibre of Elton John and added to the island's fame as a welcoming isle. The new cultural landfall cost nearly US$3 million and is a now famous venue for local and international conferences in addition to being a multi-purpose performing centre.

The reason our speaker had come to discuss the matter with us was that he was anxious to consult all the stakeholders (older readers will groan at this word). As well as people living in Montserrat, performers, DfID etc., ‘stakeholders’ includes Montserratians living in the diaspora. Which he called the diˈæspərə.

In LPD I give two versions of the pronunciation of diaspora, but in both cases with in the first syllable. They differ in stress: daɪˈæspərə and ˌdaɪəˈspɔːrə. I discussed the issue of the stress pattern of this and similar words in a posting in this blog over five years ago (blog, 20 Jan 2007)

And then there is the word diaspora. It has an etymologically short penultimate o (Greek διασπορά diasporá, ‘sowing around, scattering’), and a corresponding traditional English pronunciation /daɪˈæspərə/. But I recently heard someone pronounce it /ˌdaɪəˈspɔːrə/. The spelling doesn’t tell you whether the o is long or short: and that’s the factor that determines the stressing. Because of the Latin stress rule.

Given that we’re talking about antepenultimate stress, ought I to add to LPD our speaker’s version with i in the first syllable?

English weakens to i before a vowel or word-finally and to ɪ ~ ə before a consonant. So for strong~weak alternation in di- we can compare words such as dilemma daɪˈlemə, dɪˈlemə, direct daɪˈrekt, dəˈrekt, where weakening is not uncommon.

Exploring other words with unstressed prevocalic di-, I find diaconal, di(a)eresis, dianthus, diaphanous, diaphysis, diastole, diathesis, diatomite, diazepam, Diogenes, Dioscuri, diotic, dioxide, dioxin, and diurnal. For all these — none are what you would call everyday words — I give only daɪ, with no weakened variant. Does anyone in fact weaken to di- in any of them?

Compare however the proper names (San) Diego, Dieppe, where di- is the only possibility. But Diana always has daɪ-.

Thursday, 26 April 2012


Lipman asks about the possibilities of
syllable-initial (love that word) t-glottalisation. Also, on television, I heard somebody say nʌɪnʔɪjn nineteen. Where does this occur? The speaker didn't sound as if he artificially tried to sound folksy.

I don’t think anyone ever says *ʔiː for tea, *ˈʔɑːɡɪʔ for target, or *ʔreɪn for train, do they?

The best I can think of as slightly a less implausible candidate for down-market word-initial t-glottalling would be somethinɡ like ?ˈsiː jə ʔəˈmɒrə see you tomorrow. Has anyone ever heard that?

It’s not much different for word-internal, putatively syllable-initial t-glottalling. I don’t think anyone ever says *əˈʔæk for attack or *rɪˈʔɜːn for return (do they?). On the other hand I think I have heard ɡɪˈʔɑː for guitar, while ʔeɪʔiːn for eighteen (adjust the vowel qualities to please) is indeed widespread among the kind of speakers who use a glottal stop in words such as water and city.

Two of the teen numerals are obviously irregular: thirteen and fifteen do not fit the pattern of fourteen, sixteen, seventeen and nineteen.

The missing one, eighteen, is also irregular for RP speakers and many others, since it has single t, thus ˌeɪˈtiːn, despite the two ts of its transparent morphology (eight eɪt plus teen tiːn). On the other hand some speakers/accents, both British and American, have regularized the position by pronouncing it ˌeɪtˈtiːn.

(The teen numerals are all lexically double-stressed, and therefore susceptible to ‘stress shift’ in running speech. Let’s not get sidetracked into discussion of stress patterns, which is a different issue.)

The first t in this ˌeɪtˈtiːn is an obvious candidate for glottalling, just as the first t in nighttime. I suppose that if you manipulate the usual rule ordering so as to do glottal replacement before double consonant simplification you could explain the possibility of ˌeɪʔˈiːn. However, this would involve ordering a postlexical rule (glottalling) before a lexical rule (simplification), which cannot be right.

But then again there are speakers who have double tt (possibly realized as ʔt) in thirteen and fourteen. And, as you would then expect, there are some who reduce this cluster to a simple ʔ, giving ˌθɜːʔˈiːn, ˌfɔːʔˈiːn.

And, as Lipman noted, this can extend to nineteen, too. As well as ˌnaɪnˈtiːn, we can get ˌnaɪntˈtiːn ~ ˌnaɪnʔˈtiːn and even ˌnaɪnʔˈiːn.

With double tt ~ ʔt or bare ʔ I think you usually get pre-fortis clipping of the sonorant part of the first syllable (eɪ, ɜː, ɔː, aɪn), which might be taken to imply that there has been resyllabification, giving a fortis consonant as the coda of the first syllable, thus ˌeɪʔˈiːn, ˌnaɪnʔˈiːn etc.

I’m really not sure how to explain all this.

Wednesday, 25 April 2012

mayoral elections

London is in the throes of electing a new mayor.
Reporting on the ITV news last night, the television presenter Alastair Stewart introduced an item on this topic as being about London’s meɪˈɔːrəl election, only to segue immediately into calling it the ˈmeərəl election.

There’s a well-known disagreement about how we pronounce the stem from which the adjective mayoral is derived, namely the noun mayor. In the UK (or in England, at least), we pronounce it as a homophone of mare. Both are monosyllabic, with the just SQUARE vowel, thus meə (or some might prefer to write mɛː). In the US, on the other hand, it is commonly disyllabic and rhymes with player, thus ˈmeɪɚ; though you do also get a monosyllabic variant mer (i.e. a homophone of mare), particularly when immediately followed by a proper name.

The OED offers a lengthy historical discussion of the pronunciation of the word. The nub is that

A disyllabic pronunciation existed in Middle English, where it was a variant of a more common monosyllabic one. … The disyllabic pronunciation survived in Britain at least into the 17th cent. … as one possible pronunciation, but other sources of similar date show that this was by then highly conservative in British usage. In North America, however, disyllabic pronunciations appear to have remained current in all periods.

I’ve sometimes wondered whether Mare Street in Hackney in northeast London ought really to be spelt Mayor Street.

Back to mayoral.

  • In LPD I give the BrE mainpron as ˈmeərəl, with an altpron meɪˈɔːrəl. For AmE I give just ˈmeɪərəl. (With hindsight, I ought to have included AmE meɪˈɔːrəl too, perhaps as the AmE mainpron.)
  • CPD/EPD, 18th edition, gives BrE ˈmeərəl, AmE ˈmeɪɔːrəl (sic), with no variants.
  • OPD and the online OED give BrE ˈmɛːrəl, AmE meɪˈɔːrəl, ˈmeɪərəl.
  • My main American reference dictionary, Webster’s Collegiate, 11th edition, gives three pronunciations, the equivalent in IPA of ˈmeɪərəl, ˈmerəl, meɪˈɔːrəl.
  • Forvo has BrE ˈmeərəl, AmE meɪˈɔːrəl.

Time for a survey, perhaps; but even a preference poll isn’t going to reveal inconsistent usage like Alastair Stewart’s.

Tuesday, 24 April 2012

perfect, but accented

Writing about yesterday’s proceedings at the Leveson inquiry into phone hacking, the Guardian’s reporter Michael White describes the Russian owner of the London Evening Standard and the Independent, Evgeny Lebedev, as ‘speaking perfect, accented English’.

This made me think. Would it be possible to claim that someone spoke perfect, grammatically flawed, English? I don’t think so. Could you describe someone as speaking perfect English with inaccuracies in vocabulary? Again, I don’t think so.

How, then, is it that someone can speak perfect English, but with mistakes in pronunciation?

Why is it that ‘perfection’ tolerates L1 interference in phonetics, but not in grammar or vocabulary?

Supporters of the communicative approach to language teaching, and followers of Jennifer Jenkins and her ideas on the phonology of English as a Lingua Franca, would probably say that an insistence on native-speaker-like phonetics is irrelevant. All that matters is that you should be understood.

(But you may have seen the reports of the trial that had to be abandoned, wasting £25,000, because the interpreter's pronunciation of the word beaten was mistaken for bitten by the jury.)

No doubt the reason we tolerate some L1 interference in pronunciation is that our articulatory habits tend to be fossilized once the critical period for language acquisition has passed. We’re not good at acquiring the motor control needed to articulate a new language in a native-like way. Lexis, on the other hand, merely needs to be learnt, and we all — native speakers included — continue to extend our vocabulary throughout our lives. Grammatical ability has some similarity to doing mathematics: you just have to learn the rules and then apply them. On the whole, good adult language learners can manage this.

But this general picture isn’t valid for everyone. Thinking about my own case, I realize that I can speak German pretty fluently and with (I think) virtually native-like pronunciation. But my vocabulary range is severely limited compared with that of a NS, and I am conscious of many grammatical shortcomings, e.g. relating to the gender of nouns. (Petr Rösel, aka Kraut, can attest to whether or not this is an accurate assessment of my abilities in his language.)

I wouldn’t describe myself as speaking ‘perfect’ German. In terms of the Common European Framework of Reference for Languages I don’t think I could honestly claim to be beyond B1 level.
Can understand the main points of clear standard input on familiar matters regularly encountered in work, school, leisure, etc. Can deal with most situations likely to arise whilst travelling in an area where the language is spoken. Can produce simple connected text on topics which are familiar or of personal interest. Can describe experiences and events, dreams, hopes & ambitions and briefly give reasons and explanations for opinions and plans.
But at least I can pronounce it well.

Language ability isn’t just a matter of getting the pronunciation, the grammar and the vocabulary right. You also need to know what to say and in what circumstances: communicative ability and cultural knowledge.

I wonder if Mr Lebedev’s English really is at level C2.
Can understand with ease virtually everything heard or read. Can summarise information from different spoken and written sources, reconstructing arguments and accounts in a coherent presentation. Can express him/herself spontaneously, very fluently and precisely, differentiating finer shades of meaning even in the most complex situations.
— except for the pesky matter of pronouncing it in a native-like way.

Wednesday, 18 April 2012

a new weak form?

Petr Rösel asks
Do you still believe there’s no weak form for couldn’t (as you state in LPD)?
To which my answer is yes, I do so believe.

His query arises because he thinks he’s heard people saying kədn̩t for couldn’t.

This is entirely likely. In many English northern- and Midlands-flavoured accents you could reasonably call ə the customary pronunciation of the FOOT vowel (usually merged with STRUT). Even in RP and the south of England, the English FOOT vowel does seem to be getting more central and unrounded, more like ə.

We hear this in all the FOOT words: put, bush, full, butcher, cushion, pudding, bullet, good, wood, cook, look, wool, woman, wolf, bosom…. For the speakers we are thinking of, all these words have ə. And so do could(n’t), would(n’t), and should(n’t).

But the fact that millions of people say (ˈ)ʃəɡə for sugar does not mean that sugar has a weak form, any more than do pudding, cook, or butcher.

What it does mean is that for the handful of FOOT words traditionally described as having a weak form it may be difficult or impossible to detect any difference between the strong form and the weak form. This handful comprises could, should, and would. It means that for these speakers these words have no phonetically distinct weak form: they are always kəd, ʃəd, wəd, no matter whether the phonetic environment calls for a strong form or a weak one.

So if you hear aɪ ˈdənəu ɪf aɪ kəd ˈduː ɪt I don’t know if I could do it it might well be true that the difference between the putative strong and weak forms of could is not detectable.

But if you also hear aɪ ˈdʒəs kədn̩t ˈduː ɪt I just couldn’t do it, you can’t on that basis say that couldn’t, too, has suddenly acquired a weak form (which happens to coincide with its strong form).

We usually speak of weak forms of words only when they are paired with strong forms, the two being phonetically different. Thus we say at has a weak form ət alongside its strong form æt, and we can formulate principles for the use of one or the other.

There is perhaps a parallel in the case of the pronoun it. Most of our pronouns have paired strong and weak forms, thus he hiː — hi, you juː — ju, jə, us ʌs — əs, them ðem — ðəm (and for some speakers also əm ’em). On general grounds you might expect it to undergo weakening in the same way: yet it remains as ɪt. There is no distinct weak form. That is because in RP and similar accents ɪ happens to belong to both the strong and weak vowel systems: KIT coincides phonetically with the vowel in packet, rabbit. So you could say (though I wouldn’t) that it does have a weak form, but that it happens to coincide phonetically with its strong form.

In Australian English, on the other hand, where ɪ does not occur as a weak vowel except in certain restricted phonetic environments, and rabbit is pronounced ˈræbət, it behaves exactly as expected, having ɪt as its strong form and ət as its weak form. Pack it and packet remain homophones.
_ _ _

I’ve got a busy few days coming up. So the next posting will be on 24 April.

Tuesday, 17 April 2012


In the boys’ adventure stories I would read sixty years ago, the baddies often spoke in ‘guttural’ tones. Not only German and Arabic, but also Welsh, Hebrew, and Dutch are still occasionally referred to popularly as ‘guttural’ languages.

A hundred and fifty years ago ‘guttural’ was in use as a technical term in phonetics, as a descriptor for a place of articulation. In 1869 A.J. Ellis wrote
The guttural nasal seems to have been the regular pronunciation of ng in English.

A helpful note in the OED tells us:
By non-phoneticians any mode of pronunciation which is harsh or grating in effect is often supposed to be ‘guttural’; with this notion the designation is popularly applied by English-speakers to the German ch, but not to k or g, though technically it belongs equally to them. As a technical term of phonetics, the word was first used to denote the Hebrew spirant consonants א, ה, ח, ע; it is now commonly applied (inaccurately, if its etymological sense be regarded) to the sounds formed by the back of the tongue and the palate, as /k/ /g/ /x/ /ɣ/ /ŋ/.
(I have supplied the OED’s missing closing bracket in the last sentence, and replaced its wrong symbol ɤ by the correct ɣ. O tempora, o mores!)

As a phonetic term ‘guttural’ has now been entirely supplanted by more precise terms: mainly ‘velar’, but also ‘uvular’, ‘pharyngeal’ and ‘glottal’, as appropriate. The non-technical use of ‘guttural’ has declined in parallel, and has not been replaced by ‘velar’ or anything other phonetic term.

The OED’s first citation for ‘velar’ in the phonetic sense is dated 1876. (Its definition for the noun sense of this word, though, still reads ‘a velar guttural’.)

Here is part of the Google ngram for ‘guttural’ (blue) and ‘velar’ (red). Follow the link to view the whole graph. As you see, the crossover point at which ‘velar’ became more frequent than ‘guttural’ was as recent as the late 1960s. But if we follow the trend lines for 1990-2000 and extrapolate for 2000-2010 we might well find that it has now lost that lead.

Monday, 16 April 2012

postsibilant contraction

Why did I find this newspaper headline (right), from last Wednesday’s Metro, awkward?

As we all know, the English spelling apostrophe-s can represent not only the possessive ending but also the contracted form of is or has.

It is pronounced just like the regular plural ending.

• After a voiceless non-sibilant consonant it is pronounced voiceless, s:
    the cat’s whiskers
    the cat’s waiting to go out
    the cat’s just been sick again

• After a voiced non-sibilant consonant or a vowel, it is pronounced voiced, z
    dressed up like a dog’s dinner
    the dog’s jumping up and down
    the dog’s run off somewhere

• And after a sibilant (= one of s z ʃ ʒ tʃ dʒ) it is pronounced as a separate syllable, ɪz (or for some people əz):
    straight from the horse’s mouth

…er, but what about the contracted forms after a sibilant?
    ?%the horse’s grazing happily in the paddock
    ?%the horse’s bolted

I don’t think I’m the only one who finds the last two a little awkward. Why? Because they don’t imply any difference in pronunciation from horse is, horse has. There’s no separate contracted form in pronunciation, so we don’t want one in spelling.

Those of us with a robust weak ɪ — ə distinction will continue to use ɪz as the uncontracted weak form of is but əz as the uncontracted weak form of has. This distinction remains even if, questionably, we spell both as postsibilant ’s. That’s another reason the spelling ’s is not very satisfactory here.

My cash is mine, fine. My cash’s mine, not so sure.

Friday, 13 April 2012

implosives and ejectives

In modern phonetic terminology, an ‘implosive’ is a sound made with a glottalic ingressive airstream mechanism. That is, the airstream is initiated by sharply lowering the glottis, thereby creating negative pressure in the supraglottal cavity. However implosives are typically voiced, with simultaneous vibration of the vocal cords through pressure from the pulmonic cavities. So you could say that they have a mixed glottalic ingressive and pulmonic egressive airstream mechanism. The IPA chart lists five symbols for implosives, all voiced: ɓ ɗ ʄ ɠ ʛ. There are also voiceless implosives, produced purely with the ingressive glottalic airstream mechanism.

Examples from relatively familiar languages are the ɓ ɠ of Zulu and the ɓ ᶑ ʄ ɠ of Sindhi (the second of these is retroflex). According to Ladefoged and Maddieson (The Sounds of the World’s Languages, Blackwell 1996), the Igbo labialvelar spelt gb, often considered a voiced pulmonic plosive with double articulation, i.e. simultaneous b and g, is more accurately described as a voiceless bilabial implosive, ɓ̥.

In the WALS data, implosives are found in 75 of the 567 languages studied (= 13%), most of them in Africa or southeast Asia. (Ladefoged and Maddieson, however, say that implosives are found in “about 10% of the world’s languages”.) By far the commonest implosive is the voiced bilabial ɓ.

The OED reveals that as a phonetic term ‘implosive’ goes back at least as far as Sweet, who wrote in 1890 that
Some sounds are produced without either out- or in-breathing, but solely with the air in the throat or mouth. The ‘implosives’ are formed in the former, the suction-stops or ‘clicks’ in the latter way.

There is one other usage you may occasionally come across in older works, in which an ‘implosive’ stop is an ordinary pulmonic stop with no audible release stage, or in which the release is not taken into account, as against an ‘explosive’ one which has no audible approach stage, or in which the approach stage is not taken into account. The OED has a quotation from Bazell in 1953:
If all initial occlusives are explosive and all final occlusives are implosive, it is obvious that two distinct conventions (explosiveness of initials and implosiveness of finals) need not be postulated.
We can regard this meaning as obsolete.

We pair implosives with ‘ejectives’, sounds made with a glottalic egressive airstream mechanism. They have an airstream initiated by the raising of the closed glottis, which compresses the air in the supraglottal cavity. The OED’s first citation for this term is Daniel Jones in 1932.

In IPA ejectives are written with an apostrophe diacritic, thus pʼ tʼ kʼ etc. They are “not at all unusual sounds, occurring in about 18 percent of the languages of the world” (L & M). Although they are not contrastive in English or any other European language, many speakers in Britain use ejectives as optional phrase-final variants of p t k.

I made a Google ngram for these two terms. Rather mysteriously, it shows implosive ejective peaking in 1900, while ejective implosive does not come onto the scene until 1920, peaking in 1970. I wonder what the explanation for this is.

Thursday, 12 April 2012

booking hotels

The travel booking website Expedia is running a series of ads in the British newspapers (and possibly elsewhere — I don’t know). The aim is to encourage us to book hotels enabling us to explore different parts of Britain and Ireland.

They feature respellings such as might be appropriate for the accents/dialects of various different geographical regions. This one obviously wants to target people visiting London. It features
• rhyming slang, a well-known feature of traditional Cockney. I’ve never heard these versions of hurry and money, but they’re perfectly plausible.
• h-dropping, found throughout working-class England, but stereotypically associated particularly with Cockney.
more respelt as maw, which implies no change in pronunciation for Londoners, since both are mɔː ~ mɔə.
All these features are also found elsewhere, but happily evoke London.

But what about this?
In case we’re puzzled, there’s a giveaway: they’re talking about Dublin. But in one detail it seems to me to be somewhat inept.
• the respellings cast and tap, for cost and top, reflect the fact that unrounded vowels in the CLOTH-LOT set are typically to be found in Ireland (or of course the US and Canada), not in England. Fine.
nex for next, represents cluster reduction kst→ks before a following consonant in connected speech. This is found in virtually all kinds of English, including RP.
nuttin for nothing reflects (i) the use of an alveolar n rather than a velar ŋ in the -ing ending, and (ii) TH stopping (θ→t). The first is found as a “low” variant throughout the English-speaking world (except, apparently, among white South Africans), while the second is indeed a characteristic of Irish English (though also of Caribbean English, NYC etc).
noight for night. This appropriately suggests a lower-class Dublin ɒɪ PRICE vowel, though it could apply equally to London or Birmingham.
oyt for out. Who uses ɔɪ or similar in MOUTH words? Not Dubliners, not anyone much in the Irish Republic (sorry, the Republic of Ireland). Rather, qualities such as ɑy, ɒɨ are associated with the distinctive accent of Northern Ireland, particularly perhaps with Belfast. Not Dublin. Or am I wrong?

Wednesday, 11 April 2012

turned v

Something of a cri de coeur today from Simon Gilman, who is about to take up a new post in Voice at a university drama department in the north of England. Forgive me if I quote part of his email at some length.
Phonetics, as taught by Roach et al, still employs symbols for English RP that are at some variance with the sounds used contemporarily when set against the standard IPA vowel chart. It makes no sense to me that, for example, ʌ is used for the STRUT vowel when on the IPA chart the cardinal usage of this symbol is far further back and nearer in pronunciation to American (perhaps South-Western?). I think you previously agreed with me that the vowel that the contemporary RP user employs is far more like ɐ. Why then does Roach place ʌ in his chart more or less where one would find ɐ on the standard chart? I can’t use phonetics with my acting students and explain that ɔ is a rounded open-mid back vowel and show them on the standard IPA chart where the cardinal exists, and then explain that its unrounded equivalent is actually articulated as a near-open central vowel. The problem exists because I want to use contemporary RP as a fundamental phonetic benchmark from which to use phonetics to transcribe other accents, but I can’t fight against existing texts and teaching standards in isolation.

You’re right, Simon. Practical necessity indeed means we all have to use existing textbooks and pronunciation dictionaries. We ought not to deviate from their transcription systems without very good reason.

If you “want to use contemporary RP as a fundamental phonetic benchmark from which to use phonetics to transcribe other accents”, then don’t start by teaching students the symbols for cardinal vowels. Teach them the vowel symbols for RP.

If and when you move on to teaching the cardinal vowels, please bear in mind that they constitute an abstract reference framework and not the vowel system of any particular language or accent. They were intended to reflect the most extreme positions of which the human vocal apparatus is capable. In applying them to real-life situations we always have to compromise, since no known language is spoken exclusively with the cardinal vowels or a subset among them.

Symbols mean what we choose to make them mean. If you want your students to make use of existing textbooks and pronunciation dictionaries, you have got to teach them that the STRUT vowel in standard accents is transcribed with the symbol ʌ. Later, if you want to study the differences between Jonesian RP, contemporary son-of-RP, various kinds of AmE, and assorted north-of-England English STRUT-FOOT fudges, you can decorate ʌ with interesting diacritics, or indeed choose another symbol or symbols instead. But in doing so you will be exploring a great multidimensional vowel space, in which the cardinal vowel system can do no more than help us feel our way.

When I first studied phonetics, under John Trim at Cambridge, we were set transcription exercises, i.e. asked to convert English orthographic texts into phonetic symbols. We had been armed with a list of symbols and keywords. But as well as doing the task in the way he was expecting, I also amused myself (or was showing off) by transcribing it in accordance with various other transcription systems I had come across or invented myself. So I wrote love as lʌv as taught, but also in another version as lə́v, because I had read Gleason’s Introduction to Descriptive Linguistics, and in yet another as lɐv, following the same logic as Simon. I did the same sort of thing for various other vowels. This game is fun for beginners fascinated by phonetic symbols, as I was, but impractical for drama students who just want help with accents and dialects. (For more of my early experimentation, see my blog for 16 March 2007.)

When I grew up, I settled down to writing RP STRUT as ʌ, just like everyone else. The brief answer to Simon’s first and main problem, then, is that if he wants his students to be able to consult existing textbooks and pronunciation dictionaries he should teach them that ʌ stands for the vowel in RP/GenAm strut, love, double, rug etc. (At a rough guess I imagine that would account for over 99% of the use of this phonetic symbol in published materials, with the cardinal ‘open-mid back unrounded’ value accounting for the remaining 1%.) Anyhow, you can see how I coped with your general problem by looking at what I did in my Accents of English (CUP 1982).

Sooner or later I imagine there will be a tectonic shift in the notation of English vowels. As not only STRUT but also GOOSE, THOUGHT-NORTH-FORCE and various other vowels move further away from the cardinal qualities associated with the symbols we currently use, some author will bite the bullet and provide us with an entirely new transcription system. Geoff Lindsey (blog, 12 March 2012) has already had a go.

Until that happens, stick with what we’ve got.

Tuesday, 10 April 2012

BAAP miscellanea

More reports from BAAP:

• Calbert Graham (pictured) and Brechtje Post found that even at a low level of proficiency Japanese learners can generally assign English word stress appropriately. They can also perform “the durational and pitch feature contrasts of English stress” in much the same way as native speakers. However “results for intensity suggest a non-use of this feature in realising stress, even by advanced learners”. I don’t know whether this has any practical implications for teachers of EFL to the Japanese. I suspect that we NSs hear the stress contrasts produced by Japanese learners perfectly well, despite the missing variability in loudness.

• Arabic ḍād ض is nowadays typically realized as a pharyngealized voiced dental plosive or fricative ðˤ. In medieval treatises, however, it was described as a lateral, and may perhaps have been an ejective affricate t͜ɬ̕ˤ or voiced fricative ɮˤ. Lateral articulation of ḍād had been thought to be obsolete in modern Arabic, until in the early twentieth century scholars began to report its existence in parts of southern Arabia. Barry Heselwood and colleagues presented electropalatographic evidence from southwest Arabian dialects attesting to both and ðˡˤ realizations, the latter being a central fricative accompanied by unilateral escape.

• Kathleen McCarthy and UCL colleagues have been looking at the perception and production of English stops and vowels by London Bengali children. They call these children ‘sequential bilinguals’, since they typically start with Sylheti at home before starting to use English at about 3½ years. The phoneticians tested the children’s ability to perceive contrasts such as pea-bee, coat-goat, sheep-ship, cart-cut. At age 4 they performed significantly worse in these tests than monolingual-English controls. Their production is also “Sylheti-accented”. Surprise, surprise! We await results of the forthcoming retests to be carried out when the children are older and, presumably, more English-oriented.

• Examining the recordings in the Sound Atlas of Irish English, Elke Philburn found that glottalization of t, and to some extent of p and k, is not as geographically limited in Ireland as usually assumed, and appears to be spreading. It is well known to occur in Dublin, Belfast, and the Ulster Scots areas, but she found it ‘to some extent’ in all 32 counties. Teenagers (83%) glottalize much more than the elderly (7%).

• People acting as subjects for ultrasound tongue imaging or electromagnetic articulography tend to wiggle their tongues around when awaiting their turn to utter. How can you inhibit this ‘non-linguistic pre-speech behaviour’ or ‘vegetative wiggling’ and achieve ‘reliably still tongues’ in the run-up to speech? According to Sonja Schaeffler and Jim Scobbie, by playing them an audio prompt or making them listen to a dialogue partner. Distract them that way, and they’ll keep their tongues still.

Thursday, 5 April 2012

Scottish affairs at BAAP

Three presentations at the BAAP meeting concerned current developments relating to r in Scottish English.

• Gerald Docherty et al reported that findings emerging from a project on Accent and Identity on the Scottish/English border show that it is an oversimplification to say, as we usually do, that the English side of the border is non-rhotic while the Scottish side is rhotic. This project involved research into speakers from Gretna (purple marker on the map) and Eyemouth (blue) on the Scottish side and Carlisle (green) and Berwick (red) on the English side. Younger working-class speakers on the Scottish side, more in the west than in the east, are starting to lose their rhoticity. This confirms what has been reported elsewhere for Glasgow English.

• Arising from the same project, Dom Watt et al showed that, compared with Eyemouth, Gretna is further along the historical path whereby the ɪr, ɛr, ʌr of bird, berth, hurt merge into a single ɝ ~ ɜ. “The loss of rhoticity appears to act as a trigger for the NURSE merger. It is not a necessary precondition, however, as the merger may precede derhotacisation.” The change is also probably triggered by a change from tap ɾ to approximant ɹ as the preferred realization of postvocalic /r/.

• Eleanor Lawson et al used audio-ultrasound tongue imaging to show that two distinct ways of articulating the approximant have different effects on the preceding vowel. In central eastern Scotland, they found, working-class speakers tend to use a tongue-tip variety, which leads to retraction of the vowel, whereas middle-class speakers go for the “bunched” tongue configuration, leading to vowel centralization and completion of the NURSE merger.

Another paper relating to Scotland concerned not English but Gaelic. It was a surprise to me to hear that over half the speakers of Gaelic are to be found not in the outer Hebrides but in the Glasgow conurbation. Claire Nance carried out research at a Gaelic-medium school there, the Glasgow Gaelic School/Sgoil Ghàidhlig Ghlaschu.

Gaelic has two allophones: it is thoroughly back before lˠ nˠ rˠ but central ʉː in all other environments. Looking just at the Glasgow children’s pronunciation of the central allophone, Claire found it to be undergoing further fronting, just like the corresponding Glasgow English GOOSE vowel. Taking two groups of students divided by their degree of engagement with Gaelic and with schooling generally, she found that the studious, obedient ones keep their Gaelic vowels more distinct from their English vowels than do the ‘less school-oriented’ ones with attitude. But they all sound Glaswegian to other Gaelic speakers.

Watch and listen to her yourself here.
_ _ _

Bank holidays coming up. Next posting: 10 April.

Wednesday, 4 April 2012


No sooner had I got back from Kyiv than I was off to Leeds for the biennial Colloquium of the British Association of Academic Phoneticians (BAAP).

The presentations there were of very high quality. Everyone started on time and finished on time (thanks to tight chairing), everyone was audible, all the Powerpoint slides were readable, and as far as I could tell no one just read out a prepared text. The posters were good, too. And that’s more than you can say of some academic conferences I have attended.

At each BAAP colloquium members vote to award three prizes. One, the Peter Ladefoged prize, goes to the paper or poster that best reflects Peter’s approach to phonetics. This time it went to Adrian Simpson (pictured) for a paper on Percussives. (“The percussive manner [of articulation] involves the striking together of two rigid or semi-rigid articulators.” — IPA Handbook, p, 187) As well as the [ʬ ʭ ¡] of ExtIPA, he discussed, with wave-form support, the labial percussive that arises from the approach phase of the [p] in the [ʔp͡ t] of words such as stamped, along with the transient ‘epiphenomenal clicks’ that arise as a consequence of other articulations. Adrian is the kind of phonetician that I approve of: he personally demonstrated to us each of the sound-types he referred to. (I don’t care for phoneticians who can’t or won’t perform in public. If we want our students to make this or that sound, we’ve got to be able to do it ourselves.)

The two Eugénie Henderson prizes go to the best oral presentation and the best poster by newcomers. The first was awarded to Michael Ramsammy for an electropalatographic investigation of the weakening of Spanish l in preconsonantal environments. The second was won by Nicholas Flynn for a poster comparing twenty vowel formant normalization methods. He concluded that “vowel-extrinsic, formant-intrinsic” methods performed the best at normalizing vowel formant data for sociophonetic study. (No, I don’t understand this, either.)

Here are some new technical terms I noted among the ninety or so papers and posters at BAAP:
attriter, a bilingual whose L1 has become subject to attrition through living in an L2 environment;
enchronic, relating to the time at which an utterance is uttered;
LADO, Language Analysis for the Determination of Origin (of an asylum seeker)

I have no idea what is meant by ‘7th- and 8th-order sliding-Gaussian-window lpc analysis’ (John Esling), nor by a ‘Bootstrap Markov Chain Monte Carlo algorithm’ (John Coleman). But perhaps they had their respective tongues in their cheeks when uttering these words. (If you feel strong, look here, here, and here.)

Tuesday, 3 April 2012

macaronic signs

And what, you ask, about the public use of English in Ukraine? Is there English-language public signage to be seen on the streets of Kyiv?

Not much, compared with some other eastern European countries. And what there is tends to be masked by being written in Cyrillic.

Here’s part of a photo I took from a window in the Linguistic University. It’s a dealership for Jaguar cars. As you can see, the Jaguar logo, with the word in English, appears on the right; to the left, the Cyrillic writing says ЯГУАР СЕРВІС, i.e. ‘Jaguar service’. (I don’t know whether they would pronounce ягуар with ɦ or with ɡ.) The word сервіс ‘service’ presumably qualifies as a loanword rather than a word in a foreign language (English). The same is true for my hotel’s trilingual notice about ‘room service’. (The English version is not actually correct, since we don’t write ‘room service’ with a hyphen.)
And ‘smartphone’ has certainly been adopted as a loanword. Here it is with a Ukrainian plural ending (and a discount). Here is the ‘price list’ (прайс лист, prajs ɫɪst) offered by a Міні Хауз кафе (cf blog, 2 June 2011).
My familiarity with the Cyrillic alphabet is sufficient that if I see a sign saying РЕСТОРАН I read it to myself as restoran. But people who don’t know Cyrillic see something quite different. They see Latin letters, and read it to themselves as ˈpektəʊpɑː. Or so they tell me. What a funny word for a restaurant!

In Kyiv I saw a small political demonstration on the streets. I’m not sure what it was about, but the protesters were carrying signs reading (I thought, unthinkingly) Hi!.

Who were they greeting so warmly? Who were they saying hello to? It was only after doing a double take that I realized that the signs actually said НІ!, which is pronounced nʲi and is the Ukrainian for “no!”. No, no, no!

Unfortunately I didn’t take a photograph of this, but on the internet I found one that will perhaps do instead. If you look very carefully you will see a tiny dot on the I, which is a hint that the letter is not Latin I but Cyrillic І.
In Ukraine I also noticed several advertisements for Baxi fires. Baxi is a British manufacturer, based in Preston, and (I quote) “one of Europe’s biggest manufacturers and distributors of domestic and commercial water and space heating systems”. The name is based on the surname of the founder, one Richard Baxendale. In English, of course, we pronounce it ˈbæksi. I can’t help wondering whether Ukrainians would read it as ˈvaxʲi, since in Cyrillic B stands for the labiodental v and X for the voiceless velar fricative. Or do they recognize it as English and say ˈbeksi? There’s more here.

Monday, 2 April 2012


It was fifteen years since the previous time I had visited Kyiv (i.e. Kiev, as we used to call it), so when I was there recently people kept asking me what differences I could see around me on this, my second visit.

Physically, visually, there were obvious changes. There were quite a few new buildings — new sports stadia, new bridges across the river, new hotels, new office blocks, new (rather ugly, I thought) high-rise residential blocks. And motor traffic on the streets was much heavier, with traffic jams everywhere.

But what about the linguistic situation? That has clearly changed, too. I was struck by the fact that nearly all street names, shop signs and advertisements now seem to be in Ukrainian rather than in Russian. That’s different from the situation fifteen years ago. Only in my hotel did I see Russian signs to the exclusion of Ukrainian, but of course the hotel is mainly for visitors from elsewhere, not for locals. Likewise in the university I think I heard less Russian spoken, more Ukrainian; though there are still plenty of academics who are clearly more at home in Russian.

My hosts kindly took me to a music recital in celebration of the composer Lysenko and to a performance of a Kalman operetta Сільва (the English title of which, I have since discovered, is the Gypsy Princess). Both were entirely in Ukrainian.

How do I know this? Although I speak neither Russian nor Ukrainian, I can recognize the difference between them in writing and, less confidently, in speech. In writing, the letters ы and э are used in Russian but not Ukrainian, while є, і and ї are used in Ukrainian but not in Russian. In place of the Russian hard sign, ъ, Ukrainian uses an apostrophe.

To an outsider’s ear the two languages sound pretty similar. One difference is that Ukrainian has less vowel reduction than Russian: you get unstressed o, which in Russian would be reduced to a ~ ɐ or ə. Another is the distinctively Ukrainian consonant ɦ, which is of high frequency in running speech. ‘Programme’ is програма pɾoˈɦɾama. Canonically this is a voiced glottal fricative, though in practice it sometimes seemed to me to be devoiced.

There are a whole lot of words in which Ukrainian has the vowel і corresponding to Russian о, as in the prefix під- pid ‘under-, sub-’, which is Russian под- pod. Although the two languages share much of their vocabulary (give or take a sound change or two), there are also numerous obvious differences. ‘Language’ is язык jaˈzɨk in Russian but мову ˈmovu in Ukrainian. Where Russians thank you by saying спасибо spɐˈsʲibə, Ukrainians say дякую ˈdʲakuju.

The relationship between spelling and sound is confusingly different in the two languages. In Russian, the letters и and е imply palatalization of the preceding consonant, while in Ukrainian (if I’ve got it right) they do not. The corresponding Ukrainian palatalizing vowel letters are і and є, which are not used in Russian. The twin-dotted Ukrainian ї stands for ji. And words can begin in Ukrainian with the letter й, which is not possible in Russian.

In the picture you see the name-badge I was given. My surname is transliterated as Уелз, which would be read in Ukrainian as ueɫz. Given that the language has no initial w, that’s fine. But if read in Russian it would come out as ujeɫs, which is not so good. In Russian transliteration my name is Уэлз.