recently...

The standard descriptions of English phonetics really fall down — it seems to me — when it comes to the sequence involving a voiceless plosive preceded by a nasal and followed by a lateral, such as we have in recently.
Reading Gimson’s Introduction, even in Cruttenden’s latest revision (the seventh, 2008: §9.2.4 The Release Stage of English Plosives, (5) Lateral release, p. 167), you would imagine that in such a case we always get lateral release of an alveolar plosive, i.e. ˈriːsnt^lli. True, there’s also a section in the book headed ‘Glottal replacement in RP’ (§9.2.8 (c), p. 180), but it contains no examples of a plosive preceded by a nasal. (To be fair, neither does the panel on Glottal Stop in LPD.)
There are people who do indeed have a laterally released alveolar plosive, just as described. I do myself, in relatively formal or careful speech. But I also have another option, as I think most people do.
This is to have a glottal stop between the sonorants: ˈriːsnʔli. Listening to the party leaders debating on television last night, I was struck by the fact that not only the Scotsman (Brown) but also both of the privately educated Englishmen (Clegg and Cameron) used this variant, in this word and even more strikingly in importantly ɪmˈpɔːʔnʔli.
Actually, though, just labelling the segment a glottal stop does not tell the whole story. Once the tongue tip has made contact with the alveolar ridge for the n segment in recently, it retains the contact until the end of the l segment. What we have for the plosive between them (in the variant I’m discussing) is a double articulation, alveolar and glottal. You can tell there’s glottal closure as well as alveolar from the absence of the typical burst noise of lateral release as we move on to the lateral. The glottal closure is the more important of the two closures, since once the air stream is dammed by the closed glottis it is not available to build up pressure behind the alveolar closure, hence no noise of lateral release. If you find it helpful to have a phonetic symbol for every allophone, you could write this one [t͡ʔ] (t and ʔ joined by a tie bar).
It is not even clear that the double plosive has nasal approach. At the end of the n segment the airstream may well be cut off not so much by the rising soft palate as by the coming together of the vocal folds. Perhaps the /t/ in recently has neither nasal approach nor lateral release.
I think this is the usual pronunciation of /t/ in this phonetic environment in AmE, too.

re-

Dear Professsor Wells,
I’m confused about the pronunciation of the prefix re-. When does it have the DRESS vowel? LPD says “if stressed through the operation of a stressing rule”, but I don’t understand what that means.
XYZ

Let’s try to set out the facts about this prefix using a different approach.

1. When re- means ‘again’, it is pronounced ri: and stressed. So we have for example reapply, renegotiate, reconsider, all with ˌriː-. In words such as refit, rethink, rerun we get the familiar stress difference between verb and noun, but the first part is always riː. A ˈrefit, but to ˌreˈfit. So far, the rule is easy.

2. When re- has a vaguer meaning, it usually gets weakened and is unstressed. We have ri- or rə- (it doesn’t matter which) in words like remember, retain, remarkable, repeat. This rule is easy, too. However, if the sound immediately after the re- is a vowel, then we always have ri- (= riː- or rɪ or something intermediate). The only important example is react riˈækt and its derivatives.

3. Now we come to the tricky bit. If re- has the vague meaning AND IS STRESSED, AND IS FOLLOWED BY A CONSONANT SOUND, then it is pronounced re-. This type includes several well-known words: relative ˈrelətɪv, recognize, reference, relevance, and also for example recompense, replicate, resonate. (Compare relate, refer, where the re- is unstressed and weak.)

4. If the main stress is on the syllable after the syllable after the re- (= two syllables later), we normally get secondary stress on the re-, which therefore has a strong vowel, which again is e. So we have for example reclamation ˌrekləˈmeɪʃən (compare the verb reclaim with weak re-), recognition, recommend ˌrekəˈmend, recreation, reformation, relativity, reparation, repetition, replication, reprehensible, represent, reservation, resignation, restoration.
Types 3 and 4 are therefore the words I mean when I say “if stressed…”. They are the ones that present the greatest problem for learners of EFL.

5. This being English, some words are irregular and exceptional. For example, many people pronounce one or all of relaxation, resistivity, retardation in a way that violates the above rules.

6. These principles also apply to de- and pre-. Examples: 1. deconstruct, predetermine; preprint; 2. decide, prepare; 3. deference, preference; 4. dereliction, preparation; 5. AmE treats premature as having the explicit meaning ‘before’ (ˌpriː-); BrE treats it as having the vague meaning (ˌpre-).

Sorry, it’s still pretty complicated.

ən ɪnˈɔrgjərəl ˈlɛktʃər?

For years I had been teaching people that speakers of rhotic accents pronounce an r sound wherever there is a letter r in the spelling, and not where there isn’t. Non-rhotic speakers on the other hand — which includes me and most of the students I lectured to at UCL — don’t, since we have lost /r/ except before vowel sounds.
I would illustrate this with pairs such as court–caught, larva–lava, sore–saw, career–Korea, tuber–tuba, manner–manna, western–Weston — and dear–idea.
Then I started to notice that sometimes I would hear rhotic speakers who nevertheless seemed to be pronouncing r at the end of idea. How could this be? Anyhow, I had to start expressing myself rather less sweepingly, more circumspectly. Some words (not only colonel) seemed to be exceptions. They had no <r> but they did have [r] for some rhotic speakers. (See also my blog for 11 October 2007, and Accents of English vol. 2 p. 343.)

(We non-rhotics do sometimes say aɪˈdɪər, of course, but only before vowels: that’s our “intrusive r”, ðə ˈveri aɪˈdɪər əv ɪt. But that’s a quite different matter.)

Now the Scottish (and correspondingly rhotic) phonetician Jim Scobbie has revealed the shocking facts. He reports that he has an “extra” r in the items Chicago, unauthorised, theatre, idea and, as he has just noticed, in inaugural.

There is no historical r at the relevant point in these words. The explanation is presumably false analogy in words first heard pronounced by nonrhotic speakers. If a Scot hears an English person say ˈɔːɡən organ and knows that it corresponds to Scottish ˈɔrɡən, it’s entirely understandable that he would infer that for English English ɪnˈɔːɡjʊrəl the corresponding Scottish form must be ɪnˈɔrɡjərəl. Spelling doesn’t come into it.

In rhotic accents outside Britain, where the influence of England’s non-rhoticity is weaker — in America, in Canada, even in Ireland — you would expect this sort of thing to be less likely to happen.

creating a vowel diagram

Gwen Awbery writes

I have a longstanding problem trying to get hold of a blank, basic vowel chart which would be compatible with OpenOffice or similar, and to which I could add vowel symbols as required. Is there such a resource lurking out there somewhere in cyberspace? It seems likely, but I haven't been able to locate one. I would be very grateful for any advice on this.

I did a Google search for “blank vowel chart”, which threw up this file from Wikimedia commons.
Once you’ve saved a copy of the blank diagram to your own computer, the next question is how to add vowel symbols to it. Presumably you need to use some kind of graphics program. I succeeded in adding the symbol æ, using the Photostudio software that came bundled with my printer (which is what I use for scanning, cropping and resizing the assorted illustrations that accompany my blog). See above.
However, many graphics programs, including this one, seem to have trouble with Unicode text. Although I haven’t spent long on this, I have not so far succeeded in superimposing vowel symbols such as ɛ, ɔ, ʊ, ɑ̃.
I haven’t tried the Draw program that comes with OpenOffice.

Can someone with greater experience of graphics software offer Gwen (and the rest of us) some advice?
How do Wikipedians achieve results like the chart of AmE vowels I found here?


PS: inspired by Gerald Kelly (see comments below), here’s one I have just made in Powerpoint. To convert the result into a jpg I took a screenshot (Alt-PrtScrn) and then cropped and resized it with Photostudio.
PPS: further inspired, this time by David Crosbie, here’s one made in Word for Windows Vista.
Thanks to all.

implicational or not?

In the trailer for a television programme yesterday the voiceover invited us to join the presenter on

what’s con\/sidered to be | one of the ˈfinest ˈwalks in \Britain.

…which led me to reflect on an ambiguity in the meaning of the English fall-rise tone.
The voiceover actor, overemphasizing things as actors tend to, caused me at first to misinterpret the intonational meaning as ‘considered to be one of the finest (but actually not one of the finest)’. That is, I took the tone to be an “implicational” fall-rise (my English Intonation – an Introduction, §2.6–7).
I realized immediately that that didn’t make pragmatic sense. What was intended was merely a dependent fall-rise (§2.20), indicating only that we have not yet reached the end of the sentence, that there is more to come.
If I had had to read the voiceover script, I think I would have removed the ambiguity by saying just

what’s considered to be one of the ˈfinest ˈwalks in \Britain.

— with no accent at all on considered. But it would also have been unambiguous to use a non-nuclear accent on considered:

what’s conˈsidered to be (ˈ)one of the ˈfinest ˈwalks in \Britain.

— though, depending on speech rate, this might produce an uncomfortably long intonational phrase.
It may be the case that the falling-rising pitch movement tends to be wider in an implicational fall-rise than in a dependent one (or for that matter than in a “tentative” one). That would explain why the actor’s animatedness led to my initial misinterpretation of his meaning: he produced a fall-rise so wide that its primary meaning would be the implicational one. I would suspect that the wider the pitch range, the more likely a fall-rise is to be interpreted as implicational, and the narrower, the more likely to be interpreted as dependent. But it might conceivably also depend on the absolute pitches used.
Does anyone know of any evidence showing whether or not pitch and meaning in the fall-rise are linked in this way?

Pygmalion Cameron

This political cartoon from yesterday’s Guardian (click to enlarge) may be pretty opaque to non-Brits today and will be opaque to everyone in fifty years’ time. It is a comment on the supposed unsuccessful attempts by the leader of the Conservative party, David Cameron, to make himself sound less upper-class (blog, 8 April) in preparation for face-to-face television debates with the leaders of the other two main parties.
The cartoon is an obvious allusion to the characters Professor Higgins and Eliza Doolittle in Pygmalion / My Fair Lady (“By George, she’s got it!”).
The cartoonist shows Cameron with a condom for a head, for reasons I can no longer remember.
The hovering Mekon figure (readers of the 1950s Eagle comic will know what I mean) is William Hague, former leader of the party, definitely not upper-class, and notable for having retained an unreconstructed Yorkshire accent and therefore using a “flat” vowel in TRAP words (= an open and retracted [a], like the vowel of RP STRUT).

proven

I don’t know enough about the history of our spelling to be able to explain with certainty the oddity of having the letter o correspond to the pronunciation uː in the words move (and remove) and prove (approve, improve, reprove), but I assume we have to blame those Norman clerks who didn’t like to write u next to the then identical v. Awareness of the etymology may have played a part, too, given that these verbs come from Latin movēre, probāre.
Despite the spelling–pronunciation anomaly, there is no tendency at all to regularize the position by adopting a spelling pronunciation. No one says məʊv instead of muːv, or prəʊv instead of pruːv.
But people in Britain do quite often say ˈprəʊv^ən rather than the expected ˈpruːv^ən for proven. This is one of the two possible past participles of prove, alongside proved. To some extent it has taken on a life of its own as a prenominal adjective (as when we say “a proven track record”, though never “a proved track record”).
The variant ˈprəʊv^ən seems to be quite recent. It is not mentioned in EPD prior to Roach’s editorship. Unless someone can prove the contrary, I shall tentatively claim that LPD was the first dictionary to include it, just twenty years ago.
According to OBGP, it is particularly associated with the Scottish legal verdict not proven. Is this true, and if so, why?

pre-fortis clipping

Sergio Verdejo writes to ask about the term “pre-fortis clipping”.

Do you happen to know who coined the term and when?

I’ve an awful feeling this may be a term that I myself had a hand in inventing, although I can’t remember the details.
It is not recorded in the on-line OED, whether under “pre-fortis” or under “clipping”.
As far as I remember, it was decided on by a group of phonetics teachers at UCL (probably Michael Ashby, John Maidment, Jill House, and me), sometime in the mid 1980s. This was shortly after the retirement or death of the previous generation of UCL phoneticians (Fry, Gimson, O’Connor, Arnold, Tooley and Pring).
The reason for seeking a new term for the phenomenon was that we found the term “shortening”, as applied to English vowels, unsatisfactory. A “shortened” /iː/, as in reach, is not the same as “short” /ɪ/, as in rich. “Shortened” (i.e. clipped) /ɔː/, as in court, is not the same as short /ɒ/, as in cot. Gimson’s term “reduced” was also confusing, since reduction is generally understood as a synonym of weakening or lenition.
So, rather than speak of “shortened” or “reduced” vowels, we thought it would be better to refer to them as clipped vowels.

I did not use the term “pre-fortis clipping” in Accents of English (1982) — which I presumably would have done had it been available then — but I did use it in the first edition of LPD (1990), as well as in my article ‘Syllabification and allophony’ published in Susan Ramsaran (ed.), Studies in the pronunciation of English, A commemorative volume in honour of A.C. Gimson (London and New York: Routledge, 1990), pages 76-86, where I said of Pre-fortis clipping

This is the name which some of us have come to adopt for the rule making the /el/ of shelf durationally different from the /el/ of shelve, and the /iː/ of feet different from that of feed. (Gimson refers sometimes to ‘shortness’ of the sounds involved, sometimes to ‘reduction’. Calling such sounds ‘short’ leads to confusion when pairs of phonemically distinct vowels such as /iː/ and /ɪ/ are also categorised as ‘long’ and ‘short’ respectively; calling them ‘reduced’ is to be avoided since this term for most phoneticians denotes change of quality, a ‘reduced’ vowel being of the [ə] type. The term ‘clipping’ avoids these difficulties.)
English vowels are subject to pre-fortis clipping, then, when they are followed by a fortis consonant within the same syllable. The /f/’s in self, selfish /ˈself.ɪʃ/, and dolphin /ˈdɒlf.ɪn/ trigger clipping, but not those in shellfish /ˈʃel.fɪʃ/ or funfair /ˈfʌn.feə/. So do the /t/ in feet and the /ʧ/ in feature, but not the /p/ in fee-paying or the /k/ in tea-kettle. The vowel /æ/ undergoes pre-fortis clipping in lap, lamp, happy /ˈhæp.ɪ/, and hamper /ˈhæmp.ə/, but not in slab or clamber.

Does anyone know of an appearance in print prior to 1990?

inputting accented letters

As I have mentioned before (blog, 26 Aug 2008 and 25 Jan 2010), the Guardian newspaper is now able to print accented (= diacritic-bearing) Latin letters not only for west European languages but also for those from eastern Europe. My picture shows part of a page from the 15 April edition of the paper. In the course of an obituary for Anna Walentynowicz, “godmother of the Solidarity trade union” they managed Gdańsk and Wybrzeża correctly (though they didn’t attempt Solidarność). But look what happened to Wałęsa! The barred l was fine, but the ogonek under ę turned into a cedilla instead, giving ȩ.
I don’t know whether the Guardian journalist tried to compose the letter from an e plus a diacritic, and chose the wrong diacritic, or whether s/he just picked the wrong precomposed letter from a dropdown list. Either way, it provokes the question: what language, if any, has U+0229 LATIN SMALL LETTER E WITH CEDILLA in its orthography? If none, who uses this symbol, and for what? Presumably it must be or have been in use, because otherwise Unicode would not have admitted it. (It’s in the Unicode block Latin Extended-B, in a section headed “Miscellaneous additions”.)
Can anyone enlighten us?
Stranger still, at U+1E1D (Latin Extended Additional) Unicode also recognizes LATIN SMALL LETTER E WITH CEDILLA AND BREVE, i.e. ḝ.
You can check it out on the on-line Unicode site here.
The on-line version of the obituary has ę correctly.

Commenting on yesterday’s blog, with reference to the character ö, one contributor apologized “Sorry, I don’t have easy access to diacritics”. I was wondering just what input device he was using that can’t input an o-umlaut. Even a mobile phone can do that. On any Windows computer you can key in
• Alt+0246 (on the numerical keypad)
• or on some computers Alt+148
I am sure you can do something similar on a Mac (option-apostrophe?). If using Word you can do
• Ctrl+(Shift+):, o
• or 00F6, Alt-X
For HTML, including blog comments, you can also type
• &#246;
• or &#xF6;
• or &ouml;.
Perhaps these don’t count as “easy access”. Or perhaps people just don’t know about them. If anything, we have a confusingly large range of possibilities.

The corresponding methods for inputting æ are Alt+0230, Alt+145, Ctrl+(Shift+)& a, &#230;, &#xE6;, &aelig;.

Eyjafjallajökull

A family wedding a few days ago, causing a gap in this blog, means that we have not yet discussed the pronunciation of the Icelandic volcano that is causing such unprecedented air traffic disruption in Europe. By now everyone else has, though.

Here is the advice from the BBC pronunciation unit. The BBC’s formulation does not use IPA, relying instead on its special respelling system. Interestingly, the headline replaces the character ö with oe, which seems very lame. So do the maps below. I suppose it’s better than Eyjafjallajokull (no diacritic), which is what some of our newspapers have been printing.

For an IPA rendering you’ll have to turn to Wikipedia, which tells us that it’s pronounced ˈɛɪjaˌfjatlaˌjœːkʏtl̥.

I think, though, that many of us would find it a challenge to come up with precisely that on the basis of the sound clip supplied.

Language Log offers us not only a number of rival versions pronounced by pukka Icelanders but also a selection of feeble attempts by non-Icelanders, including a Youtube compilation of newsreaders’ efforts.

I have to say that for Icelandic final/preconsonantal ll I’ve always been tempted to write tɬ rather than tl̥. I hear quite a bit of friction in the lateral component, so that we have a voiceless alveolar lateral affricate. I have a friend named Hallgrimur, whom I call ˈhatɬkɾɪmʏɾ̥, I hope correctly.

There’s an article about Icelandic phonetics here.

Iapetus and tonotopy

I was reading a book about the geology of the British Isles when I came across the word Iapetus, the name of an ancient ocean that is believed to have once divided Laurentia from Baltica/Avalonia, and hence the future Scotland from the future England.
I read it to myself as aɪˈæpɪtəs, but was then struck by doubts. Ought it rather to be ˌaɪəˈpiːtəs?
This type of uncertainty arises whenever we meet a word from Greek or Latin for the first time, and the penultimate vowel in that word’s spelling is followed by a single consonant letter. The general rule is that if that vowel was long in Greek/Latin we’ll stress it in English, but if it was short we won’t, placing the stress instead on the antepenultimate.
So my hesitation can be reformulated as uncertainty over the ancient quantity of the e in Iapetus.
Iapetus is also the name of one of the moons of Saturn. In Greek mythology he was one of the titans.
I suppose I must have heard this word pronounced once or twice in my life. And I must have looked it up somewhere twenty years ago, because it is there in LPD. It’s also to be found in the Oxford BBC Guide.
In any case, my first surmise was right: it is indeed stressed on the antepenultimate, because the e was historically short. We say aɪˈæpɪtəs.
Greek spelling makes the history clear, because in Ancient Greek long and short e are represented by different letters: η (eta) for ē, ε (epsilon) for ĕ. And the titan was Ἰαπετός Iapĕtŏs.

Ancient Greek also distinguished long and short o: ω (omega) for ō, ο (omicron) for ŏ. The vowel in -log- ‘word, reason’ was the short one, which is why we stress words such as biology, philology on the antepenultimate. So was the vowel in -top- ‘place’, which is why I was taken aback when Sophie Scott (blog, 2 April) pronounced tonotopy as ˈtəʊnətɒpi. I would have expected təʊˈnɒtəpi.
— —
I shall be away for the rest of this week. Next blog: 19 April.

a weritable willain

Scanned image by
Philip V. Allingham

It appears that the Londoners of a century or two ago sometimes interchanged or confused the sounds v and w. Dickens certainly portrays this in his literary representation of Cockney: Mrs Gamp in Martin Chuzzlewit says on the one hand “wery best” and “[the] walley of the shadder [of death]” and on the other hand “vich” for ‘which’.
In Pickwick Papers we find

Ve got Tom Vildspark off..ven all the big vigs..said as nothing couldn't save him.

and

I may trust you as vell as if it was my own self. So I’ve only this here one little bit of adwice to give you!

(Notice that in this second passage “I’ve” and “give” presumably had the usual v and “one” the usual w.)
Writing in 1936, H.C. Wyld recalled the use of [w] in place of [v] as a jocular usage among middle-class speakers in the nineteenth century. It is utterly unknown in Cockney today. (Accents of English p. 333)
The OED (second edition, 1989) says

In south-eastern English dialects the change of v- to w- does occur, and older representations of Cockney speech exhibit a converse change of w- to v-, which recent investigators have been unable to verify as still existent.

In present-day English such occasional interchange of the two consonants has been reported from parts of the Caribbean, particularly the Bahamas (Accents of English, p. 568 and 589).

I was accordingly delighted when a workman in Montserrat, exchanging gossip about local events, told me that in his view a certain person was a ˈwɪlən (villain).
I had never previously encountered this in Montserrat, although George Irish (see yesterday’s blog) does list one such entry.

warmunce - vermin; a mean or base person

With the dubious exception of Bahamian, where the two sounds may conceivably be in complementary distribution, all core native varieties of English, including Cockney and Caribbean, appear to have a firm phonemic contrast between v and w. (In parts of Jamaica and perhaps elsewhere there may be some confusion between v and b, but that is another story.)
Indian English is notorious for often lacking this phonemic contrast, with the two consonants being merged as ʋ (a labiodental approximant). This is often perceived by non-Indians as w, so that for example a maths teacher of Indian origin in a British school acquired the nickname “Mr Wortical” because of the way he pronounced vertical.
In the EFL world Hungarians are notorious for being deaf to this contrast and for using v in place of w. This is also part of the popular stereotype of a German accent, although nowadays most German speakers of EFL seem to have mastered the contrast.

Caribbean centipedes

There was an interesting pronunciation that I observed during my recent stay in Montserrat in the Caribbean: centipede pronounced as ˈsantipi. Not only is the final d lost as compared with the BrE/AmE ˈsentɪpiːd, -ə-, but the stressed vowel is that of TRAP rather than the usual DRESS.
The only published book about Montserratian speech, George Irish’s Alliouagana Folk (Plymouth: Jagpi, 1985), is a collection of local proverbs, sayings, words and phrases. It duly records this word in its glossary.

santipee - centipede

Richard Allsopp’s scholarly Dictionary of Caribbean English Usage (OUP 1996) also records this pronunciation of the word (though with a different medial vowel) as common to the entire Caribbean English Creole region, and speculates about its origin as Portuguese.I can see that a Portuguese source could explain the loss of the final consonant — though this could also be due to the characteristic Creole uncertainty about final d (compare for instance galvanize meaning ‘corrugated metal roofing’, shortened from galvanized iron). Apart from that, though, I would have thought that Portuguese centopeia, phonetically sẽtoˈpeja, does not seem on phonetic grounds to be a more likely source than a straightforward English origin. (OK, English got it via French from Latin centipeda ‘hundred-foot’, as did the Portuguese.)

s-lessness

Adam Brown’s original query (blog, 7 April) had two components, one about [θ], which we have discussed, and the other about [s].

My gut feeling is that, of the 6,000+ languages, almost all of them, probably 90%+ have [s] (although I know that Tamil and Maori are exceptions), and that very few, probably 10% at most, have [θ].

He now returns to the question of [s].

The second question, about the non-occurrence of [s], was equally interesting, but harder to find an answer to. WALS (Ian Maddieson) doesn't investigate precisely this, but states that 8.7% of the sample contained no fricatives, and that if languages have only one fricative, it is usually [s]. In other words, the percentage of languages that do not have [s] is probably less than 10%.
WALS states that languages that have no fricatives are mostly Australian aboriginal languages, and languages of New Guinea and the interior of South America. I ploughed through phonologies of other languages, and found some Pacific languages without [s]: Hawaiian, Maori, Nauruan, Tahitian and Tokelauan. Frankly, none of the above can be described as major world languages. The only major language without [s] seems to be Tamil.
Could I persuade you to ask your blog's audience whether they know of any other major world languages without [s]?

Fine. Over to you, my readers.
But first: why does Adam think that Tamil has no [s]? Sorry, but there are thousands of Tamil words containing this fricative. Here is the text of The North Wind and the Sun in Tamil, from the IPA Principles booklet (1949).You can see that the word for ‘sun’ is transcribed suːrijanum, and that the passage contains several other words containing s. OK, Tamil s may sound a bit ʃ-like sometimes, but it still counts as s and is still a fricative.
Perhaps the source of Adam’s misapprehension is this Wikipedia article, which unaccountably omits s from its list of consonants. Go to the Wikipedia article on Tamil phonology, however, and the omission is repaired.
For what it’s worth, the Tamil writing system includes the letters ஸ (Unicode TAMIL LETTER SA) as well as ஶ (SHA) and ஷ (SSA).
But there’s no mistake with the Australian and Polynesian languages, many of which do indeed lack s. Who can forget the Hawaiian version of ‘merry Christmas’, meli kalikimaka?
Since Hawaiian has no s, the medial and final fricatives of Christmas are rendered as k.

Christmas Island in Kiribati, on the other hand, becomes kiritimati. As with the name Kiribati (< Gilberts) itself, this demonstrates the fact that in Gilbertese (like Hawaiian, a Polynesian language) there is no phoneme /s/ as such, although /t/ before i is assibilated to [s]. So English s is rendered as /ti/ and written ti; and we, in return, are supposed to pronounce Kiribati as ˈkɪrɪbæs (though we mostly don’t).

EE, yet again

Students around the world often write to me and ask for advice or help on research they say they are carrying out on Estuary English. Unfortunately many of them are not in a position to collect actual speaker data from people in the southeast of England, so their “research” has to be an armchair study.
They write to me because I was foolhardy enough fifteen years ago to launch a website on the topic of EE, with the aim of bringing together the material on the subject that I could find on line or, in most cases, put on line. A lot of it was journalism, usually more or less sensational, making wild claims about this new variety of English supposedly sweeping the country and ousting RP from its former position of preeminence. Later, when academics were able bit by bit to investigate the truth of such claims, I did my best to publish (or link to) their research findings. In my view the most important of these were those by Joanna Przedlacka, incorporated in her 2002 book Estuary English? A sociophonetic study of teenage speech in the Home Counties. ISBN 3-631-39340-7, pb. Bern: Peter Lang. This work, as I put it,

demolishes the claim that EE is a single entity sweeping the southeast. Rather, we have various sound changes emanating from working-class London speech, each spreading independently.

I summarized her findings for my UCL students here and here.

In Britain media interest in the EE phenomenon has now died down, and it is three years since I have had anything to add to the site (most of which is now eight to twelve years old, an eternity in web time).
Nevertheless, I still receive emails like this one from Ámbar Romero in Chile.

I'm interested in doing some research on Estuary English and Received Pronunciation in order to do my Thesis project… I would like to ask you if you knew or if you had any information about the percentage of usage of these 3 characteristics of EE that are mentioned in the literature i.e. L-vocalization, the use of the glottal stop in final position and before consonants, and Yod- coalescence in tonic syllables. Is there any comparison between RP and EE speakers regarding the use of this features? ...I would very much [like to] know your opinion on the current status of RP, if you think it is including or not some of the characteristics of EE and if the latter might replace it as the accent of EFL in the short term.

What can one say? Here’s what I actually said.

Please read (or reread) Joanna Przedlacka's work. The point is that there is no real definable entity “Estuary English”. You can't divide up the speakers in the southeast into those who speak EE and those who speak something else. So there can be no comparative statistics of the kind you ask for (“Is there any comparison between RP and EE speakers regarding the use of this features?”). All we can do (given money, time, and effort) is to estimate the proportions of the population of a given area who do glottalling, yod coalescence etc in given phonetic environments and in given styles of speech.

If someone uses a relatively high proportion of glottalling, you might say “Ah! This must be a speaker of EE.” If you define EE speakers as those who use a lot of glottalling, you will indeed find that EE speakers use more glottalling than RP speakers (etc). But this argumentation is circular, therefore unscientific. A scientific approach would be to divide your speakers up by social class or some other non-linguistic criterion, then establish the possible correlation of phonetic variables such as glottalling with the non-phonetic variables.

All we have is various sound changes in progress. Many sound changes seem to spread out from London and from the working class into the middle class (and defining social class is another scientific nightmare). These sound changes all move at different rates. This kind of thing has certainly been going on in English English for at least five hundred years.

The leader of the Conservative party, David Cameron, is an Old Etonian and an archetypal RP speaker. But political commentators have recently asserted (with what truth I do not know) that he has been trying to make his speech sound more popular by using glottal stops. How does that relate to "the current status of RP"?

For EE to be used as the sole or main pronunciation model in EFL someone would have first to define it clearly and then produce learning materials (dictionaries, textbooks etc) using it. I don’t see any likelihood of that happening. Rather, BrE-oriented ELT will continue to be based mainly on a modernized version of RP (aka Standard Southern British English or SSBE, the modish term at BAAP last week).

But I hope that we’re gradually getting the message across that students will benefit from being exposed to a wide range of different varieties of English, just as native speakers are.

exotic dental fricatives

Adam Brown asked me if I knew what percentage of the world’s languages have the sound [θ].

I referred him to The World Atlas of Language Structures, which is now available on-line.

Go to the WALS website, search under “features” and look for chapter 19: Presence of Uncommon Consonants. This chapter covers clicks, labial-velars, pharyngeals, and ‘th’ sounds. (It does not distinguish between voiceless θ and voiced ð.) The database covers 2650 different languages.

Dental or alveolar non-sibilant fricatives are just as rare as labial-velar plosives, occurring in just 43 (or 7.6%) of the languages surveyed, but the distribution of these languages is practically worldwide. They are found in languages as varied in location and family affiliation as Modern Greek, Albanian, Spanish and English (Indo-European), Kabardian (Northwest Caucasian), Mari and Nganasan (Uralic), Burmese and Sgaw Karen (Sino-Tibetan), Lakkia and Yay (Tai-Kadai), Swahili and Moro (Niger-Congo), Dahalo (Afro-Asiatic), Berta and Murle (Nilo-Saharan), Fijian, Yapese and Drehu (Austronesian), Ngiyambaa (Pama-Nyungan), Rotokas (West Bougainville), Aleut (Eskimo-Aleut), Chipewyan (Athapaskan), Acoma (Keresan), Maricopa (Yuman), Cubeo (Tucanoan), Huastec (Mayan), Mixtec languages and Mezquital Otomí (Oto-Manguean), Amahuaca (Panoan), Tacana (Tacanan), Cochabamba Quechua and Mapudungun (Araucanian).

We might also mention Welsh and Icelandic (both I-E, and not included in the database).
Here is part of the map that comes up. The presence of a th sound or sounds is marked by a red circle.
This means that yes, for most learners of EFL the English dental fricatives are difficult, exotic sounds. It’s not surprising that they are sounds that advocates of teaching English as a Lingua Franca want to not bother with (but that, conversely, anyone who aspires to a higher standard than ELF cannot afford not to master).

archive cineradiography

At BAAP last week there was a showing of “Highlights from the UCL Phonetics Film Collection”. As Michael Ashby explained, when the UCL Department of Phonetics moved out of 21 Gordon Square a few years ago he discovered that they had accumulated about 12,000 feet of movie film spanning the years from the early 1920s to the 1970s. Much of this footage is rare or even unique. He is now in the process of researching and restoring the material, as finances allow.
Several of the clips he showed us were x-ray sound films of short stretches of speech, taken during the window of time between the technical advances that made it possible to carry out cineradiography and the realization that exposure to the high doses of radiation involved carried serious health risks for the person being filmed.

Films made by Ken Stevens at Haskins are relatively well known (you can watch one on YouTube here).

But I was unaware of the film made by Paul Menzerath and Robert Janker in Bonn and shown at the second International Congress of Phonetic Sciences in London in 1935. This was the first x-ray sound film and, Michael thinks, probably the first to achieve slow motion. There is one other copy known to be extant. You can watch it here (where however the date is wrongly given as 1937 and Menzerath is not mentioned). As the speaker says, “besonderes Interesse verdient vor allem die Wiedergabe der Sprache” [most of all, the reproduction of speech is of particular interest].

Scottish phonetics

On days two and three of the BAAP colloquium last week there was quite a lot about the phonetics of Scottish English.

• Dominic Watt et al have been studying the surprisingly sharp accent border between England and Scotland. People in Carlisle (west coast) and Berwick (east coast) sound English; those in Gretna and Eyemouth, just a few miles to the north in each case, sound Scottish. However, judged by the (non)application of the Scottish Vowel Length Rule and the degree of VOT, the linguistic border is somewhat blurred in the west, less so in the east.

• Eleanor Lawson et al showed that many cases of /r/ that would be classed auditorily as ‘postalveolar’ are actually “bunched”, not involving a retroflex lingual configuration, not even with tongue raising. (See my blog for 8 Feb 2010.) They also confirmed that middle-class speakers in the Scottish Central Belt (Glasgow–Edinburgh) have stronger realizations of /r/ than working-class speakers, for whom “an audible /r/ articulation seems almost absent”. So the middle class are bunched, the working class are on the way to nonrhoticity.

• Glasgow Asian is by now a distinct recognized accent (used by locally-born people of Indian origin). Jane Stuart-Smith et al found that the resonance of syllable-initial l was less dark in speakers of Glasgow Asian than in those of non-Asian Glaswegian. There was less difference for those with a network of non-Asian friends and those whose religious/cultural practices were ‘modern’ rather than ‘traditional’.

BAAP day one

Seriously, though (cf. yesterday’s blog), there were indeed several interesting presentations at the BAAP meeting. I wasn’t able to attend everything, but among those that especially captured my attention on the first day were

• a study by Esther de Leeuw et al showing that native speakers of German living in Canada produce an l-sound that is darker than that of monolingual German speakers but not as dark as that of Canadian native speakers of English.

• work by Rob Drummond showing that native speakers of Polish living in Manchester can have a STRUT vowel ranging in quality anywhere from Polish a via RP-like ʌ to northern ʊ. The longer they have lived there, the more likely they are to have abandoned the open vowel they were taught at school for the closer vowel that Mancunians actually use. (More here.)

• a plenary in which Sophie Scott discussed the neurophysiology of the brain. Trained phoneticians, and those good at learning languages, tend to have a more developed Heschl’s gyrus and more white matter (connectivity) in the brain areas crucial for language. (More here and in my blog for 15 October 2008.)

More to follow next week.

a new exotic sound

BAAP (the British Association of Academic Phoneticians) has just held its biennial Colloquium here in London. As usual, there were some excellent and thought-provoking presentations.
One particularly interesting contribution concerned linguolabial sounds. Linguolabials, articulated by the tongue tip against the upper lip, are very rare in the languages of the world. Nevertheless linguolabial plosives, fricatives, and a nasal are known to occur in a cluster of languages in the island state of Vanuatu. One of these languages is Tangoa. But until now there had been no reliable report of a linguolabial trill, if we discount an extrasystemic onomatopoeic ideophone alleged to be used in Coatlán Zapotec.

However, Olaf Lipor now reports that a voiced linguolabial trill has recently been discovered to be used contrastively in Caslon and Ki-Flong, languages spoken on the island of San Serriffe. The IPA symbol for this newly attested sound-type is r with the ‘combining seagull below’ diacritic, U+033C, thus [r̼].