Archive 3 of JWL Blog


Readers who need explanations of any of the abbreviations used may find them at Section 1 of the Home Page.

RSS feed for this siteRSS Feed

14/06/2007The LPD Pron Pref Survey 2007 (iii)#030
12/06/2007Compression Anomalies#029
04/06/2007Questions of Intonation#028
28/05/2007The LPD Pron Pref Survey 2007 (ii)#027
24/05/2007The LPD Pron Pref Survey 2007 (i)#026
15/04/2007Another Mystifying Pronunciation#025
26/03/2007Spelling versus pronunciation#024
09/03/2007TrainTimes Intoned#023
26/02/2007The Gettysburg address#022
23/02/2007W. H. Auden Again#021

Blog 030

The 14th of June 2007

The LPD Pron Pref Survey 2007 (iii)

The 12th item in this series has another int'resting problem for me because I tend to feel that `protester and pro`tester are different words in a way. The more traditional pro`tester with its agent suffix added to a verb suggests to me simply an individual person who's making a protest. The more recent coinage `protester suggests a person taking part in a formal, organised or at least collective protest, a joiner in a group activity. It also smacks of trendiness for me so I don't prefer it; but simply saying that I prefer the other one hardly satisfies my feeling that it's not the sort of question that I 'm happy to be asked. So I'll pass on this one.

Item 13 / `hʌrɪkən/ or / `hʌrɪkeɪn/ gives me the feeling that people who adopt the latter version are being (too?) influenced by the spelling rather than the oral tradition so I do prefer the weaker ending. I haven't much occasion to hear normal ie non-media people using the term so I don't feel strongly about it but I can confess, as someone who lives near and often visits Harrogate, that I'm prejudiced against the version of that word as / `hӕrəʊgeɪt/ rather than / `hӕrəgət/ though I do so with some shame because I feel no-one should disparage the unselfconscious pronunciation choices of anyone else.

Item 14 ēgotist versus ĕgotist is an easy decision for me because my normal regular habit is to use the latter form: so that is preference in action. However, I don't think I do so out of a decision based on any aesthetic or scientific principle so to that extent I don't have a preference, I suppose. The corresponding choice between the basic word as ĕgo or ēgo goes for me in the opposite direction: I usually say ēgoist I believe because I rather shrink from a feeling of a ridiculous association with the irrelevant and trivial-sounding word egg and also perhaps it tends to sound ridiculous because of a vague association with made-up words like the name of the game leggo.

    Incident'ly, isn't it curious that the word egotist shd exist alongside egoist! Wherever did that t come from? Bradley in 1891 in the OED called it "intrusive t" which said nothing. However, Murray in 1884 had offered a kind of explanation: "The t is purely connective in Fr., doubtless in imitation of the mute t in words like ballot, which is sounded in ballotage" (in reference to the curious word agiotage). Yes, and like the one in "Y a-t-il" no doubt. He also speculated int'restingly when he came in 1912 to the mysterious inserted -n- in tobacconist, "perh. suggested by such words as Platonist, with etymological n." Don't words lead you in odd directions!

Blog 029

The 12th of June 2007

Compression Anomalies

John Wells says in his blog today:

"English seems to have a rule barring the compression of [i, u] to [j, w] when followed by a strong vowel, though it is fine before a weak vowel.
Hence we can compress radiant to two syllables, ['reɪdiənt] > ['reɪdjənt] but can’t really do the same with radiate ['reɪdieɪt]. We can reduce mutual but not bivouac.
At least, that’s the rule I’ve been teaching people for many years.
It also explains why sierra, fiesta, Bianca, Kyoto, which are all disyllables in the original language, get expanded to three syllables in English. Spanish ['sjerra] becomes English [si'erə], etc."

As far as I know this "rule" was invented by John: I've no recollection of any such formulation by any previous writer. However, I find his impressions in this matter are rather different from my own. I find that I'm perfectly happy saying [`reɪdjəʊ] and [`reɪdjeɪt] as two syllables and when I do so it doesn't sound to me as if I'm at all different in this respect from most other people who have roughly the sort of accent that I and John have. That's not to say that I feel that the three-syllable versions sound strange to me or that I don't ever use them myself at least in a slightly more deliberate or leisurely kind of enunciation. Probably the key to this problem lies in the concept of strong versus weak vowels. I feel that for many people /əʊ/ and /eɪ/ are not necessarily strong vowels or can vary between being strong and weak. I'm reminded of the way Roger Kingdon considered that all strong vowels tended to have weakforms (ie with different phonemic identities), specifying notably that /iː, eɪ/ and /aɪ/ could convert to /ɪ/ and that /uː/ and /əʊ/ could become /ʊ/. He so insisted that /əʊ/ could be weak that he even preferred to show it in certain books where his transcriptions indicated that it was in some positions so little different if at all from /ʊ/ that he preferred to offer that kind of symbolisation to EFL learners as a target in many words. Such learners he suggested "tend to use vowels that are too strong for unstressed syllables" (see p.14 of Kingdon's 1969 revision of Harold Palmer's A Grammar of Spoken English). He even transcribed differentially fellow and feloe as respectively /`felu/ and /`felou/ in the (uncredited) pronunciations he contributed to Michael West's An International Reader's Dictionary of 1965 (Longman). I can't say that I feel exactly as Roger did about such matters but he was no mean phonetician (Daniel Jones employed him to teach phonetics at UCL) and his views, though extreme, are worthy of consideration.
    Finally John's comment that "why sierra, fiesta, Bianca, Kyoto, which are all disyllables in the original language, get expanded to three syllables in English" is explained by his rule, while no doubt having some truth in it, should be considered in the light of the fact that the spelling in almost all such cases (Kyoto is a rare exception) suggests for English-speakers not an approximant but a syllabic vowel sound. To give but one relevant example, the Italian words Giovanni and adagio contain a syllable that could be comfortably imitated by any English-speaker because it occurs in an item like the name Joe. Even so, people seem to usually attempt those words as the quadrisyllables /ʤiːə`vɑːni/  and  /ə`dɑːʤiəʊ/.

Blog 028

The 4th of June 2007

Questions of Intonation

I hear from Tamikazu Date that he's puzzled by the following from what he quotes as English Language Services, Collier Macmillan 1967: 48:
"(1) A: Is this a book?
    B: Yes. It's a book.
(2) A: Is this a chair?
    B: Yes. That's a chair.
(3) Have you have heard from John?
    B: Yes. I've heard from John.
(4) A: Are you going to see a doctor?
    B: Yes. I'm going to see the doctor."
According to Stress and Intonation Part 2 , with respect to the the responses:
    Yes. It's a book.
    Yes. That's a chair.
    Yes. I've heard from John.
    Yes. I'm going to see the doctor.
'It's', 'that's, 'I've' and 'I'm' are all said with the High Fall, while 'book', 'chair', 'John' and 'doctor'
with the High Rise.
Does that mean the parts with the High Falls receive the nucleus?

As so often with such queries, I have to point out that some of the dificulties lie in the assumptions of the question.
It's not always straightforward to decide whether we've got a nucleus or not. I prefer the word "climax" to "nucleus" for such things and perhaps my preferred choice of terminology here can help to make things clearer because you can easily grant that one can have major and minor climaxes and even situations where whether one has a climax or not can be in doubt. I may here refer to my Website remarks abour Accentuation (§7.1.10) where I point out that speakers often use what I have long called "animation stresses" which are not necessarily used to accentuate the word on which they are placed. So these initial Falls needn't be viewed as climactic (or, if you like, nuclear tones) and the fact that one of them occurs on the very rarely accentuated word "it" seems to confirm this.
These comments seem to answer Tami's next question too:
Is it possible for the nucleus to occur on pronoun subjects in such non-contrastive contexts as above?

If so, they seem to be unusual cases of nucleus placement.
I mainly agree when he adds
I think that the Fall + The High Rise would be more normal as in:

\Yes. It's a /book.
\Yes. That's a /chair.
\Yes. I've heard from /John.
\Yes. I'm going to see the /doctor.
but I have quite a problem with accepting as more ordinary, usual or commonplace the use of a really high rise in such situations. In fact, tho' such a usage is entirely possible, it would to my mind strongly suggest a marked interrogative flavour so that the effect would be of the speaker meaning "Yes, but why ever are you asking me that?" Alternativ'ly, it might strike one as  an example of the use of the  Pacific Rim  checking tone.

Blog 027

The 28th of May 2007

The LPD Pron Pref Survey 2007 (ii)

For his 11th item John Wells asks us to say whether we stress omega on its first syllable or its second and to identify the vowel of the second. His LPD and most other reference books represent Brits as giving the word forestress (thus what I like to call "globalising" it) and Americans as looking at it more analytically –  in a way treating it as the  two-word phrase it originally was – and favouring a selection of vowels for that second syllable.
    Oddly enough I appear to be the only person in existence who's happy to call it /`ɒmɪgə/. Admittedly it's one of many words over which I vacillate and I can't remember when I picked up the habit of thinking of it with that vowel. It would've no doubt been at a time when it was pretty exclusively a bookword for me and I no doubt simply jumped to the conclusion, in my ignorance, that that was an acceptable version. I think we all have such items in our vocabulary. Only an hour or two ago I heard the distinguished scientist Lord Robert Winston say ve`hemently. And a couple of days ago I heard a Frenchman interviewed on the BBC say something was /`detəmaɪnd/ when he meant /dɪ`tɜːmɪnd/. I never cease to be surprised that so many highly fluent excellent speakers of EFL often seem to have guessed how a word is pronounced early on and never disabused themselves of the delusion even though subsequently they've constantly heard it differently and correctly pronounced by other speakers. I expect it has something to do with the very large numbers of English words which do have various legitimate alternatives.
    I've come across many examples of such xenophonisms (my coinage). In a piece of Advice to the Learner I included in my Concise Pronouncing Dictionary in 1972 I recommended reading a page or two of the text (233 pages) at something like one a day: "You will no doubt get a number of 'shocks'. Most [EFL] teachers are only too familiar with the painful experience of hearing learners mispronounce words they are supposed to have learnt in the first few months of their English studies ... You will come across many things it would never have occurred to you to look up. Did you know, for instance, that the regularly spelt plural form of house is irregular phonetically? Do you know what might be unexpected about the pronunciations of bedroom and rectangle? Are you absolutely certain that you know the right vowels for all of pretty, said, broad, bury, tongue, worse, break, height, pear or were? Should you omit the h in human? Does use have s or z?"

Blog 026

The 24th of May 2007

The LPD Pron Pref Survey 2007 (i)

It's very interesting to see that John Wells (blog Friday 11 May) has been asking volunteers again to say what their preferences are regarding the pronunciations of certain words of debatable or divided usage now that he's revealed that he's preparing a very welcome third edition of his Longman Pronunciation Dictionary. His questions are aimed at native speakers of British English specified as having lived between the ages of 4 and 14 in England, Wales, Scotland, the Isle of Man or the Channel Isles except that he leaves an entry blank for other people to fill in. (I confess to being mildly surprised at "4 to14". It's my impression that children's speech can be a clean wiped slate by any change of linguistic surroundings up to about twelve and that for most people something survives of their earlier influences after about eighteen but I claim no expertise whatsoever in such matters.) The blank entry has one wondering what sort of thing he expects to see in it. His blog commenter answered at Saturday 19 May got some useful clarification about the blank entry.

This time, so far at least, he's only asking for opinions on about thirty items – a third of the number he asked about on the two previous occasions. I hope he'll come up with a supplementary list or two because there are so many items one wd like to get people's impressions on. It's quite tempting to guess what the results will be.

His first item asks whether people prefer an aitch to be uttered saying the name of the letter H. This item was to me a surprising inclusion because I'll be amazed if more than about five percent of the kind of volunteer he's likely to get will vote for /h/ unless their background is Irish.

His second is dissect which is a very fascinating one in terms of its history. My guess is that his respondents'll mainly go for /dɪ-/ rather than /daɪ-/ which is my preference too but the history of the word offers sympathy to /daɪ-/ choosers. The spelling certainly suggests /dɪ-/ but apparently does so by a sort of historical accident due to its importation from the French after they had (on phonetic rather than historical grounds) given it a double s because a single one suggested to them a pronunciation with /z/ rather than their normal one with /s/. Perhaps the relatively recent British tendency increasingly to prefer /daɪ-/ has been due to the analogy of items like dichromic, digraph, dithematic etc but mainly of chemical terms like dioxide and dichloride.

For his third he's chosen adult. Wisely he's always given each item in the context of a sentence thus making sure that its surroundings don't prompt a particular version from the respondent. Here the context is "for children and adults". A different context such as "in adult education" might have slightly inclined respondents to stress its first syllable from a kind of rhythmic preference when ordinarily they would've more usually stressed its second. I personally vacillate with this word in just such a way.

His fourth, question, concerns the two words accept and except and here he has a curious situation. He'd no doubt best like to know how each respondent says each of the two words ie whether they begin /ək-/ or /ɪk/ or possibly /ek-/ or /æk-/ in each case but he's puzzlingly settled for asking only whether they distinguish them. (I don't suppose anyone wd be likely to say they began except with /æk-/, tho I've a number of times heard accept begun with /ek-/ ). It'll be interesting to see what he makes of the responses he gets.

I think I usually say (what I like the idea of saying) /ək`sept/ for the first and /ɪk`sept/ or /ək`sept/ for the second but I know I'm not consistent. I think these days most people in England don't differentiate them and the majority seem to me to say /ɪk`sept/ for both. I seem to've noticed that that's so increasingly in recent years and to feel we're going in that respect in the same direction as they've gone in Australia.

With his fifth item debris it'll be interesting to see what quantity of respondents have picked up the American habit of endstressing it. Few I should think.

The sixth item, on the suffix -shire, reminds me of the request "Have you stopped beating your wife – answer only yes or no" because very many people vary in their use of this suffix. In my observation practically everybody says /-ʃə/ for Cheshire and probably Yorkshire but a small minority have /-ʃiə/ for various southern counties and very many, not only in Scotland, say /-ʃaɪə/ for various Scottish counties.

The seventh item, liquorice, is a fascinating one for me. I certainly ended it with /-ɪʃ/ as a child in Cardiff (where I went to very unsmart schools) a usage with a curious history. In the 17th and 18th centuries the OED apparently accepts it as unexceptionable but labels it dialectal for the 19th. It seems to've come about by a confusion essentially with a respectable but now obsolete word lickerish (coming from Old French and cognate with lecherous) which meant something like "delicious".

The eighth item has personal undertones for me too because for decades I've been a sufferer from a mild tinnitus (in my right ear only, and able to be ignored and largely forgotten about) consequent upon attacks of the unpleasant complaint known as Ménière's disease. When Murray came to deal with it in the OED in 1912 he only knew it as /tɪ`naɪtəs/ and OED2 in1989, as so often, failed to catch up with the fact that most people now make it /`tɪnɪtəs/ despite the fact that the second vowel had been long in Latin (according to Lewis & Short). We were largely no longer a nation of Latin learners by the end of the twentieth century!

The ninth item is contribute which I don't recall ever hearing stressed `contribute before the nineteen sixties or seventies though now that seems about as common as my habitual con`tribute. This new development was unknown to Daniel Jones and was first recognised in the EPD by Gimson in 1967. It's been commented that by the rules in Chomsky and Halle's 1968 The Sound Pattern of English `contribute is the more regular stressing. Compare `compensate, `constitute, `contemplate, `prosecute etc.

The tenth item is /`dɪfθɒŋ/ versus /`dɪpθɒŋ/. The sequence /-fθ-/ is very uncommon in English and seemingly only occurs in a few Greek scientific loanwords like diptheria and ophthalmic. The term naphtha was recognised by Craigie in the OED in 1907 as alternately pronounced /`næpθə/ in educated usage. The learnèd now rare phthisis is given first as /`θaɪsɪs/ in the OED and the even rarer (hardly current) phthisic was apparently pronounced /`tɪzɪk/. Thus English speakers don't seem comfortable with /-fθ-/ so it'll be interesting to learn what the Wells respondents have to say. I guess they'll go for /-fθ-/. I do, I'm afraid, but then I'm a very self-conscious speaker as no doubt they are.

Blog 025

The 15th of April 2007

Another Mystifying Pronunciation

Working through notes I've made over some decades comparing British and American pronunciations I've just come to the word lilac. Checking on its current American forms, I find that MWO (Merriam-Webster's Online Dictionary) lists three current pronunciations:   `laɪˌlɑk, `laɪˌlæk and `laɪlək [ No. MWO don't show it in IPA. These are my "translations".]
So the one they give first can hardly be a direct phonetic development given its spelling.
Turning to the MWO etymology – and converting into IPA my interpretations of their transliterations  –  we find "obsolete French (now lilas),  from Arabic [liːlak], from Persian [nilak] ... from Sanskrit".

    Well this, with the usual development of Middle English [iː] to /aɪ/, explains the second MWO form but what about the one they suggest most educated Americans use? It took the OED to help us guess how that came about. In its historical sequence of the spellings the word has received it says that in the eighteenth and nineteenth centuries they found "(now chiefly dialectal or U.S.) laylock". Certainly this type of pronunciation is still to be heard from quite a few unsophisticated speakers in modern Britain.

    Turning to the OED etymology under "other forms" it gives "Turkish leilaq (whence possibly the early 17th c. lelacke, modern laylock." Looking back at what the great Edward Artin judged to be current American pronunciational usage in the unabridged Webster's Third New International Dictionary in 1961, I see he listed the three variants in the opposite order. Also their etymologists quoted an Arabic form "laylak". Anyway it's not all plain sailing.  Henry Bradley the 1903 OED editor makes no reference to the fact that, since Turkish has no uvular sounds, it looks odd that the letter q is used in the translation from Arabic script that his quotation must have involved. (It wasn't till 1928 that Atatürk decreed that the Turks switch to the Latin alphabet and then with of course no need of that letter.) It may be that the scribe responsible was identifying a Turkish /k/ in the vicinity of an [a] that was specially [ɑ]-like with the Arabic value of q. The old IPA Principles booklet refers to an [ɑ]-like [a] before an [l] at p.35. If such a back quality were heard by people who tried to represent it with the English alphabet, their writing the syllable as "lock" would seem quite understandable.

However that may be, the spelling "laylock" occurred in The British Magazine in 1763,  the poet Oliver Wendell Holmes called them "lalocks" in 1860 and the Cambridge graduate and novelist Sir Walter Besant used "laylock" in 1881.

So it's quite clear that in predominant American usage the standard spelling and customary pronunciation of this word have diverse origins.

Blog 024

The 26th of March 2007

Spelling versus pronunciation

I've used the word "versus" in this heading very advisedly because it tends to worry me when spelling and sounds are at loggerheads with each other as you may recall from my blog of the 8th of February. The matter that's been troubling me recently has been the startling discrepancy between /ʧӕŋgɪ`raɪ/ ie the way most news media people are pronouncing the name of the Zimbabwean leader of the opposition and the spelling Tsvangirai. Ladefoged and Maddieson at page 171 of  The Sounds of the World's Languages remarked that the Shona language has so-called 'whistling fricatives' in which there is extreme lip rounding combined with a laminal alveolar gesture. I've been assured by Catherine Sangster of the BBC Pronunciation Unit that people at the Zimbabwe High Commission and World Service broadcasters they have consulted are agreed that English "ch" as in church is the best possible approximation to the sound that Shona speakers produce corresponding to the initial cluster "tsv" that they spell the word with. However, those letters, though no doubt taken from the International Phonetic Association's romanic-based alphabet — the orthography for Shona devised early in the last century having seemingly been based on it — don't succeed in conveying anything about the extreme lip rounding involved and presumably correctly indicate a labiodental element in the articulation not a usual feature of English /ʧ/. Spelling the name in an English context with initial "ch" is one supposes not acceptable but perhaps it would be better if we permitted ourselves the anglicisation of slipping an h in and making it Tshvangirai which might have the advantage of a compromise which would make looking it up and writing it down a little less disconcerting.        

Blog 023

The 9th of March 2007

TrainTimes Intoned

In his blog a week ago on Friday the second of March John Wells commented on the way the automated train announcement he heard at a railway station used intonations he notated for us as

The 'train at platform \/three | is the four/teen | forty-/two | to \Woking.

This was too unnatural to be really acceptable, as he said, "[b]ut unfortunately speech engineers tend not to consult phoneticians".

As a Mac user I'm able to have my computer read text aloud for me, and I knew that Apple (employing British phoneticians to help with some of the research, I believe) had a while ago put some effort into making their system at least do better than most railway stations and airport pre-recorded announcements, so I was interested to hear how that sentence came out. It was something like what I like to notate as this
The ˈtrain | at platform three (i)s | the ˈfourteen forty-two to `Woking.
This was pretty satisfactorily clear and natural-sounding to me – praps even too much so for the very formal context.

The most interesting tone employed was what I have been accustomed to call a Drop ie a movement from a high to a mid pitch. Those who remember O'Connor-&-Arnold's Intonation of Colloquial English will recall that they represented such a pitch movement by a raised downward-pointing arrow something like the one from Unicode I've put in front of "three" above hoping some browsers may be able to show it. It's a very common tone, practically the semantic equivalent of the full falling-rising tone John has shown above, and very common in both British and American speech — probably more so in the latter.

 Other interesting features were the weakening (1) of "is" to /z/ - so complete that it was almost lost, and (2) of the very American-sounding conversion of the / t / of "to" to [t̬] a sort of soft/tapped /d/.

Blog 022

The 26th of February 2007

The Gettysburg address

John Wells in his blog of today says in reference to the end of the Gettysburg address
when I was a schoolboy my history teacher ... thought there ought to be a repeated accenting of the word people ...
of the 'people, | by the 'people, | for the 'people,
rather than 
'of the people, | 'by the people, | 'for the people, |

Okay, but there are other ways of saying the phrase that can  achieve both effects  simultaneously and that's what happens more often than not in five recordings of speakers available on the Web viz (1) Jeff Daniels (2)  Sam Waterston (3) Jim Getty (4) Johnny Cash and (5) "Britton Rea". Getty and (5) alone give no pitch  prominence to of. Waterston gives it high pitch while using its weakform (the only case of a weakform on any of the prepositions). On by three have falling tones and two level pitches. On for two have falling tones and two have upper level pitches with accentual effect by not linking rhythmically with the following the; the fourth (3) gives it prehead treatment. The first people has falls for three of the speakers, a level tone by (5) and a vague narrow rise from Cash.  The second people has falls (of different ranges) from all except Cash who has another narrow rise. The third people has falls from all (mostly wide) except Cash again who gives it an accentless tail value.

(1) ˈof the | ˎ people, | ˈby the ˋˋpeople, | ˈfor the    ˋˋpeople,

(2) ˈof the ˎ people, | ˈby | the ˎ people, | ˈfor | the ˎpeople,

(3) of the ˋpeople, | ˋby the ˋpeople, | for the ˋpeople,

(4) ˋof the ˏˌ people, | ˋby the ˏˌ people, | ˋfor the people,

(5) of the ˈpeople, | ˋby the ˎ people, | ˎ for the ˋˋpeople,  

My advice to the EFL advanced user is to go each time for stressing the preposition and not people to get the best shot at an idiomatic native-English-speaker-type effect. After all, people is clearly 'given  information' where  it's to be taken for granted that democracy is preferred to tyranny etc.
For access to the recordings see Wikipedia entry on the Gettysburg Address which at "External Links" refers to  actors Sam Waterston, Jeff Daniels; musician Johnny Cash; actor Jim Getty. Also at "Lincoln's Gettysburg Address" the "Britton" recording can be accessed.

Blog 021

The 23rd of February 2007

W. H. Auden Again

It's possible that the odd reader of my transcription in my last blog of the Auden poem 
The Unknown Citizen
wdve realised that it wasn't the first time I'd published a transcription of that same poem because I did so in 1970 at pp 11-12 in the last but one issue of Le Maître Phonétique before its transmogrification into the Journal of the International Phonetic Association...

However, a glance at the previous version would show at once that I've no more represented it as spoken in the same way this time than I'd have re'd it aloud in identical fashion. This time I had in mind how Auden himself wdve spoken it. His speech was very peculiar in at least one way. He in general shared the characteristics of the other members of his class and upbringing. He described himself as upper middle class and had an accent to match ie resembling that of many of his university and public-school friends such as Stephen Spender – what one might call "patrician" or "posh". However, he decided to live in America and became a naturalised United States citizen and thereupon made a curious single concession to US speech patterns by changing his previous "Broad A" pronunciations of words like bath to the so-called "Flat A". Only one example of such a word occurs in our poem viz advantages. As to advertisement, I doubt if he wdve Americanised that to adver`tisement,  but I guess he might well have used the stressing `research in Producer Research because he no doubt, with his satirical ironic intentions, invented that title and High-Grade Living regarding them as typically American. The /d/ I show in hospital is extremely common in British speech – I didn't wish to suggest it was consciously adopted. Same goes for the version of necessary I've shown. By the way, at the phrase 'our eugenists say' the sequence /-ɪs seɪ/ represents an elision which was a deliberate representation of what I regard as its normal articulation.