Archive 13 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

20/09/2008English Phonetics Practicals (ii)#132
14/09/2008English Phonetics Practicals (i)#131
12/09/2008Daniel Jones b 12th Sept 1881#130
09/09/2008The Rigidity of English Spelling#129
08/09/2008Unorthodox Derived Forms#128
06/09/2008More Wells Coinages#127
03/09/2008Georgeayna's Cavendish patois#126
01/09/2008What to Excuse#125
28/08/2008Rhoticity etc#123

Blog 132

The 20th of September 2008

English Phonetics Practicals (ii)

In my last blog I mentioned the conversational dialogues that I have the students read aloud. Here is an example of such a passage:

Hard to Please

1.    Do take something more.  That’s not enough to keep a bird alive.
2.    Well I won’t be having any meat. I’m a vegetarian, you know.
3.    Oh.  Well do have more vegetables.  More potatoes, would you like?
4.    Well, I don’t usually eat any potatoes, either. I’m on a diet, you see.
5.    I’m sure you don’t need to be. I only wish I was as slim as you are.
6.    Well, I’m not really overweight, of course, but I do it for my health’s sake.
7.    Oh!  Is this your own idea or are you following your doctor’s orders?
8.    I’m afraid I’ve never been in the habit of setting much store by doctors.
9.    Oh! Perhaps you’re like Prince Charles, into alternative medicine, then.
10.    I wouldn’t compare myself with him. I’m just sceptical about all such things.
11.    You will have a little drop of wine, though, won’t you? Red, white or rosé?
12.    Oh dear! I shan’t be having any at all. I never touch alcohol. Not of any sort.
13.    Oh well. Never mind. Can I get you a soft drink perhaps? Orange juice?
14.    No. It’s quite alright, really, thank you. I don’t care for anything like that.
15.    Perhaps you’d like your coffee now. It’s not instant. I grind the beans myself.
16.    I don’t take coffee, either, I’m afraid. Nothing at all with any caffeine in it.
17.    Ah! Now I do believe I could dig out some decaffeinated if that’d be okay.
18.    No. I’m afraid not. It doesn’t agree with me. I’ve got a very delicate stomach.
19.    Would you like a nice cup of tea, then? Or is there caffeine in that too?
20.    I’m sorry to be such a nuisance but I’d really like just a cup of boiled water.
21.    You remind me of an old aunt of mine. She drank nothing but hot water.
22.    Well it can’t do you any harm and I’m sure it’s very good for the digestion.
23a.    That’s just the sort of thing that she used to say.
23b.     I’m sure you’re perfectly right, but I’m afraid I shan’t be joining you.  
23c.    As a matter of fact I rather think that I could do with a fairly stiff brandy.

I always ask each member of the group in succession to take the next turn. (I never have a dialogue spoken between only two group members at a time. It wd leave all the others silent for too long.) I make no attempt to go out of my way to include any unusual vocabulary items. Thus attention is concentrated entirely on how the sentences are spoken. For students at a high level of achievement there'll be very few occasions when the choice of sounds for any particular word isnt known. Therefore the exact quality of the sounds produced is what receives primary attention. One of the most frequent problems is unsuitable use of the strongforms of weakform words. The mistake I find most shocking — which fortunately occurs only occasionally — is the expanding of contractions into full forms. Of course very many problems are due to the direct influence of the individual speaker’s native language. For instance many Japanese speakers with a good pronunciation in general will have persistent problems with our /l/ sound. The kinds of problem that occur are so diverse that it’s only feasible here to give a few examples as follows.

    At Turn 1 in the first sentence (if it's the first dialogue I hear them reading) it'll become clear at its last sound if the reader is one of the comparatively few who have a tendency to use a more highly rhotic pronunciation than in General British which is the ordinary target accent of the course. (This variety is more often referred to as “RP” by most of my colleagues, tho increasingly some of them alternate it with other terms). If so, I may refer to the matter but I explain that I don’t find it a big issue except that I warn them that, if that is their preference, I shall probably mention the fact if they happen to use r-sounds very inconsistently.
    From time to time certain features of sentence prosody may need to be de’lt with. The second sentence of Turn 1 involves something similar to a stress idiom. It’s not obvious, at least for many non-native speakers, that the stress climax should not in this sentence occur in its most usual position on the last content (non-function) word of the sentence but that in fact it normally occurs in this expression on the word bird with the following word treated only as the tail to its tone. This is so here presumably because we feel the wish to avoid accenting of an item which is in effect a re-occurrence.

    At Turn 2 the word won’t is the likeliest item to need attention. It’s remarkable how often students with a generally good pronunciation fall foul of this word. Only a minority fail to aim at the required /əʊ/ but many seem to have difficulty in making it properly diphthongal causing it to sound often at least halfway to want. At Turn 5 don’t is sometimes treated in a similar way.

    At Turn 3, I may perhaps with very advanced students refer to the not terribly important point that it’s normal in current usage to make only three syllables of /`veʤtəblz/.

    At Turn 4 a similar more worthwhile point may need to be made about usually which one often hears used in a form too much like /`juːʒuːəli/. The dictionaries are not very helpful on this word. They give it in first placing in a form that seems to recommend or at least sanction saying it as four syllables when in fact it’s normally spoken as three, when not two, in current usage (as my CPD clearly indicated as long ago as 1972). With LPD and ALD this is more or less an effect of having to “unpack” a rather complicated notation. ODP gives first an unusually careful or pedantic version with the diphthong /ʊə/ as its middle syllable. EPD almost fails to give the adverb the separate entry it needs: you only find one at the abbreviation usu. but at that entry it is given in a more helpfully easy-to-comprehend form than by the other main dictionaries.
    This Turn provides a quite tricky decision for students who’ve le’rnt the rule about re-occurrences not being re-accented (see on this site 8.1.2.) when they come to the word potatoes because this rule may here be overridden by the one that requires contrasting items (here in effect possibly a contrast with vegetables) to be accented (See 8.1.1). 

    At Turn 6 we have one of the fairly rare occasions where an elision (here of an /s/) can be said to be essential. Having a clear break between the final /s/ of health’s and the initial one of sake wd be too unnatural an interruption of fluency to be left uncommented on from an advanced student.

   At Turn 13 for orange juice sometimes /`ɒrɪn ʤuːs/ may crop up with students who havnt gathered that it’s abnormal not to keep both the final and initial affricates in such a sequence. 

   At Turn 18 it’s usually only weaker students who make the curiously common learner’s mistake of saying /ə`fred/.

    At Turn 21 it may be that someone is inclined to give an unsuitable accent to mine.

Blog 131

The 14th of September 2008

English Phonetics Practicals (i)

Any regular visitors to this series of blogs will've noticed that there were only a few of them last month. The reason for this was not that I was away on holiday. Actually I was enjoying myself rather more than I do on most of my ordinary holidays. This was because I was teaching under perfect (completely noise-free) conditions two blissfully small groups of very competent students (8 or 9 never normally more than 10, some quite young, others not nearly so) at SCEP ie the Summer Course in English Phonetics held annually at University College London. (The title “College” is misleading because by itself this institution is among the handful of the largest and most important universities in the UK.) It was at UCL where I laid the foundation to what I've le'rnt about phonetics sparked by contacts with Gimson, O'Connor and others beginning in the fifties. This year was the 19th successive one I've had the  pleasure of working on SCEP since my retirement from Leeds University.

    SCEP was begun by Daniel Jones, continued by Gimson, and expanded very considerably by John Wells who last year handed the torch over to Michael Ashby. This unique course extends over two packed weeks in which there are daily small-group tutorials, a morning sequence of excellent core lectures and afternoon supplementary ones catering for various special interests, extra ear-training etc. This is all accompanied by a rich collection of additional materials available to the participants when they go online.

The first of my daily classes was labelled “Pronunciation”. In this I normally begin with dictating to the students what my colleagues call "nonsense words" but what I prefer to refer to as 'synthetic English words'. This means that any of them could theoretically be an English word because they are explicitly constructed not to violate the constraints of English phonotactics even if as words they dont actually exist. Not only that. They’re designed to concisely test the participants’ ability to discriminate between various crucial contrasts of English phonology. Thus for example a number of the words contained syllables which involved one or other of the so-called “long” vowels which were in the event given the very short realisations they normally receive in pre-fortis and pre-enclitic situations. There're almost always some group members that these words bring face to face with the problem that they are misjudging which phoneme they are hearing because they are concentrating their attention only on its present length rather than its (more important for recognition) quality. For example the word /fliːptəʊ/ gets regularly transcribed by some of the participants as /flɪptəʊ/ and /ðuːpɪts/ as /ðʊpɪts/.

I like to have the students in a closely packed circle around me so that they can write what they think they’re hearing and show it to me at once individually. When any one of them has any sound wrong I repeat the word for them alternating the form of it I used with their wrong version for comparison. I repeat the words more or less as many times as they like. My synthetic words, besides sounds not occurring at all in the languages of most of them, such as /ɜː/ and /ð/, will include consonant sequences they may only encounter in English such as /gw, ʃr, str, θw/ etc.

 After “playing this game” for a ten minutes or so I change over to dictating pieces of ordinary conversational English that contain no unfamiliar words but draw attention particularly to how very familiar ones can take (to them) surprising forms in different phonetic contexts. One trusts that this exercise, by enhancing their consciousness of such forms, will help them decipher unexpectedly difficult sound sequences they may hear from any English-speaker. In such passages one introduces variant word forms that come about by assimilations, elisions and compressions etc that only specialist pronouncing dictionaries may reveal the existence of to any notable extent and then not exhaustively so. An example is the fact that the words only, alright and always lose their /l/ with great frequency in ordinary fluent conversation. Another is that a phrase like Latin American can frequently lose /ɪ/ from both words becoming /lӕtn əmerkən/. Similarly the word I’ll may constantly be heard as /ɑːl/ and temporarily as /temprəli/.

After these two exercises I then devote the main part of the session to having them read aloud in turn sentences from a variety of dialogues in everyday language. I usually ask each person to say their sentence at least twice and then ask the next one or two students to say the same sentence again before moving on. Lack of fluency can be pointed out and required to be remedied. If prosodic errors occur they wont be neglected but the passages arnt designed to contain any more than a minimum of difficulties of that sort.

It quite often happens that a sound used may be given a quality that’s so unsuitable that it’s not possible to judge which phoneme is intended. This means that it’s necessary to determine the speaker’s intention. With vowels I always make sure that they keep handy for instant reference a numbered list of simple very common keywords so that there’s no doubt or delay in their explaining their intentions. For consonants one can use their alphabetical names or occasionally ask questions such as “Is that the middle consonant in pleasure?” or “Are you aiming at the thick and thin sound or the this and that one?”

In the course of the two weeks working with them I give out three short passages of conversational dialogue (less than 100 words each) to be transcribed away from the class and then given in to me to be marked. If necessary we may discuss any problems these have involved in class but, anyway, they get them all back from me individually annotated. For more on transcription see Blogs 119 to 122.

Blog 130

The 12th of September 2008

Daniel Jones b 12th Sept 1881

Portrait of Daniel Jones

There's not much to say about Jones today, the 127th anniversary of his birth, that I havent sed already on this website in the first of my Appreciations, The Legacy of Daniel Jones at Section 2 Item 1 in the main site and also in my blog 052 on the fortieth anniversary of his death at the fourth of December last year (in blog Archive 6) but some readers might like to hear the voice of the man who was such an important figure in the field of phonetics in the first half of the last century. I've taken for this purpose his reading of his introduction to his incredibly successful Phonetic Readings in English plus the first two very brief 'anecdotes' that follow it. At this late date I'm afraid the anecdotes, with texts unchanged since their first publication half a century earlier in 1912, sound rather ridiculously like something aimed at ten-year-olds. He made the recording at the age of 82 just four years before he died. The book's main sales were in Germany. Goodness knows how long it remained in print there if it's not still so. It wd seem to've been so at least till the end of the last century to judge from p.112 of the magnificent Jones biography The Real Professor Higgins by Beverley Collins and Inger Mees (Mouton de Gruyter, Berlin, 1999).

Here's Jones's voice. The segmental phonemes are shown in the kind of transcription he eventually considered best for most purposes.

This is Professor Daniel Jones speaking on the seventh of November 1963. I'm going to read aloud a number of anecdotes. They're published in a book of mine called Phonetic Readings in English published by Carl Winter of Heidelberg, Germany.

ˈðis iz | prəˈfesə ˈdæniəl ˴dʒounz spiːkiŋ | on ðə ˈsevnθ əv nouvembə | ˈnaintiːn siksti ˴θriː.
aim ˈgouiŋ tə ˈriːd | əlaud | ə ˈnʌmbər əv ˈanikdouts. | ðeə ˈpʌbliʃt | in ə `buk əv main |  koːld | fəuˈnetik ˈriːdiŋz | in ˴iŋgliʃ | pʌbliʃt  bai ˈkaːl vintə| əv ˴haidlbeəg| ˴dʒəːməni.

ˈnʌmbə ˴wʌn | ðə ˈdog in ðə ˴meindʒə. | ə ˈbad ˈtempəd dog | ˈwʌn dei | ˈfaund iz wei | intu ə meindʒə | ən ˈfaund it | sou ˈnais ən kʌmftəbl | ðət hiː meid ʌp iz maind | tə `stop ðeə. | wenˈevə ðə ˈkatl keim ˈniə | tu ˈiːt ðeə hei | hiː grauld ən `baːkt at ðəm | tə ˈfraitn ðəm oːf. | ˈwot ə ˈveri ˈselfiʃ dog | ikskleimd wʌn əv ði oksn | hiː `kaːnt iːt ðə hei himˈself | ənd iː ˈwount | let `ʌs iːt | huː kan.

ˈnʌmbə ˴tuː | ðə ˈtravlər ənd iz ˴dog. | ə travlə | wəz ˈdʒʌst ˈgouiŋ | tə ˈstaːt on ə dʒəːni | wen i ˈsoː iz dog | ˈstandiŋ ət ðə doː | ən `stretʃiŋ imself | ˈkʌm əloŋ ju leizi dog iː sed | ˈwot ə ju ˈweitiŋ `hiə foː | ðə dog ˈwagd iz teil | ən ˈsed | ai wəz `ounli weitiŋ | fə `juː maːstə. 

The word 'anecdotes' could be said to be slightly mispronounced, no doubt unintentionally, because he produces an unsuitably strong stress on its final syllable.

One can hear nothing particularly old-fashioned or socially conspicuous about his pronunciation of [sɪkstɪ] in this rhythmic context. It's clear that the two vowels are identical: in rhythmic contexts that highlighted the latter vowel it might not be so.

The first syllables of November and phonetic are plainly diphthongal which is unsurprising in this very careful style.

The phrase number of exhibits a very lightly tapped linking /r/ as also does traveller and in the title of the second anecdote. This again is pretty inconspicuous today so we must remember that when writers refer to tapped /r/ being very old-fashioned it's to be understood that they're thinking of quite strong tapping. The whipcrack taps that could often be heard in the enunciation of Noël Coward were pretty obviously conscious affectations in his day.

Notice that /eə/ in they're and the last syllable of Heidelberg are barely if at all diphthongal. I've never been happy with the Jones Outline paragraph 448 "The phoneme ɛə may be regarded as consisting of a single member; there are no phonemic variants differing to any marked extent from the above value."

The German words, we notice, are completely anglicised. 

Number 1: The Dog in the Manger

A bad tempered dog one day found his way into a manger and found it so nice and comfortable that he made up his mind to stop there. Whenever the cattle came near to eat their hay, he growled and barked at them to frighten them off. "What a very selfish dog!" exclaimed one of the oxen! "He can't eat the hay himself, and he won't let us eat who can".

Any readers who may've wondered what it sounded like when they re'd references to (old-fashioned or socially conspicuous) varieties of /ӕ/ which have "considerable constriction in the pharynx" (Cruttenden/Gimson 2008 p.113) will hear a very marked example of such an articulation here at the word bad. It could almost be described as "bleated".

The word off exhibits the very old-fashioned choice of the /ɔː/phoneme in a number of words in which /ɒ/ is now heard. (Cf 3.1.29d on this website.) Not only is the choice of phoneme not the one normally made in the group of words to which off belongs but compared with GB speakers today the quality of the vowel is markedly opener than the average now heard, as is that used at for in anecdote Number 2. 

Number 2: The Traveller and his Dog

The traveller was just going to start on a journey when he saw his dog standing at the door and stretching himself. "Come along, you lazy dog" he said "What are you waiting here for?" The dog wagged his tail and said "I was only waiting for you, Master".

Blog 129

The 9th of September 2008

The Rigidity of English Spelling

Today’s ideas in John Wells's blog Freeing up spelling strike me as taking a somewhat new line from his past writings on the topic. I find them agreeably in tune with the suggestions I’ve recently been making in items on this website notably at my Blogs 102 Spelling Reform - Feasible or Futile? and 47 Spelling Reform. See also the warning about my spellings-in-them that precedes these blogs.

One of his suggestions was “Lets abolish the apostrophe, or at least make it optional”. In my observation it’s very widely become treated as optional already in practice except in contracted spellings like we’ll for we will but I think I’d rather keep it there than use we ll. Spaces are such vague things especially in many people’s use of keyboards. Anyway let’s stop sneering at grocers. The Times reported him as saying “It’s time to remove the fetish that says that correct spelling is a principal (principle?) mark of being educated.” I agree and perhaps principl wd be allriet. I can see his point of having only its inste’d of it’s as well: we’ve traditionally got them the rong way round anyway. I dont care for i and u but I might get to like U for you.

I notice that the Spelling Society, of which John is President, no longer advocates wholesale adoption of the system devised by its predecessor the Simplified Spelling Society and set out in the book Nue Speling by Walter Ripman and William Archer published in 1948 in its 6th edition revised by Daniel Jones and Harold Orton. It was an excellent effort but has never had any real likelihood of being adopted.

I can’t help feeling that John’s admirable realistic address to them on the 25th of January 1986 English accents and their implications for spelling reform mustv (I’m not sure I like that spelling and I wonder what John wd think of it) done as much to damp down (undiscriminating) enthusiasm for reform among many of their members as much as to encourage them in their zeal.

It’s not going to be easy to persuade people along the lines that John and I have advocated. It’s witness to people’s unwarrantable veneration for orthography, however illogical, that they tend to prefer to adopt a matching pronunciation for a bad spelling rather than make an orthographical change (thereby carrying out slow piecemeal reform). It wdve made good sense (and sound etymology) to change /sɪzm/ from schism to *scism or *sism but, no, most people have now made it /skɪzm/. Its the same story with combat, forehead, nephew, often, retch etc etc.

Harking back to my remarks on the futility of the BBC attempting to force broadcasters to toe the line on all pronunciations, especially of French names, in my blog 124. On the morning Today programme on September the 8th on Radio 4, within a short time, regarding the same short news item, I heard the newsreader (Rory Morrison), the presenter (Edward Stourton) and a correspondent (Mark Mardell) say the French president’s name Sarkosi respectively as /`sɑːkəʊzi/,  /sɑːkəʊ`zi/ and  /sɑː`kəʊzi/. They were all okay by me tho the last one is probably the most comfortable for English-speakers and the second one — the one that the Beeb are trying to force on them — is probably the least so.

Blog 128

The 8th of September 2008

Unorthodox Derived Forms

John Wells in his last two blogs has raised questions about various derived forms. How Congolese got its l is certainly a mystery but John’s explanation of Harrovian’s <v> is clearly the right one. And this may well account too for Monrovia. The American president after whom it was named of course spelt his name Monroe but this surname was earlier Munro and has been spelt by some of its bearers as Munrow etc. The coiners may well have had such a spelling in mind.

It’s not too surprising that the christeners of the district above Oxford Street from Soho shdve rejected *Fitzroia in favour of Fitzrovia nor that there’s little use of any adjectivalisation of Waugh: *Waughan or *Waughian hardly roll sweetly off the tongue. The same may perhaps be said of Kittitian but*Kittsian doesnt seem too bad to one used to Keatsian so maybe the OED suggestion of an inclination to parallel with Haitian wasnt unreasonable. Euphonistic considerations have also no doubt been responsible for the rejection of *Shanghai-ese and *Lasso-ians.

As to tobacconist and Torontonian, its worth remembering that a century ago a knowledge of Latin and even of some Ancient Greek was extremely widespread among people of any pretension to education. This wd mean that they were no doubt to some extent conscious of the fact that, since any word ending with -o wd be almost certain to form derivatives which involved an n, there might well be a tendency to follow this analogy. Witness n-extensions (for non-nominative cases and other derivatives) in names like Apollo, Cato, Cicero, Dido, Juno, Nero, Pluto, Strabo, Varro etc and words such as bubo, harundo, homo, imago, ratio, testudo, vertigo etc. This may even have some relevance to the choice of forms of Balinese, Javanese and Sundanese. (In regard to Stowe alumni being called Stoics, that’s what OED past editors mightve called “sportive”).

PS There are various adjectives like Aberdonian, Bathonian, Dundonian and Scillonian besides the oddities Mexiconian, Grubstreetonian, grumbletonian and so on. Smilar is Erpenius the Latinisation of the name of the Dutch linguistic scholar Van Erpe, a practice so common in medieval and renaissance Europe.
PPS Pharaonic seems to be another example.

Blog 127

The 6th of September 2008

More Wells Coinages

In my blog 123 “Rhoticity” which was mainly about the coining of linguistic terms I mentioned half a dozen of John Wells’s coinages. One problem with suggesting that a word is appearing for the first time is that one can’t be sure that it is so merely because one can’t trace it earlier. I don’t think for a moment that I’m the first writer to employ the word anglophonia but it’s not in OED3 nor does Google instance it earlier than my use of it in 2003 at p. 19 of Volume No 87 of The Phonetician Shortly after another spell in anglophonia, this time teaching at the Summer Institute of Linguistics at Ann Arbor...”. Similarly I have the habit of occasionally using in my private writings the term "Hispanophone" and, as I never recall seeing it used by anyone else and it's not to be found in OED3, to an extent it can be termed as my coinage, but I doubt that it's never occurred to any other writer to employ it.  Thus I'm not quite sure what to say about Wells's (standard) lexical sets.

[PS There is one notable thing to be said about these that they are all taken over en bloc for use in the key to the pronunciation of the online and in-preparation third edition of the great Oxford English Dictionary. Sadly Wells's set of symbols of LPD are not all used especially regrettably the wrongheaded failure to use his and all other sensible lexicographers' representation of the price diphthong.]

It seems likely that Wells re-invented smoothing rather than consciously or subconsciously remembered it from reading Sweet (who only used it in a couple of his least re’d books eg New English Grammar 1891 §720) as we see from his wording at its introduction at AE (Accents of English) p.238: We may refer to this monophthonging  process as Smoothing. Similarly his use of the phrases “Pre-R Breaking” and “Pre-Schwa Laxing”, like his term “D-tapping”, are certainly coinages but the phonological term “breaking” was in use by A. J. Ellis and others before him (in line with the German grammarians’ brechung), just as “laxing” was no doubt used by others before Chomsky and Halle.

Anyway, I thaut I might mention some other (presumable) Wells coinages. The one of his I hope to be forgiven for referring to as his happiest one is “the happy vowel” for the final-y vowel of vast numbers of English words for which he wisely and hugely influentially (there’s three of them for you!) accepted in his publications the LDC (Longman Dictionary of Contemporary English) innovation in its first edition of 1978 of representing these vowels with colon-free /i/.

The credit for introducing this usage in its Wellsian sense I’ve always naturally attributed to that dictionary’s initial specialist “pronunciation editor”, Gordon Walsh. Though I had many conversations with him around that time I don’t actually remember confirming the fact of it with him. Gordon was an excellent phonetician who studied under Peter MacCarthy and my predecessor at the Leeds Department of Phonetics, Beatrice Honikman, who I know thaut very highly of him. In subsequent years I’ve unfortunately lost touch with him since he left Longman but I’d be very pleased, if he or a friend sees this, if he’d make contact. Anyway, it’s long been my private conviction that he felt that his /i/ symbolisation was a good way of representing the General British pronunciation of such words avoiding both the unacceptable traditional representation /ɪ/ and the misleading notation /iː/. It was on the part of the publishers cautiously presented to the reader only as a device for economically conveying simultaneously both the General American and the General British usages in one symbol but by the second edition of LDC a decade later in 1987, when the pronunciation editor had changed and the pronunciation adviser was no longer Gimson but Wells, the representation was not merely so glossed (see p.F53).

Perhaps a less than happy AE usage is the habitual hyphenation passim “traditional-dialect” which strikes me as disturbingly irregular when it’s not attributive. In any case I shall give the rest of these coinages of his alphabetically:

consonant singling: AE p.212 “(double /dd/ in ladder becomes single /d/, and likewise other doubled consonants)

continental vowellism AE p.108 This chiefly involves giving to the spelling letters a, e, and i the values of PALM, FACE and FLEECE... (Compare Section 3 Item 7 III 4 on this main website.)

dealveolar (AE p. 58) “The (optional) RP rule of Dealveolar Assimilation is potentially triggered whenever a word ending in an alveolar stop occurs before a word beginning with a bilabial or velar stop.”

 ‘disalesce’ (AE p. 58) is an obvious nonce creation to convey the reverse of coalesce. Its apologetic inverted commas acknowledge its unorthodox formation and perhaps incomplete seriousness.

glottalling (rather than glottalising) which occurs first at AE p.102: The current British trend towards general syllable-final glottalling ... constitutes a rule simplification.”

loan-phoneme AE p.190: “Even in England /x/ can be said to hold a tenuous and marginal position in the consonant system of educated speakers ... it is clearly a loan-phoneme.”

shading (AE p. 534) was obviously introduced as a conscious coinage “... this is where what we may call Shading comes into play. This is the effect upon vowel timbre of the nature of a following consonant” .

vocalization to mean conversion (of a consonant) to a vowel (not in OED3) appears first at AE p. 20; cf p.295 “Vocalization of dark /l/ is occasionally met with in RP...”.

So we see that John was being over-modest when he referred to rhotic as "my own personal contribution to the English language" as if it were the only one.

Blog 126

The 3rd of September 2008

Georgeayna's Cavendish patois

Georgiana Cavendish, duchess of Devonshire (1757–1806), born  on the 7th of June 1757, the eldest daughter of John Spencer, first Earl Spencer, at the age of 17 married the duke in June 1774. She was the subject in 1998 of an excellent  biography by Amanda Foreman which has now served as the basis of a film The Duchess about to be released. The actress who plays the name part in it has been much interviewed on BBC etc including on Radio 5 in a series called by what hints at a Cockney version of “Daily Mail” but turns out to be “Daily Mayo”, the interviewer’s name being Mayo.

In it that actress, Keira /kɪərə/ Knightley, is questioned about the pronunciation of the Duchess’s name Georgiana when the interviewer remarks very understandably that, tho he was expecting /ʤɔːʤi`ɑːnə/, “through the movie everyone says /ʤɔː`ʤeɪnə/”. She replies “They do” and goes on to say “there was a special thing called the ‘Cavendish drawl’ which we haven’t really done in the film apart from the name ... I think they did pronounce it /ʤɔː`ʒeɪnə/”. (Incident’ly, that seems to lend support to my contention that people often don’t feel conscious of any very significant difference between /ʒ/ and /dʒ/. Compare the comments on Beijing etc in my blog 124 and the way many people say eg garage or management etc with indifferently /ʒ/ or /dʒ/.)

Just what the ‘Cavendish drawl’ amounted to isnt easy to g’age but no doubt the alternative term ‘Cavendish patois’ was nearer the mark. The form ‘George-ayna’ was probably quite widely current in the 18th century but perhaps beginning to sound old-fashioned to most people by the time she married the duke. There are many words that have exhibited a similar variation between a more English-developed and a more Continental-influenced treatment of such an ending. The  OED has the entries “ana (ˈeɪnə,ˈɑːnə), suffix and n.” and “Africana (ӕfrɪˈkɑːnə, -eɪnə)”. The order of the pronunciations given for the suffix no doubt no longer reflects accurately the current predominating preference but in the USA, and probably also still here by some who use the term, the organ stop vox humana may be heard with the non-Continentalised vowel as /vɒks hju`meɪnə/. Similar variations have occurred with names like Anastasia, Maria, Sabrina, Sophia etc.

The Foreman biography quotes various examples of the phenomenon called variously ‘Cavendish drawl’ and ‘Devonshire drawl’ both of which terms, even taking into account her mention of “bal-cony, con-template” and “cowcumber”, are no doubt less appropriate than the third one ‘Cavendish patois’. In the OED Murray remarked in 1885 “Till c1825 the pronunc. was regularly bælˈkəʊnɪ; but ˈbælkənɪ... , ‘which,’ said Samuel Rogers, ‘makes me sick,’ is now established.” I’ve updated from Murray’s symbols to the current OED3 style. OED dodgily gives the con`template stressing in second place (as an allegedly current usage) even tho retaining Murray’s century-plus-old comment that it “begins to have a flavour of age”. On cowcumber it gives his 1893 remark that such a pronunciation was “still that recognized by Walker; but Smart 1836 says ‘no well-taught person, except of the old school, now says cow-cumber ... although any other pronunciation ... would have been pedantic some thirty years ago”. Other items of the patois Foreman quotes are yellow, gold and spoil presumably pronounced as /jӕlə, guːld/ and /spaɪl/ and a “baby-talk” style you pronounced with its yod dropped. (Compare OED’s quote from the 1713 Swift Journal to Stella  I allow oo Six). Except for this last item, these were all probably merely old-fashioned. For example Wyld's History of Modern Colloquial English (1920/36 p. 239) sed of goold that “It was a very usual though by no means the only pronunciation in the seventeenth and eighteenth centuries among good speakers.

The “patois” may perhaps have been characterised by particular rhythmic and/or melodic features but these can now hardly be guessed at. As regards the above set of verbal usages reported at Foreman 1999 pp 45/46, they seem all to have been noted from the letter to her mother of the 12th of January 1784 (ie when she was sixteen and a half). So they should perhaps best be interpreted as the slight over-reactions to the speech habits of another grand family of a very inexperienced young woman in an age when many usages were less settled than they are today.

My attention was drawn to this topic, as so often, by John Wells's latest blog where he helpfully provides a link to the podcast of the broadcast I've quoted from. He makes the comparison with the Victorian /ɑː`meɪdə/ for armada which, as he points out, Daniel Jones never got round to removing from his EPD. Neither did Gimson or Ramsaran. Only when Roach and Setter took over in 1997 was it removed tho who was still using it after about 1920 one wonders. It was only withdrawn from the Concise Oxford Dictionary in 1976 in which it had remained until that time as the only pronunciation recorded for the word: see on this website (at Section 11, Item 3, No 5) my letter that was printed in The Times in 1977 which mentioned the fact. This to my profound regret much irritated the then Chief Editor of Oxford Dictionaries, the great lexicographer whom I so much admired, the late Robert Burchfield. I had mentioned it as symptomatic of the minimal attention those dictionaries paid to pronunciation matters. Such things seem a good deal better these days.

Blog 125

The 1st of September 2008

What to Excuse

John Wells’s phonetic blog of today is, as ever, thaut-provoking. It turns out to be entirely about various transcriptions produced by examination candidates of the first two words only of a passage that he “dictated in a perfectly ordinary way, as ɪkˈskjuːz mi.” I’m afraid I cdnt help wondering whether he aimed to repeat these words on the same intonation pattern each time or only adhered to the same stress pattern. [PS He's now told me that he stuck to a fall-rise on Excuse.] I’ve known occasions when I’ve felt free to vary the intonation to help the candidates determine what was being said. Anyway, one quite minor but potentially quite perplexing point of transcription is possibly involved as we shall see.

He continued:
Among the versions I was offered by the examinees were ɪˈkjus me, ɪˈkjʊz ˈmiː, ˈıkskjuz mi, and several cases of ɪkˈskjuːs mi.

The first two of these were quite seriously wrong both omitting the pair of segments /sk/. The first, besides the omissions, had /s/ for /z/ and in /e/ a seriously inappropriate final vowel.

The second, besides the omissions, had /ʊ/ instead of /uː/ and a debatable vowel value in the second word. Taking it that the examinees were required to make use of the LPD/EPD symbology, and that the prosody fitted the O&A (O’Connor & Arnold) terminology Fall-Rise, then strictly speaking one might well expect the non-accentual rising-syllable vowel to be transcribed with a weak /i/ rather than strong coloned /iː/ tho this latter choice of course wd be a very venial “error”. On the other hand if the speaker’s prosody had been consistently what O&A wdve had to represent as Exˈcuse ˏme, then a student trained to use O&A notation would have been justified as regarding “me” as accented and therefore requiring segmental notation as a strong vowel by inclusion of an IPA length mark.  The problem wd be that the O&A notation never had within its purview the existence of the complex tone which in my preferred terminology I refer to as the Alt-Rise ie a very frequent one employed on this common expression which I am accustomed to notate as Exˈˏcuse me. (A more emphatic variant cd  be Exˋcuse ˋˏme. For “Alt” see on this main website Section 8 §3.) ː

In the third example ˈıkskjuz mi the tonic syllable is grossly mistaken. Looking closely at the first letter of this transcription one is tempted to wonder if with meticulous precision the calligraphy is being shown in a way that might represent a Turkish student slipping into their national graphetics in a failed attempt at writing the IPA symbol ɪ — but that wd no doubt be too extravagant a conceit to sustain belief. No doubt it's just one of those graphological gremlins at work of the kind that wdnt let me use the original Turkish dotless i of the Wells text on this page. [PS On the contrary I find that he was indeed meticulously recording the student's use of a dotless i!]

 At the fourth ɪkˈskjuːs mi there is the common failure to observe “the s ~ z noun~verb alternation in use, abuse, excuse, refuse, house...” The rest of this ten-line aside gives a concise summary of similar types of s ~̴ z alternation which shd prove very useful to many students whose attention is drawn to it.

It was rather surprising to hear that the students found it easier to transcribe Synthetic English words than real existing ones. I think I’ve always found it the other way round. [PS I now gather that his "nonsense words" were not English but "all vowels cardinal, many consonants exotic etc" so my astonishment at hearing that the students  perform better on them is even greater tho I supose it cou'd be that more English segments than international ones conduced to a higher total of mistakes ]

Blog 124

The 31st of August 2008


The title of this present blog (Canutism /kə`njuːtɪzm/) ties in with the topic of my last one because that one was largely concerned with coinages. This expression is in fact a personal usage I've long enjoyed employing to refer (however unfairly to the monarch in question) to the fact that many people delude themselves into believing that a single authority's influence can be so great as to prevail over a wide popular tide of development. I usually have in mind linguistic processes and particularly trends in the pronunciation of individual words. This blog, then, has been prompted by the blog of August the 14th by Graham Pointon which began

It’s been very noticeable over the past week or so that almost all BBC broadcasters, from whatever department, are now saying ‘bay-jing’ for the Chinese capital. It’s been confirmed today by “a BBC employee” that a directive has been sent out by senior management that everyone must toe this line. Graham went on to refer to a 1974 booklet called BBC Pronunciation Policy and Practice, which, he said, laid down that all BBC staff newsreaders and presenters must follow the recommendations of the Pronunciation Unit adding his questionable opinion that a similar instruction now would do no harm whatsoever.

After grumbles about the arcane Byzantine convolutions of control within the Beeb he closed with the unsurprising remark that a former Controller of Editorial Policy referred to pronunciation as a “can of worms” that he wanted nothing to do with, although all he would need to do would be to accept the recommendations of the Unit — they are all very good linguists, and spend their lives researching the pronunciations which journalists often dismiss in two seconds flat. (Good!) Perhaps a start could be made by standardising the French President’s name: there is still a lot of disagreement between the second-syllable and third-syllable stressers.

My opinion is in direct contrast with that of my good friend Graham. I feel that the function of the truly invaluable BBC Pronunciation Unit shd be entirely consultative and that no broadcaster of whatever standing shd be prohibited from using any pronunciation of any word which he or she feels is appropriate for the context in which it is to be heard — or for that matter pressured to conform to a standard set up by others. The more certainly the Beeb steer clear of any tendency to control freakery the better. Attempting to force people to "toe the line" against their will or better judgment is extremely ill-advised. More detail on this matter can be seen in my review of OBG (the Oxford BBC Guide to Pronunciation) on this website at Section 12 Item 11.

I believe this partly because I feel no advisor or group of advisors, however well-equipped for such work, shd be deemed infallible in their judgment. To take the example of Sarkozi I dont think it matters much at all whether one sez /sɑː`kəʊzi/ or /sɑːkəʊ`ziː/. However, I do doubt if the latter version is likely to feel very comfortable to most people and I deplore the way it is given in OBG as the BBC's sole recommendation, especially in the light of the above-mentioned booklet on BBC Pronunciation Policy and Practice. English speakers are likely to feel that such a word pronounced as three strong syllables is too heavy: in fact I suspect very few will simply adopt that recommendation. If they attempt to do so while having the  occasion to use the word repeatedly, they'll soon be weakening the middle syllable to give /sɑːkə`ziː/ as is the common pattern in English words of three syllables the last of which is strong as with for example jamboree.

Again, I can't share Graham's evident satisfaction that a fatwa came from on high at the Beeb (no name disclosed!) prohibiting the probable majority version /ˈbeɪ`ʒɪŋ/ which OBG sez tiresomely is “not correct”. Will they be putting one out about Taj Mahal shou'd that figure prominently in the news shortly? And one wonders what the punishment was for infringement! LPD3 gives the /dʒ/ version first and observes in deprecating language (“no justification”) that it alone occurs in Chinese. EPD and ODP are both sensibly non-committal regarding which variant is the predominant one etc.

Blog 123

The 28th of August 2008

Rhoticity etc

John Wells has invited us to celebrate with him the "fortieth anniversary of the word rhotic", which, he remarks, "is my own personal contribution to the English language." While I shd not wish to be so churlish as to fail to congratulate him on its 68,000 hits in Google and so on, I have to confess that, tho I've often written about the topic it refers to, I have regularly consciously avoided using the term in his sense. My reasons for this, handy as it has proved in avoiding the clumsy term r-ful, have been twofold.

One has been the rather subjective purist feeling that it shd ideally be kept as a term to describe language varieties that contain r-sounds as opposed to ones that contain only l-sounds (lambdic, to coin a word) or like English have both (lambdic-rhotic). I dont think I cd reconcile myself to rhotiferous.

However, my main reason has been the feeling that, neat tho it might be, it was too broad for my purposes. Accordingly I've always referred to varieties of English, using instead his other coinage, as of higher or lower rhoticity. This has meant that I have referred to various kinds of southern United States English as of lower rhoticity than any other type (alternatively hyporhotic), considered General British and most east-of-England, eastern US and sub-equatorial accents as of medium rhoticity (mesorhotic), and referred to Irish, Scottish and General American etc as of high rhoticity (or polyrhotic). "Cruttenden's Gimson" (2008 p. 90) remarks "some, particularly Jamaicans, may be semi-rhotic (e.g. /r/ is present word-finally in hear but not preconsonantally in weird)". In fact semi-rhoticity is first mentioned in Wells's Accents of English p.570.

I notice that Wells's other 1968 coinage hyperrhotic hasnt been taken up much. It seems to me that it cou'd well be welcomed as a slightly less questionable expression to describe varieties that embody analogical /r/ of the type that is very often called "intrusive". Such terminology I find rather uncomfortably lends itself to being pounced upon gleefully by purists as a value judgment.

Wells's influence, especially in his prodigious Accents of English, has had notable effects on the currency of a variety of other terms including his revival of Henry Sweet's 1988 History of English Sounds use of smoothing which Sweet defined as "the levelling of the two elements of a diphthong under a monophthong". Others of Wells's often taken up terms have included TH fronting, glottalling, diphthong shift and pre-fortis clipping. I've never really cared in this last case for the clipping metaphor rather than the more usual terms shortening or reduction and I've not understood why he hasnt employed the conveniently parallel term pre-enclitic (shortening) which I've found quite useful.

Two items first employed by the present writer wdntve acquired the familiarity they now have if they hadnt appeared in his works. One of these is the term compression introduced in 1969 in my Guide to English Pronunciation (Scandinavian Universities Press) where at p. 35 it said "A compression occurs when a diphthong is reduced to a simple vowel or a vowel reduced to an approximant..."

The other is yod dropping, which I first used in my 1971 article The American and British accents of English in English Language Teaching (pp 238-48), and which, like rhotic, has now appeared in the OED which has the entry at yod n(oun)2:
1982 J. C. WELLS Accents of English I. ii. 163 ... GenAm reflects more widespread Yod Dropping than RP and most other British accents.