Archive 39 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

26/03/2012Geoff Lindsey's Great Quiz#390
23/03/2012A Patricia(n) Accent#389
21/03/2012GB Phoneme Frequencies#388
19/03/2012The Unique Word 'threepence'#387
24/02/2012Understanding Phonetics#386
22/01/2012R Dropping (i)#385
10/01/2012Rants on Odd Pronunciations#383
08/01/2012Spelling Pronunciations (ii)#382
06/01/2012Big Deal#381

Blog 390

The 26th of March 2012

Geoff Lindsey's Great Quiz

Grand Ear-Training Quiz for Non-Native Learners/Users of English

is another brilliant gift to his readers. If you havnt he·rd about it until now, leave this at once if you can, and go and try it. It's not very difficult but hugely instructive to users of English as an extra language who like to polish their performance a bit. It's here.

I notice Geoff has used the common description "Learners/Users". I decided some time ago that it's better to say simply 'users' because it's more satisfactory "politically" to use the curiously suddenly popular adverb that became almost notoriously ubiquitous in the later twentieth century in the phrase 'politically correct'. The OED defines that fairly satisfactorily, especially at the end where it sez, "often by the rejection of language, behaviour, etc., considered discriminatory or offensive".
I'm sorry to say I do sometimes use the abbreviation 'EFL' to avoid irritating readers unfamiliar with them if I use EEL or EAL by which I'd mean respectively 'English as an Extra Language' and 'English as an Additional Language' the former of which sounds more snappy but the second more dignified. EIL ie 'English as an International Language' is another I find quite acceptable. But I don't like 'English as a Foreign Language'. My non-British pupils and colleagues I've normally lookd on as my fr·ends so I'm uncomfortable about referring to them as forren or alien users of English. They're fellow users who dont happen to've begun acquiring the language at their mother's knee. I've spent plenty of years in overseas countries where it'd seemed topsy-turvy for me to be referring to my students as forren learners etc. I was the forrener, the alien, amongst them. By the way, tho I seem to've rabbited on too long about these things, I dont find Geoff's use of the expression "Learners/Users" at all objectionable: only that I think my choice more graceful. Something I do deplore as sadly infelicitous is his use of "Standard" for any accent.

Now here're some trivial comments on a few of Geoff's numbered items below:

4. If this had been an exam I'd've given full marks for either 'I work quite' or 'I've worked quite'.

5. "British vowels are clearly divided into those which are intrinsically short and those which are intrinsically long". There's quite a lot in that but I've found it not much worth mentioning to my EEL students. In fact I suspect even some of the thautful ones of taking it too simplistic·ly.
"One consequence is that lax vowels like [ɪ] may be drawn out longer than would be typical for the same context in British English". I don't think I really buy that. I'd cert·nly say that it's perfec·ly ord·n·ry for a GB speaker to draw out the is vowel quite as much as that American does here.

11. "It would be at least as accurate, and probably more helpful to foreign learners, if we transcribed Pict and picked not as [pɪkt] but rather as [pɪkd], rapt and wrapped not as [ɹapt] but rather as [ɹapd]. I liked ales is far closer to I like dales than to I like tales".
I very much agree: cf reach Dover and reached over etc. I'm not so sure whether one can make a similar statement about [katz] for cats. Seems praps more likely with some prosodies than others but `cats ˏare and `cat ˏczar pronounced with either of the two possible initial choices for czar as [z] or [s] seem not to sound wrong.
12. "Oxford University doesn’t bother with the acute accent marks, ... whereas the Royal Opera House does".
The Oxford English Dictionary has répétiteur. Geoff is being very nauty here. The failure to employ the French accents in the advert Geoff quotes is obviously only the personal responsibility of a single Oxford don, a Professor Franklin.
"On both sides of the Atlantic, it’s stressed on the final syllable. In my experience, Brits pronounce it ɹəpɛtɪˈtəə to rhyme with blur or chauffeur. Americans seem inconsistent about the French ending -eur[They sure are! JWL]; Hampson pronounces the word as if it were répétiture, which it isn’t".
A very unfair criticism. He has the company of plenty of other Yanks. Cf Merriam-Webster Online has amateur as \ˈa-mə-(ˌ)tər, -ˌtu̇r, -ˌtyu̇r, -ˌchu̇r, -chər\, saboteur as \ˌsa-bə-ˈtər, -ˈtu̇r, -ˈtyu̇r\ and entreprenure as \ˌäⁿn--n-trə-p(r)ə-ˈnər, -ˈn(y)u̇r\.

14. "Most but not all quiz respondents correctly heard carrying. Vowel merger before [ɹ] means that marry, merry and Mary are homophones with ɛɹ for many Americans. To those who aren’t sensitive to this aspect of American English, Hampson’s carrying may well sound more like caring".
He didnt comment on the fact that in GB and GA the medial yod in words like marrying, worrying, varying etc is offen, as here with carrying, dropd.

17.  I hear /dun/ ie doing and quite different from 'in the'
18.  He makes no comment on a in the sub-title for this.
20. Here tried to strikes me as completely ambiguous between try to and tried to.

It was a splendid idea of Geoff's to use this material.

Blog 389

The 23rd of March 2012

A Patricia(n) Accent

"Kraut" made available to us at his posting of the 20th of March a fascinating sound clip of Patricia Hughes whose 'dulcet tones' (as the cliché has it) were emblematic of the BBC's "culture-rich" radio service started in 1946 as 'The Third Programme' and renamed 'Radio Three' in 1970. After 'signing in' she reads the first minute of a news bulletin. Her routine self-introduction etc is delivered quite briskly, so fast indeed that at and in a few moments she elides the vowel of the word in /ӕn nə 'fjuː ˏməʊmənts/.

We're not informed of the date of the recording, which isnt at·all high-fidelity, but we gather that she retired in 1983. The way she sez 'BBC' is the completely normal one we have today in fluent speech non-finally /ˈbiː bi ˈsiː.../. In the kind of essentially phonemic transcription we see in LPD and CEPD both show it as /ˌbiː biː ˈsiː /. I think users of reference books offen fail to realise that the lexical pronunciation given first by these dictionaries isnt necessarily the one people use most offen. In this case the usual character, especially in conversation, of the vowel of the middle syllable is the weak value it has as indicated in lexical transcriptions of the final vowel of happy which is usually represented as /i/ ie not / iː /. CEPD doesnt list this variant.  LPD does.

She sez ..we'll /hjɜː/ the ˈtime ˈsignal.. pronouncing the word hear in a way that Daniel Jones always listed with the value /hjɜː/ as a common variant from /hɪə/ and I don't think she sounds in the least old-fashioned doing so. There certainly were signs in the third quarter of the last century that this formerly very common variant was becoming much less common. But I was shockt to see that Gimson, when in 1977 he braut out his only major revision of the Jones EPD, had completely removed /hjɜː/ at here, hear and various other words. So have Wells and Roach. This was quite unjustified because /hjɜː/ is still a perfectly common variant when the word isnt maximally prominent. It's not likely that you'll hear a GB speaker these days say /ɪts `nɒt `ˏhjɜː/ (ie on a Fall-Rise tone) for It's not here rather than /`ˏhɪə/, but many expressions like Here we are are still perfectly commonly he·rd today as /`hjɜː wi ˏɑː/ without people perceiving them as odd or dated. By the way, at year Gimson still gave /jɜː, jɪə/, an order reversed by his successors.

Her esses seem offen quite labiodentalised [ s͡f ] praps from a slightly idiosyncratic articulatory habit or maybe necessitated by her dentition. This is particu·ly clearly audible at the word sign. Her markedly cheerful wide low-to-high rise on "everyone" involves a paralinguistic sustaining of its final syllable that seems to contradict its intrinsic stress value enuff to possibly make some EAL users wonder if her normal accentual pattern for the word is */ˈevri`wᴧn/. It isnt. This continuity-indicating upper level tone is one of a succession she uses where, if it were conversation, she'd no dou·t use lowish rises.

She vacillates between rather old-fashioned and more modern usages as so many of us do. Notice the ending -ies has a modern fairly close [i] value in countries but not in authorities where it's much more like the recessive [ɪ] quality. She quite noticeably has an old-fashioned value for her /ӕ/ as in apparent and actions. The mainstream GB 'ash' vowel /ӕ/ today wd require in general-phonetic use, such as in dialectology, to be  recorded as [a]. The OED from 1995, after listening to advice from the distinguisht dialectologist Clive Upton, adopted /a/ for the ash vowel in its notation of British usages, regrettably but not unreasonably. The usual symbol [ӕ] for GB is one of certain items that by tradition and for the sake of continuity have been retained in most other ref·rence and text books, people preferring to avoid change and cons·quently not to conform with the practice of phonetic specialists whose work obliges them to strictly adhere to the cardinal-vowel system. If they dont, they wont be conveying to other specialists exactly what they mean.

At 0.43 where the word our is very fully strest it's noticeable that she doesnt use the disyllabic form /`aʊ.ə/ but clearly has the fully smoothed form /ɑː/. A somewhat socially conspicuous item in her speech can be noted where she slightly weakens the second element of the diphthong /aʊ/ to something like  [naɘ], ie ending not rounded and more central than [ʊ] a feature at times he·rd from the Queen and by Prince Charles who when mockt for it evidently came to avoid it. On the contrary her syncopation of particularly to /pə`tɪkjuli/ was quite modern. It's perfectly common mainstream GB today but the lexicographers are in no hurry to admit it yet. Finally the last word she sed, namely unacceptable /ˈᴧnɪk`septəbl/, featured another usage in which she was ahead of her time. The dictionaries still represent acceptable and accept as predominantly /əksept/ but in twenty-first century GB it's overwhelmingly offen begun with /ɪ/ rather than schwa.

Blog 388

The 21st of March 2012

GB Phoneme Frequencies

Denis Butler Fry (1907-1983), author of superb books including Homo Loquens and The Physics of Speech, was an outstanding language scientist and a delightful person. I regret now having never asked him for more detail than he gave us in a certain article he published in 1947. As it happened, whenever we came across each other at UCL our chats tended to be chiefly about music. Anyway, this article has become very well known to people int·rested especially in the British pronuncation of the English Language ever since Gimson, in his textbook An Introduction to the Pronunciation of English, made use of its data in the first edition (and his subsequent ones) of his classic description of General British English. This last term GB(E), by the way, he felt in his day not yet ready to adopt, tho he did remark in his third edition (1980 p.303) that “ 'General British (GB)' ... may in time supersede the abbreviation RP”.

The Fry article in question was ‘The Frequency of Occurrence of Speech Sounds in Southern English’ (not using the unlovable term 'Received Pronunciation' you notice) and it was published in the Archives Néerlandaises de Phonétique Expérimentale Volume XX. The experiment it reported consisted of counting the numbers of occurrences of each of the 44 segmental phonemes of what he referred to as "typical Southern English pronunciation" (ie of course excluding basilectal speakers, again avoiding the term 'RP') in a small set of texts. As he reported, "the material used was the conversational matter contained in Daniel Jones's “Phonetic Readings in English”... anecdotes in fairly colloquial language". In fact practic·ly all of Jones's amazingly vapid pieces began with quite an amount of narrative spoken-prose introductory matter that in the shortest items cou·d exceed the dialogue in length.

Six speakers were assigned to counting the sounds after first replacing with their own personal phoneme choices any items they found not to match them. No names or other details were given of who the counters were (probably either staff or students), or how they proceeded, but the "total number of sounds counted was just over 17,000". That me·nt presumably that each of them was assigned about 3,000 segments which wd amount to each reaching his target within well under three pages of the book. A table of the results of the count set out the recorded frequency per cent of each 'sound'. Unsurprisingly the highest scores were for schwa /ə/ at "10.74%" and the ship vowel /ɪ/ at "8.33%". As one wd expect, diphthongs and /ʧ & ʤ/ were counted as single phonemes and / tr, dr & ts/ etc as phoneme sequences. The sequence /hw/ it was noted did not occur.

This was an extremely modest piece of research which must've required less than a day's work on the part of each of the six participants. Accordingly it seems unfortunate that Gimson gave their results without any information on how they were arrived at and thereby potentially creating an impression of pretentious if not spurious precision by presenting their results to two places of decimals. An adequately meaningful and more appropriately simple summary of its results wdve been to say that the largest numbers of occurrences counted were for schwa at over 10% and for / ɪ / at over 8%. The other most frequent vowel was /e/ with about 3%. Listing them in descending order of frequency, of the remaining vowels nine ie /aɪ, ᴧ, eɪ, iː, əʊ, ӕ, ɒ, ɔː& uː / scored less than 2% and eight /ʊ, ɑː, aʊ, ɜː, ɛə, ɪə, ɔɪ & ʊə/ less than 1%. Of the consonants the two most frequent were /n/ and /t/ with respectively over 7% and about 6%. Of the rest, less than 5% was scored by each of /d, s, l, ð, m, k, r, w, z & v/ and less than 2% by /b, f, p, h, ŋ, g, ʃ, j, ʤ, ʧ, θ, & ʒ/.

In 1987 Gerald Knowles published the absolutely excellent and original but unfortunately rather neglected book Patterns of Spoken English subtitled An Introduction to English Phonetics. Ever independent, he published his own account of 'Phoneme Frequency' as Appendix 2 at page 223 of that book. This owed nothing to the Fry count that Gimson had published — not even mentioning it in fact. In it he gave the results of a count of phonemes in the same sort of accent as Fry's data were based on but: "Ten different types of text, five written and five spoken, ranging from a seed catalogue to a passage from Pygmalion to recorded interviews, were transcribed. The first 1000 phonemes of each text were counted, making a total of 10 000 phonemes. The frequency of each phoneme was divided by 100, to give an average frequency per 100 phonemes of text".

As in the Fry results, the highest percentages were scored by schwa again ie over 10% (tho below 11%) and /ɪ/ again with over 8%. Again the other most frequent vowel was /e/, with a slightly less near figure to 3%; it was joined by /aɪ/ with a little over 2%. Listing them in no order of frequency, of the remaining vowels eight  ie / ᴧ, eɪ, iː, əʊ, ӕ, ɒ, ɔː, uː  / scored less than 2% and another eight /ʊ, ɑː, aʊ, ɜː, ɛə, ɪə, ɔɪ, ʊə  / less than 1%. Of the consonants the two most frequent were again /n/ and /t/ with respectively practic·ly 8% and over 6%. Of the rest, about 5% was scored by /d & s /,  less than 4% by / l, ð, m, k, r, w & z / and less than 2% by / v, b, f, p, h, ŋ, g, ʃ, j, ʤ, ʧ, θ, & ʒ/. So we see virtually the same, happily mutually confirming, results arrived at by totally different routes.

Blog 387

The 19th of March 2012

The Unique Word 'threepence'

In his blog of the 17th of March John Maidment, indulging in a bit of nostalgia about old-time money terms, gave half a dozen pronunciations for the word threepenny as “ ˈθrepəni or ˈθrʊpəni or ˈθrʌpəni, but the ə can be elided in all three versions ”.

The noun threepence (and in this discussion I propose mainly to treat it and threepenny as the 'same' word) is uniquely variable in the English language. No other word has a sole strest syllable which occurs in so many different accepted vowel values. It dates back to the sixteenth century but it seems we only have evidence of its spoken values from the eighteenth. The two leading authorities of that period were Thomas Sheridan whose A general dictionary of the English language of 1780 gave it  only as pronounced  /ˈθrɪpəns/ and John Walker whose Critical Pronouncing Dictionary of 1791 gave it only as /ˈθrepəns/. Unfortunately, Walker made no comment on this difference of opinion with Sheridan (father of the playwright Richard Brinsley Sheridan) even though he was far from loath to do so in many other cases. B. H. Smart in his Walker remodelled: A new critical pronouncing dictionary of the English language gave only /ɪ/ in 1871. When Murray came to the word  in the OED in 1912, no dou·t knowing of both of those opinions, he gave both /ˈθrɪpəns/ and /ˈθrɛpəns/. That is the order in which they remain shown to this day in the OED tho these symbols are not Murray's original ones. This is how they appear in the OED at their orthodox spellings but if we look at them entered as "thrup(p)ence, thrup(p)enny...Repr. colloq. or dial. pronunc. of THREEPENCE, THREEPENNY" we find "(ˈθrʌpəns, ˈθrʌpənɪ; ˈθrʊ-)" in an entry that no dou·t first appeared in 1986 in Robert Burchfield's fourth and final supplementary volume. By 1966 C. T. Onions in his Oxford Dictionary of Etymology had been showing /ɪ, e ᴧ/. The ODP (2001) gave only BR 'θrɛp(ə)ns & AM 'θrɛp(ə)ns, 'θriˌpɛns.

Back in 1917 Daniel Jones in his EPD had added a third vowel to Murray's making the vowels in his entry /ɪ, e ᴧ/. Then twenty years later he added a fourth and reversed the order of the first two thus /e ɪ ᴧ ʊ/ at the same time inserting a note  saying that the BBC had recommended /e/ [to its staff] in 1935. Those vowels in that order remain to this day in the (C)EPD of 2011. LPD in all its editions has given the same items /e ᴧ ɪ ʊ/ but re-ordering them to convey the fact that /ᴧ/ is more usual than /ɪ/. I can certainly confirm Wells's decision from my continuous observations since the 1960s. The /ɪ/ version has been  a distinctly unusual form for more than a generation. I can't recall hearing it from any of the hundreds of radio and tv newsreaders and presenters whose usages I've noted with the sole exception of the long-retired Radio 3 announcer Patricia Hughes who was noted for the elegance of her (old-fashioned) speech. A public figure who had /ɪ/ was Harold MacMillan (born 1894) also noted for his patrician speech.

My collection of data in preparation for my Concise Pronouncing Dictionary (OUP 1972) gave an int·resting not to say startling result when it came to this word. My records showed a very positive preponderance of occurrences in the word not for any of the above vowels but for a fifth vowel generally referred to as never occurring at all in accented GB syllables ie for schwa. In consequence my entry was `θrəpəns `θre- `θrᴧ- `θrʊ-. I've seen no other dictionary of GB even including schwa as a variant — at least I cou·d say that until I just now lookt at a webpage headed Cambridge Dictionaries Online in which I find at threepenny the sole transcription “/ˈθrə”. My first thau·t was that this was a misprint for /e/ but UK and US pronunciations were both provided and the UK had a schwa and the US was “ `θriːpeni ” — despite the fact that the single transcription was presumably me·nt to indicate that the form was the same in both varieties! I remain apprehensive because at the same page when I searched for threepence I found again a sole transcription this time /ˈθrʌp.ənt s/ and two pronunciations again — now “/`θrɛpns/” and “/`θriːpens/ ”!

The variant /`θrʊpəns/ is a curiosity in that it really isnt a "General" British usage but very much a London area pronunciation in my experience. However, I included it in my CPD because it's evidently one of those regional items generally not perceived as being a regionalism, as witness its inclusion by various lexicographers. The usual form in Yorkshire, and it seems most of northern England, is with /ɛ/. The usual form among Scottish QGB (Quasi General British) speakers is with /ʌ/ as it is for probably much of the West Country and practic·ly ev·ryone in South Wales. Of course the words threepence and threepenny took a huge drop in currency after 1971 when our coinage went decimal but at least the latter of them survives in figurative use in the title of the Brecht and Weil musical Die Dreigroschenoper which is regularly translated as The Threepenny Opera.

Blog 386

The 24th of February 2012

Understanding Phonetics

An excellent new introductory-level book has recently appeared with the title Understanding Phonetics. It's written in a lively but completely serious style. It comes from the experienced pen of the well-known British phonetician Dr Patricia Ashby. Its 230 pages have lots of good illustrations and diagrams. For example I considered particularly effective for its audience a simple combination (at its Fig. 3.5) of a chart of the consonants of the International Phonetic Association's alphabet placed under a cross-sectional drawing of the mouth from lips to pharynx with arrows between the places of articulation and the corresponding columns showing manners of articulation.

Although naturally examples are freely quoted from the language the book is written in, it's certainly a treatment of 'general' phonetics with an impressive range of exemplifications from well over a hundred different languages among which are Amharic, Basque, Burmese, Cherokee, Estonian, Faroese, Georgian, Gujarati, Hawaiian, Japanese, Kabardian, Kurdish, Lithuanian, Navaho, Quechua, Somali, Tamil, Tibetan, Tigre, Turkish, Wichita and Zulu. Among references to English one noticed with approval that in referring to the notorious pseudo variety 'Estuary' she dissociated herself from the expression by putting it in quotes. Praps she sh·d've done the same for the term 'g-dropping' at page 6.

Anyway, when she comes to mention the most described British accent she refers to it (at page 4) as "Modern Received Pronunciation (MRP) or Standard Southern British English (SSBE) (Cruttenden 2008)". The latter of these expressions seems not to be used subsequently in the text. Their bold print seems to suggest that they are important terms that the student shd take careful note of and their accompanying abbreviations seem to suggest that they might be expected to offen reappear. It's true that one of the contributing editors of the IPA Handbook uses the latter label, unabbreviated, several times but the former one seems much less widely employed. Its abbreviation 'MRP' I cou·dnt even find in the all-knowing Google. What's more, that these two labels are followed immediately in the text by "(Cruttenden 2008)" (with no page reference, by the way) is praps unfortunate in that some readers may be inclined to interpret it as a kind of authorisation or espousing of the terms which may give a wrong impression of what Cruttenden has written. Neither of these terms or their abbreviations are referred to in that book which incident·ly is a problem to indexers. It's been so brilliantly and extensively transformed by Cruttenden that it's not surprising that people adopt that style but it might be more graceful to use "Cruttenden, A. & Gimson, A. C. Gimson's Pronunciation of English".

I can't help being pleased that, tho she includes the usual full-page setting out of the current official IPA alphabet chart, when she comes to label diagrams, she reverts to the older traditional terms 'half-close' and 'half-open' which I've always felt the IPA had no need to drop. As she sez "Both sets of terms are in current use." There are one or two very minor things I shdve preferred diff·rent in a couple of the diagrams. Certainly the positioning of the lowered centralised [ɛ] at Figure 6.12 needs shifting down a bit. It's always struck me as slightly odd that anyone, choosing to symbolise facts about vowels by means of a diagram, should represent half of them by rounded bullets (indicators too large to be described as 'dots') to which they attach the word "Unrounded" as happens at page 99. The obvious thing to do seems to make the unrounded ones an unrounded shape.

Her terminology is discretely traditional in general as befits an introductory book but she does apologise for being mildly innovative in referring to 'BOR' labelling of vowels meaning 'backness-openness-rounding" in parallel with the better known VPM ie voice-place-manner labelling of consonants. Also slightly novel-looking but perfectly welcome is her pCV and sCV before Cardinal vowel numbers as a naming procedure avoiding giving the non-primary cardinal vowels the less memorable numbers 9 to 16. This practice she explains in a note  at the bottom of page 86. Rather novel terminology occurs as well in her wording at page 112 "...Wells uses schwi (the so-called 'happy-vowel') and schwu (weak-u), replacing [ɪ] and [ʊ] .. giving [ˈhӕpiə] and [ˈdʒӕgjuə]". By the way, she refers to the voicelessness subscript circle by a happily mnemonic label as a 'tiny zero' (p. 29) ie betokening zero voicing.

This valuable contribution to the literature on introducing phonetics deserves and will no dou·t receive reviews that will do it justice. In the meantime these few trivial remarks have been offer·d as a small contribution towards spre·ding the good news of its arrival.

Blog 385

The 22nd of January 2012

R Dropping (i)

An American colleague recently sed she wonder·d what she cd say to a teacher who askt her about the pronunciation of February. I guess the teacher's main worry was that most Americans say the word with only one /r/ as /`febjueri/. So do very many of us Brits tho we dont necessarily have the /-eri/ ending with its strong vowel. It's one of those words that has a bigger than usual difference between people's lexical version of it (as in quoting it) and how they gen·ly say it in completely fluent speech. That's what I call having one or more weakforms tho my colleagues mostly dont seem to realise the sensibleness of applying that term more widely than they usually do. It was int·resting to see how some of her fellow Americans — even ones who're in the business of teaching spoken English — described their own usages. Sev·ral sed something like "My elementary school teachers were big sticklers on pronouncing the name of the month with both R's, and I remember being drilled ... in early childhood, so I say [fɛbruɛri]".

If medieval pedantry hadnt messed about trying to restore its resemblance to its Latin etymon februarius we'd now be saying it something like feverer. We took it over from the French who, resisting the pedants better than we did, now have février. Anyway, a typical General British unselfconscious isolate version of it now is trisyllabic prob·bly most offen /`febr̩i/ but we easily compress it to bisyllabic /`febri/ within fluent utterance, praps really, oddly enuff, the biggest diff·rence being loss of rounding of the /r/. Most GB speakers have tendency to round (the lips making) /r/s, the more so the longer they make them. GA speakers usually round their /ɝ/s as in fur. But I stray!

Aside from pedantic r insertions, r dropping is a very common development observable thruout the recorded hist·ry of English. If an /r/ hadnt been dropped in medieval times from the word spelt in Old English as sprӕc, today we'd've been talking about 'spreech' not 'speech' (the German sprechen is more conservative). The 24 English consonant phonemes ie our 19 plosives, (af)fricatives and nasals etc include the approximants /l, r, j, w & h/ all five of which are weaker articulations than the rest. Consequently they tend to get lost more offen than the others: their disappearances arnt so noticeable. I quite offen find it hard, playing back perfectly high-quality recordings of speakers, to tell whether they have uttered some /r/ one might be expecting.

Anyway, in British broadcasting hundreds of people daily can be he·rd to utter the very common word program with either one or occasion·ly even both of the orthographic r's not represented by a sound. The first one is the one most offen omitted. Where one only is dropt people offen put the dropping down to what's called 'dissimilation'. This 'anticipative' dissimilation is very common in all forms of English. There's no universally accepted explanation of why dissimilations happen but my hunch is that speakers produce them in subconscious avoidance of saying the same phoneme twice in one word because such repetitions tend to suggest stumbling in their speech — which they dont want to be perceived as doing. Despite the pronouncing dictionaries, the most usual form of the hugely common word prescription among GB speakers is /pə`skrɪpʃn/. This is an example of the commonest type ie those that have labial consonants before the elisional /r/. Other words frequently he·rd with such elisions include infrastructure, preliminary, prerogative, priority, progressive, pronunciation, proposal, proprietor, protractor and protrude.
(To be continued.)


Blog 384

The 20th of January 2012


John Maidment's Blog today sez "What do you call a word like controversy, which has more than one pronunciation? I don’t think there is a widely accepted term, so I would like to suggest polymorph." Now, as he observes, this is an existing word of which he remarks "As far as I know, it has not been used in linguistics". However, the fly in the ointment is that "morph" and compounds embodying it are very widely used in linguistics. As he suggests, it'd be very nice to have a term in phonetics for multi-form words but unfortunately "polymorph" strongly suggests morphemics. Accordingly why dont we consider "polyphon". This rather better suggests phonetics and has apparently only been used before in the nineteenth century for a patent German musical box and, if I remember rightly, a twentieth-century East European musical disc company. The corresponding adjective cd be "polyphonemic", as Petr Rösel has suggested, or for those who wished to avoid the musical associations of "polyphonic" it cd be "polyphonematic".

John ended his blog with a mention of the word that's probably the most polyphonematic item one can think of, namely transitionally, and invited me to say how many allopolyphons one can attribute to it. To attempt to count them I turned, some years ago, to John Wells's Longman Pronunciation Dictionary which is surely the most exhaustive repository of such things in its records of General British pronunciations. I'm quite sorry that it wou·dnt be within the rules of the game to use the compound word quasi-transitionally because that cd be viewed as embracing an amazing 3,140 allopolyphons. Anyway, the figure I find for the unchallengeably single word transitionally is 160.

I personally regard the very last phoneme in this word to be either /i/ or /ɪ/ although it's possible that some analysts take there to be three alternant phonemes involved which they identify as /iː/, /i/ and /ɪ/. For them, the total cd be sed to be 3,200. I set out below the first sixty items and indications of how I arrive at the remaining additional variants. Syllabicity sub-strokes are only supplied where ambiguity is to be avoided. There has not been the intention to indicate any esh (ʃ) as syllabic anywhere, even if some apparent intractabilities of my "Kompozer" software may have at points suggested otherwise. Please regard any syllabicity-type strokes either superimposed on or appearing below any occurrences of the esh symbol as the garbage that they truly are. Apologies are offered also for various instances similarly beyond my control of imperfect vertical alignment of any consonant with its subscript syllabicity mark.

1.    trӕnˈzɪʃənəli
2.    trӕnˈzɪʃ̩̩n̩l̩i
3.    trӕnˈzɪʃ̩̩n̩əli
4.    trӕnˈzɪʃ̩̩̩nəli
5.    trӕnˈzɪʃnli
6.    trɑːnˈzɪʃənəli
7.    trɑːnˈzɪʃn̩l̩i
8.    trɑːnˈzɪʃ̩̩n̩əli
9.    trɑːnˈzɪʃnəli
10.    trɑːnˈzɪʃnli
11.    trənˈzɪʃənəli
12.    trənˈzɪʃ̩̩n̩l̩i
13.    trənˈzɪʃ̩̩n̩əli
14.    trənˈzɪʃnəli
15.    trənˈzɪʃnli
16.    trnˈzɪʃənəli
17.    trnˈzɪʃ̩̩n̩l̩i
18.    trnˈzɪʃ̩̩n̩əli
19.    trnˈzɪʃnəli
20.    trnˈzɪʃnli
21.    trӕnˈsɪʃənəli
22.    trӕnˈsɪʃ̩̩n̩l̩i
23.    trӕnˈsɪʃ̩̩n̩əli
24.    trӕnˈsɪʃnəli
25.    trӕnˈsɪʃnli
26.    trɑːnˈsɪʃənəli
27.    trɑːnˈzɪʃ̩̩n̩l̩i
28.    trɑːnˈsɪʃ̩̩n̩əli
29.    trɑːnˈsɪʃnəli
30.    trɑːnˈsɪʃnli
31.    trənˈsɪʃənəli
32.    trənˈzɪʃ̩̩n̩l̩i
33.    trənˈsɪʃ̩̩n̩əli
34.    trənˈsɪʃnəli
35.    trənˈsɪʃnli
36.    trnˈsɪʃənəli
37.    trnˈzɪʃ̩̩n̩l̩i
38.    trnˈsɪʃ̩̩n̩əli
39.    trnˈsɪʃnəli
40.    trnˈsɪʃnli
41.    trӕnˈsɪʒənəli
42.    ̩̩trӕnˈsɪʒn̩l̩i
43.    trӕnˈsɪʒn̩əli
44.    trӕnˈsɪʒnəli
45.    trӕnˈsɪʒnli
46.    trɑːnˈsɪʒənəli
47.    trɑːnˈsɪʒn̩l̩i
48.    trɑːnˈsɪʒn̩əli
49.    trɑːnˈsɪʒnəli
50.    trɑːnˈsɪʒnli
51.    trənˈsɪʒənəli
52.    trənˈsɪʒn̩l̩i
53.    trənˈsɪʒn̩əli
54.    trənˈsɪʒnəli
55.    trənˈsɪʒnli
56.    trnˈsɪʒənəli
57.    trnˈsɪʒn̩l̩i
58.    trnˈsɪʒn̩əli
59.    trnˈsɪʒnəli
60.    trnˈsɪʒnli
+20 -ndz-
+20 -nts-
+ 60 /-ɪ/ 
Total 160


Blog 383

The 10th of January 2012

Rants on Odd Pronunciations

A related sport to 'spelling bees' (the very concept of which must mystify many people whose languages have orthographies that bear a reasonably close relationship to how they speak) is the popular game among the English-speaking peoples of deploring the ways that other people say various words. I have to confess to being a regular player myself and this note is prompted by reading rants on two successive days by two champion players of the sport Graham Pointon and John Maidment.

Graham's gambit yesterday was two-pronged beginning with bemoaning hearing "/iˈvɪskəreɪt/" from a distinguished historian, a Cambridge don. He complained that "maybe he now believes that all Latin words used in English should be pronounced as if they were still Latin". I'm afraid it was linguistic scholars like ourselves provoked that confusion in the first place by not being quiet about out how very different our Victorian habits of saying Latin words were from those of dear Julius Caesar. Tony Blair has been known to have the same problem with visceral. Graham's other victim was a Singapore-born BBC World News presenter with degrees from across the Atlantic. He noted her saying /kəˈlɛərə/ for cholera thereby putting the only stress on the one vowel that many people leave out altogether. People are prob·bly offen lulled into making that sort of mistake because there are so many words with two accepted possibilities of that kind. After all Webster Online may only give \ˈkä-lə-rə\ for cholera but for choleric it has both \ˈkä-lə-rik\ and \kə-ˈler-ik\ and the latter one is more in line with the general  'rule' of stressing the syllable immediately preceding the suffix -ic.

The other ranter, John, has today a complaint that Jeremy Paxman yesterday "perpetrated ˌdɒməˈniːkəʊ for the Italian name Domenico". I have to say that I found that pretty venial compared with various other disdainful clumsinesses I've noticed from him. Graham's remark "Shouldn’t someone in the production team be listening and persuading the broadcasters to use pronunciations which do not cause the viewer/listener to concentrate on the form rather than the content of what they are saying?" is very relevant. I've at times been scandalised by Paxman's apparent contemptuous indifference to this part of his job — not that I can stand to listen to him very offen. He in truth has more challenging things to articulate than most other broadcasters but it's shocking that he, his Producer and his Director on University Challenge shd all be so indifferent to, or oblivious of, his poor competence in these matters. He can be forgiven minor oddities like prĕfix for prēfix or /`θesərəs/ for thesaurus, but not various items that have been too gruesome to record.

We must all try to be as forgiving as possible remembering that our best frends may slip on some such bananas. This is a topic that has figured in various of these blogs eg 049. Even our most admired heroes and heroines have come croppers over the odd word. Henry Sweet confessed to 'umberella'. I remember Gimson telling me sorrowfully that Daniel Jones used to say contrăst: DJ even talked of /`swӕstɪkəz/. Churchill always referred to /`nɑːzɪz/. In recent weeks I've noted my two most admired news presenters one saying /trӕʤə`diːən/ and the other /`lɒnʤəri/ (I know the last gets LPD etc approval). Dear old Humphrey Lyttelton was capable of /ɔː`rai/ for awry. I usually don't attribute a pronunciation to anyone unless I've he·rd them repeat it clearly. This very morning I was pretty sure I he·rd dear Sarah Walker on Radio 3 say "/`kɒnsətstᴧk/" for Konzertstück (which I'd rather not have to say over the air: why do we so often torture ourselves to use the difficult-for-us-to-say German word and not simply translate it as Concert Piece?) but I take that to've been an unlucky slip of the tongue because her performance is so regularly very satisfactory (unless it was a 'slip' of my ear!). [PS OH DEAR! It was indeed a shaming slip of my ear while listening too casually. I've been able, courtesy of a frend, to hear her from a recording say with perfect clarity on two occasions not Konzertstück but Konzertsatz with a modest anglicisation of stressing that's perfectly acceptable.] The BBC music presenters are the best at forren words but even they seem to've collectively decided to mispronounce Sospiri, Elgar's Italian-titled piece, with a front stress that's un-Italian. Jonathan Dimbleby regularly sez `scintilla. The brilliant Jim al Khalili sez an`timony. I've known David Attenborough say para`sitise! And I personally freely confess to regularly saying (or at least thinking) `apparatus, o'nomato`poeic and /ɔːrə`tɔːriəʊ/. So there but for the grace...

Blog 382

The 8th of January 2012

Spelling Pronunciations (ii)

When our Blog 379 quoted Daniel Jones using the term 'spelling-pronunciation' we didnt mention then that his use of the word in 1913 was barely a dozen years after its first recorded appearance in the English language. It always strikes me as curious to think that this very well known English compound-noun term /`spelɪŋprənᴧnsieɪʃn/ designating such a vastly common linguistic phenomenon over  huge stretches of time had emerged into use no farther back in the OED's records than 1901. It's shown as the apparent coinage  of a scholar, namely "E[mil]. Koeppel" author of "Spelling-Pronunciations. Quellen und Forschungen. Strassburg. 1901". This was a monograph in German, subtitled Bemerkungen über den Einfluss des Schriftbildes auf den Laut im Englischen of 71 pages issued by the German publisher Trübner as part of a series reporting research on the linguistic and cultural history of the Germanic peoples. I'm most grateful to Prof Dr Petr Rösel for valuable information on this publication.

The linguistic process it refers to had inevitably existed for aeons before it got that name. It obvi·sly first occurred relativ·ly shortly after whenever and wherever alphabetic writing became invented. It's completely beyond dispute that all languages are constantly subject to change and a consequence of this evolution is that numbers of the "pronunciation spellings" that are the necessary form that alphabetic writing first produces become out of date. Developments such as phoneme elision lead to forms which no longer reflect faithfully current pronunciations of words that have undergone such phonetic changes. Another curiosity is that quite a lot of people seem to be under the impression that the practice is of itself reprehensible. What a staggering idea! English has countless thousands of spelling-influenced pronunciations including the great numbers that people use for archaicly-spelt forenames, surnames and untold numbers of placenames on our maps of the British Isles and other English-speaking countries. Of course some do become unfashionable and fashion changes owe precious little to logic. But the idea that a spelling-influenced pronunciation of any kind is necessarily to be deplored is preposterous.

This brings us back to the word often which was the commonest of the words we noted Jones condemning, in this case as spoken with sounding of the t of its spelling. Murray in NED [=OED1] in 1902 indicated as (then current) pronunciations of often forms of 'medial or doubtful length' as in 'soft' ie as the now obsolescent /`ɔːfn/ or as /`ɒfn/ [modernised symbolisations are given here and elsewhere for the reader's convenience]. Also he remarked in a sep·rate note "The pronunciation `ɒftən, which is not recognized in the dictionaries, is now frequent in the south of England, and is often used in singing." The curiously antiquarian comment to find in a dictionary of current pronunciations (CEPD18 p.348) we quoted previously can be seen to've had its inspiration in an OED note about the history of often as follows:

"Several orthoepists of the 16th and 17th centuries, including Hart, Bullokar, Robinson, Gil, and Hodges, give a pronunciation with medial -t- . Others, including Coles, Young, Strong, and Brown, record a pronunciation without -t- , which, despite its use in the 16th cent. by Elizabeth I, seems to have been avoided by careful speakers in the 17th cent. (see E. J. Dobson Eng. Pronunc. 1500–1700 (ed. 2, 1968) II. §405). Loss of t after f  occurs in other cases; compare soften... The pronunciation with -t-   has frequently been considered to be hypercorrection in recent times: see for example H. W. Fowler Mod. Eng. Usage (1926), s.v."

This last reference is to the Fowler entry we previously quoted. It's a little surprising that OED made no reference to Walker's 1791 (and much reprinted) Critical Pronouncing Dictionary which had no /t/ version for often in its whole existence (which even extended into the beginning of the twentieth century).

PS  Oddly enuff I've quite  often  been askt by students whether I recommended them to say /`ɒfn/ or /`ɒftən/. My usual reply has been /`aɪðə/ meaning that /aɪ ` ˏɒfn | seɪ `ɒftən | bət ðen aɪ `ˏɒftən | seɪ `ɒfn/.

Blog 381

The 6th of January 2012

Big Deal

I was following recently an email discussion centred around this well known expression whose use has spre·d widely from its original home in the US. A contributor commented "one of the attitudes conveyed via intonation is sarcasm" immediately following that by saying very reasonably "I think it helps to reflect that it takes years for children acquiring English as their first language to catch on to this". She then vividly illustrated the topic by quoting a well-known exchange in which a US "1st-grader" tells his mother something that Joey, his 3rd-grader schoolfellow, sed:
Son: Joey likes my new backpack.
Mother: That's nice, dear. How d'you know he likes it?
Son: Coz I told him it was new and he said "Big deal".
These juxtaposed comments prompted a response from a non-native-English-speaking EFL teacher who evidently, praps not very surprisingly in the circumstances, took the anecdote to illustrate what a "crucial role intonation can play in expressing a speaker's attitude". When I contacted him, it turned out that this EFL teacher had presumed that it was being suggested that the 'big deal' expression regularly if not inevitably entailed a stereotypical non-verbal form which of itself conveyed sarcasm. I was quite surprised that no-one seemed to offer to disabuse the EFL teacher of this idea. Actually, Joey's comment, tho likely to've been sarcastic, as it was merely reported not he·rd, cou·dnt be known to be one or the other. It's no dou·t possible for sarcasm to be conveyed by not explicitly verbally sarcastic language for the situation involved but by prosodic and/or paralinguistic features. I never felt inclined to devote any time in my long career in EFL teaching to such matters even with my most advanced students. It didnt seem to be a sufficiently important topic for my students to need to devote much attention to.

Mere lack of enthusiasm isnt necessarily to be taken as sarcasm, by the way.  I believe that really marked sarcasm is actually usually exprest partly by use of any of a variety of bodily gestures such as shrugging shoulders, rolling the eyes etc combined with some or other paralinguistic articulations including sighing, drawling syllables, adopting a strained or other special voice quality etc. "Big deal!" is, by the way, probably much more often, at least in the UK, used in the ironic sense than the literal one.
Looking in various, especially EFL-oriented, reference works one finds 'big deal' usually included and well illustrated with plenty of effective examples but it's one of the newest in the field that comes off best in that the MacMillan Dictionary, along with "", actually gives it a pronunciation. They both give it the perfectly acceptable neutral lexical-type treatment as /ˈbɪg ˎdiːl/ ie with a fairly high level pitch on the first word and a low falling pitch on the second. If speakers want to make it neutral in other respects but positively lively they put a high fall on the second ('climax') tone. Keeping the same neutral voice quality but making it livelier still, possibly surprised or excited sounding, many a GB speaker might 'supercharge' the fall making it what I call a Climb-Fall tone /ˈbɪg  ́`diːl/. The same effect wd be produced by many non-GB, eg GA, speakers by making the fall start higher. These last two types are likeliest to be used if irony is suggested. Pitches below the mid range on both words /ˌbɪg (ˏ)ˎdiːl/ wd tend to sound rather notably unexcited, but that wdnt exclude their accompanying something conveying sarcasm along with other features.