Archive 44 of JWL Blog

Home

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

20/03/2013Dogs must be carried.#440
18/03/2013A Brief Conversation.#439
11/03/2013Accent Shift.#438
21/02/2013The GB air/square phoneme.#437
17/02/2013More on Weakforms (ix) am, an, and, any etc#436
11/02/2013Minimally Different Words.#435
03/02/2013Accenting of a Compound Word.#434
27/01/2013John L. M. Trim.#433
09/01/2013Comparison of Transcriptions.#432
24/12/2012More on Weakforms (viii).#431

Blog 440

The 20th of March 2013

Dogs must be carried.

In his blog "carrying dogs" of Friday, 15 March 2013 John Wells' referred to a Language Log "interesting posting by Mark Liberman" containing the remark "I've never figured out a really convincing explanation for why stressing "dogs" seems to encourage the interpretation "everyone must carry a dog", while stressing "carried" encourages the interpretation "if you have a dog, you must carry it". John then supplied three prosodic variant versions of this familiar cautionary notice “ `ˏDogs | must be `carried ”, “ ˏDogs | must be `carried ” and “ `Dogs must be carried ”. It's of course obvious that the absence of accentuation of 'carried' in the third version necessarily suggests that the topic of 'carrying' is to be taken for granted as already establisht in the situation. John added "If you say `Dogs must be carried you encourage the interpretationyou can't use this facility unless you are carrying a dog”. But why?" I shd prefer a wording like 'suggest the impression' to 'encourage the interpretation' because no normally intelligent person is likely to misinterpret these words.

Anyway, I think the short ans·er is 'Coz we have a sense of humour'. I think we prob·bly all tend to hear such brief declarations in our mind's ears especially when we frequently meet them in stores and stations etc. And most of us, when repeatedly seeing such things, find that alternative prosodies pop into one's head without bidding. When such a prosody is that third Wells one it strikes us as rather amusingly absurd. That's all. This seemed so simple and obvious that I wrote to John suggesting that I shd've thaut him more likely to have saved using such an item till April the first. He assured me that he was making a 'serious point' and recommended me to look at his 'latest reply' which began:
 "I think you're ALL missing the point". (He received twenty-seven comments most of which were quite long.) It turned out that he had in mind "a text-to-speech system designed to read notices aloud ... [with] an intonation component that would correctly place an intonation nucleus ... on “carried” in “Dogs must be carried” but on “Safety” in “Safety boots must be worn”. He indicated that he longed for "an ALGORITHM" for that purpose.

I havnt much pers·nal familiarity with such things but my Apple-Mac computer has a very welcome facility for having a voice read out for me any text I select. I tried him (an American male) and I thaut he did well. For 'Dogs must be carried' he sed very clearly [ˈdɔgz | mᴧs bi ˎkӕrid] and for 'Safety boots must be worn' he gave me [`ˈseɪfti ˎbuts | mᴧs bi ˎwɔrn]. These both gave adequate emphasis to their topic words and also appropriately indicated the climactic elements of the comments by tonic accents. One of the most difficult things I imagine for designers of text-to-speech software is to build in recognition of when de-accentuation of re-occurring items is appropriate but it's something that seems to be being managed to a very reasonable extent. I habitually set my computer to murmur the time to me at ev·ry quarter of the hour. The obliging voice (American female this time), tells me from one to four and from six to eleven that eg [It's ˈone ˈforty ˎfive] but at 5.45 she always accents it [It's `ˈfive | `forty five]. Also I let my American guy read the following story: I ˈwanted to go up to the next `ˏfloor of a de`partment ˏstore, | and  I saw an `escalator with a `ˏsign | ˎsaying ˈDogs | must be ˎcarried on this `escalator | but I didn't `have a ˎdog,| so I had to ˈuse the `stairs. These intonations also seemed very effective.

The standard in such things so far reached may be much less than John wishes to get from his desired algorithm but I've found my admittedly extremely slight sampling quite impressive. People who're lissening purely for the meaning of a text that's re·d for them arnt really likely to be all that much thrown by unidiomatic stressings. We can interpret meaning effectively enuff even with quite a few incompletely appropriate or ambiguous accentuations. Many songs have musical settings that go counter to the accentuations we shd expect if their words were ordinary speech and no-one ever seems to worry about the fact. Ambiguous accentuations are very common on the stage and even the screen. Especially in drama with archaic language, I sometimes find myself wondering whether an actor has really understood the text I'm hearing. Some bits of Shakespeare are now quite incomprehensible even to specialist scholars but still get declaimed. I have a feeling that offen much of what distinguishes a really good stage or screen performance from a mediocre one can be the effectiveness of the prosodies employed.

Regarding the tonetic notation I've used above, I shd think only [] shd need any explanation. I use it to indicate a high fall restricted in its movement to approximately the top third of the speaker's voice range.





Blog 439

The 18th of March 2013

A Brief Conversation.

Most of my readers will know that I use these blog posts from time to time to offer teachers of non-native-speaking students of British English at an advanced level pieces of dialog for phonetic analysis and/or spoken practice. I've chosen this time to use one of the shorter items from my book People Speaking which was publisht in 1977 by the Oxford University Press. The sound file corresponding to it can be accessed in the main part of this website at section 4.1 where it is Passage 23 entitled English Urban Winter. I recommend copying it into Audacity or the like.

It's a conversation between two acquaintances who happen to meet at a bus stop. Because I think to study both prosodic and phonemic features at once is less easy than dealing with them sep·rately, I offer for each line firstly a prosodic notation accompanying ordinary spelling and secondly a phonemic one. Both are what I classify as "coarse" types of transcription ie almost completely undetailed. Anyone familiar with the work of transcribing relatively spontaneous speech will know that on many occasions various items might've equally well been transcribed in an alternative way. To keep things simple no alternatives are supplied on this occasion. The recordings used for my book were a mixture of scripted and spontaneous speech. The excellent professional actors who performed most of it had been requested to make the scripted items such as the present dialog sound as natural as possible. I judged them very successful at doing so. I have on occasion challenged lisseners to see if they can tell which items were spontaneous from those that were scripted. It wasnt found easy.

1. ˈHulˏlo! ˈWhat d'you think of the `weather?
This first word recalls the subject of our last blog. Not here but in other contexts 'hullo' exhibits English "alternate stress (or 'accent') preference". When used greeting a person by name, instead of being accented on its second syllable, as it normally is when used in isolation, it moves the accent to its first syllable as in `Hullo, ˏTom. Here, anyway, it's used as a quite animated (witness the initial high pitch) cordial greeting that can also to some extent suggest by its rising tone that the speaker is prepared to open a conversation.
hᴧləʊ wɒt (ə) ju θɪŋk ə(v) ðə weðə
The very first syllable, that I've shown as /ᴧ/, is a good example of how really rather uncumftably similar to each other the General British phonemes /ӕ/ (phoneticly [a] these days) and /ᴧ/ have become. It wou·d've been difficult to decide which vowel she'd used if it'd been spontaneous but she was reading a text that sed 'hullo'. The word 'do' here has no /d/ and only at best a ghost of a vowel. Much the same goes for the weak, if present, /v/ of the 'of'. It's nothing unusual for people to use /ᴧ, ə/ or /e/ in the first syllable of 'hullo' but I'd recommend students to avoid the recessive minority form with /ӕ/. Of course with /e/ the corresponding spelling is Hello but, curiously enuff, the semantic and stylistic uses of all three are the same — making them in essence three diff·rent pronunciations of the 'same' word.

2. ˈI ˈdon't `mind it.
Here we have one of the simplest types of head — two level tones with the usual small step down in pitch to the second accent. The high fall is just about the commonest of all climax tones.
aɪ dəʊn maɪnd ɪt
Here the /t/ of don't is completely elided. This is a perfec·ly common tho not invariable practice. Some speakers on some occasions wd use the /t/ of the orthographic form. Others sometimes use more extreme reductions: the /n/ might be converted to an /m/ or even dropt altogether producing what some wd call 'careless' speech.

 3.ˈSnow | doesn't `bother ˏme.
The pitch of 'snow' is rather low but not very low so it was a coin-toss decision to mark it as high and, weakly tho it's uttered, it is prominent so must be classified as accented. Some might like to describe it as in a lower 'key' from what follows. The vertical bar that we use to signify a dynamic change gives a hint of that.
snəʊ dᴧzn bɒðə miː
It's completely normal to elide the /t/ of 'doesnt' before a following plosive in fluent speech. The only position in which the final /t/s of words ending with -n't are not normally dropt in General British speech (tho they may well only survive in the form of glottal stops) is when an immediate break in rhythm follows.

4.`And I've `known it | `plenty colder than `ˏthis.
Since the word 'and' initiates a quite wide falling tone it must be classified as accented. And, since its vowel is clearly a schwa, it's another example of how misguided those who insist that GB schwa is never accented are. The vertical bar here doesnt indicate any more discontinuity of rhythm than to register a renewal of impetus produced by step up from the lower pitch of the previous second accent of the falling head. Falling tones are the most natural precursors to a fall-rise climax tone such as we have on the final word.
ən aɪv nəʊn ɪt plenti kəʊldə ðn̩ ðɪs
The word 'and' is the most frequently used word in the English language. It's relatively rarely found to occur with the /d/ of its spelling. This adverbial use of 'plenty' is classified as colloquial. Only the least elegant types of British speech drop the /t/ of 'plenty' by contrast with General American usage where it's not stigmatised. The /n/ of 'than' isnt very long but it's long enuff to make it unsuitable to suggest th·t that 'than' contains a schwa.

5. ˎYes, | but it's ˈno ˈfun | if you ˈlive | `well out of `ˏtown.
The fall-rise climax here is preceded by what I'm inclined to call a head of mixt tones rather than the more usual sequence of falling tones: the change at 'well' from level tones to a falling head-tone cou·d be sed to show a surge of vigor.
jes bət ɪts nəʊ fᴧn ɪf ju lɪv wel aʊt əv taʊn
The final /t/ of 'but' here is so swiftly and softly articulated that it's hardly classifiable as a plosive or necessarily to be described as a /t/ rather than a /d/.

6. I had to wait ˈhalf an ˈhour | for a `bus this ˏmorning.
Our bar register·d a dynamic discontinuity becoz there was a very slight hiatus as the words 'for a' dropt below the upper level of their preceding word 'hour'. The climax tone on 'bus' took the pitch range down to the lowest level, the one that the following adverbial adjunct 'this morning' began on.
ə hӕt tə weɪt hɑːf ən ɑ: frə bᴧs ðəs mɔːnɪŋ
We cou·dnt transcribe the pronoun 'I' as /aɪ/ here. It was so rapidly articulated and vague in quality that a schwa was the only thing we can call it. There was a classic pre (ie anticipative) assimilation as the final /d/ of 'had' was fully converted into a /t/ before the /t/ of 'to'.
The diphthong 'aʊə' is very offen replaced in ordinary speech with a monophthong either /ɑə/ or as here /ɑː/. It's perfec·ly commonplace for 'for' to lose its vowel in combining with 'a'. It's not usually mentioned in textbooks that the word 'this' may take a schwa weakform in the combination this morning, but it's a well known fact.

7. And there's `usually| `one every | `five or ˈten `minutes.
There's something a little unusual about the fully falling pitch on the word 'usually' here: it's notable that it's spoken very quickly — otherwise it might well sound quite odd. You'd normally expect to hear it on a fall-rise if, as here, its dynamicly ie rhythmicly separated from the following words. She keeps up a vigorous delivery for the next few words in using two falls. Then she momentarily sounds slightly more relaxed with a level tone on 'ten' but she ups the vigour agen finally with a notably high fall on 'minutes'.
ən ðəz juʒli wᴧn evri faɪv ɔː ten mɪnɪts
All the words in this sentence are pronounced in some of their most usual forms for the context including the very common schwa weakform she uses at 'there’s'. The pronunciations given for 'usually' in the pronouncing dictionaries tend to suggest that it's normal to pronounce this word as four syllables which it's cert·nly not: it's normally two, as here, or three ie /juːʒəli/. There's a little more on this at my Blog 132.

8. ˎHard ˏˌlines!
If you want to express genuine sympathy with someone's misfortunes it's safer to avoid speaking too vigorously. That's what he does here keeping to low pitches and narrow pitch movements. On the first word he has a low fall (a comparatively narrow type) and on the second an exceptionally narrow rise. I call the tone I've shown for it the Rise-Bass coz it's a rise but unlike the usual Rise tone (which goes up from the Bass ie low range of the voice into the middle range) this goes up very little, its movement being limited to that Bass range. It's a fairly common tone in rhetorical speech (once a favourite with Winston Churchill and various preachers) but hardly one you'd go out of your way to have your students practise using.
hɑːd laɪnz
Both of these words were spoken with their most usual phonemes.



Blog 438

The 11th of March 2013

Accent Shift.

Accent Shift or rhythmical accent re-distribution within words, also known (and more usually so) as 'Stress Shift' was the subject of a recent query by one of my most perceptive correspondents. He as·t, "as it's usually thir`teen in isolation but ˈthirteen `books when uttered in this phrase, what about entire? Do people say ˈentire `forest for example?" He added "There's no indication of stress shift in either LPD3 or EPD18".
It's exactly true that neither dictionary indicates completely explicitly that accent shift is not normal for words like entire but in fact both at more than one place do supply negative evidence ie contra-indication of its possibility. The EPD (Cambridge) English Pronouncing Dictionary does so by not providing an example of the word used in a stress-shifted context. The explanations of the design of the dictionary in its Introduction at p.xvi contain a twenty-five-line paragraph explaining the topic. This has some slightly cumbersome wording where it sez "when a word of several syllables has a stress near the end of the word, and is followed by another word with stress near its beginning, there is a tendency for the stress in the first word to move nearer the beginning if it contains a syllable there that is capable of receiving stress." Examples follow. It also has another, simpler and clearer form of explanation in its Glossary at page 574 that sez "The rhythm of English prefers patterns in which two stressed syllables do not come together. In order to avoid this, stress in some polysyllabic words may move to an earlier syllable when combined with another in a phrase eg..." Then follow good examples and the good news that "In this dictionary, words which change their stress in this way are shown with an example demonstrating the stress shift."

Wells's LPD (Longman Pronunciation Dictionary) tackles this problem by saying in its initial "Quick guide to the dictionary" at page xv that "Some words have different stress patterns according to whether they are being used alone or directly before a noun. (See the panel on 'Stress shift', p.784.) The symbol ◂ is used to show words which can behave in this way." At p.784 we find a large panel with about thirty lines of clear explanations illustrated by effective examples. The most important of the examples contrast the words 'funda`mental and 'Japa`nese showing the stress structure of their occurrences in isolation in comparison with their altered stress values in the combinations 'fundamental mi`stake and 'Japanese `language. (These tonetic stress marks of ours are very slightly different from the original versions in order to avoid any suggestions of low pitches and level pitches where they aren't appropriate.) The essential idea behind these stress patterns is that speakers instinctively choose to distance from each other the two strongest stresses in such units as these two-word phrases. It's one of the workings of the general basic tendency we have to alternate stresses as far as possible (with the corresponding weaker inclination to also alternate weak syllables as far as possible).

The OALD (Oxford Advanced Learner's Dictionary) edition of 2010 at page R46 explained stress shift like this:
"When two words are put together in a phrase, the main stress in the first word may shift to the place of the secondary stress to avoid a clash between two stressed syllables near to each other. For instance, ˌafterˈnoon has the main stress on noon, but in the phrase ˌafternoon ˈtea the stress on noon is missing. ˌWell ˈknown has the main stress on known, but in the phrase well-known ˈactor the stress on known is missing."

A more precise if rather less easily digestible definition is to be found in Cruttenden's seventh edition of 'Gimson' at "Secondary Accent 10.3.4" thus: "When words have more than one syllable before or after the main accent, a general rhythmical pattern is often apparent, there being a tendency to alternate more prominent and less prominent syllables. Syllables made prominent in this way will retain a full vowel; additionally, syllables before the primary accent will often receive a secondary accent involving pitch prominence."

However, the real problem my correspondent was concerned with wasnt the above as he showed by indicating that he was familiar with the stress alternations we find with thir`teen on its own and in the phrase 'thirteen `books. He was wondering in effect whether he cou·d treat the first syllable of entire as strong. For this we may look to the dictionaries for help. Here's how the three major pronunciation dictionaries represent the word entire.  
CEPD has "ɪn'taɪər, -'taɪ.ər, en- ." [Cf ˌenchi'lada, ˌener'getic; ˌenzy'mology, ˌembar'kation, em`pɔːriəm, en`demic]. 
LPD has "ɪn ˈtaɪ ̮ə  en-, ən- , § ˌen-, § ˌɪn-". [Cf (ˌ)ӕn 'tiːk]
The ODP (Oxford Dictionary of Pronunciation) has "ᵻn'tᴧɪə(r), ɛn'tᴧɪə(r)" The Merriam Webster Online notation "\in-ˈtī(-ə)r, ˈen-ˌ\" seems to cover the (lesser) possibility in GA (General American pronunciation) of a phrase strest the ˈentire `world. [Its 1961 parent edition had ˈen- first.]

What has to be clear is that, when LPD assigns one of its stress-shift in-line arrowheads ◂ to an item, it's very offen not indicating a clearcut rule but only making a suggestion of the likelihood of the occurrence of shift. The decision to use the ◂ sign essentially depends on whether or not it's considered that the first syllable can carry full stress. Notice that ˌalˈright is recognised as regularly having two fully stressable syllables whereas alˈready is only to be regarded as possibly so operated. The alternative pronunciation is indicated by the not very obvious notation ˌ.- in which the low-stress sign ˌ is followed by an online dot which stands for the first syllable of the word already and is followed by a hyphen which signals that the word's representation is incomplete.
It may be thaut that one might well have accorded a secondary stress to already because it can offen be observed to exhibit stress-shift as in I've ˈalready `seen it instead of I’ve alˈready `seen it.
This use of ◂ wasnt originated by Wells but seemingly by Gordon Walsh the former Longman phonetics editor who also introduced the /hӕpi/ type of notation (with /i/ to represent its final vowel) apparently first in 1978 in their long-gestated but excellent LDCE (Longman Dictionary of Contemporary English).


Blog 437

The 21st of February 2013

The GB air/square phoneme.

In descriptions of current General British pronunciation there're at the moment two competing accounts and thereby transcriptions of its air vocalic phoneme. The most gen·rally used of these continues to employ a traditional symbolisation reflecting its formerly virtually undisputed value as uniformly the diphthong [ɛə]. The other recognises that this phonemic unit can be sed to've evolved for very many speakers into the monophthongal long vowel [ɛː]. For the vowel phoneme in a word like fair most people — cert·nly a majority of younger ones — ord·n·rily use [ɛː] all the time. Others, mainly but by no means exclusively older people (with or without noticeably 'refined' accents), use both [ɛə] and [ɛː] mostly preferring the latter in closed syllables. All but the most elderly-sounding regularly use [ɛː] before /r/ in words like dairy but, in a word like fair, especially if it comes at the end of prosodic phrase etc, may also be found to use [ɛə] notably if it's strest. This can be observed by comparing the performances of air words accompanying the various dictionaries which in the last two or three decades have increasingly come to provide audio demonstrations from discs or internet sites notably the Longman Pronunciation Dictionary (LPD), the Cambridge English Pronunciation Dictionary, the Oxford and Cambridge Advanced Learner's Dictionaries and the Collins and MacMillan English Dictionaries.

In the pronunciation dictionaries themselves we see that in his English Pronouncing Dictionary (EPD), from its first edition in 1917 to his last in 1963, Jones always represented this phoneme by /ɛə/. When Gimson took the EPD over he continued that policy tho, at his major revision of 1977, making the minor departure of substituting /ɛə/ with /eə/. This was motivated simply from judging its simplicity pref·rable in a book consulted by a very wide public with no special phonetic int·rests. He never made such a change to his book An Introduction to the Pronunciation of English (IPE). Peter Roach and his colleagues, who have continued with revisions of the EPD for CUP, have not departed from Gimson's practice. The Wells LPD, first publisht in 1990, has maintained /eə/ for the phoneme in all its three editions. The Oxford Dictionary of Pronunciation for Current English (ODP 2001), by Clive Upton and colleagues, on the other hand has adopted the symbolisation /ɛː/ as has, from the same stable, the online third edition of the OED. Lesser Oxford dictionaries not particu·ly directed at users of English as an additional language had begun implementing this policy in 1993 with the New Shorter OED.

It seems that a century ago most GB speakers used a diphthong of the [ɛə] type in all situations. The describer of Victorian GB Laura Soames was a, praps unwarrantably, lone voice in 1891 indicating that in her opinion most speakers used what we now call a monophthongal allophone of the air diphthong in such words as dairy. However, Henry Sweet made no acknowledgment of her view. Nor did Daniel Jones who, surprisingly, in none of the editions of his famous Outline of English Phonetics from 1932 to 1956 sed anything about forms of /ɛə/ other than that it had "no phonemic variants differing to any marked extent from" [ɛə]. In his Pronunciation of English (1950:63) he made the terse remark "Occasionally one hears a monophthongal long ɛː (ðɛː, bɛː, skɛːs, stɛːz)". Gimson, in 1962 in the first edition of his IPE remarked that a "form of advanced RP uses a long pure vowel [ɛː] ... especially in a non-final syllable, e.g. careful, scarcely [ˈkɛːfɫ], [ˈskɛːslɪ]". Wells's 1982 Accents of English remarked on non "RP" varieties at his p.157 "In much English and southern-hemisphere speech and in Wales, the opposition exemplified by shed vs. shared is one of duration rather than quality..." and of "RP" itself "/ɛə/ often involves very little diphthongal movement." Additionally he remarked at its p.293 "a monophthongal /ɛə/ i.e. [ɛː], is perhaps a Near-RP northernism if in a stressed final syllable; in other environments, as careful [ˈkɛːfl], bearing [ˈbɛːrɪŋ], it carries no such connotations." A further major step onward in this evolution was to be seen in Alan Cruttenden's revision of the fifth edition of Gimson's book in 1994 when, commenting on variants of /ɛə/, he made clear his judgment that "Nowadays a long monophthong [ɛː] is a completely acceptable alternative".

In the course of providing advice to users of (British) English as an additional language, the present writer in 1969 in A Guide to English Pronunciation supplied a description of /ɛə/ as "Often narrowed to a long simple vowel before consonants and when unstressed". The present leading textbook in that field, Practical Phonetics and Phonology by Beverley Collins and Inger Mees, from its first edition in 2003 rejected the now increasingly out-of-date recommendation to non-native speakers of the diphthong in favour of the monophthong /ɛː/. With the unequivocal recognition of the predominance of this form over the recessive diphthong to appear in the forthcoming eighth edition of the universally acknowledged standard description of the General British accent of English, Alan Cruttenden's revision of the Gimson book, it'll be int·resting to consider what we may see when the next editions of the two major pronunciation dictionaries of GB appear.

Those who might like to read more on this topic than the above outline may care to look, on the main part of this website, at §5.1 which has more detail including especially bibliographic information not offered here. Readers are reminded that this account has been historical and not to be taken as expressing any lexicographical or didactic preferences.


Blog 436

The 17th of February 2013

More on Weakforms (ix) am, an, and, any etc

am: Still dealing with our first alphabetical set of items, ie those beginning a, b and c, it's time something was sed about am ie the part of the verb (to) be. This word is unusual in belonging, along with has, is and will, to the very small group of verbal weakforms which commonly give up their syllabicity in being amalgamated with a word they follow. In the case of am this only means forming the contraction I'm /aɪm/. The fuller syllabic form /əm/ occurs chiefly only in inversions of sentence order as in eg the interrogative What am I to say? /ˏwɒt əm aɪ tə `seɪ/ or the idiomatic So am I /ˈsəʊ ə`maɪ/. In relaxed conversation the initial schwa may easily be dropt eg /ˈsəʊ `maɪ/. Other examples are: Am I right? as /m aɪ ́raɪt/; How am I doing? /haʊ maɪ duɪŋ/. Especially in the context of the speaker's just previously having used the unreduced form /aɪ/ of I, agen in very relaxt style, the form /əm/ may be used, eg in /aɪ `θɪŋk əm ɔː ˏraɪt/ I think I'm alright. There's more on am at our Section 4 Item 7 ¶¶104-106. Not mentioned there is the possibility of accented Am I? /  ́əm aɪ /. Such occurrences are sometimes not believed in — especially by those who insist as a matter of faith on the non-existence of stressed occurrences of the GB phoneme /ə/. This is praps too relaxed a usage to be recommended for adoption by users of English as an additional language tho it'd be a waste of time to criticise them for using it.  

an: At our didactic account of an (at §4.7 ¶28) it was not consider·d necessary to mention that in relaxt styles its weakform /n/ may completely lose its syllabicity as in /aɪ ˈhad n̯ aɪˏdɪə.../ I had an idea...
Blog 431 mentioned some recent anomalous uses that have developt of the strongform of an.

and: We de·lt with and in Blog 057 and agen at # 403 in the first of this series 'More on Weakforms'. However, we didnt mention at either of those places the occasional occurrences of it that are notable for being so reduced that it not only loozes its /d/ — which it on·y ever has in some of that fraction of its total occurrences when it's emphasised — and also at the same time not only its vowel but also its syllabicity. This happens very freq·ently in the expression so-and-so /`səʊnsəʊ/ and, less offen, in markedly relaxt sequences such as  /ə ˈkat n̯ ə ˎdɒg/ a cat and a dog.

any: This brings us to any — which unsurprisingly didnt figure in Henry Sweet's historic 1885 first list of fifty-odd functor weakforms. It offen occurs in a weakform that has no more significance than the automatic assimilatory versions of and as /əm/ or /əŋ/. By that we mean that, when in fluent speech it closely precedes a word beginning with a vowel, its final /i/ may be desyllabised into a yod as eg when any old... may readily become /`enj əʊld/. More notably, chiefly in fairly casual speech, its initial vowel may become a schwa or very offen be elided resulting in simple /n/ — syllabic or not. So eg Got any cash? may become /gɒt əni/ or /gɒt n̩i/ or /gɒt ni ˊkaʃ/. We'll see when we come to many that it behaves in the same sort of way.


are: Daniel Jones from the 1918 first edition of his Outline of English Phonetics had a note (in his 1956 and other late editions to be found at §488) regarding the reduction of are to unsyllabic /r/ giving the example /ðə ˈʃɒps rɔːl `ʃᴧt/ the shops are all shut. This was thoro'ness rather than didactic necessity.

as: Neither of the two principal pronunciation dictionaries records any weakform for this word other than /əz/ but in fairly relaxt speech the form /z/ is not in the least unusual eg as in /əz ˈfɑː z `aɪ ˏnəʊ.../ As far as I know...


Blog 435

The 11th of February 2013

Minimally Different Words.

Portrait of Higgins
Picture of Book Cover


My fellow bloggist Kraut has already called cordial attention to my old fr·end John Higgins's unusual little book he's called nauti(c)ly "Don't Ask the Admiral to Show you his Pinnace". John has devoted most of his life's work to linguistic matters with special involvement in their applications to the teaching of English to those who use it as an extra language. At the back of its attractive shiny star-studded cover he describes his book as "A light-hearted tour of minimal pairs and some of the problems they create for those who speak English or are trying to learn it."

Regarding the book's title, he has in mind the way for example Spanish learners of English may well be inclined to attempt to say the word 'pinnace'. That is, employing the vowel sound in its first syllable that they use in their corresponding word 'pinaza' (which, in point of fact, is where we got our word pinnace from) and arriving thereby at /`piːnɪs/ ie penis. Ultimately, it seems, such little vessels were named from their being made of pinewood. It's the kind of small boat carried on board a ship for taking people ashore etc. The OED tells us that this word has been spelt 45 other ways in the past including in the sixteenth century as 'penisse'. By the way, it also tells us that at one time it was a slang word for a prostitute, a fact it illustrates with a quote from an early-eighteenth-century play by a cert·n T. Baker "Fine Lady's Airs iii. i. 26 My Dear, thou art but a whiffling sort of a Pinnace, I have been proffer'd lovely, large, First Rate Ladies for half the Mony". [Whifflers is a nice precise word we've sadly allowed to fall into disuse for those who clear the way for important persons thru a crowd by brandishing spears, swords or the like.]

Kraut sed that he particularly liked the item about a 'pole-vaulter', and so did I. It went like this: At an international athletics gathering, one participant on newly making the acquaintance of another asks him 'Are you a pole-vaulter?' and receives the reply 'No, I'm German, but how did you know my name?' This was very effective becoz the stresses, sounds and intonation that wd be employed wou·dve been exac·ly suitable also for the wrong presumption of the question as being 'Are you a Pole, Walter?' leaving only the v for w substitution typical of German speakers' English to distinguish the two versions.
This was so much better than the joke I mentioned in my Blog 063 which hinged on alleged misunderstanding of the question 'What's the bleeding time?' in a film where it was put by a senior doctor to trainees one of whom is supposed to've understood it as a swearing enquiry regarding the time of day. The joke fell completely flat for me coz the prosody was totally wrong. What they needed to do was have the question arranged to occur with a prosodically neutral effect such as wd be obtained by having the instructor say first quietly "Now the ˎbleeding ˏtime.." and then, being presumably not attended to properly, saying loudly "Who can `tell me what's the bleeding time?" That wou·dve made the possibility of a misunderstanding reasonably conceivable instead of totally incredible. The intonation ˈWhat's the ˈbleeding `time wd be the only kind that sounds in the least natural with 'bleeding' as a swearword intensifier becoz the tonic stress has to be on 'time'. By the way, 'bleeding' so used is the equivalent of 'bloody', the latter being middle class and the former lower class. Shaw's Eliza, if she'd been speaking proper Cockney, wdve sed 'No bleedin’ fear'. Saying 'Not bloody likely' praps accorded with her well known middle-class aspirations.

Minimal pair is a term much used by phoneticians referring to a couple of words exactly matching in pronunciation except at one sound. This book provides lots of mini-lessons in phonetics etc that are all good fun, illustrating entertainingly words like malapropism. Mrs Malaprop is quoted saying "Sure, if I reprehend anything in this world it is the use of my oracular tongue, and a nice derangement of epitaphs." [She gets wrong the words apprehend, vernacular, arrangement and epithets].
The book cert·nly has an agreeable 'linguistics without tears' function explaining amusingly various more and less serious terms including homographs, homophones, homonyms, phonemes outrageous rhymes and mondegreens. It strikes me as likely to make a very welcome small present for teenagers of any country learning English — not least becoz they can pick up from it a knowledge of a dozen of the nau·tiest English words — always an attraction for young students. I can imagine this taking its place on a lucky junior English learner's bookshelf alongside the Trim/Kneebone booklet English Pronunciation Illustrated just described in our Blog 433. (Illustrating the next edition of this book wd be a fine opportunity for a budding Kneebone. I can pass on any offer to do so to the author.) By the way, this jokey little book ends with yet another joke when the author refers to "my grandfather, Professor Henry Higgins".

The book (ISBN: 978-1-291-30237-0) is available internationally via Lulu.com. Price in UK is £5.50. There's an excellent review of it by John Maidment at http://blogjam.name/ in his blog of the 1st of Feb·ry.


Blog 434

The 3rd of February 2013

Accenting of a Compound Word.

A member of an internet circle of pronunciation teachers to which I belong recently put this question to the rest of us: "In an ELT coursebook, stress is marked on the first element of 'forest fire' but I've checked around and some people seem to stress the second part. This combination is not included in Jones or Wells. Any preference?" My first reaction was to think that he'd been using an American book but he only quoted the two principal British pronouncing dictionaries the LPD (the Wells Longman Pronunciation Dictionary) and the CEPD which in its Cambridge re-incarnation of the Daniel Jones English Pronouncing Dictionary is principally in the hands of Peter Roach. So, even with the provenance of the text book not stated, it seemed reasonable to presume that it was British. Since the majority of our members are teachers of American pronunciation I'd guess they'd wondered why he was worrying. I think most speakers of American English wd find that stressing normal. Not so the British. You may be able to guess why the lexicographers fight shy of giving them really full treatment — it'd probbly require so much extra space that it'd double the size of any of them. That would praps necessitate the issuing of them in two volumes. The current edition of EPD, excellent tho it is, is very considerably enlarged from what it was in Daniel Jones's or Gimson's day — taller and wider, with more pages and even in paperback heavier than the older hardback versions. Even so it's reduced in legibility partly because of the employment of more complicated types of transcription than were used in Jones's day and partly on account of the enthusiasm for displaying things that are a waste of space such as epenthetic sounds and obvious assimilations. Most of these comments apply to the LPD which, however, does benefit from a two-column layout that's somewhat more comfortable than EPD's three-column choice. The Oxford DP has not perfect but better legibility than either of these two but suffers from restricted coverage while yet being unwieldy.

The number of dictionaries which have followed the lead of Hornby's hugely successful OALD ie Oxford Advanced Learner's Dictionary (for the third edition of which I was invited by him to contribute the pronunciations) are much better at including compounds but they still don't anything like fully meet the demand for information on them as our example seems to confirm below. This book in 1974 wd pop reasonably easily into a briefcase even in its hardback form. These days, magnificent tho it's become, it's more than twice that size and weight. It, as do even more frequently its rivals, includes lots of compounds without giving pronunciations or even stressings for them. Checking online some of the most likely sources of information on our 'forest fire' example  I found the following results.

At the OALD the expression is given as an illustration of use only, but with no definition or pronunciation or even stressing. So also for the Oxford American ALD. The Cambridge equivalent was even less informative. The Longman Dictionary of Contemporary English was the same. The MacMillan Dictionary had it as an illustration of use only. The Collins English Dictionary came out far the best with (ˈfɒrɪst faɪə) and "a large, uncontrolled fire in a forest or wooded area" plus nine modern illustrative sentences. Strangely, the transcription didnt include the predominant British stressing at all. Dictionary.com had a good concise definition plus nine lines of further information and eight cross references but no phonetic information of any kind. Merriam-Webster also had a definition but no phonetic information. Tim Bowyer at Howjsay hasnt got round to saying it yet. The Random House/American Heritage didnt seem to have it. Vocabulary.com had a good short definition but no transcription tho it did have a macho-sounding American male speaker with an Eastern accent saying [`fɔ/ɒrəst faɪ.ᴧ] very clearly (if rather artificially) so with its the forestressing it was okay for its presumably American audience. The great Oxford English Dictionary has it with no definition except insofar as it has two quotations of its use, one from 1878 (with it hyphened) and the other from 1958. Finally, and not online in this case, the Oxford Dictionary of Pronunciation unsurprisingly had no entry for it.

When my colleague "checked around" it seemed to be that he was ready to accept the 'ruling' of the textbook but only with cert·n misgivings. He was quite right to have such dou·ts, I shd say, because I imagine I'dve been aware of it in all the accounts of such fires in broadcasts over many years if any number of British speakers had come up with the American stressing. Forestressing such a compound unifies its elements more than 'even' stressing does. It tends to happen among speakers who have high-frequency use of the word and they cert·nly have a lot of fires in California to talk about. Another reason why, it seems likely to me, Americans, much more offen than the British, tend to treat such compounds in this way is that considerable numbers of speakers of German and other forestress-preferring languages have settled there. You can see the German habit in their spellings: forest fire is Waldbrand in German.

 One final point is that people without the appropriate phonetic training are very offen not reliable in their impressions of what might be the usual syllable that some other speaker ordinarily selects to receive tonic stress in saying a compound word. In a fair proportion of the occurrences of a compound like forest fire that one might hear from newsreaders etc, the contrast that wdve been evident between earlier and later stressing in most situations may be dissolved out completely in some others. For example a speaker whose usual habit is to employ a forestressed version of the word as `forest fire will hear nothing contrary to their preferred stressing in a sentence such as In the `present conditions | there is `still a possi`ˏbility | of `further forest fires. In the precontext to such a sentence forest fires will have been referred to but not at all necessarily by use of that very compound word itself. Untrained listeners are quite likely to assume that the speaker shares their stressing pref·rence not becoz they actually he·rd the stressing they favour but becoz the words were spoken exac·ly as they themselves wou·dve sed them. That is to say, deprived of any stressing feature becoz of relegation to the least important part of a sentence owing to their containing no new information.


Blog 433

The 27th of January 2013

John L. M. Trim.

Portrait of John Trim


It was sad to hear that John L. M. Trim had died on the 19th of this month at the age of 88. I liked him very much and have happy memories of his hospitality at Cambridge but others are far better equipt than I am to write about his life. I'm sure we shall hear in due course from John Wells who benefited so greatly from studying phonetics under his guidance at Cambridge. However, as something in the way of a small tribute, I'll try briefly to list his writings in the field of phonetics. They were a good deal less numerous than one cou·dve wisht but they are very far from deserving to be forgotten.

In the dozen years from 1950 he made about twenty contributions to the International Phonetic Association's journal which in those days still functioned with French as its editorial language and so became familiarly known as the "m.f." short for Le Maître Phonétique. His very first piece in it was a transcription of ten lines of verse by the German poet Hugo von Hofmannsthal. He contributed seven more German items to its series of specimens called the Student Section, mainly of prose. His dozen other mf contributions included some penetrating reviews among which were his accounts of Preliminaries to Speech Analysis by Jakobson, Fant & Halle in 1952 (pp 37-40), Eugen Dieth's Vademecum der Phonetik in 1951 (pp 18-19), and Martinet's Économie des Changements Phonétiques in 1957 (pp 14-17). Besides these there was a single obituary in 1951 and a concise but valuable note on the phonemic status of German h, ç & x (p.41) also in that year. His final contribution, in 1959, was a stimulating, original article on 'Major and minor tone-groups in English' (pp 26-29).

    This last item was considered so valuable by W. E. Jones and John Laver that they selected it for reprinting in their 1973 anthology of fundamental articles commended to postgraduate students entitled Phonetics in Linguistics. It was one of four in that book taken from the 'mf' and, tho the other three were converted to ord·n·ry spelling, Trim's had to be left in its original notation becoz its form so importantly exemplified its challenging theme. Trim appeared as co-author of another of the items in that book namely 'Vowel consonant and syllable — a phonological definition', originally publisht in Word in 1953 in collaboration with J. D. O'connor. In 1964 Trim was co-editor of and contributor to the volume In Honour of Daniel Jones in which he had a substantial article, at pp 374-383, on 'Tonetic Stress-Marks for German'.

Other articles of note included a lengthy one (of nine full pages) in 1961 in the British-Council-cum-OUP English Language Teaching journal Vol. XVI. No.1. It was a thoughtful discussion proposing the use of the expression English Standard Pronunciation (abbreviation E.S.P.) for what had up to then mainly been called RP, reserving RP for the distinctively public school accent. Like so many of us he felt that the term 'Received Pronunciation' was increasingly an embarrassment. The movement away from it continues year after year and an important further step away from it will be seen in the forthcoming eighth edition of Cruttenden's Gimson. Another Trim article was ‘The Synthesis of English Vowels’ written in collaboration with G. F. Arnold, P. Denes and J. D. O’Connor publisht in Language and Speech issue No 1 in 1958 (at pp 114-125).

Book Cover

Finally we must mention, tho very few people will need reminding about it, his brilliant little 96-page paperback book of pronunciation practice materials with highly entertaining humorous illustrations by the late Peter Kneebone, English Pronunciation Illustrated. Shockingly for a publication of Cambridge University Press, my copy has no date of printing or record of previous editions. Its early editions (from 1965) for a decade or so used the Jones EPD transcription of the day. After the 1978 watershed when Gimson turned the EPD over to his preferred 'multiliteral' transcription it followed suit. It has long been accompanied by recordings and is still in print. The book offered concise and simple but phonetically sophisticated advice pointing out its potential value even to speech therapists as well as language teachers and learners. It contained eleven pages of Word indexes remarkably both 'forward' and 'reverse', that is, listed respectively by their initial and final phonemes.

In 1956 he was specially thanked for his "excellent recommendations" by Daniel Jones who, half a dozen years after his final retirement in 1949 (then in his mid seventies) was labouring over final revisions of his most famous book. The Preface to that eighth edition of his Outline of English Phonetics, gratefully acknowledged his help mentioning that a definition that appeared at p.332 had been adopted exactly as suggested to him by Trim. It was:
A 'broad' transcription may be defined precisely as one which represents only the phonemes of a language, using for this purpose the minimum number of letter shapes of simplest Romanic form (consistently with the avoidance of undesirable digraphs for 'single sounds') together with such prosodic marks as may be necessary for the avoidance of lexical ambiguity.

Especially after transferring to Cambridge to set up a Department of Phonetics and subsequently to foster the new discipline of linguistics there, Trim's preoccupations besides teaching and administrative matters, turned most intensively to subjects in the area of Language Policy and Pedagogy, the title of one of the journals to which he contributed. One partial minor return to the old phonetic int·rests seemed to be evidenced in his 1987 publication 'Daniel Jones’ “classical” model of pronunciation training' in the volume Language Topics: Essays in honour of Michael Halliday. Anyhow, he ended his career becoming for many years an active extremely highly regarded leading figure in the fields of Language Policy etc in the European Community.


Blog 432

The 9th of January 2013

Comparison of Transcriptions.


Readers of the "Kraut" blog of 27 Dec 2012 will praps remember that I welcomed it as "Another splendidly stimulating post!" I askt if any other Kraut fans wd like to c·mpare their impressions of the 40 words fr·m one of our most justly admired news presenters (Fiona Bruce) with my own as I transcribed them in some detail. They were "Well, that's it from us. There's more on the BBC news channel including a fresh look at tomorrow's front pages. But now on BBC One it's time to join our news teams where you are. Bye-bye."
I'm minded to preface the following discussion with one of the concluding comments in his epoch-making book Phonetics (1943 p.152) by the late great American authority Kenneth Pike "The fact may be emphasised that no phonetic description, no matter how detailed, is complete". We shd also remember the wise words of Sidney Wood readers of phonetic blogs will've seen about the uncert·n qualities of recordings coming from the Web. Here are Kraut's transcriptions and comments followed by my own plus a few remarks by me.

1. [wə ðæts ɪt fm ˎʌs] (This section lasts roughly 600ms. Mark the relaxed weakform pronunciations of well and from. In contrast to JWL I do hear a voiced 'th' at the beginning of that.)
[wəl `ats ɪt fm ˈᴧs |] At 'that's' I think he's prob·bly right to suggest that a tongue gesture corresponding to /ð/ was made but I expect he'll agree that a canonical /ð/ wasnt produced ie no friction was detectable. Any voicing was pretty weak and I'm surprised he doesnt detect a lateral. I think there is a tone on 'us' which is narrow and p·aps lower than I chose but I'm still sure there's no movement on it. Neither of us marked the /w/ of 'well' as underrounded.

2. [ðɛz ˎmɔː ʔɒn ə biˑbsi ˈnjʉz̥ ʧænl̩] (Note the weakform for the definite article the. I can't spot an eth in the. The whole phrase lasts about 1.38s.)
[ ðɛz ˎmɔː | ɒn ðə bibsi ˈnjʉz ʧanl ] The glottal closure before 'on' was very weak: something you cd almost say you sense rather than hear. Very weak tho it is, I seem to detect a light tongue gesture for a /ð/ which has no audible voicing or friction, so it's for me the next thing to absent altogether — which was what he plum·t for. I suppose I must've thaut it wasn· worth indicating a degree of length on the first vowel of 'BBC' but anyway diff·r·nt impressions of length like this are trivial. I didn· specify that the /z/ of 'news' was voiceless or that the /l/ was syllabic coz I'm afraid I simply took them to be automaticly so in the contexts — so there's no disagreement between us about the facts.

3. [ɪŋɣ̊lʉdɪŋ ə ˈfrɛʃ lɘk ɘth] (The /k/ in 'including' is a slightly voiced velar fricative; the vowel quality of look and at is difficult to determine because the vowel duration is extremely short. For my ears the vowels have a fairly half close character.)
[ɪŋxlʉdɪŋ ə ˈfrɛʃ ˈlək ət] I think his [ɣ] for the 'c' of including made a better choice than mine but of course there's not all that much to argue about between a devoiced voiced consonant and a voiceless one. I see we agreed that she has a schwa-type vowel rather than an /ʊ/ for 'look' but mine is the vaguer symbol and his the more precise. Kraut's marking aspirated /t/ at [ɘth] was something /sᴧnθɪn·/ I cou·dn make out.

4. [thˈmɒrz̥ frᴧnt ˎpeɪʤɪz] (For a news presenter it's a very relaxed pron of tomorrow's.)
[(ət) təˈmɒɾz frᴧnt ˎpeɪʤɪz] The first syllable of 'tomorrow’s' is extremely short but what follows the /t/ is to my ears weakly voiced rather than voiceless, hence my preference for schwa. I wou·dnt argue about the precise nature of the very weak /r/ of 'tomorrow'. I imagine we're both taking it for granted that the /z/ of 'pages' was unvoiced. He was explicit that the /z/ of  'tomorrow’s' was voiceless — I wasnt. Neither of us bothered to disclaim any intention that we me·nt the vowels [a, ӕ, ɒ, ᴧ] etc to be taken as used in their strictly cardinal values.

5. [bət ˏnã̟ʊ] (There's a low rise on now.)
[bət ˈnãʊ] If we wanted to be fussy we cd record the fact that she closes her lips immediately at the end of this 'now' which therefore we cd transcribe as [nã̟ʊp̚]. There's cert·nly a rise in pitch at 'But 'now' but I dont perceive it as on the word on 'now'.

6. [ʔɒ̃n ˊbibisi wᴧn] (with a high rise on BBC One)
[ɒn bibisi ˈwᴧn] I missed the [ʔ] before 'on' which I'm sure Kraut was justified in indicating. As he is also with the nasality of the vowel: it was too weak for me to've been anxious to record but I wou·dnt argue about its presence. And I agree this phrase has pitch ascent but I hear 'BBC' on a fairly level pitch not much above 'on'. 'One' [wᴧn] is distinctly higher but doesnt seem to me to itself move.

7. [ɪs ˈtaɪn tə ˈʤɔɪn ʔɑ ˈnjʉ siːmz̥ wɛ jʉ ɑ] (Mark the unusual change of the consonant, /t/ becoming /s/, in the sequence where news is followed by teams.) It was the alveolarising pre ·assimilation of the /m/ of 'time' to the following /t/ of 'to' that gladd·n·d the h·art of John Maidment for being an inter-word rather than intra-word example of this appar·ntly rather unusual phenomenon. It was a previous comment of John's that inspired Kraut to record this b·utifl example of relaxt speech as it'd modulated away from a necessarily very formal style. Notable as that was it was even more remarkable to find the other assimilation Kraut invited us to mark — that the initial plosive /t/ of 'teams' has surprisingly been converted to /s/ under the influence of the (subsequently elided) final /z/ of 'news'.
[ɪs ˈtaɪn tə ˈʤɔɪn | ɑ njʉ simz wɛ ˎ jʉ ɑ]. I seem to'v· overlookt another weak [ʔ] before 'our'.  I was wrong to take it for granted that the /z/ of 'seems' wd be voiceless here: there was good reason to show it so explicitly, as Kraut has done, coz it cd well've been assimilated to the following voiced consonant — which he made clear was not the case.

8. [bɐ ˈbaɪ] (The diphthong in bye has an almost whispery character.)
[ˈbɐ ˈbaɪh] Paps [ ˈbɐ ˈba̤ɪ ] wdve been more proper use of the IPA alphabet. Neither of us thaut to mention her paralinguistic  lip-spredding at 'channel' & '(BBC)1' /wᴧnː/. I felt that her very charming good-humoured smiling seemed to keep me rather specially cheerful while pondering over her linguistic performance.


Blog 431

The 24th of December 2012

More on Weakforms (viii).

In my attem·tedly alphabetic listing of weakforms (a term briefly definable as 'an articulatorily reduced word form constituting a different set of phonemes from the word's full form') I mainly hav·nt yet gone beyond those beginning with one of the first three letters of the alphabet. We de·lt last with actually at Blog 428 and before that at 425 with can, come and bedroom etc, at 420 with because, at 422 with by, at 418 with able, almost, always, any, anybody and anyone, at 403 with and, and at 397 with only along with the importantly previously essentially unrecognised we're and nearly.

Something we've not mentioned much so far is the ease with which weakforms are produced by aphesis /`afəsɪs/, the loss of a weak vowel or a consonant in the beginning of or constituting the first syllable of a word. This is a familiar process from the history of the written language with well-known examples like the weakening of esquire to squire with the latter being establisht as an independent word. In fully relaxed styles this is perfec·ly common today among mainstream GB speakers giving, especially closely following vocalic sounds, items like it's ·bout time.., high ·bove the clouds, ·f you take my ·dvice, ·fraid so, the ·lectricity's off, I ·kspec(t) so, ·cording to my idea, go ·cross the road, try ·gain, it's not ·lowed (ie not allowed, sounding exactly as 'not loud' ) also less casually /ɪts nɒt l̩aʊd/ (with syllabic /l/), years 'go etc. Various words beginning with the spelling ex- as /ɪks/ or /ɪgz/ or / əks/ or /əgz/ are increasingly he·rd with elision of the first consonant of the prefix particularly often with  the word exactly which becomes /ɪ`zakli/ etc.

We've mentioned almost and always as recognised to an extent by some pronunciation lexicographers. In fully relaxed styles their first element, the independent word all, can readily shed its consonant as for instance what we may represent informally as aw by myself, aw my life, aw the time, ·f aw· goes well, aw night long and very commonly aw right. Altho awtho exists, prob·bly the more usual weakforms of although are the two with post assimilation of its /ð/ to an ell viz /ɔːl`ləʊ/ and /ɔː`ləʊ/.

Finally, while we're still dealing with weakforms beginning mainly with  a, b and c, there's a peculiar change we may mention that has been catching one's attention increasingly in quite recent years. It concerns the pronunciation of an before notably the words heroic, historic and horrific. Both an and its alternating form a descend from the Old English word aan (meaning 'one') which by shortening produced the weakform an which in turn itself weakened to a in the context of a following consonant. These two ended having the strongforms /an/ and /eɪ/ and the weakforms of them /ən/ and /ə/. By the Early Modern English period, when you used one of the few words like historic there were two natural possibilities. You used /ə/ if you selected its form sounding the initial aitch or otherwise you chose the form with the aitch 'silent' and then you employed the other weakform /ən/. I suspect that many speakers experience something of the feeling, which I share with them, that this slightly uncomfortable hiatus which the form /ə hɪstɒrɪk/ tends faintly to suggest something of a pedantic anxiety to sound 'correct'. Linguisticly less confident speakers, one suspects, first became aware that most speakers usually plum·t for the version with /ən/ and no aitch. However, being worried about dropping any aitch — the most heavily stigmatised ever of all features of English speech — they elected for /ən/ but kept the aitch. I've also encountered this recently with hereditary and harmonic.

 All this has been leading me to remark on the curious further recent development of the last two or three decades that more and more speakers seem to be favouring the practice of originally only 'anxious' speakers who were inclined to make it clear not only that they're not dropping any aitch but highlighting the fact by using the word an in its strongform /an/. It's begun to seem that one can't any longer hear of at least any historic or horrific happening without its being preceded by /an/ or /ən/ even from a variety of mature speakers such as Channel Four's Jon Snow, Professor Diarmaid McCulloch and the BBC Television News Presenter Sian Williams.


BlogRoll