David Brazil died 25th Sept, 1995

David Brazil

credit:www.speechinaction.net

Born on the 1st of May 1925 David Brazil first became a teacher of English, then a research fellow at Birmingham University and following this a staff member in the Department of English. From 1979 to 1983 he was a full-time lecturer there. He fully retired in 1986. Among his major publications are

• The Communicative Value of Intonation in English (1997²), Cambridge University Press
  
• Pronunciation for Advanced Learners of English (1994), Cambridge University Press
• A Grammar of Speech (1995) Oxford University Press

He died on the 25th of September 1995. A more comprehensive biography can be found here. If you're interested in a bibliography of Brazil's writings, click here.


---------
There will be a short break for a few days because I shan't be able to connect to the internet. Kraut's back to the keyboard!

dissimilative elision

After non-eliding dissimilation and dissimilative addition this blog posting is about dissimilative elision. A sound is dropped because otherwise two identical/similar sounds would be too close together.

Credit must be given to Jack Windsor Lewis who coined the term dissimilative elision. It describes the deletion of "repeated sounds or syllables even when there is no logical objection to them." (source + examples here)

1. Dissimilative elision of /r/
 - caterpillar (in GA)
 - particular (in GA)
 - surprise (in GA)

2. Dissimilative elision of /l/
 - Pachelbel

3. Dissimilative elision of /h/
 - hold her hand
 - he has

I also wrote about dissimilative addition and about non-eliding dissimilation.

Lightning versus lightening

credit: www.free-extras.com

credit: skinlighteningtreatment.com

How do you pronounce these two words: lightning and lightening¹? In isolation, in their canonical forms, in careful speech probably /ˈlaɪtnɪŋ/ and /ˈlaɪtənɪŋ/. What about lightening in casual speech in a sentence like the sky was lightening on the horizon? [ˈlaɪtn̩ɪŋ] is an option - in more casual or allegro speech even [ˈlaɪtnɪŋ].
Do these two pronunciations constitute renderings of a minimal pair in the phonological sense of the term? If they do, then the syllabic and non-syllabic n-sounds seem to be allophones of different phonemes.

Let's clear this up! 
For two sounds (= phones) to be called allophones of the same phoneme two conditions must be fulfilled:

1. the distributions of the phones must be predictable and
2. no meaning difference is established if one phone is substituted for the other in the same context.

According to these conditions and assuming that we're talking here about two different sounds the two phones cannot be allophones of the same phoneme. Are they allophones of different phonemes?

Recall that the syllabic [n̩] in the casual pronunciation of lightening is the result of a structure simplification process like the change of, say, bread 'n' butter or fish 'n' chips. What's left of the weak syllable is a nasal consonant. The vowel - here schwa - is deleted and the following consonant² occupies the peak position; the consonant becomes syllabic. By this syllabic variant the syllable structure of the word is preserved. In the citation form the nucleus position is held by the schwa, the coda is occupied by the nasal. In the casual form we only have a nucleus, no coda, and this nucleus position is necessarily occupied by the /n/. As all phonemes in nucleus position are by definition syllabic, it's redundant to assign a syllabic status to the /n/. Moreover, a syllable is a syllable and a phoneme is a phoneme, i.e. they belong to different levels of analysis. 

So the two n-sounds are incommensurable from a structural point of view. I wonder if there're any auditory or acoustic differences between [ˈlaɪtn̩ɪŋ] and [ˈlaɪtnɪŋ].


¹ My thanks go to Martin J Ball for providing this beautiful example.
² The interesting question is: Which consonants can be syllabic and which cannot?

minimal pair sentences with /ʧ/ and /ʤ/

credit: www.anderslernen.com

Here's a list of minimal pair sentences containing words contrasting /ʧ/ and /ʤ/:

1. I believe he’s joking/choking.
2. The crowd jeered/cheered them.
3. What happened to your gin/chin?
4. She dropped her jello/cello.
5. They’re Jane’s/chains.
6. Is that badge/batch ready yet?
7. Marge/March is fine with me.
8. Did you see her lunge/lunch?
9. It’s a little ridge/rich.
10. They’re surging/searching now. 
11. Who's that miserable Reg/wretch
12. I prefer Jess/chess
13. I like his Joyce/choice.

Malcolm Muggeridge - sound snippet

"[...] so that to this day the BBC is thought of as the organ of the, as it were, genteel and respectable elements in society." There seems to be some flaw in the sound track that gives me impression that there is an r-sound after the vowel in 'thought'.

brɑən sjuːəl

credit: habsboys.org.uk

credit: courtauld.ac.uk

Brian Sewell was born in Kensington on the 15th of July, 1931. He is an Old Haberdasher, which means he attended HABS, a public (= independent) school by the full name of Haberdashers' Aske's Boys' School.  He then went to the Courtauld (/ˈkɔːtəʊld/) Institute of Art, a college of the University of London, where he graduated in 1957. So much for the speech-forming years of [ˈbrɑən ˈsjuːəl].

When you listen to his enunciation in various clips available on the internet, you will discover that his poshness varies quite a lot. The plummiest variant I've heard so far is to be found in his series 'Brian Sewell's Grand Tour of Italy'. Try to find a video clip by typing "last of the medici" into your browser and enjoy it.

There's even a Brian Sewell audio sampler.

Here's a list of words taken from the above-mentioned video clip:

from outside	frəm ʔaʊtˈsaɪdɦ
show	ʃ̱əʊ
Medici	ˈmedɪʧeʱ
try	tçrɑːɪ
sire	sɑːɜ
heir	ʔɛːɜ
to lead	thʊ liːːdh
alcove	æɫˈkəʊv
altar	ˈɔːɫtɑ
panniers of fruit	ˈphænɪɜˑz̥ əv̥ fru | th
grapes	ɡreɪ | ps
chin	ʧɪnː
dinner time	ˈdɪnɜː thaːɪmː
boys	bɔːɪ | z̥

Some people say that Brian Sewell speaks posher than what the Queen does.

Who is this handsome young man?

credit: BBC and PBS

It's John Christopher Wells as Alex Rotatori correctly answered. The photos are snapshots taken from a BBC series of 1986 called 'The Story of English' consisting of nine episodes. In episode 7 - 'Muvver Tongue' - John appears commenting on other people's pronunciation.

Malcolm Muggeridge interviewing John Reith

John Reith

credit: BBC 2

In his blog of the 6th of September Jack Windsor Lewis mentions an interview conducted by Malcolm Muggeridge ( = MM) in 1967 with John Reith ( = JR), former director of the BBC.

Malcolm Muggeridge

credit: BBC2

A section of that television interview was broadcast in Melvyn 'pink shirt' Bragg's BBC feature "RP RIP".

What follows is the script of an excerpt of the interview as played by Melvyn Bragg in his feature. I had to leave out two short sections which were incomprehensible to me (they are marked red). Should any of my readers be able to decipher them if they have access to the transmission, I'd be glad to add them.

MM	Your BBC men all spoke, or presumably were conditioned to speak in a sort of educated Southern English voice, which became known as the BBC accent and was one of the numerable ways in which the BBC strongly influenced social attitudes. Now why was it that you wanted them all to speak like that rather than in some rich regional way like ... as you do yourself?
JR	Do I speak very definitely Scottish?
MM	Very!
JR	What do you call the stuff that's coming down that chimney?
MM	[sʊt]
JR	All right. The Scot and the Northener would say [suːt] - as broad as that. I think I say something between the two keeping the purity of the vowel or double vowel: [sʊt]. [suːt] - [suˑt] - [sʊt]. Now I was just vehemently opposed to what variously has been called the Oxford accent or the south-eastern accent such as 'the [ˈθɪɐtɐ]', 'the [ˈfɑːsaɪd]' with more of the FEH's, you see, they are disappearing and the vowel becoming indeterminate: [ˈθɪɐtɐ], [ˈfɑːsaɪd], [θɪɐ]
MM	How did you arrive at your particular, the particular style of the BBC announcer?
JR	I had a first-class man appointed to take charge of announcers to give them all sorts of advice and instruction as to how to read inflections and everything and how to pronounce [the voice of Arthur Lloyd James is heard demonstrating the pronunciation of a few isolated words], and moreover I got a committee established with the Poet Laureate on it and Bernard Shaw and all sorts of well-known people who'd made a study of pronunciation.
MM	But the interesting point in terms of social history is that this particular accent which the BBC produced somehow identified the BBC with a certain section of society, certain social trends so that to this day the BBC is thought of as the organ of the, as it were, genteel and respectable elements in society.
JR	Anything wrong with that?
MM	Erm ... well, except that, after all, the people who speak in this standard way are in fact a minority. In the public mind how people spoke on the BBC was associated with a certain type of middle-class education, a certain type of middle-class life, ... erm ... so that it came to seem an organ.
JR	Came to seem?
MM	Yes.
JR	Came to seem!
MM	Yes.
JR	Who's responsible for that?
MM	Well ... I'm not saying anybody is responsible; I'm just saying that something had happened as a result of the decision that you took.
JR	Would you have taken anything different! Look! I ... I'm ... I'm sorry if local dialects are disappearing.

The dialogue is fascinating and amusing at the same time, isn't it?

PS: The bits marked blue in the interview were deciphered/explained by Jack Windsor Lewis, whose help is gratefully acknowledged.

PT's transcription workbook - some observations, #5

Chapter 4 - 'Allophones' -  starts with some musings on the phoneme /l/:

[...] linguists usually distinguish between two varieties, known as clear and dark [...]. The two varieties never contrast with each other in English and so are never responsible for creating differences in meaning. This is because the two varieties are distributed differently in English words; (60)

Certain varieties of English such as General American, Canadian, Glaswegian English (as well as some varieties of Australian English) predominantly have dark /l/ in almost all contexts. Cum grano salis the last sentence in the above quotation is acceptable if we restrict its applicability to what PT calls SESP (aka RP, GB, ...).

PT introduces the terms allophone and phoneme, the distinction between broad and narrow transcription and the one between slant and square brackets. Coming back to the two /l/-sounds, he describes their distribution (in SESP!):

[...] the clear [l] occurs before vowels and the consonant /j/, whereas the dark [ɫ]/[lˠ] occurs before consonants, except /j/, and at the end of words. (61)

This sentence is a bit vague as it stands, I think. Clear l occurs before vowels even if the /l/-phoneme ends a word but abuts with an initial vowel of a following word and both words are slurred. As regards the distribution of the dark variant, one must not forget to mention syllabic l, which is dark as well. One might even say that in many (all?) varieties of English there is not a binary distinction but rather a continuum from light to dark l (see John Maidment's blog of the 13th of May, 2011).

PT then illustrates various other allophone variants of phonemes such as

• aspirated and unaspirated plosives, e.g. [ph, p˺]
• glottal reinforcement and glottal replacement of plosives, e.g. [ʔp, ʔ]
• flapping of /t/, e.g. but I as [bʌt̬ aɪ]
• devoicing of canonically voiced consonants, e.ɡ. [b̥]
• fronting, backing and rounding of consonants in their respective environments
• nasalisation of vowels adjacent to nasal consonants
• pre-fortis clipping of vowels
• pre-l breaking and pre-r breaking as in [fiːəɫ] and [ˈsɪəriəs]
• smoothing of a long monophthong or diphthong plus a weak vowel as when liable becomes [ˈlaəbɫ]
• diphthongisation of /i:, u:/, e.ɡ. in [t^hɪi] or [t^hʊu]

This ends my series on PT's workbook for the time being.
I shall still recommend this book to my students, but with some accompanying comments.


Addendum: See also Alex Rotatori's critical evaluation here.

PT's transcription workbook - some observations, #4

credit: Resource International
copyright limitations here

Chapter 3 of Paul Tench's transcription workbook bears the title "Word Stress".

As to be expected of a book on transcription the concepts of primary stress, secondary stress and of non-stress are introduced and exemplified. When it comes to detecting degrees of prominence in a polysyllabic word, problems may arise as to where and how many stresses should be indicated. But do we really need to indicate secondary stress in a disyllabic word such as blackbird because the second syllable contains a long vowel?

PT also briefly touches upon the phenomenon of stress shift: fundaˈmental when pronounced in isolation, but ˈfundamental iˈdeas as a phrase. When you take a look at LPD's information box on stress shift (p. 784), you will find a similar example but with a slightly different stress pattern: ˌ fundamental miˈstake. PT claims that the words pronunciation, university and communication behave in a like manner. LPD3, however, does not indicate the possibility of stress shift by adding a wedge sign to these words. This is not to say that stress shift is impossible or even unlikely, but it illustrates that it is notoriously difficult to agree upon stress assignment.

Compound stress¹ is also touched upon briefly. Combining forms containing a Greek or Latin bound stem are included, e.g. homophone, homograph, homorganic.
Let's see how PT's workbook and the 3 pronunciation dictionaries treat these as regards stress assignment:

lemma	PT	LPD3	EPD17	ODP
homophone	ˈhomoˌphone	ˈhomophone	ˈhomophone	ˈhomophone
homograph	ˈhomoˌgraph	ˈhomograph	ˈhomograph	ˈhomograph
homorganic	ˌhomorˈganic	ˌhomorˈganic	ˌhomorˈganic	<no entry>

One final comment on this chapter: On p. 58 we read:

Notice too that a compound itself can become one part of another compound: thus, two secondary stresses are required
3.13 railway, railway lines, railway station ___________________

Do we really need two secondary stress marks? Why not ˈrailway ˌstation (see LPD3 lemma railway)?


¹ Be informed that summer heat can compound stress of expecting mothers. :)

PT's transcription workbook - some observations, #3

The next chapter to be dealt with in greater detail is number 3 dealing with consonants. The transcription of plosives is practised first. The author rightly points to the fact that the letters <p, b, t, d, k, g> may lack a sound value as in receipt, debt, whistle, handkerchief, knee, gnome. However, to say that "<k> is silent at the beginning of words before <n>" (38) disregards words such as Knesset or Knopf.

According to Tench, the symbol for the voiced post-alveolar fricative /ʒ/ is "often called" (42) zhed. I must confess that I have never ever seen or heard this term before. The one I'm familiar with is 'yogh' (Jack Windsor Lewis proposes 'ezh', a term which is admittedly more transparent than the opaque 'yogh').

There is also a section on syllabic consonants dealing with the transcription of a few consonants in syllable nucleus positions: /n, l, m/.  To be able to indicate syllabicity the syllabicity stroke is introduced. To be precise, all the words transcribed with syllabicity marks should have been set between square brackets because no meaning distinction can be established between, for example, a syllabic and a non-syllabic /n/-sound in English. 

credit: IPA

No mention is made of /ŋ/ in syllabic positions as in broken or bacon when pronounced with elided schwa. And what about February pronounced as /febrri/? There are even syllabic plosives. The latter phenomenon is discussed and exemplified at some length in Jack Windsor Lewis's blog number 239 of the 16th of December 2009.

PT's transcription workbook - the sound files

In the introduction to the workbook readers are informed that they can download recordings of all the words and discourses presented in the book from the web pages of CUP. I did this and listened to some random examples.

Most of the sound files in chapter 1 are numbered 1-1, 1-3, 1-4 etc.; so there's a minus sign between two numerals. The numbers for the examples in the workbook have a dot in between numerals, e.g. 1.1, 1.2 etc.

Sound file 1-1 is a recording identifying what the sound files are about: "Transcribing English words - Paul Tench - Centre for Language and Communication - Cardiff University", a voice says.
There is no sound file 1-2.
In sound file 1-3, which seems to correspond to example 1.1 in the book, we hear 14 English first names; the book, however lists 18 partly different ones. During the recording the speaker must have banged something against the microphone causing a noise. This is something that seemed to have happened fairly frequently. These bangs could easily have been removed by editing the sound waves before putting them online.

Sound file 1-4 contains the words 'look, loud, lure, letter, coffee". 1-4 and 1.2 do not correspond at all nor does 1-4 and 1.3. Very confusing!

1-5 and 1.2 do correspond, however, as do 1-6 and 1.6. Phew! Let's see if we can construct an equation: when you have 1.120 in the book, add 3, which gives 1-123 for the corresponding recording. Alas, there's no sound file with that number. What you find instead are numbers such as 1.258 and 1-118 (which incidentally corresponds to 1.109 in the book). Whenever I try to open 1.258 with the audio editor audacity, I get a memory allocation error and the program crashes. What a mess! The author and/or the Cambridge people must do something about it.
Until these complaints have been dealt with I cannot maintain my statement that I highly recommend the book. I was obviously too rash.

Paul Tench's transcription workbook - some observations, #2

I was asked by a blog follower to shed light on the use of the transcription symbols "æ/a" in Paul Tench's workbook (see my blog entry of the 4th of September).

The model word Tench uses for these vowel symbols is <lack>. He writes:

Lack is transcribed either as l æ k  or  l a k
You choose! The first one is traditional and is also handy to represent American accents, the second one represents most modern British accents, especially of the younger generation. By having both symbols available, you can begin to see how we can exploit them for transcribing different acccents. Get used to using one of them. (13)

I downloaded one of the sound files from the web pages of CUP to listen to the "Southern England Standard pronunciation" of the word <lack>. My impression is that the vowel the speaker uses is best represented by the ash symbol. Listen for yourself, please:

I must say I'm not particularly happy with Tench's suggestion to choose either symbol. This is too confusing for beginners. What's more: "Get used to using one of them" - what does this request entail? Toss a coin and whatever the result will be, use that symbol for the rest of your transcription career? And: If learners are granted this freedom in the case of "æ/a", why not with "e/ɛ" as well? Tench writes:

So, for comparative purposes, when, for instance, comparing the vowels of English and another language, or the vowels of two different accents of English, we need to keep the ordinary Roman letter <e>as the IPA symbol for the /e/ sound, and rely on the Greek letter epsilon, <ε>, as the IPA symbol for the /ε/ sound. (10)

The same arguments could be put forward to keep the symbols <æ> and <a> apart.


(There is more to come on this book in a future blog entry)

Paul Tench's transcription workbook - some observations, #1

I'd like to share with you some impressions of mine while going through the book.
 
The book is intended for both native and non-native speakers to be used as a training course in developing students' skills in transcribing features of English pronunciation. All the words and discourses are available online. 


The first 3 chapters are intended as a detailed introduction to broad transcription with lots of examples (maybe too many for most of my students, but I will test this). The accent which the examples are based on is termed Southern England Standard Pronunciation (or SESP for short), "a type of accent that used to be known as Received Pronunciation. But this term is no longer transparent in meaning [...]" (4). Most of you know, of course, that the accent so identified is neither restricted to Southern England nor is it a standard in the sense that other accents are non- or sub-standards. Calling it a standard is based on the author's observation "that it is recognized as a form of pronunciation that is typically used by those who professionally engage in public speech, people like newsreaders" (4). Well ... I've heard a lot of professionals in Southern England speaking publicly with an accent which is not very close to RP. But let's not split terminological hairs. We have quite a lot of terms available substituting (or offering to substitute) Received Pronunciation - PSP, NRP, GB ... - so that we actually are in no need at all of yet another one. But as the right to freedom of speech still holds within the realm of the United Kingdom, "suffer it to be so now".

Chapter 4 goes into narrow transcription, most of which I shall skip in my phonetics classes. Due to lack of time I shall also ignore chapter 5: it deals with various accents. Much more important for my clientele are chapters 6 and 7 dealing with assimilation, elision, rhythm etc. I'm afraid I can't go into chapters 8-12 because a single semester is simply too short.

Chapter 1 introduces the IPA symbols for vowels, first the short vowels /ɪ, ɛ, æ/a, ɒ, ʊ, ʌ/, next the long ones (mono- and diphthongs are treated together) and then the weak vowels /i, u, ə/.  Here they are again in chart form:

credit: Cambridge University Press

The DRESS vowel is transcribed as /ɛ/. The reason for using the epsilon rather than the Roman letter <e> is explained thus:

Some dictionaries use the ordinary Roman letter <e>, because it has a more familiar look; however, in IPA, <e> represents the sound in the German word [...] Tee and the French word thé, Italian té, Welsh tê; or in many an English accent a word like lake. That vowel sound is distinctly different from the vowel in leg. [...] So, for comparative purposes [...] we need to keep the ordinary Roman letter <e> as the IPA symbol for the /e/ sound, and rely on the Greek letter epsilon, <ε>, as the IPA symbol for the /ε/ sound. (10)

We are also informed how to draw the 'ash' or the printed 'a' symbol - students will like it:

credit: CUP

As the author introduces each vowel, he points out peculiarities in the pronunciation of individual words, e.g.

• the short vowel /ɒ/ is not used in most American accents;
• the past tense form of eat "in a British accent is usually /εt/" (10);

This is not fully corroborated by the results of the opinion polls published in LPD3 as this graph shows:

LPD3

credit: Pearson-Longman

Chapter 2 presents the 24 consonants - again pointing to individual words with a pronunciation that may come as a surprise to some learners, e.g.

• silent letters as in tomb, sandwich, gnome, debris;
• homographs such as freeze-frees; 

Chapter 3 deals with word stress. It's a tiny one - just 3 pages and an additional 9 lines. A distinction is made between primary, secondary and non-stress and the IPA symbols for the first two stress types are introduced. A special section is devoted to compound words such as blackbird or railway station.

Chapter 4 deals with allophonic variations first among consonants and second among vowels. Aspiration, glottal reinforcement, flapping, devoicing, fronting, backing and rounding of consonants are explained and exemplified. As regards vowels the reader is familiarised with the processes of nasalisation, clipping, breaking, smoothing and diphthongisation.

Skipping chapter 5 I now turn to chapter 6 which is entitled 'phrases'. Topics such as assimilation, elision, epenthesis and liaison are covered. Most of these processes are employed by native speakers of English when they engage in informal, colloquial conversation. My advice is: If non-native speakers of English want to sound less 'bookish', they should incorporate several of these processes into their ordinary, colloquial, non-formal speech. Employing these processes is an option, however, not a must.

Paul Tench

credit: Cardiff University

Chapter 7 is titled 'rhythm'. Oh yes! Very important if you want to avoid the impression of using 'machine-gun English'! The chapter is about weak forms and about syllable elision in lexical items such as February, family, police, or in phrases like for instance, for example.

I highly recommend this book to my students and anyone else interested in developing her/his transcription skills. But mind you: Reading this book or doing its exercises does not turn you into a native speaker of English. But it helps you become aware of things that so far may have escaped your attention.

accents of English - an evaluation study of 2007

Nikolas Coupland

credit: Cardiff University

Nikolas Coupland and Hywel Bishop wrote an article in 2007 on "Ideologised values for British accents", which was published in the Journal of Sociolinguistics 11.1: 74-93. Thirty-four different accents of English were evaluated by 5010 UK persons in terms of adjudged levels of social attractiveness and prestige.

The study to be commented on here is of the overt¹ questioning type,
i.e. in this study people are directly and openly questioned about their beliefs as regards language varieties. Informants give scaled responses to such varieties. In the study at hand, data were collected on people's evaluations of 34 accents of English including non-native accents. Unlike in some of the studies by Giles and collaborators in the 1970s, the 5010 informants of this web-based study were not presented with spoken guises of accents but with labels such as 'Birmingham accent', 'Queen's English', 'English with a German accent' etc. The main focus was on these two questions (i.e. semantic dimensions):

1. How much prestige do you think is associated with this accent?
2. How pleasant do you think this accent sounds?

The informants had to evaluate the 34 accents on a seven-point scale with 7 being the maximum value and 1 the minimum. Mean ratings were calculated as well as rank orderings (1 being the highest and 34 the lowest rank) for each of the two semantic dimensions. It is the overall rank orderings that I will concentrate on.

As regards prestige Queen's English ranks highest (= 1) followed by Standard English (= 2), the accent identical to the one of the informant (= 3), Edinburgh (= 4) and Scottish (= 5). The accent with the lowest rank in prestige is Birmingham (= 34); less low in prestige are Asian (= 33), Black Country (= 32), Liverpool (= 31) and Afro-Caribbean (= 30). To have a German accent in English puts you on rank position 23 on a par with Newcastle.

Standard English gets the highest rank for social attractiveness, followed by the accent identical to the informant's own accent; next come Southern Irish, Scottish and Edinburgh. Bad luck, on the other hand, for Birmingham (= 34 again), Black Country, German, Asian and Liverpool (= 30).

The upshot of my report is:
ʧøːmən ʔiŋliʃ ʔis soːʃəli ʔinʔɛtrɛktif
Dear German university students of English: Take these results to heart, will you?

¹The opposite is a covert study in which the researcher tries to reveal tacit opinions and preferences held by the informants.

new favicon

Dear blog followers,

I've added a new favourite icon (= favicon) to my blog design. Here it is:


BTW:No mention is made of the pronunciation of the blend favicon in either LPD3, ODP or EPD17.

Kraut, OAP

Please, be informed that as of today Kraut is an OAP. Not the rosiest of futures ...

credit: Kristian Sekulic - Fotolia