1. PART ONE. VARIOUS ASPECTS OF RP
1.0. Origin of the term ‘RP’
The term ‘RP’ stands for ‘Received Pronunciation’ in which ‘received’ means, as in Victorian English, ‘accepted/approved’ (as in received wisdom or received opinion). The term ‘received’ was used by John Walker (1732‑1807) for the first time in 1774, and also by him in 1791 in reference to English pronunciation, and subsequently in 1818 by Peter Stephen Du Ponceau (born Pierre-Étienne du Ponceau) (1760‑1844) who was a French-American linguist, philosopher and jurist. The term ‘received pronunciation’ (note, not ‘Received Pronunciation’) was coined by Alexander John Ellis (1814‑1890) in 1869. The terms ‘Received Standard English’ or ‘Received Standard’ (as opposed to ‘Modified Standard’) and ‘Public School English’ were proposed in 1927 by Henry Cecil Kennedy Wyld (1870‑1945) to designate the kind of pronunciation which was generally regarded the best, enjoyed the widest currency, and used by the superior class in Great Britain. The two terms, ‘received standard’ and ‘received pronunciation’, were used interchangeably by Ellis who did not believe in a uniform educated pronunciation of English.
The coinage of the term ‘RP’ is in our days often attributed to Daniel Jones (1881‑1967). Actually, as has been seen above, the term had an earlier genesis. In the 1st edition (1917) of English Pronouncing Dictionary (henceforth EPD), Jones employed the term ‘Public School Pronunciation’ which he abbreviates to ‘PSP’ (p. viii, § 8). In the Introduction, Jones says that by Public School Pronunciation is meant ‘the pronunciation used by a considerable number of cultivated Southern English people … in ordinary conversation.’ All three adjectives, viz. cultivated, Southern and English, are of equally crucial importance in defining ‘public school pronunciation’. In Ripman’s “Editor’s Preface” provided in the 1st edition of EPD, Walter Ripman (1869‑1947) mentions ‘a certain type of English speech at the beginning of the twentieth century’ (p. v) and further on, that ‘this form of speech is very widely used by educated people in Southern England … the “great public schools” where this kind of speech prevails; and the influence of these schools has been a very great … factor in rendering acceptable what Mr. Jones calls “public school pronunciation”.’ (p. v). In the 3rd edition (1926) of EPD and in all subsequent editions, however, Jones replaced term ‘Public School Pronunciation’ by the different term ‘Received Pronunciation’ (RP) which caught on.
In explaining why EPD elects to record the specific type of English pronunciation called PSP (or RP as later called), says:
… because it happens to be the only type of English pronunciation about which I am in a position to obtain full and accurate information.
The choice of RP as the type of English pronunciation to be recorded has indeed been followed up in EPD (15th, 16th, 17th & 18th editions) and also LPD (1st, 2nd & 3rd editions).
That the term ‘RP’ is very well known and its genesis is generally attributed to Jones is undoubtedly thanks to the fact that the term occurs in a number of Jones’s works on English phonetics which were/are widely read by those who are interested in British English pronunciation. Such works by Jones are among others his EPD, An outline of English phonetics (OEP) and The pronunciation of English. Also to be mentioned as equally relevant in this connection is Ward's The phonetics of English. All these works make it a practice of mentioning and explaining what RP is.
RP is frequently referred to by many men in the street as e.g. ‘Queen’s English’, ‘Oxford English’, ‘BBC accent’, ‘BBC pronunciation’, ‘public school accent’, etc. These epithets seem, however, with the exception of ‘public school accent’, to be mostly misnomers.
Bet that as it may, according to , ‘RP is often taken to have existed for a relatively long time, evolving from a prestigious accent well established in England by the 17c …’. For succinct accounts of ‘public school English’ and ‘public school pronunciation’, see .
1.1. RP as a supra-regional pronunciation
One point that Jones is at pains to emphasize is that, although essentially based on the pronunciation heard among educated British speakers in Southeast England including London, the geographical provenances in Great Britain of individual speakers with RP are largely indeterminable. In other words, RP is a non-localizable pronunciation, i.e. a supra-regional or region-neutral (or region‑less) pronunciation. As for the correlation between social-class variation and regional variation, see Fig. 2 (Relation between social and regional accents in England) in where, in a pyramid-shaped model, RP is posited at the apex where the non-existence of regional pronunciation variation and the highest social class meet. A page-long explanation is then provided by Wells. As Wells emphatically notes, all references to different types of English pronunciation in England have social and regional connotations, except in the case of RP. Note that RP is never described as a ‘classless’ accent, because it identifies the speaker as a member of the middle or upper classes’ ().
Wells is neither the first nor the only phonetician to resort to a pyramid-shaped model to explain the relationship between regional differences and different accents in British English. More than a half-century earlier, presented the correlation by making use of a cone-shaped figure instead of a pyramid-shaped figure. The major difference between Wells and Ward is that while RP is posited at the apex by Wells, it is posited by Ward at a level somewhat lower than the apex. Ward says (op. cit: 7):
At a section still nearer the apex (at b), the pronunciation will have lost all easily noticeable local differences. This type of speech may be considered as that most generally understood throughout the English-speaking world and may be called Received Pronunciation.
Writing in 1951, according to ,
English people are divided, by the way they talk, into three groups; first, R.P. speakers of Standard English ―those without an accent; second, non‑R.P. speakers of Standard English ―those with an accent; and third, dialect speakers.
I agree with when he says that RP, which is an accentless pronunciation (in England), is as much an accent as any other (to Americans, Canadians, Australians, etc.).
Also, says that ‘R.P. is a privileged accent; your social life, or your career, or both, may be affected by whether you possess it or do not.’ He mentions the term ‘accent-bar’ and compares it to ‘colour bar’. It matters in England whether a native English speaker’s pronunciation is on this side of accent-bar, and if so it is RP, but on the other side of it, then. it is any of all the other accents. Those who do not speak with RP are at a social disadvantage. R.P. is a sociolect in the matter of pronunciation. He says () that ‘Until all regional educated accents are genuinely felt in England to be socially equal with R.P., these consequences of the accent-bar will persist.’
1.2. RP spoken by people who were educated in Public Schools
According to Jones, RP is heard irrespective of whether speakers of RP hail from Southern England or other parts in Great Britain. In other words, RP is a product of public schools where it is nurtured. Jones says as follows in .
The pronunciation represented in this book is that which is I believe to be very usually heard in everyday speech in the families of Southern English people who have been educated at the public schools [i.e., the boarding schools] …
, giving a brief introduction to what RP, echoes what Jones says.
Thus, it can shortly be said that RP is typically spoken by (i) a Southerner (an Englishman who hails from Southern England) who has been (ii) educated at a public school. Jones himself perfectly meets such conditions. Jones writes about himself and gives some interesting personal details as follows.
Readers [of EPD] may like to know that my father and mother were both Londoners, and that I have lived all my life in or near London. I was educated at Radley College (Abingdon), University College School (London) and King’s College, Cambridge.
But what if a boarder (a boy) in a public school fails condition (i)? There is hope provided he meets condition (ii). According to Jones,
This pronunciation [RP] is also used (sometimes with modifications) by those who do not come from Southern England, but who have been educated at these schools … If a boy in such a school has a marked local peculiarity in his pronunciation, it generally disappears or is modified during his school career under the influence of the different mode of speaking which he hears continually around him; he consequently emerges from school with a pronunciation similar to that of the other boys.
The successful nurturing of RP on the part of the boys who go through continual education at public schools in their formative years is certainly remarkable. :
This result is perhaps due to the boarding-school system of education prevalent in England. For centuries past, boys from all parts of Great Britain have been educated together in boarding-schools.
is of the view that RP is a unique phenomenon in England, not found in other countries, because it is something cultivated and transmitted from generation to generation through the public schools which are unique in the U.K.
1.3. RP is frequently associated with upper class / upper middle class
One particular point about RP which Jones himself does not overtly mention but which is widely mentioned by subsequent phoneticians is that RP is associated with the upper class and upper middle class. This is about the question of ‘social class’ (and educatedness) of speakers with RP. In other words, RP functions in British society as a phonetic sociolect or as a phonetic badge of social distinction (and education). This relates also to the question of ‘standard pronunciation’. unambiguously puts it when he writes as follows.
There is no doubt that R.P. is a privileged accent; your social life, or your career, or both, may be affected by whether you possess it or do not … I believe it is not putting it too strongly to say that in all occupations for which an educated person is required, it is an advantage to speak R.P., and it may be a disadvantage not to speak it.
This is not surprising. Public schools are free-paying, invariably expensive, endowed schools; they are not funded by public taxes. Parents or relatives who are capable of sending their young boys for half a dozen continuous years’ education at public schools are perforce sustainably wealthy. These providers of the enormous fees to public schools are businessmen, bankers, medical consultants, military generals, bank executives, company presidents, chief executive officers (CEOs), authors, inventors, real state developers, top politicians, insurance agents, surgeons, etc. They themselves are most probably already speakers of RP.
Having completed their education at public schools at the age of 18 and being speakers of RP, many of our young adults will perhaps proceed to Cambridge University, Oxford University, London University, etc. to further their education by pursuing academic careers. Others may seek employment in various business organizations and, it is here that having RP under the belt proves to be an advantage, in interview situations to begin with. As a certain type of linguists (known as sociolinguists) often tell us, RP (rather than local dialectal pronunciations) often give interlocutors a positive impression such as trustworthiness, intelligence, persuasiveness, confidence, reliability, proactiveness, competence, creativeness, capability, and so on.
I have mentioned upper class/upper middle class in terms of their sustainable financial reserve. But it can also be seen that public schools produce those personages which go on to be what can be regarded as forming upper class in that they become military officers and top politicians. Some British cabinet ministers were/are educated at public schools.
The stamp of social prestige and educatedness often attributed to speakers of RP is summarily put in the following words by:
RP is closely associated with education and with the kind of higher social position and responsibility often associated with it … its [RP’s] speakers occupy positions of authority and visibility in English society (government and politics, cultural and educational life, business and industrial management) far out of proportion to their actual numbers.
1.4. RP is not homogeneous
So far I have said nothing about whether RP is homogenous, uniform, or monolithic, that is, whether the type of English pronunciation known as RP is not susceptible to some variation in the usage of many speakers of RP. It would indeed be surprising if RP continues to be homogeneous throughout all these years. Every language changes while being used, in fact it changes every moment. As rightly says, ‘the sounds of the language always being in process of change, there have always been at any one time disparities between the speech sounds of the younger and older generations’. Indeed RP, which has been alive many years is no exception. That RP is not homogeneous is generally accepted by phoneticians.
provides a long note in small print (almost three-quarters of a page long) to say with a number of relevant examples that RP is not absolutely uniform. I will quote here just the first paragraph.
Although those who use RP have much in common in their speech, it must not be thought that RP is absolutely uniform. Quite a number of variations are to be found in it. For instance, the qualities of sounds used in some words vary from speaker to speaker … And in the case of some words two distinct pronunciations must both be considered as belonging to RP …
suggests the following three sub-types of RP, viz. (1) conservative, (2) general, and (3) advanced, and explains them one by one as follows; ‘the conservative RP [is] used by the older generation and, traditionally, by certain professions or social groups; the general RP [is] most commonly in use and typified by the pronunciation adopted by the BBC; and the advanced RP [is] mainly used by young people of exclusive social group ― mostly of the upper classes, but also, for prestige value, in certain professional circles.’ Gimson suggests the advanced RP may move in the future towards general RP.
In , gives a concise account of different types of English pronunciations prevalent in Great Britain over the few recent centuries. He shows a historical perspective of these different pronunciations and how RP gradually came into existence. One particular type of English pronunciation used in the southeast of England where the Court continued to sit came to acquire a status of social prestige. The establishment of the Public Schools and the rise of the upper classes have aided and strengthened RP as a social badge of prestige. Gimson reiterates the classification of RP into three types, (1) conservative, (2) general, and (3) advanced, shown in as we saw above. He gives a few examples of the vowel as in the word home pronounced differently in conservative RP ([ou]), in general RP ([əʊ]) and in advanced RP ([ɜːʊ] or even [eʊ]).
uses the term ‘Modern Received Pronunciation’ (‘MRP’) and attributes the term to , though without mentioning the relevant page number. does in fact use the term ‘Modern RP’ but not ‘Modern Received Pronunciation’. Ashby mentions the term ‘Modern Received Pronunciation’ synonymously with ‘Standard Southern British English’ (‘SSBE’) which I assume to be synonymous with ‘SSB Pronunciation’, the term I am using in the title of my present paper. I understand from this that MRP/SSBE is not to be equated with RP.
In order to know what, much later as in our days, is the situation about the non-homogeneity of what is known as RP, one may turn to . He speaks of ‘Conspicuous General British’ (CGB) which he says (op. cit.: 81) is
that type of GB [General British] which is commonly considered to be ‘posh’, to be associated with upper-class families, with public schools and with professions which have traditionally recruited from such families, e.g. officers in the navy and in some army regiments.
CGB can be thought to correspond to what called ‘Conservative RP’. The term ‘GB’ which stands for ‘General British’ need not bother us for the moment. Going into a discussion about ‘GB’ just at this moment in this paper will take us away from the main point about the question about non‑homogeneity of RP. I only need to say here that ‘GB’ (‘General British’) has been proposed by some English phoneticians and adopted by many other English phoneticians as a new term with which to replace the term ‘RP’.
Cruttenden notes that the number of speakers of CGB has considerably diminished over the years and that these speakers are elderly people. This is probably something like 3 % which the proportion of speakers of RP is often said to be. He further notes ― this is frequently mentioned by many modern English phoneticians ― that CGB arouses unfavourable reactions on the part of users of non-CGB speakers who regard CGB sounding affected. gives some examples of a few phonetic features (mainly certain vowels) in CGB that give rise to such negative reactions. Some such examples are word-final vowel in e.g. city, happy, fully, etc. pronounced fairly open [ɪ̞] (without being confused with [e]); a word-final vowel in bitter, here, sure, etc. pronounced fairly open [ə̞]; [ɜː] in e.g. burn, occur, certain, etc. being pronounced opener as [ɜ̞ː]; and [æ] in mad, matter, etc. diphthongized as [ɛə], and so forth. Cruttenden says that these particular vowels in CGB result from the fact that ‘the tongue and jaw position are more open in CGB than in mainstream GB’. This is clearly a question of ‘articulatory settings’, though Cruttenden happens not to mentioned it just here. He does mention ‘articulatory settings’ (op. cit.: 302) but in connection with GB, not in connection with CGB.
One other type of RP in our days that mentions is what some English writers call ‘Modern RP’. (This term is not Cruttenden’s own coinage.) Cruttenden emphasizes that a lot of changes have happened to RP during the latter half of the 20th century, i.e. from 1960’s onwards. This situation allows the currently prevailing ‘RP’ to be called ‘Modern RP’ in which main users of RP like radio and television newsreaders, pupils at boarding public schools, (ex‑)army officers, (ex-)navy officers, etc. have gradually become somewhat relaxed. Relaxation in the requirement on the BBC’s Third Programme (1946‑1967) and its successor Radio 3 (1967‑) has seemed, however, to be a relatively slow process; this may partly be because Radio 3 specialize/s/d various forms of highbrow cultural programmes. In this sense Radio 3 distinguishes itself from Radio 1, Radio 2, Radio 4, etc. There is also another factor. The BBC has gradually adopted an inclusive policy whereby ‘incursion’ of some non‑RP speakers is being gradually allowed in, even in Radio 3, as I write these lines.
It would be fair to say that what Gimson called ‘general RP’ and ‘advanced RP’ have now amalgamated to become what some call ‘Modern RP’. Nevertheless, ‘Modern Received Pronunciation’ stands away from various non-RP features.
1.5. RP as a standard pronunciation?
says that
I am not one of those who believe in the desirability or the feasibility of setting up any one form of pronunciation as a standard for the English-speaking world. [my italics]
And also in that
RP means merely ‘widely understood pronunciation,’ [Jones’s single quotation marks] and that I do not hold it up as a standard which every one is recommended to adopt.
However, referring to early dates, says:
Three books by Jones [] … established the term ‘Received Pronunciation’ as representing standard [my italics] spoken British English.
Naturally there is a difference between RP not being intended to be a standard pronunciation and it actually being it.
It cannot be denied that, as puts it:
It [RP] enjoy[s] overt prestige [Wells’s block letters] … Such an accent [i.e. RP] is a de facto standard [Wells's block letters]. It is considered appropriate for public use ― a suitable accent for a newsreader, an ambassador, a classical actress, a barrister, a general, a society hostess.
Wells (op. cit.: 117) further says that ‘Occupations [of users of RP] perhaps most typically associated with RP are barrister, stockbroker, and diplomat.’
As RP is a widely understood English pronunciation (rather than other types of English) not only in Great Britain but even in many other parts of English-speaking world, and as RP possesses a maximum degree of intelligibility in verbal communication in English, would it not be reasonable to regard RP as a highly sought-after desideratum as a standard English pronunciation? In short, the factor that allows RP as a standard pronunciation is its widely recognized intelligibility. A language, English in the present case, is a verbal instrument of communication which should assure intelligibility to both speaker and listener.
Note, however, that, writing in 1951, does not favour the very existence of RP when he says ‘The existence of a standard pronunciation such as R.P. is, I suggest, a bad rather than a good thing. It is an anachronism in present-day democratic society.’
Whatever the argument may be that RP is a ‘standard pronunciation’ or rather ‘standard English pronunciation’, we should take note that says that ‘Other cities [than the capital city of the Commonwealth and its surroundings], notably Edinburgh and Dublin, have their received pronunciations … The English language now has unity in diversity. It has more received pronunciations than one’ and also that says ‘… Received Pronunciation [Howard’s block letters], which is supposed to define ‘educated speech’ in the south of England. Alongside RP, educated Scottish, Irish and Welsh have an equal status [as RP].’
1.6. RP under criticism
RP has come under much criticism since the middle of the 20th century, principally from the majority of English phoneticians. RP is variously described by many present-day English phoneticians as ‘archaic’, ‘old-fashioned’, ‘outdated’, ‘comical’, ‘amusing’, etc. RP‑bashing is still raging to this day. It must be taken as true that RP, especially Conservative RP, provokes unfavourable emotional reactions from those who do not use it. Such reactions are rarely aroused among non-native English speakers including myself, so far as my own experience goes.
It must be noted that there are other strong reasons for disliking RP as explain: ‘Many British people dislike Received Pronunciation, usually arguing that it is a mark of privilege and (especially among the Scots, Northern Irish, and Welsh) of social domination by the (especially) English.’
1.7. RP and the BBC
John Charles Walsham Reith (1889-1971), Director-General of the BBC, established the Advisory Committee on Spoken English in 1926 and adopted RP to be prescribed to its newsreaders and announcers in both its home service and external service. One of the Committee members was Daniel Jones. There is no denying that the BBC’s programmes which were broadcast in RP in both its domestic and external services have certainly played a substantial role in disseminating and maintaining RP worldwide over many years. At the same time RP has enjoyed a high degree of intelligibility when heard both at home and abroad. All this has, however, changed in our days. The requirement of RP to be prescribed on Radio 3 (which service replaced in 1967 the former BBC Third Programme which had been launched in 1946) and on the BBC World Service continued till 1989. This BBC policy ended when the World Service relaxed its policy on its newsreaders and announcers, and also when the policy of ‘inclusiveness’ was introduced at the BBC as a whole, bringing in the use of wide-ranging pronunciations. This new tendency is apparent even on Radio 3 which is largely perceived by listeners to be a 'high-brow' radio service which is easily associated with RP.
1.8. How many Britons speak RP?
The amount of Britons speaking with RP is very small, said to be 2 % or 3 %, and is expected to progressively decline in coming years. This may be the case with what calls ‘Conspicuous General British’ (CGB). The fact is, however, that RP is still very much in evidence.
I will merely quote here a few lines from which, already as far back as 70‑odd years, Jones wrote relevantly.
The fact that RP and approximations to it are easily understood [Jones’s italics] almost everywhere in the English-speaking world does not mean that RP is used [Jones’s italics] by a majority of English-speaking people. On the contrary, it is used by a rather small minority. There are countless other ways of pronouncing English in existence, some of them being used by large communities.
1.9. RP’S intelligibility
RP is intelligible not only in Great Britain but in wider areas in the world as well. Jones places particular emphasis on RP’s intelligibility, which is lacking in various measures in other types of English pronunciation, namely in a number of regional pronunciations in Great Britain. RP is intelligible not only in Great Britain but also in all English-speaking areas of the world, and beyond. Here precisely are Jones’s own words taken from .
I believe that the pronunciation above referred to [RP] is readily understood in many parts of the English-speaking world … this form of speech [RP] is more widely understood with ease in Great Britain than any other forms would be.
Be that as it may, also agree with the common belief (I certainly hold this belief) that RP ‘is the most widely intelligible spoken form of Standard English …’. It is for this reason that non‑English-speaking foreign desirous of learning English pronunciation is best recommended to choose RP as the model to adopt.
I am firmly of the view that arguably the most important feature of RP is its ‘intelligibility’ in English-speaking areas world-wide and, even in many non-English-speaking parts of the world. However, a contrary view exists. Here in full is what Gramley () writes.
The intelligibility argument
When, early in the century, Jones chose RP as the basis for the description of English pronunciation, one of his arguments was that ‘RP and approximations to it are easily understood [G’s italics] almost everywhere in the English-speaking world’ () [i.e. ]. It is certainly true that RP is frequently heard in the media and is therefore easily accessible to many students of English as a foreign language (EFL). Furthermore, familiarity helps to guarantee comprehensibility. Yet the intelligibility argument should not be overvalued: in the words of Trudgill [i.e. Trudgill 1975: 53], ‘Differences between accents in the British Isles are hardly ever largely enough to cause serious comprehension difficulties. In addition, it is conceivable that people in parts of the world where RP is not familiar (particularly in the sphere of influence of AmE) might find RP less intelligible than GenAm.
I personally find it difficult to follow either Gramley or Trudgill in the above quoted passage. I can mention four points. Firstly, Trudgill seems to minimize degrees of difficulty in comprehensibility among native speakers of (British) English due to their different accents. Difficulties in comprehensibility caused by some non-RP pronunciations such as ‘Brummie’, ‘Geordie’, ‘Scouse’ and ‘Cockney’ are well known, yet they are all native speakers of British English. Secondly, it is not my observation that foreigners, be they students or otherwise, seem to experience difficulty comprehending what is said in RP. Such difficulties in comprehensibility are frequently experienced by foreign speakers/learners when faced with such non‑RP as mentioned just above. Thirdly, I have abundantly been in contact with people who are native speakers of GenAm or foreign students who speak GenAm or its approximations, and these people seem to have no difficulty comprehending what is said in RP. Fourthly, it seems that RP-speakers have no difficulty understanding what is said in GenAm. It must not be forgotten that British television channels broadcast an enormous amount of programmes made in the U.S.
write that ‘the RP accent is probably rather more difficult for many foreigners to acquire than, say, a Scottish accent, since RP has a large number of diphthongs.’. I can only vigorously disagree with them who opine that a large number of diphthongs in RP can be a stumbling block for foreigner in learning RP.
1.10. RP’S Intelligibility as shown in an Opinion survey
One of the undeniable characteristics of RP is its high degree of intelligibility, whether the interlocutors be native or non-native speakers of English. RP’s intelligibility is, as we have seen, emphatically mentioned by Daniel Jones himself.
In an opinion survey aimed at 4,000 Britons which was conducted in 2013, three out of five respondents (62 %) replied that Received Pronunciation was the most intelligible type of English pronunciation compared with nine other chosen types of English pronunciation.
The survey was conducted by ComRes (a member of the British Polling Council) concerning ‘intelligibility of RP’ on the part of Britons living within the UK (England, Wales, Scotland, Northern Ireland). Respondents (whose number is unknown to me) were asked: ‘How intelligible or unintelligible do you find the following accents to be?’. Apart from RP, nine other accents were chosen for the purpose of the survey, viz. Edinburgh, Devon, Belfast, Cardiff, Manchester, Newcastle, London (Cockney), Birmingham and Liverpool. The results of the survey were originally published in a single table (not reproduced here) under eight heads, namely (i) ‘very intelligible’, (ii) ‘fairly intelligible’, (iii), ‘neither intelligible nor unintelligible’, (iv) ‘not very intelligible’, (v) ‘not all intelligible’, (iv) ‘don’t know’, (vii) ‘intelligible’, and (viii) ‘unintelligible’. I prefer to present the results in four constituent parts (named ‘four Sub-tables’) of the original single table. They are as follows:
-
‘Sub-table 1’ consists of ‘very intelligible’; ‘fairly intelligible’, and ‘INTELLIGIBLE’ ;
-
‘Sub-table 2’ consists of ‘not very intelligible’, ‘not all intelligible’ and ‘UNINTELLIGIBLE’;
-
‘Sub-table 3’ consists of ‘neither intelligible nor unintelligible’, ‘don’t know’, and N/A;
-
‘Sub-table 4’ consists of ‘INTELLIGIBLE’, ‘UNINTELLIGIBLE’, N/A and TOTAL.
-
‘INTELLIGIBLE’ refers to the percentage which is equal to the addition of the percentages of (i) and (ii).
-
‘UNINTELLIGIBLE’ refers to the percentage which is equal to the addition of the percentages of (iv) and (v).
The results of the survey were shown below in percentages which I worked out on the basis of the figures given by the pollster, representing the number of respondents who answered the different questions. Indeed no information was available about the total number of the respondents who were involved in the survey.
The results of the survey shows varying degrees of intelligibility of each of the ten types of pronunciation, i.e. each of the ten different accents. I will show below the results of the survey in four separate sub-tables, viz. Sub‑table 1, Sub‑table 2, Sub‑table 3 and Sub‑table 4.
Sub-table 1 is concerned with INTELLIGIBILITY of 10 types of English pronunciation.
As can be seen, the percentage (e.g. 31 % for RP) for ‘very intelligible’ and the percentage (e.g. 32 % for RP) for ‘fairly intelligible’ are combined to show the total percentage (63 % for RP) which is the highest percentage. In other words, sub‑table 1 indicates that RP is the most intelligible type of English pronunciation. None of the rest of 9 other different types of pronunciation achieve any percentage higher than that for RP either for ‘very intelligible’ or for ‘fairly intelligible’.
It is not surprising that Newcastle, London (Cockney), Birmingham and Liverpool should turn out to be a group of 4 types of pronunciation which only achieve the low end of percentages. The percentages achieved by the 4 types of pronunciation are, in the order of the types of pronunciation mentioned, 4 %, 3 %, 3 %, 3 % for ‘very intelligible’, 15 %, 14 %, 12 %, 12 % for ‘fairly intelligible’, and finally 19 %, 18 %, 15 %, 15 % for INTELLIGIBLE. I say ‘not surprising’ because the 4 types of pronunciation in question, known respectively as ‘Geordie’ (Newcastle), ‘Cockney’ (London), ‘Brummie’ (Birmingham) and ‘Scouse’ (Liverpool) are highly distinctive and may create variable amounts of difficulty for outsiders to understand.
Here now is Sub-table 2.
There remain the percentages, respectively, for ‘neither intelligible nor unintelligible’ and ‘don’t know’. The percentage of 3 % for UNINTELLIGIBLE (= 2 % ‘not very intelligible’, + 1 % ‘not all intelligible’) for RP baffles me, unless these responders happen to be (then) recent immigrants to the UK without much previous contact with spoken British English. Even otherwise, it is expected that the majority of Britons understand what is said in RP. These are noncommittal percentage so that they cannot be classified as either ‘intelligible’ or ‘unintelligible’. The combined percentages for each of the different types of pronunciation, which the original table does not show, will all the same be indicated as N/A for what they are worth.
In Sub-table 2, the percentages for the 4 types of pronunciation we have seen and I have commented on in Sub-table 1 are again at the low end, but this time for ‘not very intelligible’ and for ‘not all intelligible’, but unsurprisingly the percentages themselves for the 4 types of pronunciation are quite high, 19 %, 25 %, 22 %, 24 % for ‘not very intelligible’, 7 %, 7 %, 11 %, 13 % for ‘not all intelligible’. The subtle difference in meaning between ‘not very intelligible’ and ‘not all intelligible’ eludes me, but the degree of unintelligibility is probably greater in the former than in the latter. Anyhow the meaning difference in question must be correlated with difference in the two groups of percentages. The difference between the two groups of percentages loses its significance when the two groups of percentages are merged for UNINTELLIGIBLE in which the four percentages, viz. 26 %, 32 % 33 %, 37 % are close to each other and are at the low end. The group of 26 %, 32 % 33 %, 37 % for UNINTELLIGIBLE and the group of 19 %, 18 %, 15 %, 15 % for INTELLIGIBLE are, so to speak, the two sides of the coin.
Here now follows Sub-table 3.
Here finally is Sub-table 4.
In Sub-table 4 the percentages for INTELLIGIBLE, UNINTELLIGIBLE and N/A are indicated again in order to show that the addition of the three percentages for each type of pronunciation does amount to 100 %.
We now undertake an analysis of what is meant by the contents of Sub‑table 1, Sub‑table 2, Sub‑table 3 and Sub‑table 4.
As seen in Sub-table 1, RP scores the top percentage of 62 % for being judged as being INTELLIGIBLE, with Edinburgh trailing far in the second place with 38 %. The degree of ‘very intelligible’ for RP, 31 %, again at the top, is distinctly and uncontroversially superior to the percentages for all the other types of pronunciation, all of which score percentages in single digits only, even with the case of Edinburgh, 8 %. ‘Fairly intelligible’ for RP scores a respectable 32 % even though all the other types of pronunciation garner relatively high double-digit percentages, led by Edinburgh with 30 %.
But what is crucially in favour of RP as being intelligible is the excessively low percentage of 3 % for RP as being UNINTELLIGIBLE. When it is reminded that 3 % for RP is a combination of 2 % for ‘not very intelligible’ and 1 % for ‘not all intelligible’, it seems to pale into insignificance. RP judged to be ‘neither intelligible nor unintelligible’ with 27 % is a little surprising. The much higher percentages (ranging between 52 % and 43 %) which are for all the other types of pronunciation, on the other hand, are more or less expected.
1.11. Who spoke/speak RP?
But who were/are those who spoke/speak English with RP? The names which are the oftenest cited of those few speakers, now all deceased, who spoke with archetypical Received Pronunciation, known as ‘Conservative RP’, traditionally include the following: Frank Phillips (1901‑1980, actor, BBC announcer), John Snagg (1904‑1996, BBC newsreader), and Alva Liddell (1908‑1981, BBC newsreader and announcer). Incidentally, apart from ‘Conservative RP’ with which the three individuals mentioned spoke, there are two other types of RP, known as ‘Mainstream RP’ and ‘Contemporary RP’. Thus RP is not of only one kind.
If I were asked to draw up a short list of some individual Britons who are still alive and who speak with RP, I would be tempted to come up with a tentative open-ended list such as the following. Anyone, Britons included, especially those non-native speakers of English, who may wish to know how ‘Mainstream RP’ sounds like, would do so by listening to speakers like the ones listed below.
Jim Al-Khalili (1962‑, theoretical physicist), David Attenborough (1926‑, naturalist), Michael Berkeley (1948‑, composer), Tony Blair (1953‑, politician), Harriet Cass (1952‑, former BBC newsreader), David Cameron (1966‑, former prime minister), Judi Dench (1934‑, actress), Mark Elder (1947‑, conductor), Stephen Fry (1957‑, actor), Charlotte Green (1956‑, former BBC newsreader), Jeremy Hunt (1966‑, politician), Boris Johnson (1964‑, former prime minister), Ben Kingsley (1943‑, actor), Jan Leeming (Janet Dorothy Atkins) (1942‑, former newsreader with ITV and BBC), Joanna Lumley (1946‑, actress), Trevor McDonald (1939‑, newsreader, journalist), Andrew Marr (1959‑, journalist), Theresa May (1956‑, former prime minister), Richard Miles (1969‑, historian), Helen Mirren (1945‑, actress), Simon Jonathan Seebag Montefiore (1965‑, author), Rageh Omaar (1967‑, journalist), Jeremy Paxman (1950‑, broadcaster), Michael Portillo (1953‑, broadcaster), Jacob Rees-Mogg (1969‑, politician), John Sergeant (1944‑, journalist and broadcaster), Jon Snow (1947‑, journalist), Simon Michael Schama (1945‑, historian), Alastair Stewart (1952‑, journalist), David Suchet (1946‑, actor), John Suchet (1944‑, broadcaster), Justin Welby (1956‑, Archbishop of Canterbury), John Christopher Wells (1939‑, phonetician), Rowan Williams (1950‑, former Archbishop of Canterbury), Nicholas Witchell (1953‑, journalist), et al.
The people whose names I listed above happen to be celebrities in various fields in the UK, but I could go on adding large numbers of others who happen not to be celebrities but who nonetheless speak with RP and consequently qualify to appear in an open-ended list such as shown above. As long as suchlike people are alive, it would be fair to assume that RP also remains just as alive, in spite of what some would assert to the contrary.
It is easy to see that RP is not dead as I write these lines.
An additional short list, also open-ended, could be drawn up of some other people who deceased comparatively recently and who typically spoke with RP.
David Abercrombie (1909-1992, phonetician), George Alagiah (1955-2023, newsreader), Richard Baker (1925-2018, BBC broadcaster), Tom Crowe (1922-2010, BBC newsreader), Richard Dimbleby (1913-1965, journalist & broadcaster), Robert Dougall (1913‑1999), Queen Elizabeth II (1926‑2022), Margot Fonteyn (1919-1991, ballerina), Stuart Hibberd (1893-1983, BBC newsreader), Patricia Hughes (1923-2013, BBC newsreader), James Robertson Justice (1907-1975, actor), Michael Hordern (1911-1995, actor), Deborah Kerr (1921-2007, actress), Christopher Lee (1922-2015, actor), Peter MacCarthy (1912-1979, phonetician), Derek Malcolm (1932-2023, film critic), Cormac Rigby (1939-2007, BBC radio announcer & Roman Catholic priest), George Sanders (1906-1972, actor), Alastair Sim (1900-1976, actor), Margaret Thatcher (1925-2013, stateswoman), Nicola Haywood-Thomas (1955-2023, broadcaster), et al.
1.12
writes that the ‘special position occupied by RP, i.e. educated Southern British English, has led to its being the form of pronunciation most commonly described in books on the phonetics of British English and traditionally taught to foreigners.’ Those, non-native speakers of English of any age in particular, who may wish to know the phonetic facts of RP in detail can consult a few books like the following. These books systematically and accurately provide rich information from the view of articulatory phonetics about all the segmental elements like vowels, consonants and semivowels and suprasegmental elements like stress, rhythm, assimilation, elision, length, syllable division, intonation, etc. The disadvantage about these books is that, as can be guessed from their publication dates, their availability on the market is very limited. Earlier editions of the books listed below, if available, are just as much beneficial. LPD could serve as a helpful backup for checking the pronunciation in RP of any individual English words.
1.13
A few books describing English pronunciation prevalent in late 20th Century onward are easily obtainable. I will only refer to the latest editions of the books.
An Introduction to the Pronunciation of English ()
Gimson’s Pronunciation of English ()
Practical English Phonetics and Phonology ().
An Introduction to the Pronunciation of English by Gimson went through four editions, viz. 11962, 21970, 31980 and 41989. In the Preface to the second edition, writes that ‘The changing status of RP, especially amongst young people, has necessitated the re-writing of the pages dealing with this matter …’. The Preface to the fourth edition is written by (Rev Dr) Susan Ramsaran who says among others things () ‘… the tentative comments that I have felt able to add to Chapter 6 [The historical background and current changes] on current changes in RP are based on pronunciations observed and elicited in connection with lexicographical work carried out by Professor Gimson and myself.’
An Introduction to the Pronunciation of English () by Gimson can be said to have spearheaded a continuous series of books on English phonetics mainly in the UK following the long period of authoritative publications on English phonetics dominated by Daniel Jones and his disciples from early 20th century.
Gimson’s Pronunciation of English by Cruttenden continues from Gimson 41989 onward. It has so far gone through four editions, viz. 51994, 62001, 72008 and 82014, all these editions having undergone successive cumulative revisions by Cruttenden. In his Foreword to the sixth edition (62000), Cruttenden says (p. vii) that ‘The sections on the current state of RP and its variants (including Estuary English) have … been rewritten.’ Furthermore, in his Foreword to the seventh edition (), Cruttenden’s ‘a relatively flexible attitude to RP’ as he calls it is explained. I quote his words at length below.
… some form of RP remains the target for many learners. In view of the fact that this book is sometimes held to be presenting outmoded standard pronunciation, I must point out that, although RP is here considered a ‘reference accent’, the book presents a relatively flexible attitude to RP, including, for example, Regional RPs (i.e. RP with some widely acceptable characteristics of a particular region––see §7.3.(4)). With this flexible attitude, the percentage of speakers of RP cannot be claimed, as it often is, to be only in single figures; it is almost certainly much higher (there are no reliable figures anyway); In so far as RP is sometimes considered outdated and a figure of fun, the variety to which reference is usually made is one I call Refined RP (see §7.3(4)).
The Foreword to the eighth edition () has some interesting remarks by Cruttenden to note. First, he has completely abandoned the term ‘RP’ in favour of GB. This is what he writes (pp. xvi-xvii).
… I no longer regard the book as describing RP (Received Pronunciation). Despite the fact that I and other phoneticians have sought to describe changes in RP to make it a modern and more flexible standard, many, particularly in the media, have persisted in presenting an image of RP as outdated and becoming even more than ever the speech only of the ‘posh’ few in the south-east of England. For this reason I have dropped the name RP and now consider myself to be describing General British or GB.
It took Cruttenden a very long time to finally discard the term ‘RP’ in favour for the term ‘GB’, considering that ‘GB’ was launched as early as 1972 by Jack Windsor Lewis in his A Concise Pronouncing Dictionary of British and American English. Cruttenden does not yet mention ‘GB’ anywhere in his 51994, 62001, 72008. ‘General American’ is mentioned in 72008 but not yet in 51994 and 62001. Collins & Mees & Carley also eschew the term ‘RP’ (unless in direct reference to ‘Conservative RP’) in favour of the term ‘GB’ (‘General British’).
1.14
It is generally agreed that, as I write these lines, RP is on the decline while GB (General British) is gaining ground. It is important to recognize that RP and GB are currently still in coexistence. It is also important to remember that both RP and GB are satisfactory instruments of verbal communication in English and that they are intercomprehensible. There have, over the years, inevitably occurred in RP some changes in the articulation of English vowel sounds; for example, in those of the diphthong in goad, old, coke, etc., the monophthong in bus, cuff, hut, etc., the monophthong in cool, food, tool, etc., and the monophthong in bad, cat, lag, etc. The difference between RP and GB is practically limited to the vowels. This is clearly seen if we compare the vowel system (RP) /iː i e æ ɑː ɔ ɔː u uː ʌ əː ə ei ou ai au ɔi ə ɛə ɔə uə/ () and the vowel system (GB) /i e æ ɒ ʊ ʌ ə iː ɛː ɑː ɔː uː ɜː ɪə ʊə eɪ aɪ ɔɪ əʊ aʊ/ (). As for the consonants, it is enough to compare the consonant system (RP) () and the consonant system (GB) () to see that the two consonant systems are virtually the same.
2. Part Two. Dynamic synchrony
2.0
We will first see the genesis of the concept of ‘dynamic synchrony’ which is importantly put into practice in investigating changes that take place over periods of time in the pronunciations of individual languages among their native speakers of different ages.
We are all agreed that languages change; in fact, they never stop changing. says with insistence:
… personne ne doute plus que les langues changent d’une façon continue … cette certitude est une grande conquête de la linguistique …
Here is the whole of the first paragraph from .
Les linguistes depuis fort longtemps ont mentionné le caractère mouvant des langues, et Saussure lui-même dans ses Ecrits de linguistique générale, plus précisément lors de sa deuxiè me confé rence à l’université de Genève en novembre 1891, note qu’ « il n’y a jamais … un équilibre, un point permanent, stable dans aucun langage », ce qui l’a amené à poser le principe de « la transformation incessante des langues comme absolu ». Il précise encore que « le cas d’un idiome qui se trouverait en état d’immobilité et de repos ne se présente pas » (). … L‘é tude proposée [par Feuillard] se centrera … sur la notion et les faits de dynamique synchronique dans l‘optique fonctionnelle, tels que je les conçois, à partir d‘exemples empruntés au français.
Martinet famously and aptly puts it,
and
2.1
We need to know what is meant by ‘dynamic synchrony’ since it is from the point of view of dynamic synchrony that the position of RP in our days should be assessed. Here are three relevant passages about dynamic synchrony taken from Martinet’s writings. The first passage to be quoted is from .
… il convient que la description soit strictement synchronique [Martinet’s boldface], c’est‑à‑dire fondée exclusivement sur des observations faites pendant un laps de temps assez court pour pouvoir être considéré en pratique comme un point sur l’axe du temps.
The passage just quoted is also found in II-2 in all previous editions, namely .
The first passage quoted above from is immediately followed by the passage to be quoted next.
Il n’est pas impossible, il est mê me recommendé, dans une étude synchronique de relever les tendances évolutives de la langue en opposant les usages de différentes générations en présence. On dira dans ce cas qu’il s’agit d’une synchronie dynamique [Martinet’s boldface]. On parlera de diachronie [Martinet’s boldface] lorsqu’on confrontera les synchronies dynamiques successives de chaque langue.
The second passage just quoted from are also found also in II‑2 in all previous editions, viz. , but not in . Even if the second passage as such is not yet found in , this does not mean that the contents of the second passage are not already conceived at the writing of Martinet (11960). The concept of ‘dynamic synchrony’ which is already inherent in the phrase ‘la description … fondée exclusivement sur des observations faites pendant un laps de temps assez court pour pouvoir être considéré en pratique comme un point sur l’axe du temps’ in is explicated in by showing how in practice observations of linguistic evolution based on the principle of dynamic synchrony should be conducted and by actually introducing the term ‘synchronie dynamique’.
The third passage is quoted from where dynamic synchrony is explicated.
Il peut donc être indiqué d’opposer, à l’étude diachronique visant délibérément à comparer différents états successifs du même objet d’étude, une synchronie dynamique [italics added] où l’attention se concentre, certes, sur un seul et même état, mais sans qu’on renonce jamais à y relever des variations et à y évaluer le caractère progressif ou récessif de chaque trait.
The key ideas expressed in the second quotation above from are ‘relever les tendances évolutives de la langue en opposant les usages de différentes générations en présence’ while the key ideas expressed in the third quotation from are ‘relever des variations et à y évaluer le caractè re progressif ou récessif de chaque trait.’ This is because Saussure’s concept of ‘synchrony’ as against ‘diachrony’ differs importantly from Martinet's concept of ‘synchrony’, or more precisely ‘dynamic synchrony’.
A broader statement by Martinet on dynamic synchrony is found in Martinet (1990), in which he discusses dynamic synchrony in connection with not only phonology but also grammar, lexis and semantics.
2.2
At this juncture, we need to look at Saussure’s opposition between ‘synchronique’ and ‘diachronique’ in ().
… on peut parler de la science des états de langue ou linguistique statique … Est synchronique tout ce qui se rapporte à l’aspect statique de notre science, diachronique tout ce qui a trait aux évolutions. De même synchronie et diachronie désigneront respectivement un état de langue et une phase d’évolution. [Saussure’s italics]
Saussure’s own well-known illustration of synchrony and diachrony ― a tree trunk cut longitudinally (diachrony) or transversally (synchrony) ― will be recalled here.
writes:
… ce qui a amené Martinet à dépasser la dichotomie diachronie / synchronie établie par Saussure et à poser le concept de synchronie dynamique, qui permet de rendre compte des langues dans leur fonctionnement.
Saussure’s synchrony is static synchrony, and his diachrony are successions of static synchronic stages. On the other hand, Martinet's concept of synchrony is what he calls ‘dynamic synchrony’. The fundamental difference between ‘static synchrony’ and ‘dynamic synchrony’ ― in other words, between ‘static’ and ‘dynamic’ – is patent. In static synchrony, a state of a language is viewed as a fixed (static) state of linguistic evolution at any given stage in the sense that the fixed state is ‘flat’, i.e. bidimensional while diachrony consists in successions of subsequent fixed states of linguistic evolution at subsequent given stages. However, in dynamic synchrony, a state of linguistic evolution is viewed as not flat but is tridimensional and has ‘depth’ (dynamic). Indeed emphatically says: ‘J’insiste pour qu’on donne une profondeur à la synchronie. Ce n’est pas plat, la synchronie.’ It is interesting that I had precisely these visions before reading Martinet’s passage just quoted. A dynamic state at any stage of linguistic evolution is not static, fixed and frozen, but it is dynamic and unfrozen in an agitating state like, so to speak, boiling water. It has depth. In dynamic synchrony one observes the co-presence of different linguistic features that are in competition with each other, and one examines which one of the features is recessive and obsolescent and which other is progressive and innovative.
A language is constantly changing. I wrote in the past () as follows.
Linguistic evolution is an inevitable consequence of the constant antinomy between communicative needs on the one hand and physical and mental inertia on the other on the part of the speaker. Since inertia is an immutable factor while communicative needs are a variable factor, the equilibrium between the two factors varies over time, resulting in language changes.
What I said just above is nothing but what Martinet himself emphatically and consistently states in his works. See e.g. , , , and .
2.3
1996 saw the publication of a book by Martinet entitled The Internal Conditioning of Phonological Systems. The first of the 16 chapters of the book is entitled ‘Language Dynamics and Diachrony’. The first (shorter) half of the chapter is concerned with ‘dynamic synchrony’. Given the location of the publisher (in India) and possible difficulty in accessing to the book, I think that it will be convenient and helpful to the readers for me to quote below some of Martinet’s passages found here and there in the book.
‘a language changes all the time’, and if it changes, it is because the needs of its speakers change, or else a language changes because it functions. (p. 4)
… the clue to the understanding of linguistic change lies in the observation of synchronic functioning which, through our drastic formula, has been identified as the dynamics of language. (p. 4)
As a functionalist … As a realist, I think the best description is not the most attractive projection of my prejudices upon observable facts, but an ordering of those facts which conform to their actual relationships. If, in the hierarchy of relevances, I place communication first, it is because I have found it to be really decisive. (p 4)
The practical problem for the linguist is of course, how we can get informed about the trends, or, if we use Sapir’s term, the drift at work in the language. (p. 5)
… the easiest approach to the dynamics of a language consists in gathering information about people belonging to the same language community, social stratum and cultural level, but of different age groups. (p. 6)
… necessary to recall that similar research has been carried out in America by William Labov who, however, has been tempted to stress the social factors more than the chronological aspect of the phenomena (p. 7)
… no one should venture into the field of explanatory diachrony without having first studied for years the dynamics of synchronic situations looked upon as the periods when people from different regions, classes and generations cooperate by means of a language which has to be considered one and the same spite of its infinite variety. (p. 9)
2.4
refers to a linguistic feature (phonological feature in this case) in French which is described from the standpoint of dynamic synchrony. French speakers born in Paris before 1920 maintain the phonological opposition /a/ (as in patte) vs /ɑ/ (as in pâte) whereas more than 60 % of those Parisians born after 1940 do not retain this opposition and pronounce both words with /a/, while 40 % of them continue to retain /a/ vs /ɑ/. In other words, /a/ vs /ɑ/ on the one hand and /a/ on the other are in coexistence but /a/ vs /ɑ/ is a recessive and obsolescent feature while the use of the /a/ only is an innovative and progressive feature. We thus take into account the phonological practice of two successive generations viewed synchronically.
2.5
An early trailblazing survey on pronunciation preference in which the concept of dynamic synchrony was fully put into practice is surely Martinet’s survey conducted in 1941 into spoken French of 409 subjects in Weinsberg in the state of Würtemberg, Germany. A questionnaire was distributed to the subjects with a view to finding out the phonological systems of the individual subjects. The findings of this survey can be found in . As expected, the speakers’ ages were one of the factors taken into account, and age brackets were set up: 20-30, 30-40, and 41 and over. However, I am not certain whether Martinet does actually employ the term ‘synchronique dynamique’ yet in .
I should further mention a number of investigations of French phonological behaviours from the point of view of dynamic synchrony conducted by , , , , and . Also to be mentioned are the investigation conducted by which treats of what he calls SAE (standard American English). Finally, presents a short statement about a few cases of phonological variations, both vocalic and consonantal, in Persian spoken in Teheran, in preparation for a (then) future investigation of phonological behaviours in other regions in Iran.
2.6
Phonetic manifestations of languages are the external face directly observable by speakers and listeners of any individual languages. They are what are known as ‘pronunciations’ of individual languages. English pronunciation is no exception. English pronunciation is undergoing changes all the time as English is functioning as a verbal means of communication in English-speaking communities. Pronunciation changes can be seized at a given moment from the point of view of either static synchrony or dynamic synchrony.
Many words of a language have single pronunciations, but not all of them. This is why a good number of words have variant pronunciations which have arisen for diverse reasons. Some such variants are correlated with the age of speakers. In such cases it is desirable not only to just identify the variants as being prevalent at a given time but also to establish the correlation between the variants and the age of the speakers who prefer one variant rather than another. It is often the case that speakers of a language, say English, who regularly use one variant pronunciation notice immediately if their interlocutors who may be younger or older happen to use another variant pronunciation. For example, in BrE (with which we are concerned in this paper), ˈcontroversy is nowadays more commonly heard among older speakers while conˈtroversy is heard among younger speakers. (This does not apply to AmE where ˈcontroversy is the only form.)
Any researcher investigating, with the concept of dynamic synchrony in mind, a specific linguistic phenomenon which is changing over time requires a few indispensable data such as the informants’ ages (their years of birth), observing/confirming that there are differences among the informants’ comportment in the specific linguistic phenomenon being investigated. Thus, any English pronunciation preference surveys should preferably indicate how competing variant pronunciations of items correlate with the informants’ ages; in other words, the findings of the surveys show the different competing variant pronunciations from the point of view of dynamic synchrony.
3. PART THREE. WELLS’S BRITISH ENGLISH PRONUNCIATION PREFERENCE SURVEYS
3.0
John Christopher Wells, an English phonetician, conducted two major pronunciation preference surveys aimed at two separate bodies of a large number of native speakers of BrE (British English) who were resident in Great Britain. Questionnaires were sent out to prospective informants in the 1988 and the 1998 pronunciation preference surveys, conducted with a ten year’s interval. Wells specifically requested all prospective informants to reveal their ages. These two pronunciation preference surveys in BrE conducted by Wells were highly welcome. Other than the two major surveys mentioned above, there were two additional pronunciation preference surveys, one minor and the other major, the results of which were also taken into consideration by Wells. The findings of all four surveys were incorporated into and so that those who consult the two pronouncing dictionaries can easily and instantly know the age-related divergences in the pronunciations of selected English words. I should emphasize that, to my knowledge, LPD2 and LPD3 are the only English pronouncing dictionaries past and present which contain information gathered and analysed pronunciation variants of selected words from the point of view of dynamic synchrony. It is true that Wells’s two surveys can be judged to have been conducted from the viewpoint of dynamic synchrony.
To date, I know of no other BrE pronunciation preference surveys of quite the same magnitude and depth conducted by any other phoneticians than Wells. As for an American pronunciation survey, can comfortably vie with Wells’s in magnitude and depth. Findings of her survey are worth a perusal for anyone interested in the subject. As I said earlier, it is only because of my exclusive choice of dealing with BrE pronunciation that I mostly left out in my present paper. A small selective portion of Shitara’s findings are incorporated by Wells into LPD2 and LPD3.
3.1
Wells writes in LPD2 (2000: xii) as follows.
… for many words of uncertain pronunciation, LPD reports the preferences expressed in three opinion polls:
-
a postal opinion poll carried out by the author in 1988 among a panel of 275 native speakers of BrE from throughout Britain.
-
a postal opinion poll carried out by Yuko Shitara in 1993 among a panel of 400 native speakers of AmE from throughout the United States;
-
an opinion poll carried out by the author in 1998 among a panel of 1,932 native speakers of BrE from throughout Britain, some of whom answered by postal questionnaire but others by e-mail or through the World Wide Web.
Wells repeats in what I quoted just above from LPD2, but by changing ‘by the World Wide Web’ to ‘interactively online’, and adds the following.
1.7 Opinion polls For many words of uncertain pronunciation, LPD reports the preferences expressed in five opinion polls:
-
an on-line dialect survey conducted by Prof. Bert Vaux, then Harvard University, among a panel of self-selected respondents from the United States in 1999-2002; most questions received about 11,000 answers. Details of the respondents’ ages are not available.
-
An outline poll of BrE conducted by Pearson Education and myself in April-June 2007.
In the present paper in which I concentrate, in principle but not always, on BrE, I will not take account of what is mentioned above in connection with the poll conducted by Vaux.
3.2
In the pioneering 1988 poll, a questionnaire was sent out to a randomly selected 275 speakers of BrE. 1,900 responses were obtained. The respondents were asked to indicate which of the multiple pronunciations of each item in the questionnaire they preferred. Fluctuations in the pronunciations may concern consonants, vowels, or place of accent (‘stress placement’ as Wells calls it). For example, [‑θ] or [‑ð] (booth); [iː‑] or [e‑] (economic); cigaret(te) [ˌ‑ ‑ '‑] or ['‑ ‑ ‑]. Items of fluctuating pronunciations included uncertain or disputed pronunciations. The results of the poll were incorporated into LPD2 (2000) and also, later, into . does not yet contain results of the pronunciation preference poll that took place in 1988. Some further account of the procedure adopted in the 1988 poll was given in 1.6 Opinion Poll in LPD1 (p. xi) as follows.
For each of the words, members of the panel were asked which of two or more specified pronunciations they preferred. Their answers are given in LPD under the word in question, expressed as percentages of all respondents, after elimination of those who expressed no preference.
For detailed descriptions of the 1988 pronunciation preference poll, see and .
3.3
In 1998, the second poll was conducted. A questionnaire relating to approximately 100 items of fluctuating pronunciations (some of which were already used in the 1988 poll) was sent out to nearly selected 2,000 ‘speech-conscious’ speakers of BrE. One requirement was that childhood (4 to 15 years) was spent in Britain. A total of 1,932 answers meeting this requirement were accepted. The results of this poll were incorporated into LPD3 (third edition of LPD) which was published in 2008.
3.4
In 2007 (April-June), a third (minor) pronunciation preference poll was conducted by Wells and Pearson Longman. Responses of only those respondents who indicated their geographical origin as Britain (i.e. England, Wales, Scotland and the Channel Islands, but not Ireland) were accepted. The valid responses numbered round about 800‑825. The questionnaire consisted of 30 items which are shown below in boldface and non-italicized. Wells looked for the overall percentages, which he listed next to the items. I will not copy the percentages below as they can easily be found in LPD3. The results were incorporated into LPD3, along with those of the other items already found in LPD2.
accept/except, adult, applicable, Asia, careless, contribute, debris, diphthong, dissect, during [dʒ‑], during [‑ʊə‑], egotistic, electoral, H, homogeneous, hurricane, impious, kilometre, lamentable, liquorice/licorice, mischievous, necessarily, omega, poor, protester, tinnitus, tune, via, were, yours.
3.5
Meanwhile Yûko Shitara carried out a separate major pronunciation preference poll in AmE in 1993, that is, at a time intermediate between Wells’s 1989 and 1999 polls. This was the only survey conducted by Shitara. The findings of Shitara’s poll are incorporated into LPD2 and LPD3, appropriately interspersed among the findings of Wells’s 1988 and 1998 polls in BrE. It is interesting to look at the choice of multiple variant pronunciations of a number of items in AmE at a given date, i.e. 1993, and also to compare on the one hand the different proportions, among AmE speakers, in preferring some rather than other variant pronunciations of items, with, on the other, those among BrE speakers.
3.6
Here below is a complete list of items with fluctuating pronunciations (simple words and compound words) which Wells analysed and showed in LPD2 (a total of 210 items) and in LPD3 (a total of 262 items). I will present below the lists of all the items in alphabetic order; such lists are not given by Wells himself. I have checked all the pages of LPD2 and LPD3 in order to identify these items.
The complete list to be shown below consists of 375 items (262 items + 113 items) in BrE and in AmE which were surveyed and recorded in LPD3.
3.6a
Here first is a list of 262 items ― List A ― surveyed by Wells in 1988, 1998 and 2007 and recorded in LPD3. I will present each item in boldface and deliberately non-italicized. (i) An entry like ‘almond (AmE only)’ means that only Shitara’s finding for this item in AmE is shown but not in BrE simply because this item was not in Wells’s questionnaire. (ii) An entry like ‘absorb (with AmE)’ means that the findings in both BrE and AmE for this item are shown. (iii) An entry like accomplish means that only the finding in BrE is shown. (iv) Finally, at the end of each group of items, the total number of items in each group is given.
-
-
absorb (with AmE), absurd (with AmE), accomplish, address (AmE only), adult (with AmE), again (with AmE), ally (n/v) (AmE only), almond (AmE only), alto, amphitheatre (AmE only), applicable (with AmE), Asia (with AmE), associate (v), association, assume, asterisk (AmE only), ate, attitude, auction, aunt (AmE only). 20 items
-
-
baptize (AmE only), baths (with AmE), because (AmE only), bedroom, been, bequeath, booth, bouquet, brochure, broom. 10 items
-
-
capsize (AmE only), caramel (AmE only), Caribbean, casual, caviar, chance, chromosome (AmE only), chrysanthemum, cigaret(te) (with AmE), circumstance, citizen (AmE only), clandestine, coffee (AmE only), communal, complex (AmE only), congratulate (AmE only), contribute, controversy, costume (AmE only), coupon (with AmE), covert (with AmE), cream cheese (AmE only), create (AmE only), creek (AmE only), crescent, cyclical. 26 items
-
-
data (with AmE), debris, debut, decade (with AmE), defect, deity, delirious, demonstrable, depot (AmE only), deprivation (AmE only), detail (n/v) (AmE only), diagnose (AmE only), diphthong, direct (v/adj/adv) (AmE only), direction, discount (v) (AmE only), dispute (n), dissect, distribute, donate (AmE only), drama (AmE only), drastic, due (AmE only), during. 24 items
-
-
economic, ecosystem, egotistic, either (with AmE), electoral, electronic, envelope, ephemeral, equation (AmE only), equinox, evolution, exasperate, exit (with AmE), exquisite (with AmE), extraordinarily (AmE only). 15 items
-
-
falcon (with AmE), false, February (with AmE), fiance (AmE only), finance (with AmE), financial, forehead (with AmE), formidable (with AmE). 8 items
-
-
garage (with AmE), gibberish, giga-, Glasgow, gone (AmE only), gradual, graph, greasy (AmE only). 8 items
-
-
h, halt, handkerchief (AmE only), harass (with AmE), herb (AmE only), hero (AmE only), historic, homogenous, homosexual, hospitable, hurricane. 11 items
-
-
ice cream, idea (AmE only), ideology (with AmE), illustrate (AmE only), impious, incomparable (with AmE), increase (n), inherent, innovative, inquiry (AmE only), insurance (AmE only), involve, irrefutable, issue. 14 items
-
-
jumped, jury, justifiable (with AmE), juvenile (AmE only). 4 items
-
-
kilometer/kilometre (with AmE). 1 item
-
-
lamentable, lather, lawyer (AmE only), length, ‑less, liquorice, longitude, lure, luxurious (with AmE), luxury (with AmE). 10 items
-
-
maintain, mall, malpractice (AmE only), marry (AmE only), masquerade, Massachusetts (AmE only), mayonnaise (AmE only), measure (AmE only), migraine, mischievous (with AmE), Muslim. 11 items
-
-
necessarily, necessary, nephew, new (AmE only), newspaper (with AmE), niche, nuclear. 7 items
-
-
often (with AmE), ogle, omega, ominous, one, onerous, opposite, oral, orange (AmE only), ordinary. 10 items
-
-
pajamas (AmE only), palm (with AmE), patriotic, patronise (with AmE), perpetual, plaque, plastic, poem (AmE only), Polynesia, poor, predecessor (AmE only), premature, Presley (AmE only), prestigious (with AmE), presume, primarily, princess, privacy, process (n/v) (AmE only), project (n), protester, puncture. 22 items
-
-
quagmire, quarter (AmE only), questionnaire. 3 items
-
-
real (adj/adv/n), really (with AmE), regulatory, research (v/n) (with AmE), resource, respiratory (with AmE), restaurant, room (with AmE), route (AmE only). 9 items
-
-
salt, sandwich, says, scallop, schedule, schism, scone, semi- (AmE only), short cut, simultaneous, situation, soot (AmE only), sorry (AmE only), soviet, spectator, stereo, strength (AmE only), student (AmE only), submarine (with AmE), subsidence, substantial, suggest (AmE only), suit, sure, syrup (AmE only). 25 items
-
-
thanksgiving (AmE only), thespian (AmE only), tinnitus, tomorrow (AmE only), transferable, transistor, transition, translate (AmE only), tube (AmE only). tune. 10 items
-
-
umbrella (AmE only), usage. 2 items
-
-
vacation (with AmE), vehicle (AmE only), via, visa (AmE only), voluntarily. 5 items
-
-
were, white, with (with AmE). 3 items
-
-
year, yours, youths (with AmE). 3 items
-
-
zebra. 1 item
3.6b
Here next is a list of 113 items ― List B ― in AmE surveyed by Shitara in 1993 and recorded in LPD3. These items are in two sub-classes. Those items making up one of the sub-classes are presented in boldface (e.g. address), which means that only Shitara’s results (i.e. AmE) are shown. The other items making up the other sub-classs are presented in nonboldface (e.g. absorb, which means that both Wells’s and Shitara’s results (i.e. BrE + AmE) are shown. Lastly, the two numerals shown, e.g. (7 + 5) means that, the 12 items consists of (7 (AmE) + 5 (BrE + AmE)).
-
-
absorb, address, adult, again, ally (v), almond, amphitheater, applicable, Asia, asterisk, attitude, aunt. 12 items (7 + 5)
-
-
baptize, baths, because. 3 items (2 + 1)
-
-
capsize, caramel, chromosome, cigaret, citizen, coffee, complex, congratulate, costume, coupon, covert, cream cheese, create, creek. 14 items (11 + 3)
-
-
data, decade, depot, deprivation, detail (n/v), diagnose, direct (v/adj/adv), discount (v), donate, drama, due. 11 items (9 + 2)
-
-
either, equation, exit, exquisite, extraordinary. 5 items (2 + 3)
-
-
falcon, February, fiance, finance, forehead, formidable. 6 items (1 + 5)
-
-
garage, gone, greasy. 3 items (2 + 1)
-
-
handkerchief, harass, herb, hero. 4 items (3 + 1)
-
-
idea, ideology, illustrate, incomparable, inquiry, insurance. 6 items (4 + 2)
-
-
justifiable, juvenile. 2 items (1 + 1)
-
-
kilometer/kilometre. 1 item (0 + 1)
-
-
lawyer, luxurious, luxury. 3 items (2 + 1)
-
-
malpractice, marry, Massachusetts, mayonnaise, measure, mischievous. 6 items (5 + 1)
-
-
new, newspaper. 2 items (1 + 1)
-
-
often, orange. 2 items (1 +1)
-
-
pajamas, palm, patronise, poem, predecessor, Presley, prestigious, process (n/v). 8 items (6 + 2)
-
-
quarter. 1 item (1 + 0)
-
-
really, research (v/n), respiratory, room, route. 5 items (2 + 3)
-
-
semi-, soot, sorry, strength, student, submarine, suggest, syrup. 8 items (7 + 1)
-
-
thanksgiving, thespian, tomorrow, translate, tube. 5 items (5 + 0)
-
-
umbrella 1 item (1 + 0)
-
-
vacation 1 item (0 + 1)
-
-
vehicle 1 item (1 + 0)
-
-
visa 1 item (1 + 0)
-
-
with 1 item (0 + 1)
-
-
youths 1 item (0 + 1)
3.6c
Wells’s list (262 items) shown as List A and Shitara’s list (113 items) shown as List B will result in a single list (375 items) which we might call List C, but which we need not actually show.
The 113 items presented above in List B consisting of 73 items (AmE) and 40 items (BrE + AmE) are in two subclasses. The 73 items were not in Well’s questionnaire while the 40 items were in his questionnaire. As the specific purpose of this paper is to intentionally concentrate on Wells’s pronunciation preference survey in BrE only, I will subtract at this stage the 73 items while retaining all the 40 items. This means that a total of 302 items (i.e. 262 items + 40 items), viz. BrE + (BrE +AmE), form the object proper of my discussions in this paper.
3.7
The total of 113 item shown in List B that Shitara investigated for AmE in 1993 and incorporated into LPD3 obviously deserve close attention and proper discussion even if they are obliged to be left out in my present paper. This will be left for future study for me. Shitara’s findings of pronunciation preference survey are all shown in , though not on a regular basis in . For the moment it would at least be possible to say that comparison between the findings of the 1993 AmE poll and those of the 1988 and 1998 BrE polls will show the difference between British and American pronunciations of those same items which were investigated by both the BrE and AmE surveys. An example of such a comparison can be, for instance, in terms of absorb, AmE 1993, [‑'zɔːb] 75 % vs BrE 1998 [‑'zɔːb] 83 %, and AmE 1993 [‑'sɔːb] 25 % vs BrE 1998 [‑'sɔːb] 17 %. Incidentally, I understand that Shitara’s survey did not take the ages of the informants into account.
3.8
We are now ready to see how Wells’s findings in his 1988 and 1998 British pronunciation preference surveys have revealed from the point of view of dynamic synchrony. We will look randomly just a few items that Wells investigated.
3.8a
auction. tells us that the percentage of those who prefer ['ɔːk‑] is 87 % while that of those who prefer ['ɒk‑] is 13 % of the total respondents at the time of Wells’s 1998 BrE survey; in other words, [‑'ɔːk‑] was the majority variant and predominating over ['ɒk‑] at that time. Useful and interesting though this information is, the information does not reveal, in 1998, what age bracket of respondents favoured ['ɔːk‑] (87 %) and which other age bracket of respondents favoured ['ɒk-] (13 %). The information cannot be a basis on which we can even hazard a guess as to how the two variant pronunciations in competition with each other may continue to fare with each other in future. ['ɔːk‑] (87 %) may perhaps be progressive and ['ɒk‑] (13 %) recessive, particularly if we take into consideration possible effect of ‘conformity’, ‘accommodation’ and ‘socialization’ among young speakers in particular.
3.8b
applicable. (though not in LPD3) shows this item with two variant pronunciations [ə'plɪk‑] and [.æplɪk‑] in competition with each other, one in BrE 1988 ([ə'plɪk‑] 77 %, ['æplɪk‑] 23 %) and the other in BrE 1998 ([ə'plɪk‑] 84 %, ['æplɪk‑] 16 %). Clearly, there was a lapse of a decade between BrE 1988 and BrE 1998, and the attendant changes in the percentages for the two variant pronunciations. That is to say, 77 % → 84 % for [ə'plɪk] on the one hand and 23 % → 16 % for ['æplɪk-] on the other took place during the decade in question. We can only regard ‘BrE 1988 ([ə'plɪk‑] 77 %, ['æplɪk‑] 23 %)’ and ‘BrE 1998 ([əˈplɪk‑] 84 %, [ˈæplɪk‑] 16 %)’ as two separate facts of static synchrony valid in 1998 and in 1998, respectively. As presented by Wells in LPD2, we can observe two static synchronic facts, or two diachronic fact, valid in 1988 and then a decade later in 1998.
3.8c
booth. The 1998 poll revealed that [buːð] was accorded 62 % and [buːθ] 38 %. In the case of chrysanthemum, the two percentages, as given in LPD2, are [‑sæntθ‑] 61 % and [‑zæntθ‑] 39 % in the 1988 poll and, as given in LPD3, [‑sæntθ‑] 63 % and [‑zæntθ‑] 37 %, as revealed in the 1998 poll. What attracts our attention about Wells’s presentation of booth and chrysanthemum is that those informants preferring [buːð] are indicated in terms of the four parts of the UK, viz. South of England, North of England, Wales and Scotland, not at all in terms of different age brackets of informants. The variant pronunciation [buːθ] is not taken into account by Wells here.
3.9
The question of four age brackets with which Wells operates will now be discussed. There is an interval of a decade between Wells’s two British pronunciation preference surveys, viz. between 1988 and 1998, and consequently the four age brackets of respondents that Wells operates with will have shifted during the decade. The four age brackets valid in ‘BrE 1988 poll’ cannot be the same as those valid in ‘BrE 1998 poll’.
In all the cases where Wells indicates ‘Older’ ←→ ‘Younger’, it is impossible to know details of each age bracket. However, details are supplied by Wells here and there in LPD2 and LPD3. Elsewhere, in the line graph which he shows in connection with the item garage on page 3 of (3.1.2), he indicates the four equal brackets. They are as follows: ‘born up to 1933’; ‘1934‑53’; ‘1954‑73’; ‘since 1973’. Alternatively, in terms of four age brackets as at 1998, ‘65 and over’, ‘45 to 64’, ‘25 to 44’, and ‘25 and under’.
3.10
The four age brackets with which Wells operates are shown below, as reformatted by me. Wells determines the four age brackets in terms of the respondents’ years of birth, but I have added the respondents’ relevant ages. (I have named the four age brackets A, B, C and D, as can be seen below.)
3.11
It is obvious that there should be no identity between Age Bracket A in LPD2 and that in LPD3, between Age Bracket B in LPD2 and that in LPD3, between Age Bracket C in LPD2 and that in LPD3, and Age Bracket D in LPD2 and that in LDP3. What happens in reality is that the forward passage of a decade (1988 → 1998) changes Age Bracket A to (‘b. up to 1923’; 75 and over)’, Age Bracket B to (‘1924‑1943; 55 to 74)’, Age Bracket C to (‘1964‑1983’; 35 to 74), and Age Bracket D to (‘since 1984’; 35 and under)’. This means that Age Bracket A in LPD3 has shrunk (with possible deaths at the oldest) while Age Bracket D in LPD3 will have augmented (as more young speakers will have joined). When the changes delineated above are taken into account, a more realistic and (in my opinion) precise picture of age brackets in question in LPD3 may look like the following.
(where A' ← A; B' ← B; C' ← C; and D' ← D).
3.12
Wells himself provides the necessary clarifications about the problem mentioned above. He writes as follows in LPD2 (p. xii).
The four age groups referred to in the graphs are different for the three surveys. In BrE88 the categories from oldest to youngest were those born respectively before 1923, in 1923‑47, in 1948‑62 and after 1962. In AmE they were born in or before 1927, in 1928‑47, in 1948‑67, and in or after 1968. In BrE98 they were both up to 1933, 1934‑53, 1954‑73, and since 1973.
Notice that Wells refers, in the passage quoted just above, to the four age brackets in connection with the 1993 AmE preference poll conducted by Shitara. Notice also that the above quoted clarifications by Wells are not repeated in LPD3 because of the irrelevance of the identities of the years 1988 and 1998, that is, the distinction between ‘BrE 1988’ and ‘BrE 1998’.
(where A' ← A; B' ← B; C' ← C; and D' ← D)
Wells and I would completely agree with each other in setting up the four age brackets in AmE93.
3.13. Wells’s use of the labels ‘BrE 1988’, ‘BrE 1998’, and ‘preference poll(s), BrE’
One may be tempted to think that the findings of the 1988 poll are incorporated into and those of the 1998 preference poll into . Actually, this is not necessarily so.
What is immediately noticeable when consulting LPD3 is that the two different labels, viz. ‘BrE 1988 panel preference’ and ‘BrE 1998 panel preference’ which appear in LPD2 are consistently replaced in LPD3 by the simple label ‘Preference poll, BrE’. For example, the item accomplish, ‘BrE poll 1988 panel preference’: [‑'kɅmp‑] 92 %, [‑'kɒmp‑] 8 % in LPD2 is replaced by ‘Preference poll, BrE’: [‑'kɅmp‑] 92 %, [‑'kɒmp‑] 8 %, in LPD3. It may be asked whether ‘Preference poll, BrE’ in LPD3 is the same poll as ‘BrE poll 1998 panel preference’ and, if so, whether ‘BrE poll 1998 panel preference’ did actually take place and, moreover, if so, whether ‘BrE poll 1998 panel preference’ did actually yield the same percentages for accomplish, viz. 92 % vs 8 %, as those obtained from ‘BrE poll 1988 panel preference’. No answer seems to be immediately forthcoming.
3.14
Unsurprisingly, the results of the 1988 poll are incorporated into LPD2, as one may expect. This is the case with accomplish, ate, auction, bath, bedroom, been, bequeath, bouquet, casual, caviar, cigaret, etc. These items are included later also in LPD3. Below are just two examples, accomplish and ate. See (1) to (4) below. Note that the label ‘BrE poll 1988 panel preference’ appears in (1) and (3), but ‘Preference poll, BrE’ appears in (2) and (4). This is a matter of some importance as I will explain later. The pair of different percentages 92 % vs 8 %, or that of different percentages 55 % vs 45 %, represent the different sizes of groups of respondents who preferred one or the other variant pronunciation of accomplish or of ate, as the case may be.
3.15
Surprisingly, some results of the 1998 poll are already published in LPD2. This is the case with absorb, absurd,alto, Asia,associate, association, ate, booth, Caribbean, chance, etc. In all these cases, reference to the 1998 survey is explicitly made as ‘BrE 1998 poll panel preference’. Here below are just two examples, absurd and Caribbean. Note that the label ‘BrE 1998 poll panel preference’ appears in (5) and (7) but the label ‘Preference poll, BrE’ appears in (6) and (8). Again this is a matter of some importance.
3.17
At this point we need to look at a pair of different percentages, viz. 92 % vs 8 %, for the item accomplish shown in (1) and (2) in 3.14. We note that (1) relates to ‘BrE poll 1988 panel preference’ while (2) relates to ‘Preference poll, BrE’ but the two preference polls share the same pair of percentages, viz. 92 % vs 8 %. This means two things: either ‘Preference poll, BrE’ would be identical with ‘BrE poll 1988 panel preference’, hence the same pair of different percentages; or ‘Preference poll, BrE’ would actually be the same preference poll as ‘BrE 1998 poll panel preference’ which interestingly obtained exactly the same percentages, 92 % vs 8 % for accomplish. If this understanding is correct, the same can be said of ‘Preference poll, BrE’ in (4) which would be the same as ‘BrE 1988 poll panel preference’ for ate in (3) and (4) with the percentages of 55 % vs 45 %, and ‘Preference poll, BrE’ in (6) would the same as ‘BrE 1998 poll panel preference’ for absurd in (5) and (6) with 77 % vs 23 %. And finally ‘Preference poll, BrE’ in (8) would be the same as ‘BrE 1998 poll panel preference’ for Caribbean in (7) and (8) with 91 % vs 9 %.
We cannot tell, however, if ‘Preference poll, BrE’ (in LPD3) in (2) does refer to ‘BrE poll 1988 panel preference’ with 92 % vs 8 % for accomplish, or to ‘BrE 1998 poll panel preference’ which highly interestingly resulted also in 92 % vs 8 % for accomplish. I say ‘highly interesting’ because the survey conducted in 1988 and the survey conducted in 1998 involved two entirely different groups of respondents. Yet did the respondents of the two different groups nonetheless give exactly the same pair of percentages, namely 92 % vs 8 %? What has just been said applies mutatis mutandis to ‘Preference poll, BrE’ (in LPD3) in (4) for ate, (6) for absurd and (8) for Caribbean.
We know from (2), (4), (6) and (8) that the label ‘Preference poll, BrE’ says nothing about whether this label corresponds to either ‘BrE poll 1988 panel preference’ or ‘BrE 1998 poll panel preference’ or both. In other words, the label ‘Preference poll, BrE’ is undeterminable in terms of the year 1988 and/or the year 1998.
Are there cases in which the results of the 1998 poll are published in LPD3? The answer to this question can only be the following. The label that one may expect is ‘BrE poll 1998 panel preference’ but this label has already been seen in (1), (3), (5) and (7) in connection with LPD2. There remain (2), (4), (6) and (8) which must relate to either the 1989 poll or the 1998 or even both. The label attached to (2), (4), (6) and (8) is therefore either the 1989 poll or the 1998 or even both.
It is preponderantly (unfortunately in my opinion) the case that ‘BrE poll 1988’ and ‘BrE 1998 panel preference’ do not both appear in LPD3. Exceptionally (welcomely in my view) the results of both the 1988 and the 1998 polls are found in LPD2, for example, for applicable (p. 36) and chrysanthemum (p. 139) but the results of only the 1998 survey (without it being specified) are found for applicable (p. 36) and chrysanthemum (p. 146) in LPD3.
3.18
In LPD2, if not in LPD3, Wells displays the findings of either the 1988 survey or those of the 1998 survey, or those of both the 1988 survey and those of the 1998 surveys, as follows:
-
(1) Only ‘BrE 1988 poll panel preference’: for accomplish, ate (past of eat), auction, bath, bedroom, been, bequeath, bouquet, brochure, casual, caviar, cigarette, clandestine, communal, contribute, covert, crescent, cyclical, etc.
-
(2) Only ‘BrE 1998 poll panel preference’: for absorb, absurd, alto, Asia, associate, association, booth, Caribbean, chance, circumstance, ecosystem, electronic, ephemeral, equinox, evolution, falcon, false, February, finance, financial, etc.
-
(3) Both ‘BrE 1988 poll panel preference’ and ‘BrE 1998 poll panel preference’: for applicable, chrysanthemum, controversy, delirious, kilometre, etc.
I will demonstrate straightaway with the example of applicable that only the items of Group (3) are amenable to the concept of dynamic synchrony.
3.19
The item applicable has two variant pronunciations which Wells indicates as [ə'plɪk‑] 77 % and ['æplɪk‑] 23 % as revealed in ‘Preference polls BrE 1988’ and recorded in and [ə'plɪk‑] 84 % and ['æplɪk‑] 16 % revealed in ‘Preference polls BrE 1998’ and recorded also in . In Wells indicates [ə'plɪk‑] 85 % and ['æplɪk‑] 15 % as revealed in ‘Preference polls BrE’. Note that ‘Preference polls BrE’, not either ‘Preference polls BrE 1988’ or ‘Preference polls BrE 1998’, occurs for applicable in , and 85 % not 84 %, and 15 % instead of 16 %, occur for applicable in . These facts are interesting as such.
However, what I am really interested in above all in my present paper is to view the findings of Wells’s surveys in respect of how these alternative pronunciations were in competition with each other with different percentages with respect to each other through time, specifically between 1988 and 1998. (‘Preference polls BrE’ mentioned just above is judged to be actually the same as ‘BrE 1998’.) To return to the item applicable, the relevant findings by Wells in can be re-expressed as follows.
[ə.plɪk-] 77 % (BrE 1988) | vs | [.æplɪk-] 23 % (BrE 1988) |
vs | vs | |
[əˈplɪk-] 84 % (BrE 1998) | vs | [.æplɪk-] 16 % (BrE 1998) |
The above data can be looked at in two different ways, horizontally or vertically. First, it can be looked at horizontally. Viewed horizontally, we see that ‘BrE 1988’ revealed [əˈplɪk‑] 77 % vs [ˈæplɪk‑] 23 %, and that ‘BrE 1998’ revealed [əˈplɪk‑] 84 % vs [ˈæplɪk‑] 16 %. This is a view and statement from the point of view of dynamic synchrony. Secondly, it can be looked at vertically. We see on the one hand that [əˈplɪk‑] 77 % as revealed in (BrE 1988) changed to [əˈplɪk‑] 84 % as revealed in (BrE 1998), that is, the percentage for [əˈplɪk‑] went up from 77 % to 84 % during the decade from 1988 to 1998. On the other hand, we also see that [ˈæplɪk‑] 23 % as revealed in (BrE 1988) went down during the same decade to 16 % as revealed in (BrE 1998). This is a view and statement from the point of view of diachrony. We can see that, during the decade from 1988 to 1998, [əˈplɪk‑] was a progressive variant (77 % upward to 84 %) while [ˈæplɪk‑] was a regressive variant (23 % downward to 16 %). In other words, more and more BrE speakers tended to prefer [əˈplɪk‑] while, contrarily, fewer and fewer BrE speakers tended to prefer [ˈæplɪk‑]. It can be surmised that, during the decade 1988‑1998, more and more younger speakers tended to prefer [əˈplɪk‑] in 1998 while older speakers who tend to pronounce [ˈæplɪk‑] were getting fewer and fewer. It will be interesting to pursue the inquiry to see how the percentage ratio between [əˈplɪk‑] and [ˈæplɪk‑] may fare in future. Is the preference ratio [ˈæplɪk‑] 64 % vs [əˈplɪk‑] 36 % in AmE reported by Shita in 1993 (recorded in and re‑recorded in LPD3 (p. 38)) ever likely to reverse the contrary preference ratio we have seen above, [əˈplɪk‑] 84 % vs [ˈæplɪk] 16 %? Not quite likely, I think.
3.20
For the item absorb, ‘Poll panel preference BrE 1998’ [‑'zɔːb] 83 %, [‑'sɔːb] 17 %’ in LPD2 is replaced in LPD3 by ‘Preference poll, BrE’, ['‑zɔːb] 83 %, ['‑sɔːb] 17 %’. In other words, the two percentages remain the same in LPD3 as well as in LPD2 but – this is important – the label ‘Poll panel preference BrE 1998’ is replaced by the label ‘Preference poll, BrE’. The question to be asked is, with regard to absorb, whether ‘Preference poll, BrE’ in LPD3 implies that ‘Poll panel preference BrE 1988’ yielded the same percentages as did ‘Preference poll, BrE’, provided that ‘BrE 1988 preference poll’ did take place in connection with absorb.
3.21
It will have been seen that the label ‘Preference poll, BrE’, which is consistently used by Wells in LPD3, is a cover term with which to loosely refer to either ‘Poll panel preference BrE 1988’ or ‘Poll panel preference BrE 1998’, as the case may be. It would be free from misunderstanding if Wells indicated the year/s of the different poll(s) rather than resort to the simple ambiguous label ‘Preference poll, BrE’. I believe that, as employed by Wells, the label ‘Preference poll, BrE’ may do without. Following what has been saids above, I would say that anyone who wishes to be informed of the full findings of Wells’s both 1988 and 1998 British pronunciation preference surveys is recommended to consult both and where all the results will be found.
3.22
What has been discussed above in connection with Wells’s use of three labels, viz. ‘BrE 1988’, ‘BrE 1998’, and ‘Preference polls BrE’, can be traced to Wells’s decision to the effect: Polling figures in the text are no longer identified by year [My boldface]’ found in LPD3 (p. xviii). In my view this seems to be an unhelpful decision. His decision gives rise to a fair number of cases where the findings of his 1988 poll are ignored or are at least dispensed with in LPD3 and the identity of his 1998 poll is unfortunately hidden by the abundant use of his simple single label ‘Preference poll, BrE’ in LPD3, if not in LPD2. This results in the impossibility of possible instances of British English pronunciation changes as viewed in terms of diachrony which may have occurred during the decade of from 1988 to 1998, although this may probably not be Wells’s major concern. Otherwise he would mention a significant number of such cases, provided such changes have occurred. All the same, ten years, 1988-1998, cannot said to be a short period of time when we remember that languages are constantly changing. Another reason for which Wells’s persistent use of the label ‘Preference poll, BrE’ (instead of ‘BrE 1988’ or ‘BrE 1998’, as the case may be) is regretted is that the difficulty of identifying instances of dynamic synchrony in the pronunciation variants of items.
3.23
At this point I will discuss in some detail a few randomly chosen items which are found in both LPD2 and LPD3. The items to be discussed are tune, poor, perpetual, absorb, contribute, lure and garage in this order.
3.23a
Let’s first look at the item tune. The different percentages attributed to it are as follows. According to LPD2 (p. 800), ‘BrE 1998 poll panel preference’: [tjuːn] 64 %, [tʃuːn] 35 %, [tuːn] 1 %. However, according to LPD3 (p. 845), ‘Preference poll, BrE: [tʃuːn] 54 % (born since 1981, 68 %), [tjuːn] 44 %, [tuːn] 2 %. Let’s look at the ramifications of these percentages.
-
(1) We notice that, in connection with tune, ‘BrE 1988 poll panel preference’ is not mentioned in neither LPD2 nor LPD3.
-
(2) In LPD3 neither ‘BrE 1988 poll panel preference’ nor ‘BrE 1998 poll panel preference’ is specifically mentioned. Instead, the simple label ‘Preference poll, BrE’ appears.
-
(3) As is well known among British phoneticians, [tʃuːn] is definitely gaining ground among young speakers in particular. Therefore one naturally expects to see this trend reflected in the findings of the (earlier) 1988 and the (later) 1998 surveys.
-
(4) We are therefore surprised to be told in LPD2 (p. 800) that ‘BrE 1998 poll panel preference’ yielded [tʃuːn] 35%, a lower percentage than [tʃuːn] 54 % shown in LPD3 (p. 845) in ‘Preference poll, BrE’. But ‘Preference poll, BrE’ here cannot by implication be ‘BrE 1988 poll panel preference’ (which is not mentioned). We only need to recollect that another BrE pronunciation preference poll that took place subsequent to ‘BrE 1998’ is what Wells himself calls ‘LPD 2007 survey’. It is likely that ‘Preference poll, BrE’ can, at least in the case of tune, be ‘LPD 2007 survey’.
-
(5) ‘LPD 2007 survey’ is a new on-line pronunciation preference poll which was conducted in preparation for . A questionnaire of 30 items was used in this survey. I have already shown the list of these 30 items supra in 3.4. Indeed the item tune is one of the 30 items with the results [tʃuːn] 54 % and [tjuːn] 44 %, exactly as it is entered in LPD3 (p. 845), whereas in , tune appears as [tjuːn] 64 %, and [tʃuːn] 35 %.
-
(6) In conclusion, the label ‘Preference poll, BrE’ for tune in for tune is to be understood as ‘LPD 2007 survey’, not ‘BrE 1998 poll panel preference’ as one might think. Readers who miss out the information about ‘LPD 2007 survey’ would not arrive at this conclusion.
3.23b
Wells shows the findings of the item poor in and in LPD3 (p. 627). The following is what we find in LPD2 (p. 592).
Poll panel preferences: BrE 1988, [pɔː] 57 %, [pʊə] 43 %; BrE 1998, those born since 1973 [pɔː] 82 %, [pʊə] 18 %.
The above data refer to the two pronunciation preference polls conducted in 1988 and in 1998 with a decade’s interval between them. Whilst the results for poor, i.e. [pɔː] 57 %, [pʊə] 43 %, are obtained from the 1988 poll, Wells fails to refer to his 1998 poll and to indicate the percentage for either [pɔː] or [pʊə] which he must have obtained from his 1998 poll. This lack of information weakens the relevance of his information ‘those born since 1973 [pɔː] 82 %, [pʊə] 18 %’. If Wells did otherwise, this interesting information might be of full value. As it is, this information hangs in the balance, so to speak. (It would be reasonable to estimate that those respondents who were born in 1973 would have been 45 years old in 1998, and those born after 1973 would have been younger.
We read the following in LPD3 (p. 627).
Clearly, we are expected to understand that, following the 1988 preference poll, there was a significantly progressive trend among BrE speakers to prefer to pronounce poor with [ɔː] (as in saw, lawn, author), i.e. [pɔː] 57 % → 74 %. This, we are made to understand, was revealed by the 1998 preference poll, though Wells’s label ‘Preference poll, BrE’ which occurs throughout LPD3 does not overtly identify it. The substantial gain in percentage from 57 % ([pɔː]) to 74 % ([pɔː]) and the overall loss from 43 % ([pʊə]) down to 26 % ([pʊə]) should be attributed to the passage of time, a decade from 1988 to 1998, and the shift of the age brackets from ‘older’ to ‘younger’. The recessive trend in preferring to pronounce [pʊə] in 1998 with the percentage of 41 % among those born before 1942 (aged 57 or older at the time of the 1998 preference poll) is conspicuous all the more so as [pɔː] attracted 74 % overall already in 1998. The time of identically and regularly pronouncing shore, Shaw, sure, etc. may well be with us as I write these lines in 2024.
3.23c
A separate but related subject of further inquiry will be the extent to which the phenomenon of [ʊə] > [ɔː] in words ending in ‑oor can or cannot be considered regular. A possible future pronunciation preference poll similar to those conducted by Wells and Shitara may be able to answer this question. In such a future poll, one would include moor and spoor for which [ʊə] and [ɔː] coexist but the former rather than the latter is the main pronunciation, unlike conversely, as in poor, [ɔː] is the main pronunciation. The inquiry will also look at words which end in ‑ure such as cure, endure, pure, secure, sure, etc. and words ending in or containing ‑our such as gourd, tour, your, etc. Some more words may be involved, such as curious, durable, duration, etc.
3.23d
We will next look at the item perpetual in order to see how Wells shows it first in LPD2 (p. 570) and then in LPD3 (p. 602). This is how Wells shows the findings for perpetual in LPD2 (p. 570).
- (i)
BrE 1998 poll panel preference: [‑tju‿əl] (57 %), [‑tʃu‿əl] (37 %), [‑tʃəl] (5 %). (The aggregate is 99 %).
We next see how Wells shows the findings for perpetual in LPD3 (p. 602).
Those readers who consult only LPD3 cannot know that ‘Preference poll, BrE’ in (ii) actually refers to the 1998 preference poll, which fact they can only find out if they consult LPD2 as well. As can be seen from (i) and (ii), the three variants of perpetual shown are identical in LPD2 and LPD3 except that [-tʃəl] is given 5 % in LPD2 but, for some reason, 6 % in LPD3. The only reason I can think of is that Wells attempts to bring the total percentage to 100 % (= 57 % + 37 % + 6 %) rather than 99 % (= 57 % + 37 % + 5 %).
3.23e
We will look at the item contribute in LDP2 and LDP3 in turn.
The results from the 1988 poll for contribute are presented in , that is, [‑'trɪb‑] (73 %) and ['kɒn‑] (27 %). There is no indication as to how this difference in the two accentual patterns relate to age difference, say, ‘older’ vs ‘younger’. In , the results of the 1998 survey are presented, viz, [‑'trɪb‑] (59 %) and ['kɒn‑] (41 %). Wells adds a new piece of information, viz. ‘born before 1942’: [‑'trɪb‑] (84 %) and ['kɒn‑] (16 %)’. If this information derives from his 1998 survey, it means that 84 % of the informants aged 56 or more in 1998 preferred to pronounce [‑'trɪb‑] while 16 % of them preferred to pronounce ['kɒn‑]. This means of course that, conversely, 16 % of the informants aged 56 or less preferred to pronounce ['kɒn‑] while 84 % of those aged 56 or more preferred to pronounce [‑'trɪb‑].
The fact that Wells introduces the element of age factor by his specific remark ‘born before 1942’, [‑'trɪb‑] (84 %) and ['kɒn‑] (16 %) in presenting the item contribute in is welcome and is reminiscent of Martinet introducing the two timepoints, 1920 and 1940, in connection with the phonological opposition /a/ vs /ɑ/ in speakers of Parisian French. In the case of a few other items, too, Wells introduces the element of the informants’ ages in much the same way as in contribute in the following items: dissect, during, gradual, h/H, hurricane, lamentable, liquorice, mall/Mall, mischievous, omega/Omega, poor, protester, questionnaire, restaurant, schedule, sure, tinnitus, tune and via.
3.23f. We will now look at the item garage in and .
The item garage which has multiple variants in BrE with respect to vowels and consonants and accentual patterns, receives virtually the same detailed presentation by Wells in both and . The multiple variants of garage as indicated by Wells are as follows: ['ɡærɪdʒ], [ɡə'rɑːdʒ], ['ɡærɑːdʒ]. The second and third variants may be expanded as follows: [ɡə'rɑːdʒ] ([ɡə'rɑːdʒ], [ɡə'rɑːʒ]) vs ['ɡærɑːdʒ] (['ɡærɑːdʒ], ['ɡærɑːʒ]). It is appropriate to see two different variants of the item garage, namely ['ɡærɑːdʒ] and [ɡə'rɑːdʒ], not as four different variants, namely [ɡə'rɑːdʒ], [ɡə'rɑːʒ], ['ɡærɑːdʒ], ['ɡærɑːʒ]. The different percentages for the different variants of garage which were obtained in the 1998 survey (this survey being mentioned in LPD2 but not explicitly mentioned in LPD3) are, as indicated in LPD3 (p. 332), as follows: ['ɡærɑːdʒ] (56 %), [‑dʒ] (31 %), [‑ʒ] (25 %), ['ɡærɪdʒ] (38 %), [ɡə'rɑːdʒ] (6 %). Wells previously, in LPD2 (p. 316), indicates ['ɡærɑːdʒ] (56 %) ([‑dʒ] (31 %), [‑ʒ] (25 %), ['ɡærɪdʒ] (38 %) and attributes (5 %) to [ɡə'rɑːdʒ], not (6 %).
Here is what (3.1.2) has to say about the three variants of garage.
In RP, garage is traditionally called a /'ɡærɑːʒ/ or /'ɡærɑːdʒ/; popularly, though, it is a /'ɡærɪdʒ/. … In the survey, the proportion of those preferring initial stress plus strong second vowel was 56 %, a weak second vowel 39 %, and final stress 5 % [This is what we have already seen]. … /'ɡærɪdʒ/ being particularly favoured by the young (66 %) … and particularly disfavoured by the old (13 %).
3.24
Below is a list of 64 items selectively drawn from LPD3 each of which has two or more variants indicated by Wells. For example, absorb (2) means that this item has two variants, and chance (4) has four variants.
absorb (2), Asia (2), ate (2), chance (4), contribute (2), crescent (2), debris (2), delirious (2), dissect (2), during (8), ecosystem (2), exquisite (2), false (2), February (2), financial (3), forehead (2), formidable (2), gradual (2), h (2), halt (2), harass (2), hurricane (2), incomparable (2), issue (3), lamentable (2), length (4), liquorice (2), longitude (2), lure (6), mall (2), migraine (2), mischievous (3), nephew (2), newspaper (2), ogle (2), omega (4), one (2), ordinary (3), patriotic (2), perpetual (4), poor (2), primarily (2), princess (2), protester (2), questionnaire (2), real (2), really (2), salt (2), scallop (2), schedule (4), short cut (2), situation (3), stereo (2), suit (2), sure (2), tinnitus (2), tune (3) usage (2), vacation (2), via (2), voluntarily (4), white (2), with (2), zebra (2).
3.25
Wells makes use of line graphs in which, on the y‑axis (the vertical axis) are shown different percentages representing different ratios of preference for one variant over the other(s) of individual items, while, on the x-axis (the horizontal axis) are shown the different ages of the the respondents in terms of different ‘age brackets’, that is, in terms of ‘Older ← Speakers → Younger’. It is emphasized that only by means of line graphs is it possible to present dynamic aspects of the choice of variants of items in competition with each other from the viewpoint of dynamic synchrony.
In what follows, I bring back 30 items randomly chosen from among the 64 items in 3.24. We shall see how facts of dynamic synchrony are implied with regard to the multiple variants of the individual 30 items.
-
(1) February (p. 301): [‑ru‑] (61 %), [‑ju] (39 %). It is the variant [‑ru‑] with the lower percentage that is represented by a rising line. One can see the rising trend among young speakers to pronounce [ˈfebju‑].
-
(2) h (p. 360): [heɪtʃ] (16 %), [eɪtʃ] (84 %). It is [heɪtʃ] with the lower percentage that is represented by a rising line. We see that the trend to pronounce [heɪtʃ] is spreading among young speakers.
-
(3) incomparable (p. 407): [ˌ‑ ‑ ˈ‑ ‑ ‑] (41 %), [‑ˈ‑ ‑ ‑ ‑] (59 %). It is the variant [ˌ‑ ‑ ‑ ˈ‑ ‑ ‑] with the lower percentage that is represented by a slowly rising line. It may mean that primary accent on third syllable will move gradually towards to (primary) accent on second syllable.
-
(4) issue (p. 426): [ˈɪʃ uː] (49 %), [ˈɪs juː] (30 %), [ˈɪʃ juː] (21 %). The line graph definitely shows [ˈɪʃ uː] is preferred among young speakers.
-
(5) length (p. 459): [leŋθ] (49 %), [leŋkθ] (84 %) ([leŋθ] (48 %), [leŋkθ] (36 %)), [lentθ] (16 %) (but not [lentθ]). The percentages Wells attributes to these variants are not quite straightforward. The percentage of 16 % is unproblematically attributed to [lentθ]. On the other hand, the percentage of 84 % is given to [leŋkθ], but this percentage is divided into [leŋθ] (48 %) and [leŋkθ] (36 %). (In other words, 48 % + 36 % = 84 %; and 84 % + 16 % = 100 %.) Note that Wells sees the division of the variants of length as [lentθ] on the one hand, and in addition, [leŋθ] + [leŋkθ] on the other hand. Also note that Wells separates [leŋkθ] and [leŋkθ] from each other. I am not sure how Wells will explain the auditory difference between [k] in [leŋkθ] and [k] in [lenkθ], but presumably the former is less audible than the latter. Note further that Wells does not reckon with [lentθ] while reckoning with [lentθ]. The progressive trend for [lentθ] results from regressive assimilation [ŋ] > [n] caused by [θ] and the occurrence of optional epenthetic [t ](correctly [t̪]), during the transition from [n] and [θ]. The line graph shows that [leŋkθ] (84 %)(= [leŋθ] (48 %) + [lenkθ] (36 %)) is rising sharply to 100 % so that the youngest speakers will definitely pronounce length [leŋkθ]
-
(6) lure (p. 475): represent [‑ʊə] (72 %), [‑ɜː] (17 %), [‑ɔː] (11 %). It is noticeable that the line for [‑ʊə] is mildly descending but descends rather sharply for young speakers. In contrast, the lines for [‑ɜː] and [‑ɔː], respectively, are low‑lying through from ‘older to younger’ but make a mild rise for young speakers. This seems to indicate that all of [‑ʊə], [‑ɜː] and [‑ɔː] may well remain 3 variants with young speakers for some more time. Wells also indicates ‘with [j]’ (58 %) and ‘without [j]’ (42 %), i.e. ‘with [j]’ (58 %) (= [ljʊə] + [ljɜː] + [ljɔː]), and ‘without j’ (42 %) (= [lʊə] + [lɜː] + [lɔː]).
-
(7) mall (p. 485): [mɔːl] (50 %), [mæl] (50 %). The two variants have an equal percentage, 50 %, but a single line is provided for [mɔːl] only. The line rises sharply to almost 80 % with young speakers. In the absence of another line for [mæl], it is impossible to guess how [mæl] fares, particularly whether it follows the trend for [mɔːl]. It would be interesting to see how the 2 variants with an equal percentage, 50 % in this case, will fare through ‘older ← speakers → younger’.
-
(8) migraine (p. 507): [ˈmaɪɡ‑] (39 %), [ˈmiːɡ‑] (61 %). A single steadily rising line represents [ˈmaɪɡ‑], but no line for [ˈmiːɡ‑]. The line for [ˈmaɪɡ‑] (39 %) continues to reach about 70 % in ‘younger’ bracket. We certainly need a line for [ˈmiːɡ‑] which one may guess has much lower percentage among young speakers.
-
(9) ogle (p. 560): [ˈəʊɡ‑] (76 %), [ˈɒɡ‑] (24 %). It is [ˈəʊɡ‑] that is represented by a single line. It may not be surprising if the predominance of [ˈəʊɡ‑] is hard to beat for quite some time. The variant [ˈɒɡ‑] in BrE is a fairly recent innovative one.
-
(10) omega (p. 562): [ˈəʊmɪɡə] (52 %), [əʊˈmɪːɡə] (28 %), [əʊˈmeɡə] (12 %), [əʊˈmeɪɡə] (8 %). Of the 4 variants, it is [ˈəʊmɪɡə] that is the only one represented by a line. [ˈəʊmɪɡə] has the highest percentage and also is the only one to bear accent on the first syllable. The single line steadily descends from just above 80 % (at ‘older’ end) to 20 % (at ‘younger’ end). All three other variants have accent on the second syllable, the vowels in the second syllable vary from each other ([ɪ], [e], [eɪ]), though all 4 variants share the identical diphthong [əʊ]. Clearly [ˈəʊmɪɡə] (52 %) with accent on the first syllable is definitely recessive and is replaced by the other three variants with accent on the second syllable.
-
(11) ordinary (p. 568): [‑ri] (34 %), [‑eri] (34 %), [‑əri] (32 %). A single line for [‑eri] rises m 34 % to just over 50 % at the end of ‘Younger’. One would like to see two more lines representing [‑ri] and [‑əri] which share very similar percentages.
-
(12) patriotic (p. 592): [ˈpætr‑] (79 %), [ˈpeɪtr‑] (21 %). The single line for [ˈpætr‑] definitely rises from 5 % for ‘Older’ to nearly 40 % at the end of ‘Younger’. [ˈpeɪtr‑] with [‑eɪ‑] is clearly favoured in AmE, and that is why it achieves a low percentage.
-
(13) perpetual (p. 602): [‑tju‿əl] (57 %), [‑tʃu‿əl] (37 %), [‑tʃəl] (6 %). In other words, [‑tʃu‿əl] (37 %) + [‑tʃəl] (6 %) = [‑tʃ] (43 %). That is, 57 % relates to [‑tj‑] and 43 % relates to [‑tʃ‑]. This is all correct. Bizarrely, Wells writes that 57 %, not 43 %, is attributed to [‑tju‿əl] in connection with Preference poll, BrE. Incidentally, [‑tʃəl] (6 %) is to be understood as both [‑tʃəl] and [‑tʃɫ̩], which means two variants. A single line for perpetual is for [‑tju‿əl] with non‑coalescent [‑tj] in which [t] and [j] are not fused into [‑tʃ]. It is significant that [‑tj] is retained and reinforced in popularity among young speakers. The single line for [‑tju‿əl] keeps its upward line from just about 20 % at the ‘Older’ end steadily to just under 60 % near the end of the ‘Younger’.
-
(14) princess (p. 640): [ˌ‑ ˈ‑] (60 %), [ˈ‑ ‑] (40 %). The single line is for [ˈ‑ ‑] which shows a rise, gaining over [ˌ‑ ˈ‑] which has a higher percentage. This means that [ˈ‑ ‑] is progressively favoured by young speakers.
-
(15) protester (p. 647): [ˈ‑ ‑ ‑] (31 %), [‑ ˈ‑ ‑] (69 %). The single line represents ‑ ˈ‑ ‑, i.e. accent on the second syllable.
-
(16) questionnaire (p. 659): [ˌkwe‑] (94 %), and [ˌke‑] (6 %). (Unfortunately, [ˌkwe‑] and [ˌke‑] are misprinted as [ˈkwe‑] and [ˈke].) [ˌkwe‑] (94 %) is the predominant variant over [ˌke‑ ] (6 %). The dominance of [ˌkwe‑] is no doubt related to [ˌkwe‑] which is the only form in question, questionable, questioner. The line graph for [ˌkwe‑] demonstrates it, as it reaches 100 % for young speakers.
-
(17) really (p. 671): [ˈrɪəli], [ˈriːəli], [ˈreəli]. Wells describes the three variants as ‘rhyming with neither freely nor frilly; rhyming with freely; rhyming with frilly’. He gives the percentages for the three variants as follows: rhyming with neither freely nor frilly [ˈreəli] (80 %); rhyming with freely [ˈriːəli] (19 %); and rhyming with frilly [ˈrɪəli] (1 %). Yet Wells presents [ˈrɪəli] as the main pronunciation (in boldface) which seems to have only 1 %. The line graph for really has a single line for freely [ˈriːəli] which consistenly rises from ‘Older’ to ‘Younger’. Comparison between really and real (p. 670) is interesting in that real [riːəl] (same as reel) accounts for 55 % and [rɪəl] 45 %, while really [ˈriːəli] accounts for 19 % and [ˈrɪəli] 1 %. Yet, the single upward line for real [riːəl] and the single upward line for really [ˈriːəli] are palpably similar. Both real [riːəl] and really [ˈriːəli] share similar upward lines. This means that both [riːəl] real and [ˈriːəli] really have been increasingly favoured by young speakers.
-
(18) schedule: another complex one in that, on the one hand, two pairs of variants exist, viz. [ʃ‑] (70 %), and [sk‑] (30 %); [‑dj‑] (79 %), and [‑dʒ‑] (21 %). There result 4 variants altogether, viz. [ʃ‑ dj‑], [ʃ‑ ‑dʒ‑], [sk‑ dj‑], [sk‑ ‑dʒ‑]. [sk‑] which has a lower percentage than [ʃ‑] steadily rises, reaching nearly 65 % (born since 1973). Wells says in LPD2 (p. 678) and LPD3 (p. 716): ‘The AmE pronunciation with [sk‑] is increasing heard in BrE [Wells’s italics].’ The single upward line definitely indicates that [sk‑] will dominate at the expense of [ʃ‑] in the not too distant future, particularly among young speakers.
-
(19) situation (p. 746): [‑tj‑] (64 %), [‑tʃ‑] (35 %), [‑t‑] (1 %). The single line representing [‑tʃ‑] rises sharply reaching just under 60 % with young speakers. Still, [‑tj‑] (64 %) seems surprisingly stable without succumbing to what Wells calls ‘yod coalescence’ ().
-
(20) stereo (p. 775): [ˈster‑] (90 %), [ˈstɪər‑] (10 %), and the sharp single line which represents [ˈster‑] resembles very much that for [‑tʃ‑] in the item perpetual but reaches nearly 100 % in the young age bracket. The variant [ˈster‑] will well continue to predominant with young speakers.
-
(21) suit (p. 790): [suːt] (72 %), [sjuːt] (28 %). The line for [suːt] (72 %) sharply rises and reaches near 100 % in the younger age bracket, as, we saw, do [ˈster‑] (90 %) in stereo and [‑tj‑] (64 %) in situation.
-
(22) sure (p. 794): [ʃɔː] (46 %), [ʃʊə] (54 %). The single rising line represents [ʃɔː]. Wells adds [ʃɔː] ‘(born since 1973… 60 %)’ in connection with [ʃɔː]. Indeed 60 % for [ʃɔː] is situated at the highest end of the rising line. This means that [ʃɔː] is steadily favoured by young speakers at the expense of [ʃʊə]. No line is provided for [ʃʊə]. It would probably be right to guess that [ʃʊə] progressively loses its percentage in favour of [ʃɔː]. Words ending in ‑sure do not necessarily develop [ʃɔː] along with or without [ʃʊə]. The oft quoted items sure/shore/Shaw for [ʃɔː] is the exception rathan than the rule.
-
(23) tinnitus (p. 826): [ˈ‑ ‑ ‑ (ˈtɪn‑)] (82 %), [‑ˈ‑ ‑ (ˈnaɪt‑)] (18 %). The single rising line represents [‑ˈ‑ ‑ (ˈnaɪt‑)], which is shown to be progressively and considerably favoured by middle‑aged to younger speakers. Note that the difference between the two variants is complex in that the place of accent is different in them and also that some segments are also different, i.e.[ˈtɪnɪtəs]/[ˈtɪnətəs] for the first variant and [tɪˈnaɪtəs]/[təˈnaɪtəs] for the second variant. If another line were provided for [ˈ‑ ‑ ‑ (ˈtɪn‑)] (82 %), one would see this variant disfavoured among young speakers.
-
(24) tune (p. 845): [tʃuːn] (54 %), [tjuːn] (44 %), [tuːn] (2 %). The single rising line represents [tʃuːn]. It rises sharply with younger speakers in particular to nearly 70 %. This variant [tʃuːn] contains [tʃ] which is the result of ‘yod coalescence’, viz. [tj] >[ tʃ], which is progressively popular among young speakers. This reminds us of items like perpetual and situation we have seen above. The single lines for [tʃ‑] in tune and [‑tʃ‑] in perpetual are remarkably similar to each other.
-
(25) vacation (p. 868): [veɪ‑] (61 %), [və‑] (39 %). A single line represents [veɪ‑]. This line steadily rises to reach almost 90 %, which shows that [veɪ‑] (61 %) is definitely much favoured by young speakers. It is possible that the high percentage, 90 % which is attributable to this variant may have been influenced by AmE pronunciation of vacation in which [veɪ‑] attains 91 % (cf [və‑] (9 %)) according to Preference polls, AmE (LPD3, p. 868).
-
(26) via (p. 875): [ˈvaɪə] (88 %), [ˈviːə] (12 %). A single line represents [ˈvaɪə]. This slowly descending line is interesting in that, unlike all the other lines we have seen so far, it hardly descends for about two‑thirds’ distance from the beginning, but then does descend, showing that young speakers are progressively disfavouring [ˈvaɪə‑] but progressively favouring [ˈviːə]. It is worth investigating if this trend has been going on well into the 21st century. It would also be interesting to see if [ˈviːə] continues to maintain a low percentage also in the 21st century.
-
(27) voluntarily (p. 880): Wells indicates the different variants of this item is as follows: [ˌvɒl ən ˈter əl i, ‑ˈteər‑, ‑ˈtær‑, ‑ɪ li], [ˈvɒl ənt‿ər əl i, ‑ɪ li]. He presents four different variants with four different percentages as follows: [ˈter] (41 %), [ˈteər] (15 %), [ˈtær] (12 %), [ˈvɒl] (32 %), the total being 100 %. [ˈter], [ˈteər] and [ˈtær] all relate to the pronunciations (68 %) which start with [ˌvɒl], that is, secondary accent on the first syllable and primary accent on the third syllable, while [ˈvɒl] relates to those (32 %) which start with the first syllable accented. This means that voluntarily pronounced with [ˌvɒl] is decisively favoured among young speakers. Wells provides 2 lines, one of which is a rising line for [ˈter] while the other is a falling line for [ˈvɒl‑]. In fact, Wells could do with a rising line for all of [ˈter], [ˈteər] and [ˈtær] since it is undoubtedly more relevant that accent falls on the third syllable than that different vowels [e], [eə] and [æ] occur. The 2 lines, one rising and the other falling, intersect each other at a point corresponding to the middle point of the x‑axis. The falling line for [ˈvɒl‑] shows that [ˈvɒl‑] which the oldest informants preferred is progressively less maintained by fewer and fewer (and less and less old) informants while [‑ˈter]‑ (along with [‑ˈteər‑] and [‑ˈtær‑]) gains ground more and more with young speakers. The case of voluntarily reminds us of e.g. customarily, militarily, necesssarily, temporarily, and so on.
-
(28) white (p. 898) [waɪt]: (77 %), [hwaɪt] (23 %). The difference between the 2 variants of white is the presence or absence of h initially. It remains to be seen if this difference applies to all pairs of wh‑words One wonders if 23 % for [hwaɪt] is due to the inclusion of some Scottish respondents in the surveys. Wells writes in ‘ʍ voiceless labio‑velar fricative/approximant. Scottish wh’. The single rising line represents [waɪt] and shows that this variant is definitely favoured by young speakers.
-
(29) with (p. 904): [wɪð] (85 %), [wɪθ] (15 %). Wells considers that [wɪθ] is ‘BrE non‑RP’. He provides a line for [wɪθ] which stays very low throughout, meaning that this variant is not favoured in BrE irrespective of ages. Wells provides no line for [wɪð] (85 %) which is definitely favoured in both young and old speakers in BrR. He additionally provides another line for AmE for [wɪθ] which stays high around 80 % throughout, meaning that [wɪθ] is definitely favoured in AmE for all ages. BrE with [wɪð] (85 %) and [wɪθ] (15 %) and AmE with [wɪθ] (84 %) and [wɪð] (16 %) is an interesting mirror‑image contrast.
-
(30) zebra (p. 918): ['zeb‑] (83 %), ['ziːb‑] (17 %). A single line shows ['zeb‑] which start from about 65 % at the ‘older’ end and rises continuously till it reaches just under 100 % towards and at the ‘younger’ end. If Wells provided a line for ['ziːb‑] as well, which continues at a low percentage throughout.
3.26
In bringing to a close Part Three entitled WELLS’S BRITISH PRONUNCIATION PREFERENCE SURVEYS, I wish to emphasize the fact that Wells’s present work was conducted in the spirit of ‘dynamic synchrony’, the concept which is well known and characteristically practised by a circle of linguists known as functional structuralists. Two decades ago, in 1991, at the end of presenting a paper () at a linguistics conference, I referred to Wells’s LPD2 (1990) in these words: ‘… it approximates at least in spirit what we functionalists may be encouraged to pursue with our own expertise in the future.’ At that time I had not yet consulted LPD2 (1990) with a fine comb, and I am happy to have had an opportunity to do just that in writing my present paper. As I already said earlier in this paper (3.0), to repeat, ‘To date, I know of no other BrE pronunciation preference surveys of quite the same magnitude and depth conducted by any other phoneticians than Wells.’ I will add here that LPD2 and LPD3 which contain the results of Wells’s British English pronunciation preference surveys outshine other English pronouncing dictionaries on the market at present. Minor shortcomings in Wells’s presentation of his British English pronunciation preference surveys have been duly pointed out here and there in the course of Part Three. The only major shortcoming, as I see it, is Wells’s decision to adopt in LPD3 the simple label ‘Preference survey, BrE’ by abandoning ‘BrE 1988 poll panel preference’ or its equivalent variant labels or ‘BrE 1998 poll panel preference’ or its equivalent variant labels, in all of which the year of survey, 1989 or 1998, as the case may be, is mentioned. These longer labels were quite rightly used in LPD2. Every use by Wells of ‘Preference survey, BrE’ is accompanied by the absence of another label indicating 1988 or 1998 as the case may be. This unfortunately hinder any serious potential investigation of the results of Wells’ meritorious surveys in the spirit of dynamic synchrony. It is hoped that these faults may be addressed in a future survey. Wells’s pronunciation preference surveys were a welcome step towards eventual investigation of British English pronunciation preference precisely in the spirit of ‘dynamic synchrony’, the like of which we have not otherwise so far observed. I also wish to pay tribute to the American pronunciation preference survey conducted by Yûko Shitara, although I need to add that regrettably Shitara did not take into account the ages of the informants, which reduces the quality of the final results from the point of view of dynamic synchrony.
4. PART FOUR. SOUTHERN STANDARD BRITISH PRONUNCIATION
4.1
The year 2019 saw the publication of a book written by Geoffrey Lindsey (hereafter GL) entitled English After RP: Standard British Pronunciation Today. GL characterizes RP as ‘the British upper class accent of the last century [my italics], Received Pronunciation (RP)’ (p. vii). Notice that the very title of Lindsey's book emphatically features the word Afterin italics, suggesting that RP is dead and buried in the 21st century. Besides, according to GL, the demise of RP happened precisely at the end of the 20th century. RP is ‘out of date’ (p. vii). The title of the Introduction (pp. i‑v) reads ‘The Fall of RP’ [my italics], as opposed to ‘The Rise of RP’ also in the Introduction.
4.2
GL asserts that the dominant pronunciation in Britain at present is that of London and the Southeast, and he says that this pronunciation can be called Standard Southern British pronunciation which he abbreviates as SSB pronunciation. SSB pronunciation should not be confused with RP.
GL writes as follows (p. 4).
Accents of the South, particularly middle and upper-middle class accents, are heard more often than others in public life … Southern speech of this type is a natural teaching standard for ‘British English’ … It’s an accent of England [as distinct from that of Scotland, Ireland, etc.]
As seen from the above quoted passage, GL makes a case for choosing SSB pronunciation as a model of British pronunciation to be taught to foreign learners of English pronunciation.
As for the passage from RP to SSB pronunciation, GL has the following to say.
4.3
Yet GL’s belief in the alleged demise of RP at a specific point in time is explicitly indicated in his following passage (p. 5).
A line was finally drawn under the British Empire over twenty years ago, with the handover of Hong Kong in 1997. The turn of the twenty-first century might be taken as a convenient point from which RP can be referred to in the past tense.
I reject such a statement as inappropriate, unrealistic and unacceptable. I do not believe that such an alleged cut-off point existed at the turn of the century. Granted that the handover of Hong Kong to the People’s Republic of China in 1997 was a significant political event, I disagree with GL that this political event caused a sudden demise of RP any more than did the tragic death in a car crash of Princess Diana of the UK which happened in Paris in the same year, 1997.
GL seems to assume that RP is currently non-existent as it is dead and buried. He describes RP as ‘the British upper class accent of the last century …’ (p. vii). He further writes: ‘… in the terms of its classical definition, it [i.e. RP] no longer really exists’ (p. viii). He indeed consistently makes references to RP in the past tense throughout his book.
4.4
It has to be admitted that RP is in decline but we need to reckon at the same time with factors that largely work in the direction of sustaining RP and its variants. First of all, RP is very much alive in the speech of those some of whom I have earlier identified in 1.11. Their speech is widely heard on the radio and television and in conversations in personal contact with them. As long as they continue to live and speak, RP will still continue to be alive. Secondly, in the UK, i.e. England, Wales, Scotland and Northern Ireland, there are a total of approximately 500 boarding schools such as Eton College, Harrow School, Rugby School, to mention just a few of them. Boarding schools, which are public schools (in the British sense) are residential schools where pupils live and study during the school year. The teaching staffs are generally speakers of RP or its variants. Pupils at boarding schools who hail from various localities in the UK come under the influence of their teachers’ RP. This results in sustaining and continuing RP.
4.5
The contemporary state of RP is such that it is still in existence not only after 1997 but also in 2019 when GL’s book in question was published, and furthermore, in 2024 when I am writing these lines. One cannot and should not ignore the continual existence of people who speak with RP. RP did not suddenly die at the turn of the 21st century. In fact, GL in his book seems ambivalent here and there about the continued existence or otherwise of RP in our days. I would suggest that RP should be looked at from the point of dynamic synchrony. RP and other types of English pronunciations are in coexistence and are in competition with each other before and after 1997 regardless of certain well-known events GL and I refer to. We need to look at RP and non-RP (SSB pronunciation) in terms of being recessive (RP) and progressive (non-RP (SSB pronunciation)).
4.6
GL’s intention in his book is to show us how SSB pronunciation differs from RP which he thinks is a thing of the past. In his book, GL provides what calls ‘Mini Dictionary’ (pp. 109‑145) a list of differences between RP (referred to as ‘Older’) and SSB pronunciation (referred to as ‘Newer’) for them to be compared with each other. Notice that GL’s use of a pair of words ‘Older’ (not ‘Old’) and ‘Newer’ (not ‘New’) which imply that RP and SSB pronunciation are both alive and coexist, that is, the former has not been replaced by the latter.
4.7
Here are just a few selected items of significant differences between RP and SSB pronunciation chosen from ‘Mini Dictionary’ which consists of 118 examples listed in alphabetical order. Presenting the following items in tabular form is mine. My presenting ‘Older’ first followed by ‘Newer’ is the reverse order adopted by GL. I will present in phonetic notation, e.g. [skuːl] and [skoːl] what GL presents in phonological notation, e.g. /skuːl/ and /skoːl/.
drip | drunk | drive | true | strong | drawing | halt | whole | |
---|---|---|---|---|---|---|---|---|
‘Older’ | [drɪp] | [drʌŋk] | [draɪv] | [truː] | [strɔŋ] | ['drɔːɪŋ] | [hɔːlt] | [həul] |
‘Newer’ | [dʒrɪp] | [dʒrʌŋk] | [ʤrɑjv] | [ʧrʉw] | [ʃʧrɔŋ] | ['droːrɪŋ] | [hɒlt] | [hɒl] |
4.7a
Various types of difference between ‘Older’ (RP) and ‘Newer’ (SSB pronunciation) such as shown above are actually quite limited in number, hardly any more. Practically all the other types of difference exemplified in ‘Mini Dictionary’ by GL consist of multiple variants of pronunciation of words which coexist already in RP. For example, (see Table 2), [hɔːlt] and [hɒlt] coexisted/coexist in RP, the former as the main variant and the latter the sub-variant. What has happened in our days is that, whilst both variants subsist, more speakers (principally young speakers) prefer to pronounce [hɒlt] while fewer speakers (older speakers) continue to pronounce [hɔːlt]. The balance between ‘Older’ and ‘Newer’ is thus progressively shifting in favour of ‘Newer’, but ‘Older’ is still in existence. ) gives 52 % for [hɒlt] and 48 % for [hɔːlt]. GL says that RP, i.e. ‘Older’, prefers [hɔːlt] while ‘Newer’ largely prefers [hɒlt]. For another word, salt(not in my Tables), indicates 57 % for [sɒlt] and 43 % for [sɔːlt] while GL says that RP, i.e. ‘Older’, prefers [sɔːlt] but ‘Newer’ largely prefers [sɒlt]. There is thus agreement between GL and Wells in the case of halt and salt. However, in the case of yet another word, false (not in my Tables), indicates 52 % for [fɔːls] and 48 % for [fɒls] and GL admits that, though there is a recent tendency to pronounce [fɒls] but that ‘THOUGHT pronunciation of false [i.e. fɔːls] is still widely heard.’ (p. 121). It seems that in certain cases GL’s generalized statement about the preference of [ɒ] and [ɔː] before /l/ (as in halt and salt) may not be altogether valid (as in false). GL says that the occurrence of [ɒ] or [ɔː] as we have seen is before /l/ followed by /t/ or /s/ in particular (p. 124) and cites words like ‘halt, assault, cauldron, false, salt’ (p. 125). We see that GL’s generalized statement is not always valid. It is more important to note that the difference is between the two percentages given by Wells in all such words cited is minimal.
4.7b
What is specifically striking is that cool, school and pool (see Table 1) are pronounced with [‑uːl] in ‘Older’ but [‑oːl] in ‘Newer’ so that e.g. pool and Paul, cool and call, you’ll and yawl, are presumably three homophonous pairs. [‑oːl] instead of [‑uːl] in the example words can only be a detrimental element in terms of intelligibility in general conversation in English. GL writes that [‑oːl] instead of [‑uːl] occurs in AmE before ‘dark [l]’ as in cool, pool, school, you’ll in many young people’s SSB pronunciation. He further writes (p. 28) that ‘… the backing of GOOSE merges it with the THOUGHT vowel, so that they [many younger speakers] use a quality in the region of [oː] and [uː] for pool and Paul, cool and call, and fool and fall’. He does not specifically say if [‑oːl] in question occurs in word‑final position, but it presumably is so. As for GL’s allusion to and description of [‑oːl] in cool, pool, school, fool and you’ll in SSB pronunciation, I personally cannot comment on them as the only example I have (often) heard is AmE pronunciation of school. Nor can I say if [l] of [‑oːl] is or is not pronounced in school.
4.8
Let’s now look at some linguistic impact of certain features of SSB pronunciation shown by GL. Let’s look at both [ʧ] /ʧ/ and [ʤ] /ʤ/ which occur in SSB pronunciation. In the table above are shown two of the examples, tube and institute (see Table 1), cited by GL, viz. tube ‘Older’ [tjuːb] vs ‘Newer’ [ʧʉwb], and institute ‘Older’ ['ɪnstɪtjuːt] vs ‘Newer’ ['ɪnstɪʧʉwt]. Also in the table above are shown two of the examples, during and graduate (see Table 1), cited by GL, viz. during ‘Older’ ['djʊərɪŋ] vs ‘Newer’ ['ʤɵərɪŋ / ['ʤoːrɪŋ], and ‘Older’ ['gradjʊɪt] vs ‘Newer’ ['graʤuːət].
4.8a
In RP, [ʧ] occurs word‑initially (e.g. chess), word‑medially (e.g. hatchet), word‑finally (e.g. pitch), and also in intraword position (e.g. cant you?). Likewise, [ʤ] occurs in e.g. jam, magic, edge, and could you?, in the respective contexts. In SSB pronunciation, [ʧ] additionally occurs in e.g. tube [ʧ‑] (instead of [tj‑]), attune [‑ʧ‑] (instead of [‑tj‑]), and in can’t you? [‑ʧ‑] (instead of [‑tj‑]). In a closed compound in which one stem ends with [t] at the end of one stem and the following stem begins with [ʃ] (e.g. courtyard < court + yard), [tj] remains [t +j] without coalescing into [ʧ]. Furthermore, in SSB pronunciation, [ʤ] occurs as in dew [ʤ‑] (instead of [dj‑]), as in endure [‑ʤ‑] (instead of [‑dj‑]), and as in could you? [‑ʤ‑] (instead of [‑dj‑]). In RP too, [‑ʧ‑] and [‑ʤ‑] can also occur in e.g. can’t you? and could you? One might think that there exist English closed compounds in which a phonetic phenomenon happens in which [dj] remains [d+j] without coalescing into [ʤ]. My research for such examples produced no results. The phonetic phenomenon [tj] > [ʧ] or [dj] > [ʤ] is known as ‘yod coalescence’. The additional forms [ʧ] and [ʤ] produced through yod coalescence (cf tube and during seen above) instead of [tj] and [dj] (which does not undergo ‘yod coalescence’) are bound to increase the occurrences of [ʧ] and [ʤ] in SSB pronunciation, both lexically and textually, as compared with those in RP.
Notice that LPD1, LPD2 and LPD3 consistently put [tjuː] (primary variant) and [ʧuː] (secondary variant) in items like contuse, intubate, tulip, tutor, etc. while they consistently put [ʧuː] (primary variant) and [tjuː] (secondary variant) in items like mature, statute, virtue, etc.
4.8b
What is the impact of the use in SBB pronunciation of [ʧ] rather than [tj] and of [ʤ] rather than [dj]? First of all, this gives rise to a small number of pairs of homophones. For example, chew [ʧuː] vs Tue [ʧuː] (< [tjuː]); dew [ʤuː] (< [djuː]]) vs Jew [ʤuː]; June [ʤuːn] vs dune [ʤuːn] (< [djuːn]); juice [ʤuːs] vs deuce [ʤuːs] (< [djuːs]); juke[ʤuːk] vs duke [ʤuːk] (< [djuːk]); duel ['ʤuːəl] vs jewel ['ʤuːəl] (< ['djuːəl]), and so on. However, suchlike homophones are not thought to cause serious problems in daily conversation in English. After all, there is always context which helps to disambiguate. From another point of view, what will be the phonological impact brought about by homophony in pairs of words adduced above? For one, the functional load of /tj/ vs /č/ and /dj/ vs /ǰ/ will be reduced. Also, phonotactically, /t/+/j/ and /d/+/j/ do not occur at least word‑initially.
4.8c
What attracts our particular attention is that GL presents the sequences [ʧr‑] (as in [ʧrʉw] true) instead of [tr‑] ([truː]),[ʃʧr‑] (as in [ʃʧrɔŋ] strong) instead of [str‑] ([strɔŋ]), and [ʤr‑] (as in [ʤrɑjv] drive) instead of [dr‑] ([draɪv]), as occurring in SSB pronunciation. GL’s explanations are as follows, which rather baffles me.
In RP, /tr/ and /dr/ clusters were pronounced as postalveolar affricates, but at a phonemic level [? my italics and my question mark] began with /t/ and /d/. Today, many younger speakers pronounce tr and dr words with clusters of /tʃ/ and /dʒ/ plus /r/. (p. 142)
In RP, the cluster /tr/ was pronounced as a postalveolar affricate, but at a phonemic level [? my italics and my question mark] began with /t/. Today, many younger speakers pronounce tr as a cluster of /tʃ/ and /r/; these speakers may pronounce str clusters with /ʃ/ rather than /s/. (p. 139)
4.8d
Before explaining why GL mentions [ʧr-], [ʃʧr-] and [ʤr-] happening, I need to quote warning strongly (and rightly) that ‘The distinction must be insisted on between … /ʧ/-/tr/, /ʤ/-/dr/ …’. in this connection cites near-minimal pairs (for /ʧ/ vs /tr/) cheese vs trees, chip vs trip, chap vs trap, chew vs true and chain vs train; and (for /ʤ/ vs /dr/) jest vs dressed, jaw vs draw, Jew vs drew and jam vs dram.’ Putting aside [r] in trees, trip, trap, true, train and dressed, draw, drew and dram, we can say that non-observance in SBB pronunciation system of /ʧ/ vs /tr/ and /ʤ/ vs /dr/, and the above-mentioned homophones chew vs Tue, dew vs Jew, dune vs June, deuce vs juice, duke vs juke, and duel vs jewel, makes SBB pronunciation a less effective verbal communication system than RP.
4.8e
GL (p. 161) writes that ‘RP allowed postalveolar /r/ to form clusters with the alveolar stops /t/ and /d/, but not with the postalveolar stops /tʃ/ and /dʒ/. Many speakers now use all-postalveolar clusters /tʃr/, /dʒr/.’ I believe that /tʃ/ and /dʒ/ are both palato-alveolar, not post-alveolar (cf ) and that /tʃr/ and /dʒr/ which GL describes as postalveolar are more likely to be palato-alveolar. On this second point, It need to be pointed out that Jones says () in connection with [tʃ] and [dʒ] that ‘the main part of the tongue is shaped nearly as for ʃ (Fig. 99) … the removal of the tongue is performed in such a way that the effect of the homorganic fricative ʃ is audible…’
4.8f
In connection with /dr/ and /tr/, writes:
When /d/ precedes /r/, the allophone of /r/ is fricative [my italics], the /d/ closure being released slowly enough to produce friction, e.g. in drive, tawdry and, in rapid speech at syllable or word boundaries, e.g. headrest, bedroom, wide road. … A completely devoiced fricative [r̥] may be heard following accented /p,t,k/, e.g. price, try, cream, oppress, attract, across.
The above quoted passage applies to drive and true in GL’s example (see supraTable 2).
4.8g
GL himself offers no explanation about why [tr] is pronounced [ʧr] and why [dr] is pronounced [ʤr]. A tentative explanation could be as follows. Fortition caused by [r̥] (< [r]) (voiceless post-alveolar fricative) results in [ʃ] (voiceless palato-alveolar fricative), so that [tr] (post-alveolar) changes into [ʧ] (palato-alveolar affricate), hence [tr] > [ʧr] which is a cluster of all three palato-alveolar consonants. Likewise, [dr] > [ʤr], another cluster of all three palato-alveolar consonants.
4.8h
GL’s statement that [str‑] (as in [strɒŋ] strong) being pronounced [ʃtr‑] (hence [ʃtrɒŋ] strong) is corroborated by Wells who says (in ) as follows.
‘Some speakers of BrE assimilate s to ʃ before tr and tʃ, thus strong strɒŋ → ʃtrɔŋ, student 'stjuːdənt → 'stʃuːd‑ → 'ʃtʃuːd‑. This is not shown in LPD. The EFL learners should not imitate it.
This statement is repeated in except that ‘This is not shown in LPD. The EFL learners should not imitate it.’ is changed to ‘This is not shown in this dictionary.’ What does the omission of ‘The EFL learners should not imitate it’ signify? I do not know.
4.8i
Interestingly, the sequence [ʃʧr] shows a new phonotactic development in English. Presumably, GL will phonologically notate [ʃʧr] as /ʃʧr/. He would probably suggest that, in SSB pronunciation, all words orthographically beginning with str- (such as Strabane, strabismus, Strabo … Strymon, Strzelecki) would tend to be pronounced [ʃʧr] instead of [str‑]. Most probably, in SSB pronunciation, words orthographically beginning with sht- (not shtr‑) such as shtetl, shtick, shtook, shtum and shtup are pronounced [ʃtetl], [ʃtɪk], [ʃtʊk], [ʃtʊm] and [ʃtʊp], respectively, with [ʃt] not [ʃʧ], since [ʃt] is not followed by [r]; English has not words beginning with shtr‑.
4.9
Let's now turn to some more selected items in GL’s ‘Mini Dictionary’. Those items show further aspects of SSB pronunciation. We will compare GL’s presentation of two variants of each item with the different percentages LPD3 attributes to them.
4.9a
distribute: ‘Older’ [dɪs'trɪbjuːt] vs ‘Newer’ ['dɪstrɪbjʉwt]. indicates [dɪs'trɪbjuːt] (74 %) vs ['dɪstrɪbjʉwt] (26 %). Considerable discrepancy seems to exist between GL and LDP3. Note that Wells says (ibid.): ‘The '‑‑‑, although disliked by many, is widely used in BrE.’
4.9b
applicable: ‘Older’ ['æplɪkəbl] vs ‘Newer’ [əp'lɪkəbl]. indicates ['æplɪkəbl] (15 %) vs [əp'lɪkəbl] (85 %). The substantial preponderance of ‑'‑‑‑ over '‑‑‑‑ is recognized by both GL and Wells. This is in sharp contrast with '‑‑‑‑ (64 %) over ‑'‑‑‑ (36 %) in AmE shown in LPD3 (p. 38). The preponderance of ‑'‑‑‑ in BrE (85 %) can hardly be said to be an influence of ‑'‑‑‑ (36 %) in AmE.
4.9c
harass: ‘Older’ ['hærəs] vs ‘Newer’ [hə'æs]. indicates ['hærəs] (68 %) vs [hə'ræs] (32 %). Wells adds a long note (ibid.) in the case of harass as follows (the italics in the quotation are Wells’s): ‘The traditional RP form is 'hærəs. The pronunciation hə'æs, which originated in the US, was seemingly first heard in Britain in the 1970s. In time it may predominate in BrE, as it already does in AmE. Meanwhile, it evokes negative feelings among those who use the traditional form.’ It is interesting in this connection to see that in Pronunciation Preference polls on AmE, the percentages are -'- (87 %) vs '-- (13 %). This is precisely the opposite from the percentages we have seen for BrE.
4.9d
primarily: ‘Older’ ['praɪmərɪlɪ] vs ‘Newer’ [praj'mɛrəlɪj]. indicates ['praɪmərɪlɪ] (51 %) vs [praj'mɛrəlɪj] (49 %). The difference between (51 %) for '‑‑‑‑ and (49 %) for ‑.‑‑‑ is minimal so that it is hard to see any clear‑cut transition from ‘Older’ to ‘Newer’ in this case. In connection with primarily, GL says (p. 136) that ‘Today this ['‑‑‑‑] has been almost completely replaced by a pronunciation with stress on the second syllable [‑'‑‑‑] … as in America.’, and he cites some other relevant examples, i.e. arbitrarily, necessarily, temporarily and voluntarily. GL’s statement and his examples are in agreement with what LPD3 indicates about necessarily '‑‑'‑‑‑ (68 %) (with '‑‑‑‑‑ (32 %) as a second variant) and voluntarily, '‑‑'‑‑‑ (68 %) (with '‑‑‑‑‑ (32 %) as a second variant), if not, though, in arbitrarily '‑‑‑‑‑ as a primary variant (with '‑‑'‑‑‑ as a second variant) and temporarily ' ‑‑‑‑‑ as a primary variant (with ,‑‑'‑‑‑ as a second variant).
As seen above, 4.9a., 4.9b., 4.9c. and 4.9d. are all concerned with matters of placement of accent affecting ‘Older’ and ‘Newer’ pronunciations. In 4.9e. and 4.9f. which will follow, I deal with two more items which are concerned with some segmental units.
4.9e
dissect: ‘Older’ [dɪ.sekt] vs ‘Newer’ [dɑj'sɛkt]. indicates [dɑj'sɛkt] (87 %) vs [dɪ'sekt] (11 %). The preponderance of [dɑj-] over [dɪ-] is recognized by both GL and Wells.
4.9f
various: ‘Older’ ['veərɪəs] vs ‘Newer’ ['vɛːrjəs]. shows ['veərɪəs] as the primary variant and ['værɪəs] as the secondary variant. What is shown as [eə] in [.veərɪəs] by both Wells and GL is the same vowel as some other British phoneticians (e.g. Jones, Gimson) represent as [ɛə]. This means that both [e] of [eə] and [ɛ] of [ɛə] represent the same vowel, that is, the same quality as Cardinal Vowel No. 3 [ɛ]. Consequently, ‘Older’ ['veərɪəs] is actually the same as ['vɛərɪəs] whereas ‘Newer’ [.vɛːrjəs] consists both in the disappearance of [ə] of [eə] and in the lengthening of [ɛ] which results in [ɛː]. In other words, the diphthongal [ɛə] changes to the monophthongal [ɛː]. This characteristic change is noted by, among other English phoneticians, , who gives examples pair, there, chairs, cared (with long [ɛː]) and scarce (with reduced [ɛˑ]). Cruttenden (ibid.) says that ‘Older speakers of GB [General British] may have a diphthong [ɛə].’
4.10
All four items we will see next, i.e. crescent, exit, aversion and Asia, involve the same phonetic feature in common to be discussed, namely, [s] vs [z], [ks] vs [gz], [ʃ] vs [ʒ], in short, voicelessness (in [s], [ks] and [ʃ]) vs voicedness (in [z], [gz] and [ʒ]).
4.10a
For crescent, GL indicates ‘Older’ ['kres(ə)nt] vs ‘Newer’ ['krez(ə)nt] compared with ['krez(ə)nt] (55 %) vs ['kres(ə)nt] (45 %). For exit, GL shows ‘Older’ ['eksɪt] vs (‘Newer’) ['egzɪt] compared with ['eksɪt] (55 %) vs ['egzɪt] (45 %). For aversion, GL indicates ‘Older’ [ə.vəːʃən] vs ‘Newer’ [ə.vəːʒən]. indicates [ə.vəːʃən] as the primary variant but [ə.vəːʒən] as a secondary variant which is ‘BrE non-RP’.LPD3 gives no percentage figure to either variant. Finally for Asia, GL indicates ‘Older’ ['eɪʃə] vs ‘Newer’ ['eɪʒə]. Compare this with ['eɪʒə] (64 %) vs ['eɪʃə] (36 %). Wells adds (p. 47): ‘those born before 1942, ['eɪʒə] (32 %), ['eɪʃə] (68 %.)’ From what LPD3 says, then, it is clear that the use of ['eɪʒə] has risen sharply and steadily over the years among young speakers.
4.10b
A possible reason for the increased use of the voiced version over the voiceless version of crescent, exit, aversion and Asia we have examined, though perhaps with the exception of crescent, may be influence of American usage for these items. This seems particularly to be the case for Asia. The overwhelming difference in AmE between the percentage for ['eɪʒə] and that for ['eɪʃə] as shown in is 91 % and 9 %, respectively. AmE pronunciation, recorded even 70 or so years ago in A pronouncing dictionary of American English (by Kenyon & Knot) shows [ə'vɝːʒən, ‑ʃən] (p. 34), and ['eʒə, 'eʃə] (p. 28), ['ɛɡzɪt, 'ɛksɪt] (p. 157) with the two variants in each case placed in this order. Only crescent is given ['krɛsn̩t] (but not ['krɛzn̩t]). GL says (p. 116) that Americans prefer /s/.
4.10c
forehead: (‘Older’) ['fɒrɪd]/['fɒred] vs (‘Newer’) ['foːhεd]; see Table 4. ['fɔːhed] (65 %) vs ['fɒrɪd] (35 %). GL does not give ['fɒrɪd] (even as a second variant) so that ['foːhεd] is understood to be the only (‘Newer’) pronunciation. In AmE, does not show what would correspond to ['fɒrɪd] in BrE, but shows two variants, viz. ‘with h’ (88 %) vs ‘no h’ (12 %). This latter pronunciation may probably be notated ['fɔːed] (or also ['fɔːr-ed] and/or ['fɔːrʔed]?).
4.11
A possible reason for the occurrence of ‘Newer’ pronunciations in some of the items we have seen and in a good number of other items we happen not to have seen because not listed by GL is what is known as ‘linguistic accommodation’. In the matter of pronunciation differences of the sort we have seen above, linguistic accommodation is a powerful factor among young speakers in particular. One can imagine various situations in which this may happen. Suppose, for instance, in the company of young speakers (English speakers in this case), one of them ― call him/her A ― is in the habit of pronouncing forehead [.fɒrɪd] while all the others ― call them B ― pronounce ['fɔːhed], What may happen? Both A and B immediately notice that they do not pronounce the word in the same way. ['fɔːhed] is what B, the majority, pronounce while ['fɒrɪd] is what A, the minority, pronounces. Camaraderie may bring about the situation in which A abandons ['fɒrɪd] in favour of ['fɔːhed] because A feels uneasy, embarrassing and self-conscious by standing out by pronouncing forehead ['fɒrɪd] not ['fɔːhed] because ['fɒrɪd] which is normal to A may be thought by B to be ‘old-fashioned’, ‘anachronistic’, etc. The conflict produced by the two alternative pronunciations of forehead, viz. [ˈfɒrɪd] and ['fɔːhed], can best be resolved by A dropping [ˈfɒrɪd] and joining B and learning to pronounce ['fɔːhed]. In other words, A accommodates himself/herself to B.
The following lines about accommodation will be both informative and interesting. They are taken from who briefly writes about ‘accommodation (in linguistics)’.
… where people alter their style of speech or accent to make it similar to that of their interlocutor, perhaps as a sign of identification. So the British Prime Minister Tony Blair was noted as using ‘Cockney’ glottalization, unusual for him normally, in a TV chat-show interview with the comedian Des O’Connor. The context, and the speech-act frame of anecdote, probably also influenced ‘street-wise’ informality.
The crucial point to Tony Blair being exemplified is the fact that Blair, as a young boy, attended Fettes College which is a public school and speaks RP.
4.12
We have had a look at some of the 118 items listed by GL in his ‘Mini Dictionary’. I would conclude that both RP and SSB pronunciation, which GL refers to as ‘Older’ pronunciation and ‘Newer’ pronunciation, respectively, form co-existing parts of the whole panorama. This panorama is not stably static but fluctuatingly changing and is best seized from the point of view of dynamic synchrony. I firmly reject GL’s idea that RP is succeeded by SSB pronunciation and that even at a certain date (GL suggests the year 1997).
4.13
This paper has revolved round the theme of ‘dynamic synchrony’ practised by André Martinet and his colleagues and disciples for many decades up to the present. Dynamic synchrony differs from and should not be confused with ‘static synchrony’ professed by Ferdinand de Saussure, who put forward the dichotomy ‘synchrony’ vs ‘diachrony’. The universally acknowledged fact that every language is changing every moment as it is used for communication makes languages a dynamic living ‘creature’. The dynamic nature of languages affects all aspects and domains of languages. The phonetic manifestation of languages is directly observable in human verbal communication. In this paper I am concerned with this particular area of language, specifically how native speakers of English exhibit preference among alternative pronunciations of individual words in terms of variants. The question relating to ‘RP’ and ‘non-RP’ is an exciting one for academicians and non-academicians alike. Pages and pages have been written by those who are fascinated by this question.
I started this paper with various aspects of RP and emphasized that RP is neither monolithic nor immutable. Dynamic synchrony plays a major part in what is happening to RP or non-RP. I have provided a relatively short account of dynamic synchrony. A sizeable group of French and Francophone linguists are working in that field, but this does not seem to be the case with Anglophone linguists. This is why I limited myself to a number of passages written in French on dynamic synchrony. The well-known Longman Pronunciation Dictionary by John Wells contains numerous interesting results obtained from the two pronunciation preference surveys he carried out in 1989 and 1999. These were surveys conducted in the spirit of dynamic synchrony. Finally, I turned my attention to Geoff Linsey’s English after RP which describes RP as moribund (as he perceives so) and highlights the emergence of Standard South British (SSB) Pronunciation. He abundantly illustrates RP and SSB pronunciations with many examples in which the contrast between the two is shown.
The readers will, I hope, have seen that RP and non-RP provide a rich field for different views and lively discussions from both sides. As a non-native speaker of English nurtured in RP, I have my own view which would be better kept to myself.
References
2
3
5
Akamatsu, Tsutomu (2001): “The development of functionalism from the Prague school to the present”. In S. Auroux et al. (eds.): History of the Language Sciences: An International Handbook on the Evolution of the Study of Language from the Beginnings to the Present. Berlin / New York, Walter de Gruyter, vol. 2, 1768‑89.
6
7
8
Akamatsu, Tsutomu (2023): “RP and dynamic synchrony”. In J. B. Coyos (ed.): Actes du XLIII. Colloque International de Linguistique Fonctionnelle. Hizkuntzalaritza Funtzionalaren 43. Nazioarteko Jardunaldiko Aktak (2022ko urriaren 4-7 / 4-7 Octobre 2022. Donostia / Saint-Sébastien). Académie de la langue basque / Euskaltzaindia, 155‑61.
13
15
16
17
18
19
20
23
24
26
28
30
31
32
33
34
35
Jones, Daniel (182011 [172006, 162003, 151997,141977, 131967, 121963, 111956, 101949, 91948, 81947, 71945, 61944, 51940, 41937, 31926, 21924, 11917]): An English Pronouncing Dictionary. London: Dent. Note: 151997, 162003, 172006 and 182011 were published as Cambridge English Pronouncing Dictionary by Cambridge University Press.
36
37
39
42
43
45
46
47
48
49
51
52
53
54
Oladipupo, Rotimi Olanrele & Aderonke Akinola (2022): “Nigerian English Pronunciation Preferences: A corpus-based survey of pronunciation variants”. Cogent Arts & Humanities 9/1. https://www.tandfonline.com/doi/full/10.1080/23311983.2022.2061104.
55
58
61
62
64
65
66
68
69
71
72
73
74
76
77
80
Notes
[1] This paper is a substantially expanded version of Akamatsu (2023). Many thanks are due to two anonymous reviewers for their helpful advice given for stylistic improvement of the original manuscript of the paper.
[3] A Critical Pronouncing Dictionary and Expositor of the English language (London, 1791). writes: ‘Walker concluded that the pronunciation of London “is undoubtedly the best”: “that is, not only the best by courtesy, and because it happens to be the pronunciation of the capital, but best by a better title, that of being more generally received.”’
[5] Although the term ‘accent’ is actually not mentioned in these terms, they do specifically refer to the accent known as Received Pronunciation.
[8] Following the decease of Queen Elizabeth II on 8 September 2022, is the appellation ‘Queen’s English’ to be ever replaced by ‘King’s English’ after King Charles III’s coronation that took place on 6 May 2023? Most probably not.
[10] says: ‘We may extend this model to cover Wales as well as England without serious modification. But we cannot extend it to Scotland or Ireland, or of course to outside the British Isles.’
[11] : ‘I have used the word dialect here to mean any form of English which differs from Standard English in grammar, syntax, vocabulary, and, of course, in pronunciation too …’
[15] This is from the age of around 11 (or even earlier) to 18, no doubt an impressionable period of language learning.
[16] The public schools exist, in the main, in England, Scotland, Wales, Northern Ireland, but also abroad as well nowadays.
[17] : ‘… the higher a person’s social-class standing, or alternatively the greater the length of someone’s formal education, the greater the likelihood that he or she will speak RP …’
[18] For the years 2022/23, for instance, Gordonstoun School (Scotland) charged £33,900 for Day fee and £46,650 for Boarder, while Cedars School of Excellence (Scotland) charged £7, 950.
[19] I have reservations for what are known as ‘sociolinguistics’ and ‘sociolinguist’. I am in agreement with who says: ‘La sociolinguistique et la pragmatique, qui se présentent ailleurs comme des disciplines distinctes, sont, chez nous, partfaitement intégrées dans une seule et même pratique linguistique.’
[20] Here are a few examples, Boris Johnson (Eton College,), Tony Blair (Fettes College), Rishi Sunak (Winchester College), Alec Douglas Home (Eton College), and Douglas Hurd (Eton College). Back in 2019, two‑thirds of cabinet ministers of the United Kingdom had been educated at such fee‑charging schools (public schools).
[22] refers in this connection to , , and . employs the term ‘Modern Received Pronunciation’ which she immediately abbreviates as ‘MRP’. I find that where she writes about ‘Attitudes’ to various (mainly English) accents does not actually mention people’s attitudes to ‘MRP’, though Cruttenden refers to . Modern RP is also said to be mainstream RP.
[23] (The following information is largely based on .) The other members were Robert Bridges (Poet Laureate and a founder of the Society for Pure English), Arthur Lloyd James (the Committee’s honorary secretary and a Welsh phonetician at the School of Oriental and African Studies, University of London), Sir Johnston Forbes-Robertson (actor), Logan Pearsall Smith (naturalized American scholar), and George Bernard Shaw (Irish playwright and critic). Daniel Jones was at that time Professor of Phonetics at University College London. The Committee was dissolved at the beginning of World War II in 1939, and day-to-day advice work fell on the shoulders of Gertrude M Miller (Assistant secretary to the defunct Committee) and Elspeth D Anderson, who were both U.C.L.‑trained phoneticians; they were both Scots. The Committee was not reactivated, and it became known in 1940’s as the BBC Pronunciation Unit, which was staffed by Miller and Anderson. As time went on, there have since been changes in the membership involving Graham Pointon, Hazel Wright, Lena Olausson and Catherine Sangster, et al., if not necessarily in this order.
[24] says that ‘A common form of Received Southern English contains twelve essential diphthong phonemes. Nine of these are … ei, ou, ai, au, ɔi, iə, ɛə, ɔə, uə [Jones’s boldface]. In addition to the 9 essential diphthongs, Jones adds 3 non-essential diphthongs, viz. ĭə, ŭə and ŭi. However, Jones hastens to say (ibid.) that ‘Of the above diphthongs two, ɔə and ŭi, may be ignored by the foreign learner, the first because many Southern English people (including myself) never use it but replace it by ɔː … and the second because it can always be replaced by disyllabic u-i …’. This will bring down the number of the essential diphthongs from nine to eight, and the number of the non-essential diphthongs from 3 to 2. The total of the diphthongs (both essential and non‑essential) in Received Southern English will be 10. On the other hand, according to Abercrombie (1979), Scottish Standard English has 13 vowel phonemes, of which 10 are monophthongs (/i e ɛ ɪ a ʉ ə o ɔ ʌ/) and 3 are diphthongs (/ai ʌʉ ɔe/).
[26] : ‘The great majority of the people of Britain do not speak RP, though all of them understand this variety of accent because they hear it as the form most commonly used on radio and television and from public figures.’
[27] Burchfield (1986: 141) has the following to say about , i.e. The Pronunciation of English (1st ed.): ‘In sixty-nine pages, with the utmost clarity, he described the careful conversational style, or StP (standard pronunciation) as he called it, with which he was concerned.’ Jones’s The Pronunciation of English has since grown in size and the 4th ed., , consists of 223 pages.’
[28] In her Preface to the Fourth Edition, i.e. , Susan Ramsaran writes (p: vii): ‘… I have undertaken as one who over seventeen years was successively a student, colleague and collaborator of Professor Gimson..
[29] I quote here at some length what : ‘… this dictionary excludes any British pronunciations which are associated specifically and only with a public boarding-school or any socially conspicuous background … This most general type of educated British pronunciations … It is a convenient parallelism with the term General American and a welcome avoidance of the ‘less than happy’ archaic-sounding term ‘Received’ to abbreviate this to simply General British pronunciation (GB).’
[30] In 4Collins-Mees-Carley: 63, of [ts], [dz], [tr], [dr], [tʃ] and [dʒ] all of which presents as the English affricates, [ts] and [dz] do not appear but [tr] and [dr] do in the consonant system; on the other hand, only [ʧ] and [ʤ] are presented as the English affricates.
[31] : ‘… l’impression qui se dégage de son [Saussure’s] enseignement est qu’une description synchronique est nécessairement statique, c’est-à-dire qu’elle doit éliminer toute référence à quelque changement en procès.’
[32] has this to say, quite appropriately: ‘… ce qui a amené Martinet à dépasser la dichotomie diachronie/synchronie établie par Saussure et à poser le concept de synchronie dynamique, qui permet de rendre compte des langues dans leur fonctionnement.’
[34] Note specifically Martinet’s mention of Age at the beginning of his questionnaire () and that he writes thus (op. cit: 14): Il est à peine besoin de justifier la question sur l’âge du sujet, puisqu’il est un fait acquis que, dans un milieu donné la prononciation varie selon les générations.’ But see further also (op. cit.: 33‑34).
[35] : ‘… ce sont les concepts de dynamique linguistique et de synchronie dynamique, dont André Martinet nous a fourni une première évidence dans La Prononciation du français [contemporain …] et dont s’occuperont ici même Tsutomu Akamatsu, Horst Geckeler, Cecilia Hare, Alphonse Leguil, Pierre Martin, Gillette Staudacher-Valliamée, Agustin Uruburu Bidaurrazaga et Henriette Walter …’ I myself briefly touched on the concept and term of dynamic synchrony in the past in , , , , and .
[36] Though this paper by Moulton is highly interesting, I am far from happy about his recourse to ‘underlying form’, ‘phonological rule’ and ‘deletion (of a phonological rule)’.
[37] : ‘… ce soit dans le domaine de la description phonologique qu’aient apparu les premières tentatives pour dégager, en synchronie, des tendances évolutives.’
[38] The survey conducted in 2007 by Pearson Education and Wells was on 30 selected words including contribute, dissect, kilometre, liquorice, omega, tinnitus and tun. For AmE (American English) there exists a survey conducted in 1993 by Yûko Shitara. Also for AmE, another survey was conducted in 2003 by Bert Vaux, on words including almond, asterisk, caramel, chromosome, Presley, Thanksgiving and thespian.
[39] I wish to mention that in the past I dealt in some detail with Wells’s British pronunciation preference surveys in and if only briefly touched on it in .
[40] I should add that, though falling outside the confines of this paper of mine, there exists an excellent corpus-based survey of Nigerian English pronunciation preference which was published very recently, i.e. . The goal of this survey is, as the authors put it, ‘to develop a framework for the codification and standardisation of Nigerian English.’ This extensive work is also worth a perusal.
[41] Details of the breakdown of the 275 native speakers of BrE are given in LPD1 (1990: xi) as follows. ‘The panel comprised 62 academic phonetics/linguistics specialists, 51 school or college teachers, 40 students (mainly of speech sciences or linguistics) at London University, 31 broadcasters (mostly radio announcers with the BBC) and actors, 13 speech scientists and engineers, and 78 other volunteers from the general public, with occupations such as High Court judge, diplomat, advertising copywriter, and stained glass craftsman. They were recruited either by personal contact or in response to a general invitation in a Sunday newspaper. They ranged in age from 15 to over 80. Most of them (254, or 92 %) described themselves as English, and of these 103 had grown up in the north of England, 133 in the south (and 18 had moved around); 10 described themselves as Welsh and 11 as Scots. 149 (54 %) were male, 126 (46 %) female. The survey was carried out between November 1988 and February 1989.’ One thing Wells does not and cannot say about his informants is what proportion of his informants were speakers of RP. This is inevitable since his surveys were conducted by means of questionnaires sent to the informants. Judging from Wells’s description of the panel of 275 informants, it can be surmised that a reasonable amount of the informants were probably speakers of RP.
[42] Bert Vaux survey on words including among others the words almond, asterisk, caramel, chromosome, Presley, Thanksgiving and thespian.
[43] This survey was on 30 words including contribute, dissect, kilometre, liquorice, omega, tinnitus and tune.[
[44] accept/except and liquorice/licorice count, respectively, as one item, and during is counted twice, as the two variant forms are [dʒ‑] / [dj‑] in one case and [‑ʊə‑] [‑ɜː‑]/[‑ɔː‑] in the other.
[46] Actually, there are items which were already also recorded in LPD2, but I deliberately refer to only LPD3 which is so far the final edition of LPD published.
[47] For instance, in LPD2, absorb, address, again, ally (v), baptize, etc. are entered, On the other hand, almond, amphitheater, asterisk, caramel, chromosome, etc. are not entered in LPD2. All the items (all other citable items) are entered in LPD3.
[48] In order to show in LPD3 the results of his pronunciation preference surveys conducted in 1989 and 1999, Wells resorts to diverse means. First of all, verbal information which is without exception provided to each item, for example, ‘Preference poll, BrE: [‑.sɜːd] 77 %, [‑.zɜːd’] 23 %’ for absurd. He also resorts to 3 types of graphs, namely, pie graphs, bar graphs and line graphs. For the purpose of my present paper, I will only take into account Wells’s line graphs which alone faciliate to view his data from the point of view of dynamic synchrony.
[49] In LPD3 (p. 5) Wells merely mentions (deliberately) ‘Preference poll, BrE’, but we know it actually refers to ‘BrE 1998 poll panel’ mentioned in LPD2 (p. 5). This point is raised and discussed infra in 3.13. Wells’s use of the labels ‘BrE 1988’, ‘BrE 1998’, and ‘Preference poll(s), BrE’.
[50] Wells only says ‘Preference poll, BrE’. He says in LPD2 that the information derives from ‘BrE 1998 poll panel preference’. The item absurd appears also in LPD3 though, curiously, Wells fails to say whether the information derives from the 1988 poll or the 1998 poll.
[51] In LPD2 the result of the 1998 survey for Asia is indicated as follows: ‘[ˈeɪʃə] 49 %, [ˈeɪʒə] 51 %.’ However, In LPD3 the result is significantly changed as follows: ‘BrE: [ˈeɪʒə] 64 %, [ˈeɪʃə] 36 %; those born before 1942, [ˈeɪʒə] 32 %, [ˈeɪʃə] 68 %; See chart on p. 48.’
[52] For applicable, BrE 1988, [əˈplɪk‑] 77 %, [ˈæplɪk‑] 23 % and BrE 1998, [əˈplɪk‑] 84 %, [ˈæplɪk‑] 16 % in LPD2 are changed to BrE’:, [əˈplɪk‑] 85 %, [ˈæplɪk‑] 15 % in LPD3. For chrysanthemum, BrE 1988, [‑ˈsæntθ‑] 61 %, [‑ˈzæntθ‑] 39 %; BrE 1998 [ˈsæntθ‑] 63 %, [‑ˈzæntθ‑] 39 % in LPD2 are changed to BrE: 1998 [‑ˈsæntθ‑] 63 %, [‑ˈzæntθ‑] 37 % in LPD3. It is clearly seen that in LPD2 (but not in LPD3), both BrE 1988 and BrE 1998 are indicated for applicable and chrysanthemum. This is satisfactory. On the other hand, In LPD3, for both applicable and chrysanthemum, just BrE (but neither BrE 1988 nor BrE 1998) is indicated. This is not satisfactory.
[53] A line graph takes the form of an x-axis and a y-axis which are joined at a right angle at the extreme left end of x-axis and at the bottom of y-axis, thus forming an ‘L’ shape.
[54] Wells operates with four age brackets, viz. ‘born up to 1933’, ‘1934-53’, ‘1954-73’ and ‘since 1973’ in his BrE 98 survey.
[61] Notice Lindsey’s use of a singular form. Contrast this with Lindsey's use of plural form ‘accents of the South’, which he uses at the beginning of the passage quoted here.
[62] GL himself writes ‘dark /l/’ which should correctly be ‘dark [l]’. He also writes ‘dark /l/ [ɫ]’.
[63] [ʉw] as used by GL represents a diphthong which corresponds to [uː] (a monophthong) as in [ʧuːb] in RP. [ʉ] stands for close central rounded vowel.
[64] The phonetic symbol [ɵ] (close-mid central unrounded vowel) should not be confused with [θ] (voiceless dental fricative).
[66] This is probably why I cannot find any relevant statements in phoneticians’ English phonetics manuals.
[67] Jones (41956: 841a) writes: ‘Contexual assimilations may be described by the following formulae: … (ii) (for coalescent assimilation) the sounds A and C influence each other and coalesce into the single sound B [Jones’s italics]. … t and j influence each other and coalesce into the single sound tʃ.’ Jones does not employ the term ‘coalescent assimilation’. Also see LPD1 (p. 47) under Assimilation where yod coalescence is also called coalescent assimilation, and see the examples tune [tjuːn] → †[ʧuːn] and endure [ɪn.djʊə] → †[ɪn.dʒʊə]. Notice that the symbol † which Wells places in front of [ʧuːn] and [ɪn.dʒʊə] means ‘BrE non-RP’. Wells says in LPD1 (p. xii) that ‘Pronunciations widespread in England among educated speakers, but which are nevertheless judged to fall outside RP, are marked with the special sign †’. Wells in LPD2 (p. xiii) writes that ‘Speech with local features of the southeast of England is often referred to as Estuary English … and, among other such features, cites [tj] > [ʧ] and [dj] > [dʒ] and mentions ‘use of [ʧ] and [dʒ] in place of [tj] and [dj], thus tune [ʧuːn], reduce [rɪ ˈdʒuːs] (= yod coalescence …).’ In LPD2 Wells replaces the symbol † used in LPD1 by a different symbol § which means ‘BrE non-RP’. In LPD2 (p. 50) where he explains yod coalescence, Wells says y-coalescence as in tune [tjuːn] → §[ʧuːn] and endure [ɪn.djʊə] → §[ɪn.dʒʊə] ‘can frequently be heard in BrE, although not in careful RP.’ However, Wells’s previous treatment in both LPD1 and LPD2 of [ʧ] and [ʤ] when they are products of yod coalescence has drastically changed as he writes in LPD3 (p. xiii): ‘Following research into contemporary RP by Bente Hanniasal [University of Bergen], I have removed the § sign from forms with [ʧuː] and [dʒuː].’ His earlier statement in LPD2 about one of the features of Estuary English, viz. the use of [ʧ] and [dʒ] resulting from yod coalescence is retained in LPD3 (p. xix). All the same, Wells in LPD3 (p. 52) repeats and retains what he said in LPD2 (p. 50) about [tjuː]→ [ʧuː] and [djuː] → [dʒuː] in yod coalescence are ‘not in careful RP’. In the final analysis, Wells’s stance on the status of [ʧ] and [dʒ] (RP or non-RP?) resulting from yod coalescence seems ambivalent.
[68] This monophthongal [ɛː], and [üː,ɯ̈, ɨɯ] (unrounded and centralized) instead of [uː] (as in food), personally strike my ear conspicuously while listening to young speakers’ speech.