This is an updating and revision of an article first published in 1987 in the Journal of the International Phonetic Association
The massive growth of English as an international language has meant that many among the very large numbers of those who are engaged in teaching it have been eager to possess a full description of the features of the spoken variety they employ. The first really full account of a suitable British model was found in Daniel Jones's Outline of English Phonetics, notably as it appeared in 1932. For thirty years this was unrivalled. Despite numerous features that have become outdated and others which were never entirely suitable treatments of their subject matter, this book can perhaps be said to have still not quite been entirely superseded. It was reprinted in 2002 in a collection of his works edited by B. S. Collins and I. Mees.
In 1962 A. C. Gimson's over-modestly titled Introduction to the Pronunciation of English immediately showed itself to be a better organised and more up-to-date account but, initially at least, markedly less oriented to the needs of the EAL (aka EFL) teacher and learner. Originally, less than five percent of its content consisted of advice for this audience. Its second revision in 1980 more than doubled its EFL coverage by the addition of an appendix specifically on 'Teaching the Pronunciation of English'.
However, this served to highlight how greatly EFL considerations had been neglected in the main body of the book. Purely as a description of the facts, Gimson's book was in some ways less complete and up-to-date than J. C. Wells's three-volume account of mothertongue English worldwide, Accents of English which was in no way aimed at teachers or learners. Both of these books were quite difficult for the reader who had no general grounding in phonetics. At any rate, the Gimson work has been further improved in EFL terms in the course of its various subsequent editions, especially since 1994 when it was taken over and extensively re-written by Alan Cruttenden who has continued to revise it with ever-increasing admirable improvements.
There are two major varieties that are in most parts of the world accepted as the most useful broad types to serve as models of pronunciation for learners of English who do not have the language as their mother tongue. Neither of them can be delimited with any great scientific exactness but the most characteristic features of each are fairly easy to state. One is the variety most commonly referred to as General American (GA). This is most simply described, with an acknowledged looseness which nevertheless has its convenience, in a negative way. That is to say it is the pronunciation which is not marked by any regional features of the southern or of the Atlantic seaboard states of the USA and Canada. Its most striking and pervasive single feature is what may be called its relatively high 'rhoticity' meaning that virtually every 'r' letter of the spelling is pronounced regularly by its speakers. It has at least two other major features which may be mentioned. One of these is that it has the same vowel type in words like stop and palm, unlike most other forms of English. The other is its preference for the hat-type vowel in words like bath, ask and chance.
The other principal EFL model does not have such a universally accepted title. In the language-teaching world and most widely in recent decades it has been best known by the initials 'RP' which stand for the expression Received Pronunciation. Despite its widespread use there have been a number of objections to this term. It employs the word 'received' in a sense which had become more or less obsolete even in Victorian days, that of 'socially accepted'. This is in such an application unfortunately question-begging. It suggests to many that the speech of anyone who does not conform to it is being condemned as not socially acceptable. This is a widely offensive and indefensible suggestion. From a worldwide view the expression is not only 'complacent' as H. L. Mencken, the author of The American Language dubbed it, but embarrassingly parochial. Despite its propagation by such important writers as, in the first place, Daniel Jones (from 1926) and subsequently David Abercrombie, A. C. Gimson and J. C. Wells it is surely a regrettable usage that is better avoided. Terms that have had some essentially informal currency like Standard or Educated pronunciation are open to similar objections. It was pointed out for the first time in Windsor Lewis 1972 (p.xiv) that, by a convenient parallelism with General American, the kind of English pronunciation most widespread throughout Great Britain could well be designated 'General British' (GB).
GB is not the only 'educated' variety of English in Britain. Its spread was due to the influence of London as the social, legal and commercial centre of the country. Some writers use terms like 'Southern British Standard' but these are plainly unsuitable when in the rest of the country it is possible to find considerable numbers of people with in at least most respects GB accents who have spent virtually the whole of their formative years without residence in the south. It is perfectly true that the further one goes from London the smaller becomes the proportion of speakers with no trace of local influence in their speech. Indeed the importance of GB, as also of GA, lies not at all in the numbers of speakers who have the accent with absolutely no regional admixture but in its function as a common denominator for huge numbers of people who may or may not have it with a slight or more considerable admixture. Anyone who speaks with one of these two accents unmixed will be likely to be referred to popularly in the respective country as having 'no accent'. This neutrality of association with any region within either country is what makes each of GA and GB so eminently suitable as a model for users of English as an additional language.
The other accents of English, again only very loosely delimitable, employed by non-dialect-speaking people in the British Isles include the Irish, Scottish, Welsh, Western, Midland and Northern educated varieties. Like GA the Irish, Scottish and West-of-England varieties exhibit higher rhoticity than the rest. Most of these are in more or less general agreement in treating bath, ask and chance words like GA rather than GB. Many Northern varieties treat words like cupboard similarly to most GA speakers in having the same vowel in both syllables. The unique characteristic of most northerly varieties is to have a "clear" vowel rather than a schwa in the 'strong' but unstressed syllables beginning with the prefixes con-, ad- etc. See §7.4 ¶¶ 4-13 on this website. Most Scottish varieties are strikingly independent in making no contrasts between the vowels of good and food or between those of Sam and psalm. Like some forms of GA they may not have different vowels in caught and cot. Australian, New Zealand and South African varieties are mainly more like GB than GA.