1. The Daniel Jones English Pronouncing Dictionary (EPD), so deservedly for so long held in such universal high repute, first appeared in 1917 but the transcription it embodied was not invented by Jones. For about thirty years it had already existed in one or other of several very slightly different forms. It had been born in effect in 1888 when the principles adopted for their new international alphabet were agreed on by the International Phonetic Association. Their alphabet was essentially Henry Sweet's "Broad Romic" alphabet with the use of the symbol [ʌ] (with a different value from that which Sweet had attached to it) proposed by Otto Jespersen the Great Danish scholar at whose suggestion the alphabet was drawn up, and also the convention of employing a colon placed after any symbol to suggest that its sound was a long variety. Sweet himself had doubled the symbol for vowels to indicate greater length. These were only the arm and fur vowels – still those for which some length suggestion is most appropriate.
2. The precise set of symbols of the famous "EPD" transcription were used by Jones in new books only in the five years from 1912 (Phonetic Readings in English) to 1917 (EPD) in the earlier part of which he wrote his Outline of English Phonetics, whose delayed publication took place in 1918. Before these, in his two other phonetic readers Phonetic Transcriptions of English Prose (1907) and Intonation Curves (1909) and in his Pronunciation of English (1909) he used transcriptions with by no means exactly the same vowel symbols. Afterwards, in his book The Phoneme (1950) and his complete re-writing of the long-out-of-print Pronunciation of English in 1950 and in various of his contributions to Le Maître Phonétique (the Journal of the International Phonetic Association as it is now called) he used yet other sets of symbols for the English vowels. It was the remarkable success of the three books in EPD transcription which largely prevented his changing the transcription in them. He regretted having used it in them but he seems to have been too busy meeting the demand for new editions and reprints of them to do anything about the matter.
3. From about 1916 Jones used a type of transcription which appeared in print first – under his auspices – in L. E. Armstrong's English Phonetic Reader (1923) and subsequently in much the same form in the two most important descriptions of British English pronunciation since Jones, namely The Phonetics of English (1929) by Ida Ward and A. C. Gimson's Introduction to the Pronunciation of English (1962 etc). It appeared also in David Abercrombie's English Phonetic Texts (1964). In the Gimson book alone of these, the colon is employed as a "blanket" length mark – a fact actually deplored by Daniel Jones. Jones's ultimate preference for a transcription to be used in the teaching of English as a foreign language was for the set of symbols he referred to as the "Simplified Transcription" of English. He devised it around about 1930 and encouraged its use in phonetic readers by N. C. Scott (English Conversations 1942) and E. L. Tibbits (A Phonetic Reader 1946). It was used in certain works by A. S. Hornby including the Oxford University Press English-Reader's Dictionary (1952) from which it was abandoned in the second edition. It was the "house" transcription of English Language Teaching from 1946 but for some years.
4. Certainly a far better transcription than the EPD, it was admirable from the purely typographical point of view, but it is now little used. Its form hinged on a decision which Jones knew to be controversial and is no longer acceptable. Up to the time when his health still permitted him to deal with such matters, in the mid fifties, although fully recognising that the "tendency to lengthen 'short' vowels" appeared to be "on the increase" he still retained the impression he had had when he first referred to this phenomenon in 1932 that the "normal educated speech" of Britain, unlike that of America, did not as he put it lengthen the "traditionally short" vowels even in situations such as when eg yet or this is uttered with a falling-rising tone. He recognised the possibility that a new development of the language was "beginning to take place, by which the ... distinctions of quantity combined with quality" would "eventually give place to distinctions of quality only". Hardly any phonetician today would uphold the view he took, no doubt rather idiosyncratically even of his own Victorian English, that the length can be said to constitute the fundamental difference. And most would probably agree that the lengthenings he mentioned as, according to his impressions, still minority usages are now at least as characteristic of the most general types of British English as the older shorter usages.
5. The distinguished authority on the rhythm of British English, the late Professor David Abercrombie, in a private conversation with the writer, expressed the view that the type Jones regarded as minority usage was by the seventies clearly the usage of the majority. It is noteworthy that Jones's opinion was that "If such a new system of vowel-length were to become the regular usage in educated Southern English, it would become necessary to modify the method of vowel representation in practical phonetic transcription by introducing special letters to distinguish the pairs of sounds now sufficiently well distinguished by length-marks" (0utline §879).
6. This brings me to the circumstances in which I planned the pronunciation representations for my 1972 Concise Pronouncing Dictionary of British and American English and the 1974 edition of the Oxford Advanced Learner's Dictionary of Current English. First of all, the old EPD transcription had been rejected by Gimson in his admirable Introduction and even more importantly he had (i) made one modification of the EPD transcription already in his interim revision of 1967 viz /ou/ to /əu / (ii) he had firm plans to bring the EPD much more in line with his Introduction in his next and fuller revision of that dictionary.
7. In the 1967 EPD Editor's Preface Gimson said "it seems probable that the first pronunciation of such words as fire and power will have, in an early edition, to be shown with a diphthong or monophthong rather than with a sequence of three vowels". For the specialized phonetics-orientated public of his Introduction such a change would have been quite unexceptionable: for the non-technical general public whose convenience I had chiefly in mind my pedagogical dictionary work it would have been less easy to justify. In fact Gimson never carried through such a procedure.
8. Gimson's revision of the EPD ultimately retained the colons of his Introduction. This proved a very popular choice in Germany and Scandinavia etc where they were part of a long pedagogical tradition. As a theoretical choice I had misgivings about their redundancy and very frequent inappropriateness. Pedagogically speaking, to retain them I felt risked diverting attention from the all-important quality contrasts to the very uncertain length contrasts of General British English pronunciation. He also decided to retain the then rather unfamiliar "sawn-off b" shape for the got vowel / ɒ / in the interests of the specialist student of phonetics. In terms of pedagogical effectiveness and legibility I was reluctant to use a transcription which entirely eliminated one of the only five vowel letters of the normal English alphabet in favour of a letter so unfamiliar to the general public. Gimson in the EPD, though not in his Introduction until it was re-edited after his death by Dr S. Ramsaran, generalised one kind of a to the beginnings of the five and now diphthongs and avoided also his Introduction (and the EPD's) unfamiliar Greek e [ɛ] in the hair diphthong. In the interests of his specialist public again he maintained a different a for the arm vowel [ɑ] from the one he used for those diphthongs and for the hat vowel for which he retained the a/e ligature /ӕ /. I certainly went out of my way to follow his Introduction – still the best description of current British English – as far as one possibly could without disadvantage to the general reader who merely wants information on how words are usually uttered. To say that my CPD/ALD symbols were harder to learn than the old EPD set was hardly true. I used three new symbols but at the same time discarded two EPD symbols that could be dispensed with. The total was one extra vowel symbol to learn. The advantage was that the user was enabled to give up the false idea, responsible no doubt for many poor pronunciations from Germans and other nationalities, that some vowels are always long and others always short in English.
9. The traditional habit of reference to some English vowels (ie vowel phonemes) as long and others as short is too easily misunderstood. There is a complicated set of length relationships operating in the English system of vowels but at its most powerful (when vowels precede the "sharp" consonants [p t k ʧ f θ s ʃ ]) it serves to identify the following consonants rather than the vowels themselves. I can recall cases of total misunderstanding where a non-native speaker of English failed to shorten a so-called "long" vowel such as when an intended beat is produced in a form one hears as bead. I once, for example, thought a whole conversation with a native speaker of Danish was about a Spanish-speaking person called Cuba when it was meant to refer to an Englishman called Cooper. (For more detailed discussion of some of the above topics, see the writer's article 'The Notation of the General British English segments' in the Journal of the International Phonetic Association Vol. 2 No. 2. December 1972 a version of which follows immediately on this website.)
10. Finally the Kingdon-style "tonetic" type of stress notation of the CPD and 1974 ALD was much more realistically simple than the old EPD system in that it had three symbols instead of two. What is more, in such a system of stress notation, the vitally important principal stress stands out better by being preceded by an angled stress mark which takes up more space laterally than the less important upright mark for subordinate stresses. On a numerical basis which includes stress markings this made the old EPD and CPD/1974 ALD systems to my reckoning exactly comparable in difficulty.
11. By the end of the twentieth century the freedom to use a wide range of phonetic symbols without excessive embarrassment to a numerous printers brought about by the availability of computer typesetting led to the matter of such debates being seen in a very different set of contexts. In any case such developments as occurred from the late seventies onwards were led by the commercial requirements of publishers rather than the opinions of phonetics specialists. Since then at least as far as phonemic representations of British English are concerned we have had for the most part for an unprecedented sequence of decades a beneficial consensus in such matters.