In the case of A. S. Hornby's Oxford Advanced Learner's Dictionary of Current English we have a completely revised version of this well-known and highly-regarded general dictionary aimed specifically at the foreign learner. Our concern is here with the pronunciations given for some 100,000 items for which Mr J Windsor Lewis (WL) is responsible. In this dictionary the purpose has naturally to be prescriptive (or, at least, in the great majority of entries, to recommend a single typical form) rather than descriptive. Such a task is in many ways more difficult than that facing MW [the French authors A. Martinet & H. Walter whose Dictionnaire de la Prononciation Française dans son Usage Réel had been reviewed in the previous pages] since the recommendation of a form must be taken to imply that it is demonstrably the most widely used and generally accepted pronunciation. In his search for a valid standard model (the 'best known' variety of British English abbreviated to GB and defined as 'the variety of English most associated with broadcasting and least restricted in its geographical distribution'), WL has for many years been observing the usage of BBC newsreaders, and the conclusions which he reaches must be taken to reflect the results derived from his data. It would be to the advantage of us all if WL would publish the findings of his personal survey, at least as a provisional source of information until a survey similar to that of MW [Martinet & Walter] for French can be completed. In the event, there are on the whole few of WL's recommendations which one would dispute, though some teachers of English may be alarmed by certain of his telescoped (albeit well-attested) forms e.g. government /`gʌvm̩ənt/. There is no doubt that in many controversial areas of choice WL shows the newest form in the familiar style of the younger generations. One such area in which there is an undoubted pronunciation shift concerns the replacement of traditional /ɪ/ by /ə/. It is true that these phonemes, in opposition only in unaccented situations, are in sufficient phonetic proximity for a shift to be difficult to identify in the intermediate stages. WL maintains the /ɪ/⁓/ə/ distinction in such endings as -es⁓ers, e.g. boxes v. boxers, though in many forms of English the opposition is not kept. It is also evident that a termination such as -ity (as in quality) is nowadays far more often /ətɪ/ than the former 'received' /ɪtɪ/. Yet some of WL's conclusions are less incontrovertible. Are the -less, -ness endings in GB quite so certainly with /ə/ rather than /ɪ/ — as in hopeless, endless, goodness, business? Is it generally the case that preface, palace, solace have /əs/ whereas surface, furnace have /ɪs/; or that pamphlet, toilet, omelette have /ət/, whereas planet, wallet, lancet have /ɪt/; or that palate, obstinate have /ət/ whereas private, climate have /ɪt/; or that hostess, goddess have /ɪs/ whereas actress, wardress have /əs/?
An innovation in this new edition of the dictionary is the inclusion of American English pronunciations (GA) where they differ from GB. Again, a choice of an American model has to be made. There is naturally some difficulty in transcribing with symbols that are designed to represent British phonological categories: phonetic correspondences have to be either inferred, e.g. the different values to be assigned in the forms of the two vowels in hot, or derived from a note in the Introduction, e.g. hurry is shown as /`hʌrɪ US `hɜɪ/. A more detailed comparative statement in the introduction would have been helpful.
WL rightly abandons the type of vowel transcription used in previous editions (which followed Jones in EPD) in favour of one which emphasizes the qualitative rather the quantitative oppositions of English vowels. There has been much controversy over this matter over the last half-century, but I think it is time that we recognized that the important cue for distinguishing bead from bid or beat from bit is one of quality. Vowel duration, of course, remains significant in a contextually determined sense in a pair such as bead and beat. WL's inventory of vowels is symbolized as /i, ɪ, e, ӕ, ɑ, o, ɔ, ʊ, u, ᴧ, ɜ, ə; eɪ, əʊ, ɑɪ, ɑʊ, ɔɪ, ɪə, eə, ʊə/. From the practical (user's) point of view, it can be argued that /i/~/ɪ/ and /u/~/ʊ/ are not easily distinguished in small type. I am inclined to retain the length mark in spite of the redundancy involved. Another innovation introduced by WL is his use of the symbol ‘o’ for the vowel in cot as opposed to ‘ɔ’ for the vowel in caught. This has much to be said for it on the grounds of simplicity and symbol economy, but phonetic purists who are attached to the traditional cardinal values of symbols may object that the cardinal values are by this symbolization reversed. Some may also disagree with WL's choice of ‘ɑ’ to represent the first element in the diphthongs of both high and how. One other innovation: primary (tonic) accent is shown with ` and secondary with ˈ. This too has advantages, since it associates primary accent with pitch movement, but again it might be objected that this is a misuse of the high-fall tonetic mark, where the traditional ˈ and ˌ might equally well have been retained to mark the two types of stress accent. The foreign teacher does not usually welcome innovations of transcription (even when they are theoretically justified), but it must be admitted that he is not likely to be unduly upset by this change of stress marking. What may puzzle him rather more in this as in other dictionaries (such as EPD) is a certain apparent inconsistency in the distribution of secondary stresses, e.g. WL has has unearned with a secondary stress on un- but unbeaten without [This last item did in fact have a secondary stress on its un- in the text. JWL] .
Mr Windsor Lewis is to be congratulated on the vigorous and honest contribution which he has made to this new version of the Hornby dictionary. His more radical innovatioins may be questioned both at home and abroad, but he has obliged us all willy-nilly to reflect upon the real state of English pronunciation in the second half of the twentieth century. A survey of contemporary British English pronunciation usage, at least as extensive as that carried out by Martinet and Walter for French, is overdue.