The Daniel Jones Legacy

(Adapted from "Daniel Jones (1881-1967)" Zeitschrift für Anglistik und Amerikanistik 29/4 pp 343-8: 1980)

1. Daniel Jones came of a well-to-do family. His father was a barrister and one of the founders of the All England Tennis Club at Wimbledon. His mother's uncle was Richard D'Oyley Carte of the Gilbert and Sullivan operettas. He was sent to a preparatory school from which he went to Radley College at Abingdon in Berkshire half a dozen miles (10 kilometers) south of Oxford. Later he proceeded more happily to the well-known University College School in north-west London. After that he entered King's College, the richest of the Cambridge foundations, where he studied mathematics, graduating with distinction in 1903. He also acquired a legal qualification (to comply with his father's wish) though he never practised as a barrister. The reason was that his Continental summer holidays had infected him with a passion for languages and particularly their sounds at the making of which he discovered he was extraordinarily gifted. He studied German at Marburg under the tutelage of William Tilly, an eccentric scholar of Australian extraction, who deeply influenced him. By 1904 he had joined the International Phonetic Association and in 1905 he went to the École des Hautes Études at Paris to be able to study under its first president Paul Passy. He stayed at the home of a relative of Passy's and eventually married Passy's niece, Cyrille Motte, in 1911.

2. In 1907 he began his career in phonetics by giving a course of lectures at University College London on the phonetics of French for English school teachers. They were so successful that he soon obtaincd a full-time post and by 1918 he was able to appoint his first full-time assistant, a Manchester-born teacher of French, Lilias Armstrong. In 1914 he had been appointed reader and in 1921 he was appointed to a Chair of Phonetics. In this period he had worked intensively to broaden the scope of his subject. He became the editor of an important series of phonetic readers some of which (Cantonese 1912, Sechuana 1916, Sinhalese 1919) he wrote himself usually in association with a pupil of his who was a native speaker of the language. His first phonetic book had been a translation of Passy's Les Sons du Français as The Sounds of French (1907) in collaboration with the Oxford Professor D. L. Savory. In the same year he brought out his Phonetic Transcriptions of English Prose. Two years later this was followed by a phonetic reader containing passages in English, French and German in phonetic transcription along with their pitch patterns which he titled Intonation Curves (1909). That year also saw the first version of a book of his, The Pronunciation of English, which in a totally re-cast form appeared again in 1950 and was revised yet again in 1956. The year 1912 saw the appearance of his Phonetic Readings in English which also long remained in print its last revision being in 1956. In 1913 he brought out in collaboration with Hermann Michaelis, a German schoolmaster of Biebrich, A Phonetic Dictionary of the English Language in which the entries rather eccentrically were given in alphabetical order according to their sounds! It was never reprinted. His great textbook of English pronunciation for the foreign student, An Outline of English Phonetics, was completed in 1914 but not published till 1918, and then by a Leipzig publisher. Germany at that time was the foreign country ahead of all others in studies of the English language.

3. In 1917 Walter Ripman, a senior colleague at London with a number of books on phonetics to his credit, induced Dent to bring out the first edition of his later to be world-famous English Pronouncing Dictionary. He was so busy revising all these that he subsequently published few new ones, the volume In Honour of Daniel Jones (D. Abercrombie et al. eds 1964) lists over 100 items under the heading 'Articles and Pamphlets' by him. However, a book came out at the end of the forties of which he was most proud, his The Phoneme: its nature and use. Nevertheless it was with his Outline, especially from its extensively revised third edition of 1932 onwards, that he made his greatest impression. By the time of The Phoneme his theoretical outlook was widely felt to be dated especially in America where so much new thinking on phonetic theory was in progress.

4. With the Outline he undoubtedly did more than any other writer (except perhaps Leonard Bloomfield) to spread knowledge of the phoneme theory. It was certainly the most influential book in the history of English phonetics if not in English linguistics as a whole. He showed in it a break with the direct line of influence of Henry Sweet. Although Sweet's great work, especially his epoch-making Handbook of Phonetics (1877), was the powerful inspiration of the Continental progress in phonetic studies by Passy, Jespersen and others in the late 19th century, Jones did not draw directly upon it but absorbed it through filtration from the Continental scholars. Jones told the writer that, though he made contact with Sweet and attended Oxford study weekends given by him, that was after he had imbibed a Continental phonetic outlook. His books do not contain Sweet's boldly saxon terminology, even less so in later versions (he dropped "mixed" vowels in the 1932 third edition of the Outline). The content of them quite eclipsed the brilliantly condensed observations of Sweet by its unprecedented wealth of detail.

5. His Outline of English Phonetics was Jones's sixth work on English pronunciation. His third such book The Pronunciation of English had first appeared in 1909, with 69 pages of theory, 59 of transcriptions and half-a-dozen mouth diagrams. It was the first book worthy of comparison with Sweet's Primer of Spoken English (1890) which had had 44 pages of theory and 53 of texts. Sweet's books on phonetics were all quite brief though full of original observations; they contained no diagrams. Five years after his Pronunciation appeared Jones had completed in his Outline probably the most widely known and influential book in the history of phonetics not merely of English phonetics. Jones took nearly twenty years to revise it and, in the form it finally took, it dated essentially from 1932 when it was "re-written" fundamentally. Over the next thirty years he made minor changes and additions from one edition to another but nothing of substance was added after the appendix on transcription and the short final chapter on syllable-separation of 1956. It never contained continuous transcribed texts. The magnitude of the work can be partly gauged by considering that the 1956 edition was approximately twice the size of the original volume; it contained about 250 pages of matter not in the first edition which had itself been a not inconsiderable 220 pages of which 50 or so were later dropped apparently at the publisher's request (though Jones was evidently becoming less enthusiastic about the types of instrumental investigations not retained from the earlier version) for reasons of "economy". The volume of new observations it contained on the substance of English pronunciation was remarkable.

6. Even though Gimson's Introduction to the Pronunciation of English was in general better organised than the Outline and quite superseded it as a description of the characteristics of British pronunciation, it didn't entirely replace it as a source-book for the EFL specialist. Various analytical procedures and opinions of Jones's were to become no longer acceptable or applicable, certain assertions of fact were not attested by later investigators and in many places various forms were presented which had become old-fashioned, obsolescent or no longer in use at all, but the Outline still provides a large amount of information and advice nowhere else available. (See Zeitschrift für Anglistik und Amerikanistik 4/78 for an account of 'Some Changed Usages'.) The whole of the Outline was, as the opening chapter made quite explicit, directed at the user for whom English was not the mother tongue and at those who engaged in teaching such pupils. Contrast Gimson's terms of reference which, in a book about four-fifths of the extent of the Outline devoted more space to the historical background, the acoustic nature of English sounds and their historical sources than he did to EFL problems: the explicit "Advice to the Foreign Learner" sections in his IPE were mainly rather general and occupied no more than five percent of the text until the not very extensive 1980 Appendix. Daniel Jones never wrote a book on general phonetics as such, although his book The Phoneme deals at great length with a most fundamental concept of phonological analysis. On reflection this was perhaps not surprising because he had in effect written such a book in writing The Outline which dealt with a very wide range of the topics of linguistic phonetics. Apart from a token couple of pages (Chapter IV) on "Experimental Methods" and a short preamble on "Types of Pronunciation", the book had four main topics: (i) the description of the articulations of the English segmental sounds; (ii) their theoretical phonological analysis; (iii) the impacts on each other of the constituents of connected speech; and (iv) the prosodic suprasegmental features of English.

7. In Chapter II, "Types of Pronunciation", the suggestion that the kind of speech described in the book ("based on my own") was fairly uniform in Public Schools and "independent of their locality" has become increasingly open to question. His remark in 1932 that the term Received Pronunciation is "often used to designate this type of pronunciation" was rather disingenuous in view of the fact that its currency was no doubt mainly due to his own putting it into circulation and especially his adoption of it in the third edition (1926) of his English Pronouncing Dictionary in the earlier editions of which the term 'Public School Pronunciation' and its abbreviation 'PSP' were used. At any rate, neither term was satisfactory and adopting the term 'Received Pronunciation' and in effect saddling British phoneticians with such a clumsy term for generations, Jones made one of his least happy judgments. Its parochiality in world terms and its complacent insensitivity sociologically (seeming to imply for those who did not use it that their speech was other than acceptable) when any geographical indication, which was surely necessary, was absent from it has remained an embarrassment for anyone who uses it and at all ponders on its signification. Previously 'rp' had been one expression among many used by Alexander J. Ellis in learned works on historical phonology more than a generation earlier. H. C. Wyld used the term "Received Standard" of the speech of public-school-educated people (not merely their accent). An outdated if ever justified assumption seems to underlie the whole book that its users were most likely to have verbal contacts chiefly with upper or middle-class people from the south-east of England in that the detailed account of their variants occupies space that might perhaps have been accorded to more general regional characteristics of English pronunciation. It is noteworthy that of the two conversational "styles" mentioned the "slower colloquial" and not the "rapid familiar", is the style indicated throughout ... except where "the contrary is indicated". An interesting footnote to this section refers to some variants within "RP" and the important point is made – which the book as a whole can be accused of neglecting – that a type of English in which certain distinct phonemes including /æː/ and /aə/ and some other features do not occur is perfectly adequate as a basis for study directed at "enabling foreign students to learn to pronounce English like English people". (This variety of accent seems to be more or less what he used a term "common pronunciation" for at p 341 fn 21.)

8. The descriptions and classifications of the vowel and consonant sounds of English constituted the largest and most sucessful section of the book at Chapters VI to IX and XIV to XXV. They did, however, contain a certain number of inconsistencies, discrepancies etc. For example at §71 he said "The extremity of the tongue is called the tip or point, and is included in the blade", but at §709 of /s/ that "This sound is articulated by blade (or tip and blade) of the tongue against the teeth-ridge". In §183 "the principal English r" is given as the example of a "frictionless continuant" consonant whereas in §747 "The most usual English r" was referred to as a "fricative" consonant. This last comment seemed to spring from academic over-subtlety, a tendency of which there are various other signs. A markedly fricative /r/ would sound un-English in most situations. The name Henry Kenrick would sound like Hendry Kendrick, the word gunnery in its common disyllabic form would become indistinguishable from the name Gundry and so on. Of course there is no very sharply defined point at which fricatives can be delimited from frictionless sounds (cf §680) but there can be no doubt that such a presentation in a pedagogical work was ill-advised. A similar confusion was caused by the either unjustified or over-subtle assertion in §756 that "generally a flapped variety" of /r/ was used for the linking /r/. Again there seems to be partial self-contradiction or at least vagueness at §750 where a flapped variety of /r/ was referred to as used "in unstressed intervocalic position" by "many" speakers as opposed to a fricative variety "quite commonly used in such cases". A strongly flapped /r/ from most speakers (not necessarily in all Northern or Scottish accents) tends to sound markedly "precise". The word very can be uttered quite emphatically with a non-fricative /r/; with a flap it was likely to sound quite old-fashioned by the mid twentieth century. The recent developments in the description of paralanguage have made us better able to distinguish non-prosodic suprasegmental features of speech. We now have a framework into which we can fit items like what Jones called a "peculiarity of individual speakers" who can be heard to use retroflex varieties of /r/ (§ 830) (a "wincing" type of movement of the tongue) rolled varieties of /r/, unrounded varieties of the good vowel and so on.

9. The academic over-sublety just referred to was also demonstrated in his approach to vowel diagramming. He was so attached to the idea of "scientific accuracy" in distinguishing one sound from another that he displayed the English simple vowels, although his point was simply to show a set of phonemes in opposition, on a diagram, of curiously excessive "precision". It was, however, not quite as unfortunate as the other diagram which was always a frontispiece to the EPD: that not only showed the simple vowels but also, with some inconsistent oversimplification, the starting points of the diphthongs. While reiterating that it represented "roughly" the "approximate" tongue-positions of the "average" English vowels, he placed upon it dots of a size appropriate to the representation of the finest distinctions the trained ear can possibly perceive. However, he did offer (Fig 35) a "Simplified Chart" more useful for "practical teaching of the language" (whatever implications this may have had for the kind of instruction conveyed by the Outline) which instead of enlarging the minuscule dots simply pushed them away from usefully contrasted truer positions to either falsifying or at least unhelpful peripheral ones. The most valuable pedagogical application of all of the diagram – to display the close relationship of /ɪ/ and /ʊ/ with /ə(ː)/ – was thus jettisoned. The next most useful factor – the display of clear contrast between /ɔː/ and /ɒ/ was also much reduced in its effectiveness. The passage of time made pushing the [o] of his diphthong /ou/ (Gimson's /əʊ/) back rather than forward rather unfortunate: on the other hand there was the gain that /ʌ/ was pushed forward to a less old-fashioned value. It is noteworthy that this and the EPD were his only books of which the diagrams were never re-designed after 1932. In his others he used only the "practical" trapezium after the early thirties, even in highly theoretical works like The Phoneme. The alleged scientific basis of his cardinal vowels system was investigated searchingly by Peter Ladefoged who wrote of such phoneticians' attempts "to describe the positions of the vocal organs during the productions of different sounds" as actually only succeding in "providing categories with which to describe their auditory impressions ... It is probably incorrect to consider that points on vowel diagrams describe tongue positions for any speaker, even approximately" (Three Areas of Experimental Phonetics, pp. 68/70, 1967). Nevertheless, in spite of over-optimism about the degree to which it could be claimed to have had a purely scientific basis, the Cardinal Vowels system enormously improved the effectiveness with which practitioners of phonetics could communicate with each other on paper about vowel qualities, as Ladefoged fully conceded. It also enabled users to display very effectively the relationship between the various elements of the vowel system of almost any language.

10. Just as Jones oddly asserted that the typical English /r/ was fricative in spite of Sweet's flat statement (1907 §122) that it was "quite free from buzz", he took a similarly unfortunate different line from Sweet on the typical values of /iː/ and /uː/. He referred to the diphthongal varieties of them only as variants "many English people use". As regards both these and /r/ Gimson resumed the right road indicated by Sweet. On the vowel /ɪ/ Jones's suggestion that any value for "the final vowel of city, heavy etc" which is closer than the ordinary /ɪ/ is "dialectal" was to be contradicted by Gimson (1980 p. 105). The wise advice (§ 282) that in attempting /æ/ it is "better to err on the side of [a] rather than on the side of [ɛ]" is even more timely these days when anything more than half way up to Cardinal [e] sounds old-fashioned and realisations of /æ/ slightly retracted or centralised from Cardinal [a] are perfectly common The verbal description of /ɒ/ shows a slight inconsistency with its diagram representation not improved upon in Gimson's IPE. For /ɔː/ the description "between half open and open", as Gimson warned, if taken precisely as a prescription was by the mid 20th century in danger of producing a quality "which can no longer be said to be typical of RP": the Jones diagram represented a value distinctly nearer to Cardinal 6 than 5 whereas it had become safest to direct learners to a value at least as close as Cardinal 6 by the mid century. A revealing difference between the 1956 edition and earlier versions of the Outline was in the wording of the description of /ʌ/ which reflected his consciousness that his own value of it was becoming less usual than that of the "many who use a more 'advanced' and less ə-like vowel".

11. A striking example of Jones's excessively academic approach to helping "a person wishing to acquire an acceptable pronunciation" (§54) was to be seen in his lengthy account of four numbered allophones of /ə/ and the various related sounds [ʌ] (sic) and [ɐ] etc including "sounds intermediate between" the first and second types of /ə/ specified which "are also common". It started at §355 with the remark that "It is sufficient for practical purposes to distinguish three varieties" and at §368 continued with "The foreign learner therefore need only learn two of the numerous varieties of /ə/ occurring in English". Of the diphthongs Jones said /ɔə/ "may be ignored by the foreign learner" long after the advice needed to be "should be ignored". Cf my Concise Pronouncing Dictionary pxvi §E. At the treatment of the diphthongal vowel No.14, Jones added to the 1956 edition the significantly cautious "as I pronounce it" to his description of it. However, he described no variant as such and, amazingly enough, only revealed his awareness of the very much more fronted varieties which were already the norms indirectly in making such remarks as that "it is not far removed from ... a sequence of English vowels Nos 11 and 8". There was more excessive academicism in his description of the variants of the /aɪə/ and /aʊə/ sequences. As with his remarks on final /ɪ/, his comment that a variety of the diphthongal vowel No 18 /ɪə/ beginning "nearer to long /iː/ than to short /ɪ/ is due to the influence of some dialect" was even in his day very questionable: the modern norm is probably about midway between these. Similarly out of date was his lack of any reference to the very common levelling of non-final /eə/ and widespread smoothing of /ɪə/ and /ʊə/ in all but final positions (on the first of which see Gimson 1970 p. 144). More misplaced academicism occurred in the distinguishing of rising and falling variants of the diphthongs /ɪə/ and /ʊə/. The same comment applies to the listing of well over 100 weakforms of over 60 words: more than half of the weakforms could be totally ignored by the learner and a perfectly satisfactory pronunciation acquired.

12. In the generally excellent treatment of the English consonants, besides the curious assertion with regard to /r/, there are the remarks – now hardly acceptable – that some degree of aspiration normally accompanies /p, t, k/ when preceded by co-syllabic /s/. These probably arise from his lack of a sense of what was "normal", the "unmarked" articulation and what was a paralinguistically "marked" variant. This showed itself oddly in his dropping the absurdly mutually uncontrasting photographs of lip-postures in the first two editions of "normal" articulations to offer in the third edition onwards only the 20 photographs which had been previously labelled as pronunciations 'with exaggerated distinctness', and with no longer any reference to their abnormality. The suggestion (§755) that the schwa-less form of quarrel was less usual than that with /ə/ had become no longer acceptable. Similarly the account of the occurrence of linking r no longer corresponded to current usage. The suggestion that it was not normal for linking /r/ to occur in phrases such as a roar of, rare animal, area of and he opened the door and walked in was almost certainly well out of date if ever true. Omission in such cases is clearly in present usage a "marked" feature. In the last type r-link is not merely usual but by no means unknown if there is a distinct pause even. See the writer's article 'Linking /r/ in the General British Pronunciation of English' on this website. In regard to the segment descriptions it should be pointed out that, of however little value the lip photographs may be, the "sagittal" mouth-cross-section diagrams and the drawings of palatograms were extremely effective in bringing out the natures of the types of difference which distinguish articulations. The diagrams of his Pronunciation of English (1950) were distinctly superior to those of the Outline. The phonemic theory accompanying the segment descriptions was the first widely known use of this technique of analysis which Jones not only took up but played no small part in developing: he seems to have invented the term "diaphone". His book The Phoneme proliferated coinages quite uninhibitedly! His preference for an analysis of the English vowel system leaning heavily on its tenuous length contrasts (he treated the vowel pairs /i:/ and /i/, /ɔː/ and /ɔ/, /uː/ and /u/ , /əː/ and /ə/, and in his Simplified Transcription /a:/ and /a/, as allophones of single phonemes) has found favour with no later analysts. It was even silently contradicted by his own numbering of the English vowel sounds in what corresponds to what all other writers take to be the English vowel phonemes. Compare his rigid refusal to recognise the existence of neutralisations of phonemic contrastiveness eg of /p ~ b/ after co-syllabic /s/ etc.

13. Jones was far from unaware of the imperfections of his Outline. In a discussion of the book with the writer a couple of years before he died he referred to it in his characteristically modest manner as "not a good book". A perfect book it was not but we may surely call it fittingly "a great book". Many items of terminology that have become the stock-in-trade for various later writers on (particularly English) phonetics appear to have begun to be used (in English at least) mainly as coinages by Jones. Various of these are no doubt not realised to have so originated by many who now use them. These include, besides the regrettable Received Pronunciation (& its abbreviation RP), Cardinal Vowels, intrusive r, linking r, similitude, diaphone, di-phonemic, ejective, uniliteral & multiliteral (of phonetic transcriptions), toneme and variphone. He has hardly been followed in his use of some of his coinages such as chrone, chroneme and the curious formation stroneme. Apparently, since they did not appear in OED in his lifetime, he may have even first used creaky voice, free variant, frictionless continuant and flapped consonant. Altho Jones didnt coin the word phoneme, he was the first writer in English to do more than quote the term from another language, to incorporate it integrally into his own work and to promote its institutional use.

14. In JIPA 36/2 of 2006, at p. 138 in an account of an interview with the late Peter Ladefoged conducted by Alan S. Kaye, Ladefoged said: "I enjoyed writing a review of the biography of Daniel Jones, the first Professor of Phonetics in Britain, by Collins and Mees. It's a great book, a wonderful account of the beginnings of twentieth-century phonetics in Britain and its antecedents on the Continent." (See Obituaries on this website for one on Peter Ladefoged.) So for anyone who wants a complete account of Jones's life and achievements the reference is: Collins, B. S. & Inger Mees (1989) The Real Professor Higgins: The Life and Career of Daniel Jones Berlin: Mouton de Gruyter. [Readers may be mildly amused to note at its p. xxiv how the book got the its title.]