The Official IPA Vowel Diagram

from the Journal of the International Phonetic Association.

Vol 6 No 1 June76 (slightly adapted)

The eagerly awaited revision of the Association's Principles will no doubt incorporate a less controversial statement than the present (paragraph 11) remark that THE positions of THE tongue in the articulation of cardinal [i, a, ɑ] and [u] have been determined. The old form of diagram introduced by Daniel Jones and still surviving as the only official IPA shape must surely be now discarded. Its form implied both excessive optimism both about the universality of the applicability of the X-ray photography data on which it was based, and about its practical usefulness. It can only be said to have flourished for about the decade 1919-29. Since the early thirties it has been overwhelmingly rejected in favour of Jones's second choice which, many would be perhaps surprised to hear, has as yet absolutely no explicit official IPA sanction. This seems to have first appeared in 1929 in Ida Ward's The Phonetics of English. Jones himself never made use of the present [1976] official diagram in any new or fully re-written book after 1932 not even in The Phoneme. So far as one knows he never formally abandoned it but his discarding of it seems to leave little doubt that he himself would have been in sympathy with authorisation in the new edition of the de facto current version. If anyone still considers the older version preferable it would be very interesting to hear their arguments and to hear of any evidence on the basis of which such an opinion were held. Now is clearly the time for speaking up on this topic.

The main point of the present note is to make a plea that the Association should state its vowel diagramming policy clearly in regard to the relative dimensions of the parts of the diagram. The editors of this Journal (under which I include Le Maître Phonétique) have allowed some pretty grotesquely distorted variations of the diagram to appear in these pages in the past. It would seem an appropriate policy to make recommendations like those in respect of the Association's alphabet with its clearly specified shapes for letters. Jones indicated quite clearly in his historic Outline of English Phonetics (footnote 9 to § 149) the shape that has generally been approximately confirmed to. This is in effect a rectangle three units deep and four wide with the bottom left-hand quarter removed (ie minus the portion to the left of a line running from the middle of the [rectangle's] base to the top left-hand corner to produce a 'trapezium'. The base, the right-hand upright side and the top [of the diagram] are thus in the proportions 2:3:4. (See eg Abercrombie 1967 Fig. 3.) [Jones] also specified that the inner triangle, within which vowels may be said to be central in quality, should have its apex at what I have found it convenient to call the semi-half-open line, ie halfway between the Cardinal 3 and 4 levels. I should suggest that this perceptually based division should be retained rather than that we should subscribe to the transatlantic minor heresy of three roughly equal divisions vertically. Not that that is illogical but because the typical central quality is non-open.

However, the new Principles could well make the point that unfortunate ambiguities have occurred in the past when the term CENTRALISED has been used. It has not always been clear whether a writer is referring to a vowel value removed from an open cardinal vowel only horizontally or both horizontally and vertically, ie in the direction of the LINE down the centre of the diagram or in the direction of the centre (point) of the vowel area. It would be well for users of our authorised terminology and diacritics to employ only ADVANCED/RETRACTED for the former sense at least when specifying opener vowel varieties. For examples of possible confusions see the texts accompanying them and Figs 12-15 etc of Gimson 1962.

Some readers may have noticed that (in Windsor Lewis 1969 etc) the present writer has found it convenient to subdivide the diagram further into a grid of about thirty slots as first suggested for broad comparisons of vowel systems in Windsor Lewis 1966 ... Such diagrams are less convenient for quick drawing but they help the eye to take in relationships very effectively especially in comparing one diagram with another ... It is no doubt best that the IPA authorised diagram should be the basic traditional one with lines across at half-open and half-close and internal triangle with, or preferably without, central dividing line, but the grid version is offered as a perfectly legitimate extension of it in no way clashing with the basic conception.

It might also be helpful if one or two diagrams were included or some remarks were made in the new Principles which would make it clear to users of the Associations authorised vowel diagram that the micro-dots appropriate for indicating the 'mathematical' locations of the cardinal vowels are not necessarily of appropriate dimensions for showing other vowel information. Vowel charts have in some cases been published with vowel indicators so small that their implication was that a trained ear could discriminate 6,000 vowel shades. Unless it is a question of showing something such as the allophones of a particular vowel in an idividual's speech, it is very doubtful whether vowel indicators need be any less bold than one-hundred-and-fiftieth of the space the diagram provides for them. Jones's diagram showing his own simple vowels in Jones 1950 (Fig 12) displays approximately this relationship thus suggesting the possibility of about 300 distinguishable vowel types (with two lip postures postulated for for each of the 150). This seems more realistic than the approximately 800 suggested by the relationships in his other works and in eg Gimson 1962. All too often we see these works imitated slavishly in contexts where such degrees of precision are inappropriate to the point of absurdity.

[PS In the event the projected revision of the Association's 53-page booklet of its Principles was replaced in 1999 by a 204-page book the Handbook of the International Phonetic Association published by Cambridge University Press.]


Abercrombie, D. (1967). Elements of General Phonetics. Edinburgh: Edinburgh Univ. Press.

Gimson, A. C, (1962). An Introduction to the Pronunciation of English. London: Arnold.

Jones, D. (1949 anonymously). The Principles of the International Phonetic Association. London: IPA

Jones, D. (1950a). The Pronunciation of English. London: CUP.

Jones, D. (1950b) The Phoneme. Cambridge: Heffer.

Jones, D. (1932). An Outline of English Phonetics. 3rd edition. Cambridge: Heffer.

Windsor Lewis, J. (1966). The Symbolisation of Vowels by Diagram. Språk og Språkundervisning. Oslo.

Windsor Lewis, J. (1969). A Guide to English Pronunciation. Oslo: Scandinavian Universities Press.

Ward, Ida. (1929) The Phonetics of English. Cambridge: Heffer.