[Journal of the Simplified Spelling Society, 1987/3 pp17-21 later designated J6.]
[See Journal and Newsletter articles by Edgar A Gregersen.]
The Best Way to Write the Vowels of Moon and June in a Rational Spelling of English.
Edgar A Gregersen.
Edgar Gregersen is Professor of Anthropology at Queens College and Graduate Center of the City University of New York. He has a special knowledge of accents of English, Egyptian hieroglyphics, the alphabetization of West African languages, and the spelling of Norwegian.
1 BACKGROUND.1.1 <oo> or <uu>?
The great majority of spelling-reformers at present believe that the vowels in moon and June should both be written with <oo> (thus: moon, Joon or joon). Until 1971, the Simplified Spelling Society had proposed the use of <uu> in its Classic Nue Speling (muun, Juun), but in that year it too changed to <oo>, using <uu> for the vowel in book, foot.
This is a plea to restore the earlier Nue Speling value. I make this plea because <uu> (or a variant involving <u> such as <uw> or <û>) is the optimal spelling, at least for the vowel in moon. The vowel in June has problems I shall discuss near the end of this paper in section 2.8.
1.2 Phonetic/phonemic variation
In the following discussion, several vowel sounds will be mentioned because how they are represented has some bearing on the issue. I shall represent them tentatively with fairly well-known symbols from the International Phonetic Alphabet (=IPA).
/uw/ (also written by some as /u:/) as in moon
/u/ (phonetically [ω]) as in put, book
/ʌ/as in hut, cut, love
/ə)/ the unstressed vowel of alone, China
/juw/as in cube /kjuwb/, music /'mjuwzik/
/iw/ a variant of /juw/ found in several American English dialects as in new /niw/, most Americans say /nuw/; elsewhere in the English-speaking world the pronunciation is /njuw/, which is also found in the USA.
The vowels of cooed /uw/, could /u/, and cud /ʌ/ (or of moon, book, and hut) are differentiated in nearly all of the standard dialects of modern English. The major exception is traditional standard Scots English where moon and book are pronounced with the same sound, a centralized vowel (IPA [
1.3 Phonemic definition.
All dictionaries that indicate 'correct' pronunciations indicate the vowels of these words as different. A rational spelling for English would also have to indicate these distinctions. In 1985, the Working Party of the Simplified Spelling Society recommended that both /uw/ and /u/ be written with <oo>. This means that cooed and could would be written identically, as would pool and pull, fool and full. I urge that this recommendation be rejected.
I do so because I believe that the fundamental principle for a rational spelling is that it be essentially 'phonemic'. In practical terms, this means that words pronounced differently should not be written the same. For various reasons an orthography may have to write differently some words that are pronounced the same (at least by some people): it is important to keep the language community together and not break it up by showing dialect difference. So, for example, the words father and farther, different for most speakers of English, would have to be spelt differently even though they are identical in RP and in the pronunciation of all Australians, New Zealanders, South Africans, and even many Canadians and (eastern) Americans.
A decent spelling may have to show more differences than are phonemic for any one dialect. It can never show fewer.
So the /uw/ - /u/ - /ʌ/ distinctions must be shown.
1.4 Previous suggestions for writing /uw/
How they should be shown is another matter. The majority of English spelling reformers seem to accept that they be represented as <oo, uu, u> respectively: kood: kuud: kud for traditional cooed: could: cud. This is the present official position of the Simplified Spelling Society in their 1971 Revised Nue Speling. It is also the position incorporated into the orthographics championed by Dr Godfrey Dewey in World English Orthography and Edward Rondthaler's Typographic Council for Spelling Reform in American Spelling, and generally by BEtSS (Better Education thru Simplified Spelling) headed by Ayb Citron.
Edward Rondthaler and Edward J Lias, in their Dictionary of American Spelling (1986, New York: The American Language Academy) call the Classic Nue Speling use of <uu> for /uw/ and <oo> for /u/ "typographic insensitivity... doomed from the start". They are particularly concerned about the ending <-ful> (as in beautiful: Classic Nue Speling writes it as <-fool>, which they consider "ridiculous" and add "Critics of reform merrily regarded this as utter foolishness, which indeed it is" (all quotes are from p. 304). Rondthaler and Lias themselves write this ending exceptionally as <-ful> (in their system this should otherwise rime with hull; the independent word full they write as fuul.
Of course, a few reformers have not gone along with the majority position. The Simplified Spelling Society ov Canada, headed by Ted W Culp, proposes that /uw/ be written <u> finally: thru, tu, (=to, too, two), yu (=you). It is not clear how it would be written elsewhere. In The Times ov Toronto, a newly formed newspaper published by Culp in reformed spelling, we find spellings such as grupe (group), whuze (whose), cluus (clues but the singular is clu), fuud (food), and exceptionally bambuu (bamboo). Both /u/ and /ʌ/ seem to be written as <u>: cud (could), buk (book), luv (love), becum (become). Presumably, therefore, cuud: cud: cud for cooed: could: cud.
F C Wingfield in his Fonjmik crthqgrafi reassigns values for certain letters. He writes <w> for /u/, which frees <u> for /uw/. For /ʌ/ he uses either <a> or <v>; for the vowel of bat he uses <ae> (IPA æ). Hence, kud: kwd: kud.
Newell also proposed <w> for /u/. He felt a single letter was reasonable here because the vowel is phonemically short. However, he accepted the spellings moon and hut. Hence kood: kwd: kud.
A number of other suggestions have been made by people outside the spelling reform movement proper. The linguists Martin Joos and Charles Trager, who see no hope for reform but find the matter of the ideal spelling for English a kind of intellectual game, have both gone along with long-standing linguistic conventions and so agree that /u/, the vowel of book, would be written as <u>. However, their agreement ends there. Joos wants <û> for /uw/ and <ö> for /ʌ/. Hence, kûd: kud: köd. Trager writes <uw> for /uw/ and <a> for /ʌ/ (he uses <æ> for the vowel in cat).
So much for the background. What are the arguments for using <uu> for /uw/?
2. THE ARGUMENTS FOR <UU>.2.1 The internationalist argument. As often as possible, unless some other compelling reason exists, letters should be used with 'international' values. For example, it would be absurd to show the sound /s/ with the letter <q>, all other things being equal. Of course, a perfectly consistent spelling could be invented using <q> to mean /s/, and such a system might be admirable in all other ways. But there is no reason to go out of one's way to use obscure rational values. One of the problems with the otherwise excellent current Hungarian spelling is that <s> represents a <sh> sound, and /s/ is represented by <sz>.
In the great majority of languages using the Roman alphabet, an /uw/ or /u:/-like sound is written with <u> as in Italian, Spanish, Portuguese, Polish, Hungarian, Turkish, Danish, German, Swahili, Zulu, Yoruba, Hawaiian, Vietnamese, Navaho. In the spelling of Hausa used in Niger, long /u:/ is shown as <uu>. The letter <u> is comparably used in official romanizations of Chinese, Japanese, Russian, and Hebrew. No language uses <oo> to indicate an /uw/-like sound, though French uses <ou>, and Dutch <oe>. In Swedish and Norwegian, unlike Danish, the letter <u> formerly indicated an /u:/-like sound, but this sound has moved forward in the mouth (to /
The International Phonetic Alphabet or its variants the Africa Alphabet and World Orthography have sometimes been used in the creation of practical orthographies. In all of these systems, the vowel of moon would be written with <uu> or <uw>. Short /u/ as in book is written <u>.
To write moon, buuk agrees with no international principles. To write (as in Classic Nue Speling) muun, book at least gets us one international value.
2.2 The classicist argument.
In English words of Latin origin, the letter <u> often has the value /uw/ (or sometimes /juw, or /iw/). Such words include the following, with the original Latin form given in parenthesis:
ablution (ablūtīo), brute (brūtus), crucifix (crucīfixus), crude (crūdus), frugality (frūgālitās), judicial (jūdiciālis), lunar (lūnāris), plume (plūma), pluvial (pluviālis), prudence (prūdentia), prune (prūnum), rude (rudis), rudiment (rudīmentum), solution (solūtiō).
These are only a few of hundreds of similar words. When the <u> in Latin was long (here shown with a macron, <ū>), the Romans themselves sometimes wrote <uu>
If such English words were written with <oo>, the relationship to the Latin forms would be further obscured. Altho I do not believe that a decent spelling should necessarily display etymology, I don't see why etymology should be totally ignored if it can easily be accommodated. Simply choosing <uu> as the spelling for /uw/ in general would do the trick.
However, because of much dialect variation, it might be best to spell such words with <ue> - which would also accommodate the etymology. This dialect variation will be discussed later on in this paper in the section on June.
These first two arguments considered here - the 'internationalist' and the 'classicist' - are external ones, that is, arguments not prompted by reverence for English traditional orthography or the facts of English pronunciation.
2.3 Reducing dialect variation.
Several internal arguments derived from the facts of English itself also occur. All the following fall into this class.
A fairly common sound sequence in English is a <y>-sound or yod (IPA /j/) followed by /uw/, as in cue /kjuw/ or few /fjuw/. In many systems of spelling reform, including Nue Speling, this sequence is written as <ue>; hence kue, fue. The weakly stressed or unstressed sequences /ju/ and /jə/ are normally also written as <ue> as in monuement 'monument', kalkuelaet 'calculate'. For the rest of this discussion it will be assumed that these values - including also /iw/ (which may be phonetically a fronted [
Considerable dialect variation occurs with regard to the use of <ue> and /uw/. For example, in many dialects in the United States and Canada, <ue> cannot occur in a stressed syllable if it would be preceded by <t, d, n, or θ)> (the sound represented by <th> in thin) and /uw/ is used instead, as in tune, due, new, enthusiasm. (But in weakly stressed syllables, at least after <n>, <ue>, does occur, e.g., in annually.) We can call such dialects, yod-dropping. Elsewhere in the English-speaking world the standard form is almost invariably <ue>. We can call such dialects, yod-pronouncing. In Merriam Webster dictionaries, the notation <(y)ü> has been adopted to show such dialect variation.
Even the yod-pronouncing dialects may in fact drop the yod after<s, z> and especially after <l> as in suit, presume, lute, revolution. The use of <ue> in such words represents the older upper-class pronunciation still much used on the stage in serious drama (e.g. Shakespeare). The facts are complicated and usage varies from word to word. In RP sue usually has <ue>; Zeus and lurid always have <ue> even tho other words may vary. In the United States, among yod-pronouncers, lute almost never has <ue> but revolution occasionally has, especially in the speech of actors and radio announcers on 'sophisticated' stations (e.g. WQXR or WNCN in New York City).
In the United States, yod-pronouncers sometimes use <ue> where the rest of the English language community does not, e.g. in maneuver (elsewhere also spelt differently: manoeuvre). Some yod-droppers even use <ue> where it is not traditional, e.g. after the /k/ sound in coupon. On the other hand, in some other words after a /k or g/ sound the traditional <ue> alternates with /uw/, as in lacuna, cuprous, gubernatorial, lugubrious (with this word even RP shows such variation); or when unstressed (<ue> = /jə/), as in executive, lucubration, inaugurate, configuration. The American pronunciation of figure with unstressed <ue> as opposed to the Commonwealth pronunciation with /ə/ is; another instance of such variation.
Clearly a great number of words with alternations involving <ue> and /uw/ exist. I have made an informal count of all words with <ue>-/uw/ alternates as given in Webster's New Collegiate Dictionary (8th edition, 1973) where the variation is usually indicated as <(y)ü> or <(y)u>-the <ü> indicating /uw/, the <u> /u/. I don't guarantee that my count is absolutely accurate; my eyes started to glaze over after the first 30 pages and there were over 1300 to go. Also, what should count as a separate item was a vexing question, which I dealt with in probably inconsistent and certainly ad hoc fashions.
Altho this dictionary does not indicate <ue> pronunciations after <s, z> and <l> at all, even so the count I arrived at came to 829 items where <ue>-/uw/ alternates exist This suggests that a total figure including <s, z> and <l> forms is well over 1,000. Other problems exist, which will be taken up later with reference to June in section 2.8 below.
The important point is, that a sizeable part of English vocabulary would have alternate spellings if both <ue> and /uw/ pronunciations are shown. The Commonwealth countries would probably all adopt the <ue> spellings - and let us assume that <ue> would be the actual form used. If <oo> is adopted for the /uw/ pronunciation, the relationship between the dialect variations would be obscured. Alphabetization would be thrown off. On the other hand, adopt <uu>, and the <ue - uu> alternates will be less noticeable and alphabetization would at least be confined under a single letter <u>.
As a matter of fact, my own view is that in order to preserve the unity of the English-speaking world, the yod-pronouncing forms should preferably be used in all these instances. Yod-dropping pronunciations would be predictable because, for example, the sequence <due> could mean only <duw> in such dialects. I shall not press the matter, however.
2.4 Morphophonemic alternations with 'short u' /ʌ/.
Systematic relationships occur between <ue> or /uw/ and the vowel in but, here shown with the compromise symbol /ʌ/. In RP this short sound is phonetically [ɑ]; in General American, it is much more central and could be written as either [ʌ] or [ə]; in some western American dialects, it is higher, IPA [ɣ]. The sound /ʌ/ is usually written as <u> in most spelling reform systems (altho, as we've seen, <a> has also occasionally been proposed - or else an altogether new letter has been suggested). My own preference is a new letter, either & lt;ə> or <ʌ > or <
In this discussion, I shall assume that /ʌ/ is written with <u> or with a letter that would be lumped together with <u> in alphabetization.
The relevant fact is that, in effect, <u> counts as the stressed shortened form of /uw/ or <ue>: assume: assumption; induce: induction (and similarly with all the -duce/-duction words); duke: duchess, duchy; and commonly in the USA at least, quadrúple with /uw/. quadrúplet with <u> (in RP the stresses and vowels are different: quádruple: quádruplet). To write <oo> for such words would obscure this relationship; e.g. dook: duches, (better duuk: duches, best duek: duches).
Edward Rondthaler suggests that in such words <ue> be written altho normally the variety of English he uses as the basis for his system is an extreme y-dropping American form. He finds himself with this fairly arbitrary solution because he uses <oo> for /uw/. If he adopted <uu> for /uw/ then no makeshift compromise would be necessary since the morphophonemic relationship would not be obscured.
2.5 Morphophonemic alternations depending on stress.
Even among 'yod-droppers', the yod may be restored when stress shifts. The yod is dropped in a stressed syllable but not in an unstressed, e.g. voluminous (with /uw/, less often <ue>): volume (always with <ue>); diminution (with /uw/ in USA, Canada; <ue> elsewhere): diminutive (with <ue> everywhere). Again, to write <oo> for /uw/ would obscurethese systematic relationships: voloominus: voluem is clearly not so desirable as voluuminus: voluem, etc.
After <t> and <d> the situation becomes somewhat more complicated because before the unstressed <ue>, <t> becomes <ch> (IPA ʧ) and <d> becomes <j> (IPA ʤ) at least in American English; cf. constitute with a moderately stressed ending <-tuut> (USA) or <-tuet> vs constituent with <-stich-> (USA); assiduity with <-duu-> (USA) or <-due-> vs assiduous with <-sij-> (USA). In RP at present forms with <t> and <d> have apparently usually been restored from the spelling. The <t, d> forms - not the <ch, j> ones - should probably be adopted in a rational spelling since <ch, j> forms can be deduced from the <t, d> spellings but not the other way round.
2.6 Weak forms and word signs.
In some instances for some very common words, it would be useful to have shortened spellings. For example, the words do, to, who, and you which when pronounced in isolation have /uw/, often are unstressed in connected speech and then have shortened or reduced vowels. Ideal representations for these words are du, tu, hu, yu. In Nue Speling they are rather inconsistently spelt as duu, to, huu, U, respectively.
The spelling thru for older through or thro' is universally known already. Although this word is normally always pronounced with a full /uw/, nevertheless it is convenient to use thru as a word sign. It would certainly be a pity if thru were given up for the less desirable throo or thruu.
Some reformers have wanted all instances of final /uw/ to be written as <u>, e.g. <shu, kru, kanu> for shoe, crew, canoe. I do not, because a special rule would have to be added to restore the full <uu> before plural endings and the like: shuuz, etc. Furthermore, if we keep <tuu> for <too> - which is never unstressed - we can keep it separate from <tu> for <to> which normally is stressed.
Writing <oo> for /uw/ does not readily permit such reductions. Note particularly that the spelling <du> for <do> lines up very well with Nue Speling duz for does.
2.7 The diphthong /uj/ (or /ui/) and /uw/ before other vowels.
In ruin, doing, bluish speakers vary as to whether a sequence of /uw/ + /i/ in two syllables is used, or a monosyllabic pronunciation with a diphthong /uj/ (also written as /ui/). If <uu> were adopted as the main representation of /uw/, a reduced form <ui> could be used as a compromise spelling to accommodate both pronunciations. In fact, as in Classic Nue Speling, /uw/ before any vowel could generally be written as <u> as a convenient abbreviation. If <oo> were adopted for /uw/, such an abbreviation would not be possible.
2.8 The problem of June, rude, chew, etc.
Most spelling reformers have assumed that the vowels of moon and June, rule, chew, etc. are identical. This is probably so in a great many dialects. But in so many others, involving millions of speakers, this is not the case.
In large part, dictionary makers have been guilty of confounding the facts, in part because they don't want their pronunciation guide to seem to be too complicated (I suppose). For example, Webster's New International Dictionary, Second Edition (1934) - which was long held to be the definitive dictionary by many Americans, and was championed by the New York Times in a special editorial launched against the Third Edition of 1961 - wrote <oo> to indicate both classes of word but noted in the introduction that there are really differences, even for the yod-dropping pronunciations of tune, duty, new:
It is to be observed, however, that the oo in these words [e.g. tune, duty (EAG)] is the advanced variety [u+] [=So far, this dictionary adds nothing new to our discussion but simply recapitulates one of the points I've been making: that the use of <ue> is more widespread than may be realized, since <ue> often shows up as /
uor iw (EAG)], and that suit, duty, thus pronounced are not accurately represented by the spellings "soot", and "dooty", and do not exactly rhyme with shoot, booty. (p. lii)
More significant for the present section are the sentences that follow in Webster's:
... many Americans use the modified [iu] [= /fiw/ (EAG)] sound of u after consonants where the y sound has disappeared or is rare, as after r, l (after a consonant in the same syllable), ch, j, and less often sli, zh. Thus many Americans preserve the historical distinction in sound between rude and rood, rheum - room, threw - through, brewed - brood, chews - choose, chute - shoot, lute - loot, luce - loose, and the vowels of glue - gloom, blue - bloom, etc. (ibid.)In short, the problems encountered with do: dew etc. are even more extensive than most spelling reformers have considered. Martin Joos is an exception: not only did he distinguish between do: dew (as dû: diu respectively) and show the yod-pronouncing distinctions consistently after, <n, t, l>, (e.g. niu 'new', tiun 'tune') - he also noted the distinction at least after <r>. Thus we find rude: rood, threw: through distinguished as riud: rûd and thriu: thrû.
The Oxford English Dictionary often recognizes alternate pronunciations after these residual consonants with a compromise symbol <1ū>: chew is shown as ʧ1ū, Jew as ʤ1ū, glue as gl1ū or glū, blue as blū or bl1ū. The OED is not consistent, however, and writes all the following with ū not 1ū: June, rheum (= room), chute (=shoot), threw (= through), rude (= rood).
Daniel Jones in the third edition of his The pronunciation of English (1950, Cambridge: University Press) maintains that Welsh speakers of English use /ju:/ or /iu/ after /r, ʧ, ʤ/ or consonant plus /l/. Thus, blue as /blju:/ or /bliu/.
My general rule of thumb has been that - in order to preserve the unity of the English language community - dialect differences should be resolved by showing maximal differences. I am tempted to invoke this rule here as well. But the fact that dictionary makers have virtually never shown the /uw/: <ue> distinction after <r, ch, i>, or consonant plus <l> creates something of a problem.
A comparable problem occurs for the so-called 'short <a>' in several dialects of English. Daniel Jones always noted for RP that some speakers had a phonemic contrast between a long and a short 'short <a>' which he wrote as /æ:/ and /æ/, respectively. Thus, in his An outline of English phonetics (1962, ninth edition, Cambridge: W. Heffer & Sons) he noted that sad, bad generally had /æ:/ but lad, pad had /æ/. In his pronouncing dictionary, he recorded several minimal pairs, e.g. bad /bæ:d/: bade /bæd/ (also pronounced as /beid/). He noted that for some speakers, jam actually represented two words, one pronounced /ʤæ:m/ meaning 'fruit conserve', the other /ʤæm/ meaning 'crush, wedging'. Later editions of this dictionary edited by A C Gimson after Jones's death, dropped this distinction.
However, in the USA in many areas, such a distinction is kept - tho it may not be used in the same words that Jones listed. Very commonly can 'be able to' is /kæn/, whereas the noun can 'tin' or the verb can 'to put into a tin' is /kæ:n/. This is my own pronunciation. Some speakers similarly distinguish have /hæv/ from halve /hæ:v/. In other dialects, particularly in the New York City area, /æ:/ has changed its quality and is pronounced with a much higher vowel, that of yeah, and could be shown as /e:/. Noam Chomsky and Morris Halle in their influential book The sound pattern of English (1968, New York: Harper Row) describe the large vowel as 'appreciably tense' as opposed to the shorter one, which they describe as 'negligibly tense'. They add that
This distinction is fairly common, and almost completely predictable, in many American dialects, but the contexts in which it appears vary. (p. 68 fn.)Apart from Jones and Webster's Third New international dictionary of the English language unabridged (1971; Springfield, Mass.: G. & C. Merriam), where /æ:/ or /e:/ is noted as a secondary pronunciation and written with /aaə)/, dictionary makers have never shown a difference between these varieties of 'short <a>'. In Webster's Ninth New Collegiate Dictionary (1985), which is derived from the Third unabridged, the distinction is discussed in an introductory section on pronunciation but ignored elsewhere in the text. The editors justify their practise by maintaining that "this distinction is sufficiently infrequent that the traditional practise of using a single symbol is followed in this book" (p. 34).
The fact that several million educated speakers differentiate the two /æ/ s and several million others differentiate /uw/ from <ue> after <r, ch, j> etc, makes me uneasy about ignoring these phonetic contrasts so cavalierly. I could very readily agree to showing all of these distinctions in the maximal form of English that I think an official written standard should be based on. Or else, they could at least be optional variants. Neither issue is crucial to the main point of my paper, however, and so I shall not belabor them further.
To review that point and the arguments for it: <uu> (or some variant involving <u> such as <uw> or <û>) is the optimal way of writing the vowel in moon. The digraph <uu> is certainly superior to <oo> (which should be abandoned with this value) because <uu>:
(1) agrees with international values of letters,
(2) preserves the etymology of Latin loan words,
(3) also preserves relationships between closely related forms in English itself both across dialects and with regard to morphological relationships within a single dialect, and
(4) is compatible with convenient abbreviated forms with <u> before other vowels (as in ruin) or finally in words of high frequency often having weak forms (as in thru).
Spelling reformers at present are almost all pushing for <oo> for /uw/. But the only thing <oo> has going for it is its relative frequency in the traditional spelling. This frequency is hardly overwhelming, however, and is a specious argument, to be discarded.
In short: spell moon as muun!
Back to the top.