[Spelling Progress Bulletin, Winter 1982 pp4-7]
[Walter Gassner: see Anthology, Bulletins.]

SSS Conference 3: Development of Improvement in English Orthography continued.

"How to reconcile conflicting principles for a reformed English spelling,"

by Dr. Walter Gassner*

*Randwick, N.S.W. Australia. Deceased Dec. 4, 1981.


The conflicting principles are:
(1) Consistency in the use of letters and minimizing deviations from traditional orthography. The latter involves: avoiding the introduction of new letters or written accent signs: moreover, avoiding, within the limits of practicability, unfamiliar use of letters; and finally, providing a means to distinguish words with the same sound but different meanings which are distinguished by distinctive written forms in the traditional system. (It does not involve limiting arbitrarily the number of occurences of letters whose sounds are established unambiguously, such as k and z, and replacing them by c and s respectively, just to conform with established practise). Any system that would aim at preserving more than, say, 30% of traditional written forms would, necessarily, be encumbered with complicated rules, and such a situation should be avoided.

Ways to reconcile the conflicting principles: (a) allowing certain sounds to be represented by letter groups, such as sh, ng, ai; (b) allowing certain letters to be used for different sounds in different positions, such as y for the consonant in "yes" when occuring in front of a vowel, and for the diphthong in "hydrant" when occuring in front of a consonant (where the other sound cannot occur); (c) allowing certain vowel letters to be used for different sounds in stressed and unstressed syllables - as "a" for the sound in "hat" in stressed syllables, and for the Neutral Vowel in unstressed syllables (This has the prerequisite that the system is one in which the location of stress is clearly indicated).

(2) Basing the system on one particular pronunciation or not taking into account variant regional pronunciations. The speech of Southern England is used as a basis (not for any inherent superiority, but on the grounds of intelligibility and acceptability and the possibility of checking it in reference books). Deviations from it are established, where a substantial number of speakers make a distinction that speakers of Southern English do not make and such distinctions are backed by distinctions in traditional orthography - as the use of ar, or, and er.

(3) Differences in styles of speech: Using as a basis careful pronunciation of educated speakers, but rejecting fanciful pronunciations that exist only in the minds of persons who are influenced by traditional orthography. Words with distinctive "strong" and "weak" forms rendered so as to suggest the strong form, the form used in isolation. This for the sake of consistency.

(4) Rejecting inordinate emphasis on time, saving and etymology. A brief description of the author's Consistent Evolutional Spelling pointing out the features through which the best possible compromise between conflicting principles is achieved. Occasional reference to the position in other languages.


At this and previous conferences, a large number of spelling reform proposals have been presented. The common goal of all these proposals is to replace the present traditional system of orthography - which is full of arbitrary features and inconsistencies - with a system that is simple and straightforward, easy to learn and easy to use. Most spelling reformers - the exceptions being, of course, those who want to introduce a completely new system of writing-stress the need to limit departures from existing practice to the absolute minimum, and this is justified in view of the need to make transition easy and to obviate a situation in which it will be impossible for future generations to read and understand anything that has been written before the reform. These two principles are in direct opposition to each other; and whilst most reformers try to bridge the gap between the two conflicting principles, it lies in the nature of things that they arrive at different results.

This is an attempt to determine what is essential and to arrive at a solution which is effective and workable.

In order to determine what is essential, we have to remind ourselves of the plight of youngsters who are faced with the task of learning to read and write a language in which sounds and letters do not agree. They have to memorize long lists of words by rote, an exercise which is stultifying because it is devoid of any stimulus to logical thinking. It is also wasteful from the educational point of view because the time required to become literate is an obstacle to proficiency in other studies, the range of which is continually increasing. Where a language is phonetic, all that a person who knows how to speak has to do is to learn the alphabet and their sounds and, possibly, if the alphabet is deficient, a few groups of letters. Attempts at framing rules for English spelling have, of course, been made and some ingenious teaching devices have been proposed, but they do little to alleviate the situation. A rule that is riddled with exceptions becomes useless if the exceptions appear arbitrary. It is only if this situation is radically remedied that spelling reform becomes effective; when such a reform is implemented, it will no longer be necessary to ask the question, "How do you spell (such and such a word)", for the answer would almost invariably be, "As you pronounce it" or, possibly, "As you hear it pronounced" (say, by the B.B.C. or some other authority) - subject to qualifications only in variations in pronunciation and the need for clarity.

Likewise there should be no need to be provided with pronouncing dictionaries or to have the pronunciation of words indicated in an all-purpose dictionary. If one takes account of the fact that there are numerous systems of imitated pronunciation, varying from one dictionary or reference work to another, with which users in most cases are not familiar, one can see that here, too, a lot of time and effort is wasted, which will be saved if a phonetic system of spelling replaces the existing one.

From what has been said, it follows that "scratching merely the surface," that is, eliminating certain glaring arbitrary written forms, or omitting redundant letters, would not sufficiently alter the present position, for learning by rote would still be the necessary thing to do; a limited reform of that type would not be worth the upheaval. It is only when the written forms of words are a reliable guide to their spoken forms and vice versa, that the staggering rate of illiteracy will disappear in English-speaking countries and that one can expect foreign students to arrive at an acceptable and intelligible pronunciation. (I am, of course, thinking of persons who are able to express themselves flawlessly in writing and yet distort the words when expressing themselves in speech). Obviously, the new spelling, as I conceive it, will not prevent foreign or regional accents from continuing to exist, but in general, every user, whether English-speaking or otherwise, will be able to deduce the written form from a spoken form with which he is familiar, and vice versa.

If consistency and effectiveness are the essential requirements, it follows that the reverse principle, that of limiting departures from existing practice to a minimum must play a minor role. Yet, in certain respects, it is an essential principle. The important feature of a system thus conceived is not the number or percentage of words that remain unchanged, but the degree to which words are recognizable without special instruction to persons accustomed to the traditional spelling. Thus, there is no point in insisting that the letter c should continue to be used for the k-sound and the s-sound, because we have the letters k and s at our disposal. If, accordingly, the letter c were to be eliminated, the words in which a c is replaced by a k would evidently be recognizable at sight; indeed, only persons who are adverse to any change should be shocked at an increase of the occurrence of the letter k. What we must reject, however, is the idea of abandoning the Latin alphabet or augmenting it with new letters. Implementing an entirely new system of writing, as George Bernard Shaw suggested and provided for in his will, would sever the ties of the English language not only with the past, but also with the languages of the greater part of the European continent, and, indeed also of the other continents - seeing that even for Chinese the Latin alphabet has - been allocated certain functions. And these remarks are applicable not only to a completely unrelated system of writing, but also to an alphabet which is essentially the Latin one, but is augmented by additional letters or diacritical marks. No matter how much ingenuity is evident in the designing of these additional symbols, they would impair the readability by the uninitiated and the acceptability from a world-wide point of view.

The Latin alphabet has 26 letters, and there is some agreement that the English language has at least 40 distinctive sounds, or rather phonemes. This is evidently an area where there is conflict between fundamental principles, but these are easily bridged if one agrees to the attitude that for some sounds, we can continue to use groups of letters (digraphs or trigraphs, referred to as "compound symbols") - chiefly for long vowels and diphthongs. There is quite a choice of such compound symbols among those used in traditional orthography, and this makes it possible to reflect one important aspect of the pronunciation of words: stress. Many spelling reformers will refrain from indicating the stress in their proposed systems, arguing that a notation of that kind is impracticable or unnecessary. Suggestions are made to the effect that stress might be marked, especially in books for children and foreigners, by underlining, bold type or written accent signs - but that for general use, the indication of stress can be dispensed with. I hold the opinion that indication of the stress should be incorporated in the system in common use - thus avoiding the additional expense and effort to have books especially marked for certain types of users. However, none of the devices mentioned would be practicable and they would be deviations from existing practice. Indications of stress can, indeed, be effected by making available a second set of vowel symbols in addition to the ordinary ones and, in a limited range of instances, by using double consonants after short vowels. Of the various ways of indicating the stress (I have experimented with several), the most appropriate one is to use the second set, as referred to above, to mark the stress on a syllable that is not the initial one, on the understanding that where only "ordinary" symbols appear in a word, stress falls on the first syllable.

Once it has been made clear which syllables in a word are stressed and which are not, it is possible to represent certain obscure vowel sounds - sounds that can only occur in unstressed syllables - by vowel letters which have a different function in stressed syllables. The sound that requires special attention among obscure vowels is the so-called neutral vowel, sometimes referred to as "schwa" or the muttering vowel sound. It is the sound most frequently occurring in unstressed syllables and is, under the existing system, represented in a variety of ways (a in about, e in silent, o in develop, u in circus, ou in grievous, ia in parliament, iou in precious, oi in tortoise). In the International Phonetic Alphabet it is represented by an inverted e (thus ə). Earlier spelling reformers ignored the existence of this sound. Their schemes were based on the assumption that the words involved contained the sounds which these letters have in stressed syllables, either leaving the written forms of the unstressed vowels as they are in the traditional system, or with minor arbitrary simplifications. The effect would have been a continued need to memorize the spelling of a lot of words. Then came some spelling reformers who did take account of the existence of the neutral vowel, establishing newly invented symbols for it. As mentioned before, extending the alphabet is extremely undesirable - even if only a single symbol is added to it. Apart from the costs and inconvenience of adapting all printing fonts and typewriters in English-speaking countries, there would have been the additional problem of printing English words in countries in which other languages are spoken; and the need to print English words throughout the world is obvious in view of the position of English as a world language. At one stage I toyed with the idea of using the letter "q" for the neutral vowel - not a new letter, but one that in the existing system serves no useful purpose. Later on I abandoned this idea in. view of the strange appearance given to the most common words.

It is actually because the second set of vowel symbols is used as a stress market on syllables other than the initial one, that an obscure sound such as the neutral vowel can be represented by a letter which is used for a different sound in stressed syllables. The letter best suited for this purpose is the letter "a" - which in stressed syllables represents the sound occurring in the word "hat"; this chiefly in view of the frequent occurrence of this letter for the neutral vowel sound in the initial and in the final position - almost to the exclusion of other ways of representation. (In about, afraid, along, the sound occurs in the initial position; in banana, data, China, villa, the sound occurs in the final position). The cases in which the traditional system uses a combination involving the letter r at a word end or before a consonant - such as river, tailor, sugar - are different; here the symbol er is used. It is not possible to show the occurence of the neutral vowel in words of one syllable, and it is actually not necessary to do so because such pronunciations of monosyllables with the obscure vowel sound, as can in the expression: "I can do it" without any emphasis on can - have to be considered as incidental to sentence stress. It depends upon the desires of the speaker as to whether he wants it to be stressed or not. And sentence stress is something that is not practical to be indicated in the spelling. All such words as at, from, of, have, must are represented the way they are pronounced in isolation.

Two letters of the alphabet - "y" and "w" - can be used both as consonants and as vowels, as they are in T.O. The consonants are those occurring in yes and wind, and the vowels are those occurring in hydrant and put, how (which will be written pwt, how). This double use is not an infringement on the principle of concistency if the rule is established that the two letters in question are consonants when followed by a vowel, and vowels when followed by a consonant, or used as part of a diphthong.

Another source of differences of opinion is the variety of pronunciations in various parts of the English speaking world. A certain form of speech has to be taken as the standard, and it has to be a form that is reflected in currently used pronouncing dictionaries, in dictionaries in general use and in foreign language dictionaries in which the pronunciation of the English words is shown for the benefit of foreign students. This standard is Southern British, sometimes referred to as "the Queen's English" or "Received Pronunciation." Deviations are allowed for in cases in which a substantial body of speakers uses an alternative pronunciation which is clearly backed by use in traditional orthography. Thus, certain vowel sounds are split up into "cases with r" and "cases without r" (such as aa and ar), the letter r in such cases being almost silent in Southern British speech, but sounded in Scottish speech. Also the vowel in such words as "ask" (pronounced with the a in "father" in Southern British speech) is represented in a special way. Generally speaking, where different pronunciations are used in different styles of speech, the system is based on careful pronunciation used by educated speakers, but fanciful pronunciations that exist only in the minds of persons who are influenced by traditional orthography are left out of account.

The suggestion has often been made that spelling reform should be put into practice gradually, the idea being that changes of a trifling nature would be more easily accepted, and that with each successive change, resistance would decrease. My chief objection to implementing a spelling reform in a large number of small steps is that the intermediate steps would, of necessity, be unphonetic (they would even in some instances deprive the written forms of that modicum of consistency they might appear to have) and that each step would require re-editing of dictionaries and reference works. However, following a frequently heard demand, I will make two suggestions for a spelling reform step-by-step: the first, in fact, chiefly to reject it and to demonstrate why; the second is one that is practical, simple to understand and reasonably extensive. The first is based on the assumption that we, the alphabeteers, have come to an agreement and that we know what the final outcome is going to be. The first step is that proposed by Mr. Lindgren of Narrabundah, Australia, viz. that the letter e should be used for the vowel sound in "bet" to the exclusion of all other ways of representation. In my opinion, this makes sense only if we also refrain from using this letter for other sounds and use the symbols that truly represent these sounds. Thus: ee in "lever", i in "pretty." Now with each subsequent step the same procedure would have to be followed, and if each step takes 10 years to carry out, it would take over a hundred years to arrive at the final shape. It is self evident that a step-by-step spelling reform of this type would require a public throughout the world (whether English-speaking or not) endowed with an infinite amount of patience and docility, ready to replace their dictionaries frequently with new ones and assimilate the changes gratiously.

The other suggestion for a gradual approach does not assume that there is complete agreement among alphabeteers. And only one intermediate step is required - the rule is simplicity itself: to each of the usable letters - all except c and q - one sound is allocated, and whenever this sound occurs, that letter is used. (But the letters continue to be used also for other sounds - adjustment of this and handling sounds that are represented by compound symbols and other features have to be left to the second and final step - which can be taken only after spelling reformers have come to an agreement). In the proposed intermediate system, the vowel letters would have their "short" values - as in hat, bet, sit, hot, and hut. Among the consonants, k and s will replace c; z will frequently replace s; f will replace g wherever it is thusly pronounced, leaving to g the duty of representing the "hard" sound (as in get); z represents only ks, and f replaces ph and sometimes gh (as in tough).

Once this intermediate system is introduced, a definite effort should be made by all spelling reformers to come to an agreement. The final goal-effectiveness in learning in the sense that learning lists of words would no longer be necessary - should not be left out of sight.

Whilst, generally speaking, a reform in steps is undesirable, one which would not interfere with arrangements in dictionaries for the intermediate step is not so. The German language is much closer to being phonetic than English, but if it were to attain the same standard as envisaged for English, a lot of alterations would be necessary. But there is one change that could be carried out prior to a large scale spelling reform - and there is a strong movement in Germany in favour of it: abolition of the capitalization of nouns in general, limiting capital letters to proper names, as in other languages. This step would not interfere with the arrangement in dictionaries and could be carried out in advance of a more thoroughgoing reform.

It is hoped that those who oppose a spelling reform for fear that it would destroy a valuable inheritance will rest assured that such a sacrifice will not be required. Those in favour of radical reforms, introducing a new alphabet, or augmenting the old may consider that all the advantages they envisage can with equal ease be achieved by staying within the limits of the existing alphabet. (see example)

Examples from Consistent Evolutional Spelling

From "A Krismas Karal, by Charlz Dikinz. Marli'z Goast.

Marli woz ded, tw bigyn wi'th. Thair iz noe dout whottever about that. The rejister ov hiz berial woz siend bie the klirjiman, the klark, thi undertaiker, and the cheef moerner. Skrooj siend it. And Skrooj'iz nain woz gwd for enithing hee choaz tw pwt his hand tw.

Oald Marli woz az ded az a dornail.

Minde! Ie doant meen tw say that ie noa, ov mie oan nolij, whot thair iz pertykywlerli ded about a dornail. Ie might hav been inklinde, mieself, tw rigahrd a kofin'nail az the desist pees ov iemmunggari in the traid. But the wizdam ov our ansisterz iz in the simili; and mie unnhaeload handz shal not disturb it, or the Kuntri'z dun for. Yoo wil thairfor permyt mee to repear, emfatikali, that Marli woz az ded az a dornail.

Skrooj neu hee woz ded? Ov cors hee did. How kwd it bea utherwiez? Skrooj and hee wir partnerz for ie doant noa how meni yeerz. Skrooj woz his soal egzekywter, hiz soal admynistraiter, hiz soal asine, hiz soal rizydyweri legatea, hiz soal frend, and soal moerner. And even Skrooj woz not soe dredfwli kut up bie the sad ivent but that hee woz an exalant man ov biznis on the veri day ov the feunaral, and solamniezd it wi'th an undowtid bargin.

The menshn ov Marli'z feunaral bringz mee bak tw the point ie startid from. Thair iz noe dout that Marli woz ded. This must bea distynktli understuud, or nothing wunderfwl kan kum ov the stauri ie am goaing tw rilayt. If wee wir not pirfikrli konvvnst that Hamlit's faather died bifoer the play bigaen, thair wwd bea nuthing mor rimahrkabl in hiz taiking a stroal at night, in an eesterli wind, aponn hiz oan ramparts, than thair wwd bea in eni uther midl'aijd jentlman rashli timing out affter dark in a breezi spot - say Snt Paul'z Chirchyear for instans - literati tw astonnish hiz sunn'z weak minde.

From the monolog in Akt III, Sean I ov "Hamlit" by Wiliam Shaikspeer.
Tw bea, or not tw bea; that iz the kweschan.
Whether 'tiz noabler in the minde tw suffer
The slingz and aroaz ov outrajas forchan,
Or tw taik armz agenst a sea ov trublz,
and bie opoezing, end them? Tw die: tw sleep;
noe mor; and bie a sleep tw say wee end
the hahrtaik and the thouzand nachwral shoks
that flesh iz ehr tw; 'tiz a konsumayshn
divowtli tw bea wisht.

(Editor's comment):

 Gassner is not very consistent in the use of his system(?). In the 2nd line, he spells about as abowt. Yet in the 2nd line, 2nd paragraf, it is about. He spells hee yet bea, and in 1st line, noe but in 2nd line, 5th paragraf, noa. Also wwd for would, yet understuud.

Back to the top.