[Spelling Progress Bulletin, Winter 1982 pp10-13]
[Valerie Yule: see Book, Journals, Newsletters, Media, Personal View 10, Anthology, Bulletins, Web links.]

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

"A Research-Developed Reform for English Spelling"

by Valerie Yule.*

*Dept. of Psychology, Univ. of Aberdeen, Aberdeen, Scotland, U.K.
*This paper was published in Revista Canada de Estudios Ingleses, Univ. of La Laguna, Tenerife, Spain. No. 4, Apr. 1982.


An international modernization of English spelling has been held up by conservatism backed up by mistaken assumptions - that spelling reform is a purely domestic matter, that ideally it must be purely fonemic ('spelling how you speak'), that the appearance of English print would need drastic change, that immense costs would outweigh the immense savings, that the requirements of the literary elite have priority and are irreconcileable with the needs of learners, foreigners, or the ordinary average public, and that reforms can be argued out or in, regardless of careful technological and psychological research.

This analysis of the international and national requirements for English spelling today accepts none of these assumptions, and suggests directions for investigation of the possibility of a Chomsky-style 'morfo-fonemic' reform that shows the pronunciation of words, conveys their meaning quickly through its visual form, and has simple, consistent rules for learning that takes account of learners' abilities and difficulties.


As 'international' English becomes more important than local 'native' English, it is conceivable that economic, social, and scientific interests could back an international English spelling reform that by-passed home conservatism and was introduced first into international communication and overseas education - brought in first as parallel alternative spellings which then substituted for the old through common preference, in the same way as internal spelling reforms have succeeded in countries like Korea.

However, the old (and still present) assumption must be abandoned, that the whole matter of spelling reform can be argued out in armchairs at an academic level. Spelling is part of modern communications technology even more than a shelf in a corner of an Arts of Education faculty, and it requires the same approach of Human Engineering and inventive testing that has revolutionised the rest of audio-visual communication in the past three decades. There are volumes of research on what is wrong with people who cannot spell; now we must look at what is wrong with the spelling that so meny people cannot learn it - and how to change it so that they can. All the directions for reform outlined in this paper have been suggested by converging findings in independent research in cognitive psychology, education, linguistics, and electronics communication, making the multi-disciplinary approach evident at the Edinburgh conference.

2. The script for English spelling cannot be one applicable to English alone as reformers have tended to assume. At some date, technological change may make a radical super-efficient orthografy practicable or necessary, but it will affect the whole world, and be adaptable for all languages. For the foreseeable future, however, the Latin alfabet must be the basis, since it is the common medium for most modern languages.

3. There has been almost universal assumption that English spelling reform must be purely fonetic, that is, going backwards - reaffirming the principle of the original alfabet break-thru, that symbols represent speech-sounds. However, 'spelling how you speak', the purely fonemic reform, today faces problems of regional variations in English dialect between country and country and even within districts and cities, problems of homofones (words that sound the same), of clumsy polysyllables, of how to represent slurred vowels, discontinuity with present spelling, and the degree to which such a spelling would lose visible relationship to other modern languages. (See Appendix 1).

We now have the benefit of a century's experience in the design of new orthografies for developing countries and languages, and of spelling reforms on other modem languages. This experience shows that in practice, plain sound-symbol correspondence must be modified in consistent ways to make reading for meaning faster and more accurate, and learning to write easier. Without such modification, some of the most theoretically perfect fonemic designs for new spelling for tribal languages have proven disastrously impractical. (O'Halloran, 1981)

The most important differences between an easy spelling and a difficult one may be that the rules must be few and consistent, not multiple and unpredictably applied, and the exceptions to rules must number at most a few dozen words, not thousands.

The rearguard action agenst improving English spelling thinks up meny arguments which all assume that eny reform must be purely fonemic. Today the understandable desire to retain familiarity is rarely dressed up as an 'aesthetic' argument, and few people have the scholarship to be able to benefit from the 'etymological' argument, but at present the main thrust of conservative academic argument is to admit that a fonemic spelling such as Pitman's 'initial teaching alphabet' is proven to be easier for learners, but to claim that the visual appearance of present English spelling is better for users, especially skilled readers.

There are two types of 'visual appearance' arguments, one sponsored by Chomsky, and the other by Albrow and Sampson. It is easy to see by simple examination that neither argument applies very well to present English spelling, but they could be used to support directions for spelling reform.

Chomsky's claim (Chomsky and Halle, 1968, and Carol Chomsky, 1970) is that English spelling shows the 'lexical form' of words which underly their surface pronunciation. It is still frequently repeated, although now disproved. Less than 3% of the irregular spellings in a school book of 6000 words was justified as linking word-families and helping to decode new words by showing underlying lexical form. (See, for example, Francis, 1970, Sampson, 1975, Yule, 1978). However, when one considers how much visual similarity of words across languages aids learners and readers in other languages, Chomsky's idea becomes an exciting possibility for English spelling reform, if applied more consistently - that is, a 'morfo-fonemic' spelling reform that shows the core 'word form' as well as pronunciation.

Albrow (1972) suggests that readers can scan sentences for meaning faster if meaning-bearing words are longer than function words, so they are more easily distinguished, and if grammatical inflexions have invariant spellings, e.g., the plural s in cats and dogs, although the spoken form is closer to /cats/ and /dogz/. If these ideas are proved by research to be valuable, they could easily be part of English spelling reform - for most function words could easily be made shorter still, and the grammatical markers could be made more consistent.

Sampson has suggested that fast scanning in reading is aided by visual distinctiveness in the spelling of English words, and he implies that this is achieved by the bewildering variety of spelling patterns for words - 318 different ways to spell the 20 English vowel sounds, and 226 ways to spell 23 consonant sounds (See Appendix), and by the redundancy of extra letters.

However, agenst that there is the evidence that the most familiar spelling sequences are recognized more easily than the more unusual ones. Reduction of choice in spelling sounds could be a beneficial reform - and could also reduce decision time - for learners, particularly, when sound patterns overlap, as in should/shoulder, were/there/here. The 'redundancy' that is valued in speech or in the content of writing is all related to the message, to ensure that the message may get through even if some of the information is missed; however, redundant letters in the spelling of a word do not shore up the form of a word - they are only 'noise.' Research can easily prove or disprove whether English words are actually more distinctive if streamlined down to essentials - or barnacled with surplus ink. (Would words in the preceding paragraf become more or less recognizable if cut down to size?, e.g., sujested, acheved, ar, mor, lerners, riting, mesaj, thru, misd).

It is possible that the shorter the word, the more visually distinctive it may be, and the easier to scan for meaning. The compact mixed script of Japanese and the almost equally fonemic scripts of Indian languages are reported to be faster to read than English spelling, while early experiments by Beech (1981) and Yule (in progress) are indicating that literate adults can need only a few hours' practice in order to reach their normal reading speed when tested on reformed English spelling systems that use few but consistent rules which remove the irregularities and 'redundancy' which are claimed to be an advantage. With more practice, the subjects might well become faster than they are with the present English spelling.

Streamlined consistent spelling is also likely to serve the interests of learners too, since it avoids the problems of a purely fonemic spelling - of longer polysyllables and risking everything on auditory discrimination. Recent research has been finding differences between people who prefer a 'Chinese' strategy of visual clues in reading and writing, and those who prefer a 'Phoenician' alfabetic and fonic clues. (Baron and Strawson, 1976) It has been suggested that Phoenician may be the better method for learning or writing, but Chinese is the better for faster, efficient reading, and that the two interests conflict. (Frith, 1981)

However, research on learning and learning difficulties is tending to suggest that the more economical the representation of a word, and the shorter it is, the closer it is to a 'Chinese' type of compact visual gestalt, and the easier it also is to use auditory analysis and synthesis of a Phoenician type, in view of the nature of short-term memory and its limitations. The more decorated and lengthy the spelling, on the other hand, the harder for 'Chinese' operators because the gestalt is weaker and the basic structure less visible, while 'Phoenician' operators take longer to scan or to resynthesise a sentence, and find poorer linking to the spoken word. (Sometimes I think the natural spelling of five-year-olds should be the model - c.f. "I hav ben t th epot and ther ws a plan ful ov pepl nd lugaj.")

The ideal spelling might be shown to have a fonemic basis, for learning and writing and speaking and to ensure that the primarily visual activity of reading had the slower strategy of 'sounding out' words as an essential back-up technique to decode new words. However, this fonemic base would be mediated not purely by direct sound-symbol correspondence, but by a limited number of modifying rules which pack as much information into the appearance of a word as possible, to transmit word meaning and sentence meaning, avoid confusion with other words, and promote faster visual scanning. It would still be possible to derive the spoken language from the written and vice versa.

4. The assumption that eny real reform of English spelling would require such radical revision that everything now in print would become obsolete, and impossible demands would be made on the present literate generation faced with a completely new spelling. However, if you look closely at a printed page of English, you will observe that 70-80% is regular in the crucial sense that it is predictable from rules. Only 20-30% needs 'cleaning up.' As it is, this 20-30% wrecks the whole system, because you cannot tell in advance what is going to be predictable and what is not. If there are only three booby-traps on a road, it is still the whole road that is unsafe.

An international perspective on the visual-versus-fonemic issue, however, restates the dilemma in a form research can tackle. What weight must be given to the need for an international standard for sound-symbol relationships, and to the need for visual similarity of similar words in different languages that give them different pronunciations? (e.g. theatre, imagination, or machine). To what degree could the apparently conflicting demands be reconciled?

Mosterin's recommendation (1981) of the universal adoption of the International Fonetic Alfabet (IPA) for all national spelling reforms is interesting but there are meny difficulties. The present IPA letters are not well suited for everyday use in print and handwriting, are not generally available in printer's fonts and on typewriters, the large number of symbols is unwieldy for our present technology and would require too much variation of keyboards from country to country; the symbols are designed for precise representation of sounds whereas an English spelling would do better with 'diafonic' representation, conventions that allowed some dialectal range in their pronunciation rather than quibbling about whose speech would be the 'standard.' The major question remains: is it more important to preserve visual similarities between languages, or to clarify fonic differences in their spoken forms - or can the two be reconciled?

At present, the relationship of consonant letters and sounds in English spelling is basically close to IPA and international usage, and only needs 'cleaning up' the exceptions. These are the sounds represented in English spelling by j, ch, th, wh, x, sh, ng, c, qu, or the sound of zh. However, English use of the five Latin vowel letters a, e, i, o, u differs from Continental usage because English has a different set of paired long and short vowels, which can alternate systematically within word-families.

Switching to IPA vowels and their Continental usage would change the appearance of English text dramatically, with 22 vowels required, of which only one would be retained as at present, and four extend their occasional representation. (e as in bet, a as in car, i as in police, o as in solo, u as in tabu).

However, observation and experience suggest, and experiment could test the opinion, that as long as shifts in sound-symbol representation are systematic and limited in range and number, and the fonemes are close enough to existing repertoire, learners of different languages adapt quickly to some variation in the values of letters and letter combinations - usually in the first lesson. Sets to speak a different language can change like a shift in gears; so can set to read one.

At this stage it may be appropriate to give an illustration of what a 'morfo-fonemic' spelling could be like, that included in its charter the requirements that have been discussed, of continuity with present spelling and international recognizability, of economy, of minimum 'special cases' for distinguishing confusable homofones or abbreviating the commonest function words, or providing grammatical markers, of operating within an IPA framework or towards one, as far as it seems practicable, with no variability in consonant representation except for nine special cases (described below) and rules that govern a limited range of vowel representations:

"How, cd yu expect me not t'be wurrid at whot that antiqated lejislater thay caul th' public wil say when it sees me now, aftir al these years I hav been sleping in th' silens o oblivion, cum out with al my years on my bak, with a tale as dry as a rush, barrin o invension, devoid o stile, poor in wit an laking in al lerning and instruxion , without qotasions in th marjins or notes at the end o th' book; wheras I see uthir werks, nevir minde how fabulus an profane; so ful o sentenses from Aristotl, Plato an th' hoel herd o filosofors as t' impress thair reders an get thair authors a reputasion for wide reding, erudision an eloqens? (Prolog, Don Quijote)


1. The principle of representing the 'form of the word,' despite sound changes, has been achieved through the simple technique of extending the existing use of 'silent e' to indicate that a preceding vowel is long. The corollary, absence of a silent e or use of double consonants, indicates when the preceding vowel is short. This can cover most cases except some initial vowels. e.g. slepe/ slepd/ sleping not slepping); profane/ profanity, long vowel shown: antiqated, legislater, these, stile, minde, etc. short vowels shown: wurrid, aftir, barrio, uthir, nevir.

2. Vowel representation modified by place, in word and length of word:

thatwhenwilnot cum



This vowel scheme allows for some regional variation in pronunciation, some flexibility in further reform (e.g., towards Continental vowel representation) and choices are generally rule-governed so that the reader can know how to say what he reads, and the writer can know how to spell what he writes.

3. Special cases. Shortening of function words, e.g. cd, tb, t, o, etc. Distinction of confusable homofones only, e.g., hoel/hole, -sion, -tion, -zion, suffixes as conventions for pronunciations /-sion/, /-tshun/, /-zhun/ to preserve continuity (although sn, tn, zn, might serve better)

4. Grammatical markers. -s as plural and verb ending; single nouns may end with -se or -ss. -d, -n as verb participle endings. (e.g., grone is a noun, groen is a verb)

5. Consonants. Transisional retention of velarplosives c, k, q (not qtr) with rules for their use. j still with English pronunciation until international agreement on j, y, etc. is established by research. Formal spelling of words like nature, special, which are slurred in actual speech.

6. Visual distinctiveness and speed of reading. A transcription in other languages is given in Appendix, and I would welcome reports of timed tests, using each subject as his own control with a time interval, and alternation of order of presentation to subjects. (The reformed spelling is 5%-10% shorter, as shown by the indication of omitted letters in the illustration.)

7. Towards an international English spelling. The illustration can be compared with other transcriptions (Appendix) for resemblance to the Spanish original. In a comparison of the spellings of 100 'transnational' words from the commencement of that passage and later paragrafs, findings for closest resemblance were:

39% present spellings, 26% 'morfo-fonemic' Spellings, with 35% other spellings identical.

The words in which present spelling has the visual advantage are of course all words which present difficulty in pronunciation and spelling to the foreigner, and it remains to be tested, indeed, whether the advantage is actually complete, i.e. whether there still remain in the 'reformed spelling' version sufficient visual clues for transnational recognition, as well as sufficient fonemic clues for transnational pronunciation according to the English key.


All the ideas put forward here are subject to testing by empirical research. They may be substantiated, modified, or refuted, and are in no final form. But we should learn from metrification the hazards of implementing eny ideal system without thorough practical testing first. In spite of what most alfabeteers say, most of their systems have not been tested adequately.

Geoffrey Sampson has suggested (1980) that failure to initiate English spelling reform may be linked with the loss of national self-confidence. "We see ourselves now as following the lead of others rather than as the model to which foreigners aspire; since even the Eurocrats of Brussels have not yet presumed to reform our own language for us, we instinctively suppose that change must be inappropriate or impossible."

I am of course being provocative in suggesting that the 'Eurocrats' or rather the 'cosmopolitans' may change, not the language, but the spelling. However, the first steps that can safely be taken within all the foreseeable possibilities for future English spelling can be taken, by all and enyone now - tacit adoption as alternative spellings of the international usage of f for ph and e for the short e sound, as in bet (as has been used throughout this article). Both are now appearing in the English-speaking press, often as much unintentionally as by design, and Australian publishers are putting out books and journals which use 'Spelling Reform 1', the short e reform.

Further directions for reform can be tested out now - what spelling can best help adult illiterates, dyslexic learners, fast readers, transliterating computers? There is exciting research needing to be done on the question of international convergence in sound-letter conventions, and how the spellings of one language can be most accessible to the speakers of another.

There may be some future technological break-thru in a completely new direction, but an internationally useful English spelling reform is needed now, that could be gradually introduced by the existing route of co-existing alternative spellings. The features put up for research and discussion are

1. Consistent use of the present alfabet, with reduced choice of vowel spellings regulated by few consistent rules, to maintain continuity with present spelling, accomodate regional dialect variation and allow a future transition to a spelling system applicable to all languages, if required. The 'silent e' and double-consonant techniques are extended to preserve basic word-forms that have sound-shifts between English long and short vowels.

2. Maximum compactness, including condensed function-words, to aid visual distinctiveness and efficient scanning for the 'Chinese-strategy' reader, ease of production and decoding for the 'Phoenician-strategy' writer and learner.

3. Minimum rules and exceptions to modify basic fonemic spelling, to meet the needs of reading, learning, pronouncing, writing, and electronic communications technology.

We have dramatic evidence all around us of human capacity to adapt to change. Stone Age Papuans entering the modern age in a few decades, modern cultures risking reversion to the Stone Age. Research is also showing that intelligent, literate people can even adapt to changes in their spelling system in hours rather than generations, and all the resources of modern communications research give guidelines on how the transition can be made.

The spelling of a language is an example of the importance of very small things, often ignored as much as the air we breathe. A social elite can use it to maintain its own superiority by claiming that the convenience of the most educated users must be its major determinant, or it can be an instrument for universal literacy and fuller development of all the potential intelligence in a population. In this day and age, the reform of English spelling could be a sign that hope is possible. "The letter killeth, but the spirit giveth light."(Paul's 2nd letter to the Corinthians)


Spanish compared with two English translations in spelling reform and with present English spelling.

1. Spanish. "Porqué ¿cómo queré is vos que no me tenga confuso el que dira el antiguo legislador que llaman vulgo cuando vea que, al cabo de tantos años como ha que duermo en el silencio del olvido, salgo ahora, con todos mis años a cuestas, con una leyenda seca como un esparto, ajena de invención, menguada de estilo, pobre de concetos y falta de toda erudición y doctrina, sin acotaciones en las márgenes, y sin anotaciones en el fin del libro, como veo que están otros libros, aunque sean fabulosos y profanos, tan llenos de sentencias de Aristóteles, de Platón y de toda la caterva de filósofos, que admiran a los leyentes y tienen á sus autores por hombres leí dos, eruditos y elocuentes?" (Don Quijote, Alhambra edition, 1979, pp 56-7)

2. An English translation using a fonemic spelling reform ("World English Spelling") with a consistent relation between sounds and letters. (Spell as you speak) "How cuud yoo ekspekt mee not too bee wurid at whot that antikwaeted lejislaeter thae caul the publik wil sae when it seez mee now, aafter aul theez yeerz Ie hav been sleeping in the sielens ov oblivion, cum out with aul mie yeaz on mie bak, with a tael az drie az a rush, barin ov invenshun, devoid ov stiel, puur in wit and faking in aul lerning and instrukshen, without kwoetaeshunz in the marjinz or noets at the end ov the buuk; whaeraz Ie see uther works, never miend how fabyoolus and profaen, soe full ov sentensez from Aristotul, Plaetoe and the hoel hurd ov filosoferz, az too impres thaer reederz and get thaer autherz a repyootaeshun for wied reeding, erudishun and elokwens?

3. A fonemic spelling for English using the Roman alfabet but within the guidelines of the International Ponetic Alfabet.
"Haw cud yu: ekspekt mi not to bi wurid aet whot that aentikweited ledzhisleite: thei co:l th publik wil sei when it si:z mi nau, a:fte: o:l thi:z jie:z Ai haev bi:n sli:ping in th sailens ov oblivion, cum aut with o:l mai jie:z on mai baek, with a teil aez drai aez a rush, baeren ov invenshen, devoid ov stail, pue: in wit aend laekin in o:l le:ning and instrukshen, without kwoteishenz in th ma:dzhinz o: nots aet thi end ov th buk; weiraez Ai si: uthe: we:ks, neve: maind haw faebjulus aend profain, so ful ov sentensez from Aeristotel, Pleito aend th hol he:d ov filosofe:z, aez tu: impres the: ri:de:z aend get the: o:the:z a repju:teishen fo: waid ri:ding, erudishen send elokwens?"

4. Translation in present English spelling for comparative tests of readability.
"How could you expect me not to be worried at what that antiquated legislator they call the public will say when it sees me now, after all these years I have been sleeping in the silence of oblivion, come out with all my years on my back, with a tale as dry as a rush, barren of invention, devoid of style, poor in wit and lacking in all learning and instruction, without quotations in the margins or notes at the end of the book; whereas I see other works, never. mind how fabulous and profane, so full of sentences from Aristotle, Plato and the whole herd of philosophers as to impress their readers and get their authors a reputation for wide reading, erudition and eloquence?"


Albrow, K. H. (1972) The English Writing System: Notes towards a description. London: Longman.

Baron, J. & Strawson, C. (1976) "Use of orthographic and word-specific knowledge in reading words aloud. Journal of Experimental Psychology: Human Perception. 2 .386-93.

Beech, J. R. (1981) "The Effects of Spelling Change on the Adult Reader." Paper presented at the 1981 SSS Conf.

Cervantes, M. Don Quijote de la Mancha. Spanish Ed., Alhambra, Spain, 1949. Eng. tr. by J.M. Cohen, Penguin.

Chomsky, C. (1970) "Reading, writing and phonology." Harvard Educational Review 40, 287-309.

Chomsky, N. & Halle, M. (1968) The Sound Patterns of English. Harper & Row, New York.

Francis, W.N. (1970) "Linguistics and Reading" in Levin, H. & Williams, J.P. (Eds.) Basic Studies in Reading, Basic Books: New York.

Frith, U. (1981) "Cognitive Processes in Spelling and their Relevance to Spelling Reform." Paper for SSS Conf.

O'Halloran, G. (1981) "The Road to Eng. Sp. Reform in the light of the development of Orthografies for other Languages." Paper in absentia for SSS Conference.

Mosterin. J. (1981) "Spelling Reform in International Perspective." Paper presented at the SSS Conference.

Sampson, G. R. (1980) Schools of Linguistics: Competition and Evolution. Hutchison: London.

Sampson, G.R. (1981) The Advantages of English Spelling. (in press)

Yule, V. "Is there Evidence for Chomsky's Interpretation of English Spelling?" S.P.B. 18,4,1978, pp. 10-12.

Back to the top.