[Simplified Spelling Society Newsletter Spring 1986/1. p27,28 Later designated Journal 2]
See Journal and Newsletter articles by David Stark.

Phonics and Spelling.

David Stark.

[David Stark aims to combine skills and experiences as an Adult Literacy Tutor and as a designer, to help formulate a practical spelling reform. This article is the second in a series, of which the first Alphabetic Consistency in Reading, appeared in the Summer 1985 Newsletter, pp.18-22. He has also contributed to most other recent issues of the Newsletter.]

In my last article I argued that it was understandable that the educational establishment could ignore the benefits of a more alphabetically consistent orthography for reading, since the reading process is basically morphographic. That is, the workings of our brain require 'that whole words and groups of words be processed at a time, these forming potentially meaningful units which can be combined to reveal the meaning of text. Also, where an unfamiliar word does require to be broken down into smaller units in order to identify it by its pronunciation, it is more efficient to use syllables rather than individual phonemes.

Educationalists can argue that it is easier to make up for any deficiency in English orthography by extra teaching, which is their speciality, than to contemplate tampering with the established orthography, thus causing incalculable confusion for existing literates and children caught in the transition period.

Many educationalists employ similar arguments for the writing process. Judging reform to be a nonstarter, even if it is considered, they point to the fact that the children who are best at spelling have developed the ability to visualize spellings before writing them, and they advocate the teaching of skills which encourage this. They play down the phonic approach which they, quite correctly in traditional orthography, regard as unreliable and limited.

One must share a certain amount of sympathy with such an approach for a number of reasons. Firstly, as with reading, it is more efficient to process whole words or at least syllables at a time. When one is writing, one must hold in one's short term memory the individual words in sentences, which form part of the overall message, long enough to write them down. If one has visually memorized the spellings of individual common words, and all the common syllabic elements for uncommon words, one will be able to write a sentence before its part in the overall meaning is forgotten. One might argue that plodding through individual phoneme/grapheme relationships would place a greater burden on the short-term memory, with a tendency to allow only less complicated or shorter messages to be written at a time.

Secondly, the only reason one ever writes anything is for it to be read, and if reading is more efficient when it contains morphographic clues, it follows that these have to be built into text when it is written. The only way to remember homophones is to relate meaning directly to the visual transcription of words, and if inflections are morphographic, the writer will have to ensure that his spellings are meaningful, rather than following an imagined pronunciation which might be unreliable or perceived differently by the reader.

Thirdly, spelling requires much more precision than normal pronunciation can give. Actual pronunciation is too variable and imprecise to act as a wholly reliable standard for an orthography. (A fuller discussion of this follows later.) In any language this will render the oral word memory somewhat deficient placing some reliance on our ability to remember what words look like.

One uses one's knowledge of the appearance of words and possible letter strings or groupings to check if spellings look correct once they are written. However, when I was being taught spelling, I was not allowed to write down a word during a dictation exercise to see if it looked correct, and then score it out and try again if it did not. I had to get it right first time or it was considered a mistake. This encouraged me to develop my recall of the visual appearance of words, or as it was said, 'be able to see the word on the inside of my eyelids'.

I learned many new words by their visual appearance rather than their pronunciation, and in the literary tradition of language learning, I was taught French in a similar way, with less importance being given to pronunciation. In the past, before widespread travel between countries became easy, it was the norm to learn to read the literature of a foreign language without much concern for the living pronunciation. Perhaps this was just as well for French and English where the relationship between the written word and pronunciation is unreliable.

When I decided to learn conversational Italian for travel abroad, I found I was unable to learn it 'the natural way, as children learn it', by absorbing the pronunciation of words. The literary approach was so much fixed in me that I had to see a new word written down first in order to memorize it. The pronunciation was derived from the appearance of the word or learned after it.

There may be an advantage in reforming English orthography to aid learning of English as a foreign language by being taught pronunciation first and then relating, this to the written word. However, it would most certainly be highly beneficial for native speakers of English to relate their wide knowledge of the oral language more directly to the written word. Educationalists who advocate that traditional orthography can be easily and effectively learned using primarily visual methods are ignoring the fact that, in spite of its irregularities, English orthography is overwhelmingly alphabetic. Children work this out without being taught it, even if they cannot learn by themselves how to use all the alphabetic clues in the best way.

In practice we use a mixture of phonic and non-phonic cues when we spell English, the non-phonic ones being used in words or bits of words which we have learned are not phonetically regular. For example, one might use phonics to remember the Bs and T in the words bough and bought, but the <ough> is remembered as a letter string, independently of pronunciation.

Spelling reformers point to the fact that there are at least 2 million adults in Britain who can be described as illiterate, and the vast majority of die rest have problems at some time with spelling. They argue that since all the efforts to teach our way out of the situation, including a 10 year programme of Adult Literacy Training, have been completely ineffective in reducing the overall problem, the only solution is to reform the orthography so that phonic cues can be increased and non-phonic ones decreased. The premise is that children remember pronunciations of words more easily than their visual appearance.

If we, as spelling reformers, can lead sceptics through the minefield of argument to reach this conclusion, we will be rewarded with entry to another. Spelling reformers have spent a great deal of time in the past inventing sets of graphemes for revised orthographies when the greater problem involves the definition of a set of phonemes and a pronunciation standard to link to the lexicon.

Up until about the 1950's, reformers on this side of the Atlantic could argue that British Received Pronunciation (RP) should be used for this purpose. It had been widely studied and defined for some time, and had become established as the verbal currency of the BBC. As such it reached the homes of most Britons, and through the British Empire and in connection with the Second World War it was heard in many countries throughout the world.

However the days are long gone since the BBC hired only RP speakers as newsreaders and presenters. A large number of accents are heard over die media as the central sources of broadcasting become more democratized, local broadcasting develops, and the influx of American and other non-British accents continues. There is no standard accent which prevails in the world and there is no trend towards one. To try and impose an accent like RP as the standard for a revised orthography is totally impractical. Apart from anything else, one would have to force media presenters, teachers and all others who could be used as a point of reference to take elocution lessons so that their accents became standardized and frozen.

In addition to this, even if such a standard accent could be established, it would not be an exact model for an orthography. No writing system is a phonological record of speech, nor could it be. The variability of human speech, displaying emotions stress and intonation, does not produce a machinelike consistency. When we hear human speech, we are so used to this variability that we only use the sounds heard as a rough guide to the words spoken. We rely as much on our hypothesis of whether the possible words made sense in context in order to identify them. The individual jigsaw piece only makes sense when we know die meaning of a large section of the puzzle.

This is why voice recognition computers are so ineffective at present. To program powerful 5th generation computers to deal with processing human speech, linguists and computer scientists will have to define what is meant by 'meaning', and teach the computer how to recognize it. In the statement 'the man played (a round/around) with his secretary', a computer would need to appreciate context to determine whether to write 'a round' or 'around', or deduce whether the subject referred to is a keen golfer or a dirty old man.

We would have the same sort of problem transcribing a foreign language which we did not understand. Even if the person we were listening to spoke the standard dialect and the orthography were as alphabetically consistent as it could be, our transcription would be hopelessly wrong. We would not even be able to separate individual words effectively.

If one studies the definition of the pronunciation of words using the International Phonetic Alphabet as displayed in a large dictionary, one would gain the impression that phonetics is a precise science. However, the precision of a scientific device like a voice spectrograph in studying a particular accent is useless when confronted with the variability of a human voice. Phonetic researchers do not even trust tape recordings of speech, preferring to study subjects at first hand, so that they can observe the various parts of the mouth and vocal tract as words and phonemes are pronounced. Even a trained linguist will be afraid that his own preconception of how he thinks a word should sound will affect the way he perceives the word being spoken.

The conclusions to be drawn from all this are: that it is impractical to choose and fix a standard living pronunciation; that normal speech in such an accent would not be consistent enough to translate directly into written text; and that people do not possess an ability to define individual phonemes accurately, having found the development of comprehension skills as rewarding a way of identifying words.

To find a practical way forward for a revised orthography, we must explore the nature of the pronunciation of words which can be used in a Phonetic approach to spelling, and how this relates to die various living English accents. However that must be left for another article.


Jorm, A F, The Psychology of Reading and Spelling Disabilities, Routledge and Kegan Paul, 1983.

Peters, Margaret, Spelling: Caught or Taught? Routledge and Kegan Paul, 1985.

Back to the top.