Reducing the worst irregularities of English spelling
English spelling is in need of reform because it makes learning to read and write too slow and difficult. It entrenches educational underperformance and inequality, because it leaves literacy acquisition too heavily dependent on home background and help from parents. But after centuries of putting up with a shambolic system, the main aim of reform should now best be just to make learning to read easier. Spelling ability is becoming less important, because most adults now write only on electronic devices which automatically correct the majority of mistakes. On mobiles phones, writing has become largely just a matter of recognising the right word and tapping on it. Reading remains essential, for education, life and work.
Learning to read English could also be made easier with much less change to the English spelling habits than learning to write. It could be greatly speeded up by merely undoing some of the worst changes inflicted on it since it was first adapted from Latin in the 7th century by:
|Reducing exceptions to short /e/ (e.g. head, said, friend), short /u/ (come, once, touch) |
and /ee/ (speak, even, believe, weird, people …),
Reverting to sensible consonant doubling (without inconsistencies like ‘finish dinner; announce annual’)
and dropping surplus -e endings (gone, have) which undermine their sensible use (bone, save).
All common words with the above irregularities [along with respellings which show how they would look if spelt systematically] can be seen in chapter 11, “Amendments which would help learners most”, in my blog English spelling and its effects, entitled .
One of the strongest reasons for reforming English spelling is to make the early school years less difficult and demotivating. Its irregularities make the lives of young children much harder and less enjoyable, and much more dependent on help from adults, than in other languages. And one of the biggest causes of this are the irregular spellings for a, e, i, o and u in around 200 much used words, like ‘said, any, other’. Any reduction in their number would be helpful.
Nearly all children make good progress with both learning to read and write when they first start, because they are mostly taught only with regularly spelt short words, like ‘dad met him on the bus’. If /a/, /e/, /i/, /o/ and /u/ were always spelt like that, their reading and writing would continue to improve rapidly. It becomes markedly slower only when they begin to be introduced to words in which some of the letters with which they first learned to read are used for different sounds, such as ‘was, he, kind, once, push’.
The irregular spellings for a, i and o do not impede progress as much as those for e and u because: 1) Short /a/ is spelt irregularly just in ‘plait, plaid’ and ‘meringue’. 2) Short /i/ is irregular in only six common words (build, built, English, pretty, busy, women). The other 39 are less common and hinder learning much less (abyss, crypt, syllable, synchronise, system). 3) Short /o/ is spelt abnormally in only 33 words, mostly with a after w and qu (was, want, squash…). Only a few more recently coined words use o (wobbly, wombat), but wa and qua cause some reading difficulties (swan swam; was wagging) and would ideally be improved too.
The irregular spellings for /e/ and /u/ obstruct progress worst of all, because they occur in many of the tricky words that children begin to meet soon after they first start school. Improving at least most of them would make a clearly noticeable difference to young children’s literacy progress. They also make strong candidates for reform because most of them were made irregular with deliberate changes to the original English spelling system.
The spelling of short /u/ was diluted chiefly in the 9th century, with the adoption of o next to n, m, v and w (son, mother, love, wonder). This was partly because the letter v (or Latin number 5) was back then used for four sounds (i.e. vp, vse, even, vvith). Spellings for short /e/ became irregular mainly in the 15th century, with the adoption of ea in 51 words (dread, head, thread) and for /ee/ in 156 words as well (dream, heal, treat). The reading difficulties which this change introduced are epitomised by ‘must read’ and ‘have read’. They could be much reduced by at least dropping the surplus a from the 51 words in which ea spells the short /e/ sound.
Adoption of regular spellings for the /ee/ sound would be of great help too, to both young and older learners. The current 12 different spellings for /ee/ (leave, sleeve, believe, these, weird, police, people…) are used in 459 words and pose one of the biggest hurdles in learning to write English. But they cause many reading difficulties as well, because nine of the 12 different spellings for /ee/ are used for other sounds too:
|Treat – great, threat, react; even – ever; ceiling – veil, eider; fiend – friend, died, diet; he – then; key – they; machine – define, engine; people – leopard, leotard; ski – hi.|
Reducing the 12 unpredictable spellings for the /ee/ sound to just ee, would reduce the current total of 4,219 unpredictably spelt common English words by more than a tenth and make learning to read much easier too. Even adopting ee for just the 156 words in which /ee/ is currently spelt with ea (lead, to read) would make a substantial difference. Completely regular spellings for /ee/ would have the additional benefit of reducing the 335 differently spelt, labour-intensive English homophones like ‘heel/heal’ by 47. They became standardised with the publication of Samuel Johnson’s dictionary in 1755, although they hardly ever serve any useful purpose. At least 2,500 other homophones get by perfectly well with just one spelling for their different meanings (bank, boot, trunk, act, play…). The US adoption of ‘practice’, for both noun and verb, has caused no problems either. The perverse English differentiation between ‘to practise’ and ‘a practice’ generates endless ‘misspellings’ (cf. to service, a service; to notice, a notice).
Johnson was also responsible for weakening the English method of distinguishing between short and long vowels, as in ‘made, madder’ or ‘hidden hideout’, by introducing hundreds of exceptions like ‘radish, shadow, study’. There is no good reason why his whims should continue to be obeyed and why regular consonant doubling – only after stressed short vowels in words of more than one syllable – should not become as acceptable as its current random usage. Children’s ‘errors’, like ‘annimal’, ‘fammily’ and ‘holliday’’, show that the original purpose of doubling is easy to grasp. It causes endless ‘mistakes’ only because it is used unsystematically. Adoption of systematic doubling would reduce the time currently needed for learning to read and write English very considerably. But unfortunately, it would also change the present look of hundreds of words. If adopting completely systematic consonant doubling should be considered to be too much change for one reform, it should at least be adopted when regularising irregular spellings of short /e/ and /u/. A few dozen of them have no doubled consonants after their irregular spellings (e.g. many, dozen – cf. penny, buzzer).
Some words with irregular /e/ and /u/ have surplus letters as well, along with omitted doubling (e.g. honey – cf. runny). They were added mainly in the 15th and 16th centuries, to help printers earn more money, as they were paid by the line. All they do now is to make learning to read more difficult. Surplus –e endings hinder reading progress because they undermine the use of -e for indicating long vowels, as in ‘home, alone’ (e.g. bone – gone, five – live), just like omitted consonant doubling. They should at least be scrapped when their short vowel spellings get regularised (e.g. done → dun, some →sum, honey → hunny).
There is much else wrong with English spelling, but some irregularities would be more difficult to regularise, because they have no clear best spelling (e.g. her, bird, turned). They also create predominantly only writing difficulties and are therefore less detrimental to overall literacy progress (e.g. ‘pardon, certain, urban, truncheon’).
The irregular spelling for /e/, /u/ and /ee/, inconsistent consonant doubling and surplus uses of –e are most responsible for English-speaking countries continuing to have high levels of functional illiteracy. Improving them would be a more certain and permanent way of speeding up literacy acquisition and raising overall educational attainment, than changes in teaching methods or increases in educational expenditure. Ideally, all five irregularities would get corrected. But even just making the uses of e, u and ee more regular would make a very substantial difference to learners.