Some forty per cent of modern English words are actually foreign imports; loan words. Some Anglo-Saxon words were displaced, but many remained. French and Latin words entered the language, and do so now, as does Greek in particular, as new technologies and new ideas crave linguistic expression. We have far greater choice of words today as a result. ‘Bit’, for example, is from ‘bita’ (old English), ‘part’ is from ‘pars’ (Latin) and ‘piece’ is from ‘piece’ (French). And we go on borrowing: ‘yoghurt’ is Turkish, ‘taxi’ is French, ‘potato’ is Spanish, ‘ski’ is Norwegian, ‘pyjamas’ is Urdu. Sometimes we invent new words altogether: ‘Diesel’ was a German engineer, ‘petrol’ (petroleum) is a hybrid of Latin & Greek: petra = rock (Latin), oleum = oil (Greek); and so is ‘television’: tele = far off (Greek) and videre = to see (Latin).
Our characteristic posture vis-a-vis English spelling tends to be exasperation, or even rage and despair. Considering derivations and families, connections and relationships, semantic or grammatical roles can go some way towards alleviating such hypertensive and counter-productive emotions. It may also help to fix spellings in the mind and language in the heart. English is a language of quirky personality and much pleasure is to be had from learning a little of its ways. I always have a nice dictionary incorporating etymology to hand and use it whenever a student wants to spell, say, ‘catastrophe’, ‘charisma’, ‘fascinate’, ‘malignancy’, ‘centigrade’ or ‘vinaigrette’. (And see an interesting consideration of the usefulness of English language history in the classroom by Invernezzi & Hayes 2004.)
A whirlwind spin through some of the principles of English spelling follows, remembering that we think morpheme rather than letter.
Most spellings are phonically regular, or very close. All spellings give some information about pronunciation.
Those spellings which are phonically irregular are either idiosyncratic, with a one-off, often historical, explanation or are being spelled according to a different beat altogether - for example a syntactic or semantic beat.
For example ‘catastrophes’ is not spelled ‘cattastrofiz’ for two reasons - the ‘ph’ is idiosyncratic and historical (the word is Greek-derived) and the ending is ‘es’ because it is simply the plural ‘s’ ending attached to catastrophe. All regular plurals, and most are regular, are s endings. This is all quite understandable. Likewise, the word ‘many’ is, apparently oddly, spelled with an ‘a’ but this is because it derives from Norse and Old English ‘manig’ (and where English has -y endings Scandanavian languages (the Vikings spoke Old Norse) have -ig). This is also an idiosyncratic and historical explanation, but it helps - particularly if you go on to look up ‘any’ and find that it derives from Old English ‘anig’. Before the ‘great vowel shift’ they were pronounced differently, of course. That ‘a’ would have been more phonically regular, then, not that anyone was much concerned at the time.
Words are often spelled in such a way as to indicate their grammatical role even, if necessary, at the expense of exact letter-sound correspondence.
The past tense marker ‘-ed’ is a good example. Batted, bowled and watched end in the sounds ‘id’, ‘d’ and ‘t’. However, since they are all the same grammatical construct, similarly affecting their attached verbs, carrying the same meaning, they are spelled the same in order immediately to convey this extremely important information to the reader (who is presumed already to know how he will say the words, and does). The fluent reader instantly understands that a past tense is being represented; the speller has to learn, for the reader’s sake, to produce it reliably. The reader can tell at a glance that ‘feted’ is the past tense of something while ‘fetid’ is not.
Words which seem phonically irregular in one form are very often phonically regular in another.