English as she is writ

A detour into spelling

Pronunciation and spelling

My rant about the non-word ‘brassic’ in the previous post has led me to consider further the issue of the extent to which spelling should be influenced by pronunciation. The answer is, of course, not at all. I know: I’ll have to explain my reasoning. And yes, we will get on to punctuation eventually – in fact the pronunciation/spelling question will lead naturally into the subject of punctuation.

There are in this world idiots who hold the view that spelling should be phonetic – that the way a word is spelt should reflect how it is spoken. This is, at first sight, a reasonable opinion: it would, it is suggested, make the spelling of words much easier to learn. English is a notoriously non-phonetic language; the proponents of phonetic spelling dislike the fact that the ‘ough’ formation of letters represents so many different spoken sounds. Consider rough, though, thought, bough, cough, thorough – in each case ‘ough’ is pronounced differently. Spelt phonetically the same words might appear in writing as ruff, thoe, thort, bow, coff and thuru.

Wikipedia – as usual – has an informative page on the subject of spelling reform: https://en.wikipedia.org/wiki/English-language_spelling_reform.      

There are a number of arguments against reforming English so that its spelling more accurately reflects speech. One is that the existing spellings often display clues as to the origin and therefore the meaning of words. Another is that in some cases the fact that two homophones have different spellings helps to indicate that they have different meanings (for instance night and knight).

However it seems to me that the most obvious objection to the principle that spelling should follow speech is that spoken English is as inconsistent as English spelling.

Take, for instance, the word ‘bath’. Having been brought up in the Home Counties in southern England I pronounce this word ‘barth’. In the north of England is has a short a: ‘bath’. In the southern states of the USA it is pronounced ‘bayth’. In South Africa and New Zealand, ‘beth’.

Even the shortest, simplest words have huge differences in pronunciation. I say ‘yes’. In North America it’s more like ‘yairs’, or in some areas ‘yus’ or ‘yurs’. In parts of Scotland it’s also something like ‘yairs’. In South Africa, Australia and New Zealand it’s ‘yis’ – in fact in New Zealand it’s sometimes even ‘ears’.

So the question for the phonetic reformers is: which pronunciation do you choose as the basis for spelling? Whichever variant you select – spelling based on pronunciation in Surrey, say, or in Cincinnati – is automatically wrong, in terms of phonetics, for speakers of all the other variants.

At the root of this apparent difficulty is a truth that the would-be reformers of spelling do not acknowledge: writing and speech are two different forms of communication.

I accept that both writing and speech employ words; also that speech evolved long before writing. However there is no evidence that writing was first conceived of as a way to record speech; on the contrary the earliest examples of writing are lists – of goods in storage, of battles won, of kings.

Until the invention of punctuation writing consisted of nothing but words and numerals. Speech, on the other hand, consists of words, pauses, non-verbal utterances (um, er), intonation and, when conducted face to face, gestures and facial expressions. In casual conversation between individuals who know each other well the words are the least significant element; the tone and performance of a political speech may well be more persuasive than the words it contains.

Speech is therefore, I contend, a method of communication that has very little to do with writing.

And writing has very little to do with speech. Writing is an upstart ability: while human speech is estimated to have arisen between 350,000 and 100,000 years ago, the earliest known writing occurred only about 5000 years ago.

Writing consists of nothing but words (and, relatively recently, punctuation). It can be read aloud, and there is evidence that in some historical eras it was usual to read aloud (in his Confessions, written in the fourth century AD, Augustine remarks on Ambrose‘s unusual habit of reading silently).

However for the past few centuries, at least, it has become usual to read silently, and this has emphasised the divergence between speech and writing: in order to understand writing one does not need to hear, even in the privacy of one’s own mind, the spoken sound of the words. Indeed in order to read quickly it is necessary for the words to travel from page to eye to brain without troubling the mouth, the ear, or any part of the brain devoted to registering sound.

Writing is, then, a system of symbols – letters – that are arranged into groups – words, sentences, paragraphs – that convey meaning when seen (or touched, as in Braille). It is a visual medium. Written text can be read aloud, but writing does not represent sound. Writing is independent of sound. Letters arranged into a word – let us say, for example the word ‘freedom’ – convey the meaning: lack of restraint, liberty, and so on. They do not prescribe how the word ‘freedom’ should be pronounced when spoken.

Therefore I conclude that there is no justification for spelling to be bound by pronunciation. Clearly there are advantages to there being a relationship between spelling and pronunciation: it would be difficult to learn to read if written words bore no resemblance to their spoken forms. But any attempt to force written text into a rigid phonetic scheme would be impractical – in fact, I think, doomed to failure – and more importantly based on a misapprehension of the nature of writing.

I must return briefly to the example of the word ‘boracic’ spelt as ‘brassic’. ‘Boracic’ is a real word: it’s the adjectival form of the word ‘borax’, in the same way that ‘sulphuric’ is derived from ‘sulphur’. It has meaning: it conveys in its spelling its connection to the substance borax. It contains other information: it suggests that usage of the word was common at a time in the past when borax was used as an antiseptic. And once one has learnt that in those days borax was applied in a dressing known as boracic lint, one can appreciate the verbal inventiveness of the Cockney rhymers who conceived of saying ‘boracic’ to mean skint. ‘Boracic’ is a cornucopia of meaning. ‘Brassic’ conveys none of this. It is rootless; meaningless. One might assume that it has some connection with the metal brass, but of course it hasn’t. The only route it might take you along, therefore, is a dead end. You can pronounce it however you like; but to have any meaning it must be spelt ‘boracic’. 

I should also emphasise that I do not subscribe to the belief that there such a thing as correct spelling, any more than there is correct punctuation or grammar. All of these are matters of convention. Spell ‘sulphur’ as ‘sulfur’ by all means, if that’s the convention in your part of the world; I don’t mind. My argument is that the way a word is spelt often provides information about its meaning, its derivation, and other words that are associated with it, and that is one of the reasons why it is a mistake to spell words as they are pronounced.

Having dealt with spelling, I will turn next to punctuation – which, you will not be surprised to learn, also has nothing to do with speech.  

The only way to deal with an unfree world is to become so free that your very existence is an act of rebellion.