The REGULARITY of our written language

Please contribute past postings from the message board that you have have found particularly inspiring.

Moderators: maizie, Lesley Drake

Post Reply
User avatar
Vicki Lynch
Posts: 200
Joined: Fri Oct 31, 2003 12:18 am
Location: Kent

The REGULARITY of our written language

Post by Vicki Lynch » Sat May 06, 2006 11:29 pm

A personal favourite of mine, that i am always searching for, is as follows:

Dave Philpot - Posted: Sun Feb 22, 2004 10:21 pm
Post subject: Debbie H's response to DfES phonics & Greg Brooks report


On page 9 of the article it is noted that Quote:
Prof Brooks estimates that the spelling system of English is about 75% regular
. This is typical of the myths that exist about English spelling and underpin arguments about the necessity for whole-word approaches either on their own or in tandem with 'phonics'. These myths need to be taken head-on and demonstrated to be untrue. English is actually spelled in a remarkably accurate and consistent manner. For those who are interested I shall outline my arguments that lead to this conclusion.

In order to examine the regularity of English spelling it would make sense to start by developing a basis for examining texts and analysing the spelling therein. This I have done using the following ideas.

a) Include any grapheme that is used to represent the same sound in more than one word. (The grapheme uy is therefore included because it represents the phoneme / ie / in the two words buy and guy, but no others as far as I am aware. One off spellings such as ough representing the sound / oo / {moon} as it does in the word through are not included.)

b) Assume that, as spoken English contains no silences other than those used to draw breath or insert a dramatic pause, every letter in every word will either be a grapheme in its own right or else be part of a grapheme composed of two, three or four letters.

On this basis, regular English spelling contains about 135 to 140 graphemes and 210 to 220 correspondences. We only need to consider two situations, viz

1) Each sound may be represented by more than one symbol as the phoneme / n / can be shown by the five graphemes n nn ne kn and gn in the words no runner gone knot and gnaw.

2) Each symbol may represent more than one sound as the symbol a represents the seven vowels / a / / ae / / o / / ar / / or / / e / and the schwa in the words cat lazy was spa also any and Coca-Cola.

It is a fairly simple process to inspect texts underlining each sound to look for both regularities and irregularities. To demonstrate I shall now do this with the previous sentence.

I t....i s....a....f air l y....s i m p le....p r o c e ss....t o....i n s p e c t....t e x t s....u n d er l i n i ng....ea ch....s ou n d.... t o....l oo k....f or....b o th....r e g u l a r i t ie s....a n d....i rr e g u l a r i t ie s.

You can now count that this sentence comprises 86 sounds represented by 85 regular phonemes, ie this sentence is spelled with 100% regularity.(NB the letter x codes for two phonemes!)

I have analysed various chunks of text from novels, newspapers, academic research,etc, by taking a random point and reading the next 500 sounds. So far they all turn out to be between 99.6 and 99.8% regular, ie on average only one or two sounds per 500 are not spelled regularly.

This myth about English spelling being irregular seems to me to have come about for three reasons.

1) The incorporation of foreign words into English whilst retaining some or all of their foreign spelling, eg one, two and yacht. (Currently we can see that a new English correspondence is being created for the ai digraph by the use of far eastern words such as bonsai and Kawai where ai is used to show the phoneme / ie /.)

2) Chances in pronunciation over the years not being matched by changes in spelling, eg Wednesday.

3) Incorrect analysis of spelling in traditional phonics by introducing the nonsensical idea of silent letters for a language that contains no silences,
eg in the word know, the k is silent but the w isn't. Logically, either kn and ow are both digraphs, or else both k AND w are silent! This idea of silent letters is a graphemic one and has no place in any real phonic programme!

However, when you examine texts, although the above three points may refer to thousands of individual words, their prevelence in text is totally overwhelmed by the preponderance of words that are spelled in a regular fashion.

What needs to be understood here is that English spelling, although it is uneccessarily complex, is extremely regular. This in fact is why real phonic tuition is effective. For most pupils only about 175 actual phoneme-grapheme correspondences need to be taught. The rest will be assimilated, if appropriate, from reading text.

Posts: 1638
Joined: Sun Nov 20, 2005 7:56 pm

Post by dickschutz » Sun Oct 01, 2006 7:11 pm

That is a "classic" and classy explanation! Thanks, Vicki and Dave!

Another thing that is typically overlooked, is that you don't need to use all f the corpus of the language to teach children how to read. There is more than an ample number of words to chose from and you can pick whatever simple code and code combinaations in any sequence you like.

I think it's ironic that "computers" can read any text (True, they cheat) and can also spell-from-speech with much greater accuracy than can most kids. If we went about programming computers in the same way we go about developing and using programs with kids, a high percentage of computers would be "dyslexic" and "poor spellers."

Post Reply

Who is online

Users browsing this forum: No registered users and 2 guests