a word marking system

Moderators: Debbie Hepplewhite, maizie, Lesley Drake, Susan Godsland

Post Reply

a word marking system

Post by guest » Sat Apr 03, 2004 1:31 am

Following is a letter that I wrote to Dr. Groff today.

My interest in the use of diacritics stems from my work on various
systems aimed at the respelling of English to make it easier for
children to learn to read. I have produced the SRS, SRS3, and SRS4g

The simplicity of the English alphabetic system makes it fairly
trivial to produce a respell system, but there are many choices to be
made. Perhaps the biggest choice is the selection of a way to
indicate that a vowel has the traditional 'long' vowel sound. I at
first used digraphs but later decided that here was a good place for
the use of an accent mark. I selected accent acute. It has been my
finding that just the indication of the long vowel sound renders
English much easier to read. n√°tion and fashion, simply and impl√Ω,
etc. In this matter there is long precedent. The macron was used
when I learned to read.

But there is a problem when extending the use of diacritics. If one
does not have the luxury of respelling the language (and I don't
think one realistically does, at least for the near future) then
using diacritics to indicate up to seven different sounds for one
letter takes a lot of diacritics. Furthermore the diacritics must
have different meanings with regards to different letters. This is
not practical, though I know it has been tried.

In early 1998 (as you may remember from our correspondence) I
conceived of a new way of showing how a letter should be pronounced.
This did not involve the traditional diacritics. I decided I could
just slightly alter the shape of the letters themselves to indicate
the sounds that they represented. I found that the worst case was
the letter 'o'. I had to be able to mark: code, tomb, done, wolf,
and cost. I also had to indicate whether the letter had assumed the
schwa sound, and whether it should be ignored in the pronunciation of
a word.

Unlike the use of diacritics, changing the shape of a letter is
equivalent to the generation of another distinctive graphic symbol.
All that is required is an alteration sufficient for a child to see,
yet one that does not obscure the original shape of the letter. I had
decided that I did not want to depend on any measure of the usual
respell procedures to indicate sound.

With modern word processing capability it is entirely practical to
alter the shape of a letter. I think it would be wise for the RBG3
group to take a look at a system of spelling which would avoid the
necessity of teaching dozens of ground rules and the memorization of
many hundreds of 'sight' words in order to obtain a sufficient set
of 'decodable' words to allow literature to be appropriately keyed to
a child's vocabulary. I would thus negate one of the supposed
advantages of the disastrous 'whole word' or 'whole language' system.

In learning to read and write I would not ask that the child write
marked text. The words he would copy would be marked however, at
least at first. In the later stages of the learning-to-read process
the facing page technique would introduce unmarked text.

It seems to me that a system which clearly shows the difference in
vowel pronunciation between: bomb, tomb, comb, done, bone, gone,
move, wolf, golf, starry, tarry, fuse, dues, fruit, guise, etc.
should have some place in the world of education with respect to the
teaching of reading.

My new converting dictionary is capable of supplying proper marking
to over 90% of the words in a typical children's story.

(The converter properly converted 89% of the words in the fairly
technical literature represented by the text in the paragraphs above.)

I am presently attempting to find a good set of thirty or forty
downloadable children's stories which I could mark using the
converting dictionary and supply as an appendage to my primer. Any
suggestions (or comments) would be welcome.

Bob Boden

Dave Philpot
Posts: 14
Joined: Wed Dec 03, 2003 8:46 pm

Post by Dave Philpot » Sun Apr 04, 2004 5:48 pm

It's sad to see so much hard, well intentioned work going in the wrong direction. If we want to make English as easy as possible to read we only need to invent a further 17 letter symbols and use each of the (now) 43 symbols to uniquely represent one English phoneme. Alternatively we could create 17 digraphs that each uniquely represent one sound and do the same thing. However, all those who have achieved good literacy and make a living from it are unlikely to agree to any such change. Children therefore have to master the system we have. Teaching things that later have to be unlearned is generally not a successful teaching approach. Furthermore what is being advocated by the use of diacritics is focusing on the graphemes and is therefore essentially not a phonic approach - although on the surface it might look as though it is. Those of us working with linguistic phonics are finding that spelling is a far less problematic issue than has traditionally been believed.


decodable text

Post by guest » Mon Apr 05, 2004 2:59 am

Hi Dave,

It is hard for me to see that I am not following a phonics approach when I specify a different vowel shape for the 'o' in 'code' than that which I specify for the 'o' in 'done'.

I don't believe you really understand what I have proposed. I'm sorry I was not able to make myself clear.

As far as unlearning is concerned Diane McGuiness feels that precise vowel sound indication becomes less important as proficiency is reached.

Bob Boden

Dave Philpot
Posts: 14
Joined: Wed Dec 03, 2003 8:46 pm

Post by Dave Philpot » Mon Apr 05, 2004 7:30 pm

Hi Bob.

What you're doing is quite straightforward, but complex given 19 different vowel sounds in English. If I remember rightly the Vietnamese use 7 diacritic marks to modify each of their consonants with one of the seven vowel sounds in the language - thereby creating a written language where each grapheme stands for a CV diphone. We have a language that uses 26 letters in a variety of 1,2,3 & 4 letter graphemes that are used to represent the 43+ sounds. There is enormous redundancy resulting in there being about 175 common graphemes and a further 40 to 50 less common ones. Your system will cause pupils to attend to the diacritic marks as well as the letters. For some pupils this will be a not insignificant increase in information to be learned beyond the 26 letter symbols. Assuming they master this system then at some point the diacritic symbols will need fading out. What is likely to happen then is that the children will end up trying to memorise words as sight vocabulary, ie a visual graphemic strategy rather than a phonic one. We are finding with linguistic phonics that pupils can master the alphabet first time round at a faster rate than their increase in knowledge of the spoken language. They learn the various common sounds that the symbols (like a) represent and practise the skills to help them 'normalise' the word if their initial decoding of the vowel sound is incorrect.The problems of grappling with reading initial texts is therefore largely being resolved quite early on, ie by about the middle of Y1. If your approach is successful then it will culminate in the same unlearning problems that bedevilled ITA.

Dave P.


reading stratigies

Post by guest » Tue Apr 06, 2004 12:59 am

Hi Dave,

I showed Steve Bett one of your messages. He said he would like to have your email address.

Dr. Bett is the manager of the Saundspell group which attempts to produce improved ways of teaching reading by appropriate respelling of English, either as a full respell or as an initial teaching alphabet.

His email address is: Bettstevet@aol.com

Bob Boden

Dave Philpot
Posts: 14
Joined: Wed Dec 03, 2003 8:46 pm

Post by Dave Philpot » Thu Apr 08, 2004 4:08 pm

Hi Bob!

The postmaster tells me that the address bettstevet@aol.com has permanent fatal errors. Steve can contact me via


Dave P


Steve Bett' email address

Post by guest » Fri Apr 09, 2004 3:16 am

Hi Dave,

Something is wrong with the email. I am also having trouble with Steve's address. I will try to send your address to Steve via Saundspel.


Bob Boden

Post Reply

Who is online

Users browsing this forum: No registered users and 67 guests