Shawalphabet YahooGroup Archive Browser

DShep and Paul,

Dyslexics that have trouble with p/b will probably have trouble with the
rotated Shavian characters.

Not all people labeled as dyslexic have the same difficulties with rotations,
inversions, and mirrored character forms.

Paul's generalization from one case, however, is probably correct. Mirrored
letterforms [e.g., d/b or Shaw so-zoo] may cause the most difficulty.

--Steve

Hi Shep

Let me answer you question on Dylexia.
I happen to be familar with someone with a mild case of Dylexia.
He says (Most people with Dylexia are male by the way) that the
format of the Voiced/Unvoiced pairs does not cause a problem for him.
rotating a letter 270 degrees and lowering creates a distinctly
different letter to his eyes.
Not to say that he doesn't have any problem with the Shavian
Letters. A lot of the Vowel pairs are mirror images of each other.
He especially has trouble with some of the short Letters that are
mirror images. (i.e. mime/nun Ash/Egg/Ado/On Age/Ice)
Insufficient redundancy for him.

dshep wrote:
> Pitman's insistence that each letter be recognizable in isolation
> implies that they weren't as yet at the time, and this might have
> provided the inspiration for spinning them round, the second
> member of each pair, that is. I've always thought this arrangement
> brilliant. Someone earlier however pointed out that this
> complicates matters for those suffering from dyslexia, but I
> wouldn't know.

DSHEP: This would be an argument against happy as hap-ee, wouldn't it?
Even if someone gave the final syllable a little extra emphasis, as
dictionaries are now recording, this final sound would still not be
as distinct as that in 'be'.

I think this is another issue unrelated to representing <for> as Shavian
/far/.
If you are representing segmented speech rather than continuous speech,
you can sidestep relative word stress within a sentence.

Should happy be reprsented as hapF or hapi?
This is your question.

The issue is... How do you represent schwi?

It can be handled either way, it is an orthographic convention since there
is no unigraphic way in Shavian to represent a short unstressed /F/ sound.

If you want to emphasize shortness hapi is the choice.
If you want to emphasize vowel quality hapF would be the choice.

I think PAD McCarthy's Androcles convention is to use hapi

--Steve
--- In shawalphabet@yahoogroups.com, stbetta@a... wrote:

> dshep,
>
> SB: If you are transcribing continuous speech you have to
> worry about sentence stress.
> If you are transcribing segmented overpronounced words,
> you don't.
>
> A phonemic transcription needs only to represent relative
> stress in multi-syllable words. Relative word stress can be
> ignored.
>
> SB: If you are transcribing continuous speech you have to
> worry about sentence stress. If you are transcribing
> segmented overpronounced words, you don't.



regards,
dshep

Paul,

You raise some good points regarding Spell Checkers.
Most spell checker's are based on look-up tables rather than phonetics.
I would be interested on finding one that was not.

On some spell checkers and dictionaries you will receive a list of possible
spellings.
If you spell the word by sound, these programs will be able to list the word
you want in the list of possibilities about 90% of the time.

If you key in happee or hapi, happy will be one of the guesses.
You can try this at www.m-w.com

I have suggested to Franklin, producers of the Spelling Ace, that they include
a sound-spelling spelling option. sapIna should generate sub poena as the
first choice.
Currently, it not recognized as a valid spelling. However, supeena probably
is.

--Steve
Hi Steve
A side point.
I noticed that all these alternate alphabets rendered on my screen
in Roman letters are vaguely recognisable as English, due to a
sensible Keyboard mapping of upper and lower case characters.

Teaching a child a phonetic spelling is easy for us (Shavians) to
interpret, but I was wondering if any of the commonly used spell-
checkers are phonetically based.

Generally speaking, humans are better at pattern matching than
machines. We occasionally even notice patterns outside of our
programming.

If an ordinary Spell-checker would consistently bring forth the
correct T.O. spelling, from phonetic text, then the question of
whether learning Shavian Spelling first, or Pitman's I.T.A. first
results in you being a poor T.O. speller becomes moot.
One of the Unifon converters will give you the correct traditional spelling
when you enter the correct Unifon sound-spelling.
Even if the i.t.a. did actually fail to assist in teaching many students how
to
spell English in the T.O.

I think spell-checkers are based on more on statistical rather than
Phonetic principles.

Anyone familar with them?

Regards, Paul V.

Here's my thought on this (from the silence of left field):

h0w 0f73n d0 y0u s33 num834s a5 pa47 0f a w0rd?

--Star

--- dshepx <dshep@...> wrote:

>
> --- In shawalphabet@yahoogroups.com, "paul vandenbrink" wrote:
> >
> > Hi Shep
>
> > As far as the numeric digits go, I am sure it would be easy to
> > find a recognisable form of the digit 0 and digit 1, that does
> > not in any way resemble the letters tot and Oak. For example
> > a zero (0) with a dot in the centre and a straight up slash or
> > bar for a one. I have to consider this a red herring.
> > Regards, Paul V.
>
> Could be. As you point out , a naught with a slash is sometimes used
> in serial numbers for example. In Germany, and perhaps other European
>
> countries, a one is written (at least in handwriting) as a two-stroke
> figure with a forward-leaning diagonal attached to the head of the
> vertical stroke. So, an alternative already exists for the numerals,
> and it really doesn't matter which of the two, letters or numerals,
> defers to the other to avoid confusion. It is only the widespread use
>
> of the binary code, which will almost certainly not be changed, and
> the use of optical text-readers, which argues that it should be the
> letters instead that are altered, especially if one is introducing
> and
> encouraging the use of an entirely new alphabet upon the world.
>
> regards,
> dshep
>
>
>
>
>


====http://www.livejournal.com/users/wodentoad

Numfar! Do the Dance of Joy!



__________________________________
Do you Yahoo!?
The all-new My Yahoo! - Get yours free!
http://my.yahoo.com

Hi Steve

On a spell checker that gives a range of possible matches, one that
is phonetically based would not be out of place.

Because the Shavian keyboard retains some correspondences, and the
spelling checker doesn't differentiate between upper and lower case,
the spell checker does work surprising well on Shavian text. This
relationship could be enhanced by setting up a keyboard mapping that
enhances the correspondences between the Shavian Alphabet and the
Roman Alphabet.

You would have the primary and alternate pronunciation of each Roman
letter, mapped to the lower and upper keys for typing the Shaw
characters.
For example, instead of using Key c for the Shaw ch sound, have key
c map to Shaw Letter "K" and Capital or upper case "C" map to the
Shaw Letter for "ch". You would end up with a few duplication, as
both the "c" key and the "k" key would map to the same Shaw letter,
but you would just pick one of the keys to type with.
Both variants of the "n" sound work that way already.
Both variants of the "t" sound work that way already.

Also, this would enhance the readability of the raw Shaw letters,
if you didn't have a Shavian Font installed.

Regards, Paul V.

P.S. Ideally, you would want a real Shavian to T.O. conversion
program. The suggestion discussed above would just be a workaround.
I don't know if we need a T.O. conversion program at this point.
Anyway, it is not as urgent.

--- In shawalphabet@yahoogroups.com, stbetta@a... wrote:
> Paul,
>
> You raise some good points regarding Spell Checkers.
> Most spell checker's are based on look-up tables rather than
phonetics.
> I would be interested on finding one that was not.
>
> On some spell checkers and dictionaries you will receive a list of
possible
> spellings.
> If you spell the word by sound, these programs will be able to
list the word
> you want in the list of possibilities about 90% of the time.
>
> If you key in happee or hapi, happy will be one of the guesses.
> You can try this at www.m-w.com
>
> I have suggested to Franklin, producers of the Spelling Ace, that
they include
> a sound-spelling spelling option. sapIna should generate sub
poena as the
> first choice.
> Currently, it not recognized as a valid spelling. However,
supeena probably
> is.
>

Can the keyboard mapping of Shavian be improved?

Paul,

I too think that keyboard Shavian could be improved.
However, this would be part of an improved Shavian which would be called Shaw
inspired but not Shavian II. I do not wish to do anything with classic
Shaivan or usurp the name.

Hugh is right, Androcles Shavian and the keyboard norm have been adopted by
100's of users and a majority do not want change any more than those who have
learned traditional English spelling want to change to a more consistent and
rational spelling.

I doubt that keyboard Shavian can be "understood" by spell checkers.
The arbitrary assignments such as F to /aI/ are not currently interpreted.
Such conventions could be added to the program but I think your idea of
trying to optimize the recognizability of keyboard letter strings should be a
first step.

Here are some attempts at making keyboard Unifon more recognizable. --Steve

Can the keyboard mapping of Unifon be improved?
U2 and ENgliS are attempts to make words written in keyboard Unifon more
recognizable.
U2 keeps the schwa-c, ENgliS uses a schwa-a
Neither U2 and ENgliS make diphthongs unigraphic. This makes Qq available
for reassignment.

Purpose: wot wos Dc purpcs ov cnuDcr skEm simclcr tU Ynifon?
The purpose of U2 was to see if one could use the keyboard transcription more
readable.
Could new assignments make the unigraphic notation more readable?
A related effort to come up with an ASCII system that did not have to be
converted was called ENgliS. This notation used a schwa-a instead of a schwa-c.

Differences between Unifon and U2:
Unlike ENgliS, U2 keeps the schwa-c

The purpose of U2 was to improve recognition of some transcribed words.
hQt dogz qr gvd instead of Classic Unifon: hot dxgz or gCd.
Traditional SpellingUnifon
1950 ClassicUnifon
2000 Rev.U2
Unifon 2EngliS
Unifon 3IanspelWebster
Latin 1Truespel
1998
orOr xrOr xrorororôror
awed costxd kxstxd kxstod kostod kostod kostôd kôstaud kaust
all arexl orxl orol qrol qrol aarâl äraul aar
herderhcrdcrhurdcrhurdcrhRdcrh3da'hørdørherder
very niceverE nIsverE nIsvery nIsveri nYsveri naisveri nýsveree nies
alms spaomz spoomz spoqmz spqcmz spcaamzämz späaamz spaa
odd spotod spotod spotQd spQtQd spQtaadäd spätaad spaat
add attackad ctakad ctakad ctakqd atqkcd atckad øtakad uttak
loyal boillQl bQllQl bQlloiycl loyal boilloilloil boilloiyul boil
owl outql qtql qtowl owtowl owtaul autaul autoul out

www.foolswisdom.com/~sbett/u2-unifon.htm
Hi Steve

On a spell checker that gives a range of possible matches, one that
is phonetically based would not be out of place.
I am not sure I get your point. A phonetic spell checker would be a good
idea.
However, I do not know of any.
Because the Shavian keyboard retains some correspondences, and the
spelling checker doesn't differentiate between upper and lower case,
the spell checker does work surprising well on Shavian text.
If you checked www.m-w.com, you will notice that it does not recognize many
words written in its notation. It recognizes sed for said but not pAl for
pail.

Keyboard Shavian starts with mostly Continental or Latin based sound
assignments
where the assignments are Unigraphic. It does not always start with the most
frequent
English spelling of the sounds. nFs for nice and nEl for nail would not be
recognized. However, nIs and nAl might be.

This relationship could be enhanced by setting up a keyboard mapping that
enhances the correspondences between the Shavian Alphabet and the
Roman Alphabet.
SB: Which Roman Alphabet the shifted English one or the Continental Latin
based one?
You would have the primary and alternate pronunciation of each Roman
letter, mapped to the lower and upper keys for typing the Shaw
characters.
For example, instead of using Key c for the Shaw ch sound, have key
c map to Shaw Letter "K" and Capital or upper case "C" map to the
Shaw Letter for "ch".
SB: Did you check this out on any spell checker or dictionary?
Do they recognize cip or Cip for chip?
You would end up with a few duplication, as
both the "c" key and the "k" key would map to the same Shaw letter,
but you would just pick one of the keys to type with.
Both variants of the "n" sound work that way already.
Both variants of the "t" sound work that way already.

Also, this would enhance the readability of the raw Shaw letters,
if you didn't have a Shavian Font installed.
SB: Are the raw Shaw letters are what I have called KEYBOARD Shavian?
(with DISPLAY shavian used to refer to the actual Shavian characters).

Hi Star
You are right. The only time I see numbers mixed with letters, is
when you are counting. 1st, 2d, 3d, 4th 5th, 5th 7,th 8th, 9th,
10th, 11th, 12th, 13th, ..., 21st, 22d, 23d, 24th, etc.
And typically, you write out numbers up to ten or twenty in T.O.
That's just a convention to avoid in confusion in sentences like
"I had 1 of my brothers over." or "I'd like to see you 2." (You dual)
For large numbers, you do use Digits in T.O.
For Shavian, I would expect because the aim is to minimize the
number of letters and other characters we write, I would always use
Digits. In T.O. they use the # sign or the Abbreviation "no." in
front of an arbitrary sring of numbers used or a name or address.
I don't know what conventions we should have for Shavian?

Regards, Paul V.


--- In shawalphabet@yahoogroups.com, Star Raven
<celestraof12worlds@y...> wrote:
> Here's my thought on this (from the silence of left field):
>
> h0w 0f73n d0 y0u s33 num834s a5 pa47 0f a w0rd?
>
> --Star
>
> --- dshepx <dshep@g...> wrote:
>
> >
> > --- In shawalphabet@yahoogroups.com, "paul vandenbrink" wrote:
> > >
> > > Hi Shep
> >
> > > As far as the numeric digits go, I am sure it would be
easy to
> > > find a recognisable form of the digit 0 and digit 1, that
does
> > > not in any way resemble the letters tot and Oak. For
example
> > > a zero (0) with a dot in the centre and a straight up
slash or
> > > bar for a one. I have to consider this a red herring.
> > > Regards, Paul V.
> >
> > Could be. As you point out , a naught with a slash is sometimes
used
> > in serial numbers for example. In Germany, and perhaps other
European
> >
> > countries, a one is written (at least in handwriting) as a two-
stroke
> > figure with a forward-leaning diagonal attached to the head of
the
> > vertical stroke. So, an alternative already exists for the
numerals,
> > and it really doesn't matter which of the two, letters or
numerals,
> > defers to the other to avoid confusion. It is only the
widespread use
> >
> > of the binary code, which will almost certainly not be changed,
and
> > the use of optical text-readers, which argues that it should be
the
> > letters instead that are altered, especially if one is
introducing
> > and
> > encouraging the use of an entirely new alphabet upon the world.
> >
> > regards,
> > dshep
> >
> >
> >
> >
> >
>
>
> ====> http://www.livejournal.com/users/wodentoad
>
> Numfar! Do the Dance of Joy!
>
>
>
> __________________________________
> Do you Yahoo!?
> The all-new My Yahoo! - Get yours free!
> http://my.yahoo.com

Numbers for sound-signs
Number shapes that resemble letter shapes.

Star, Dshep, Paul, and others,

Comments inserted below. --Steve
Star wrote:

h0w 0f73n d0 y0u s33 num834s a5 pa47 0f a w0rd?

SB: There can be confusion particularly with alpha-numeric serial numbers.
I do not have any problem with the similarity of letterforms and numberforms
in traditional text. Was this your question?

SB: Numbers have been used to augment the sound-signs.
For instance: 5 for Sh, ingli5 5ip, 2 for zh, iluu2n
8 for theta, thý 8ý. 8 for B does not augment the alphabet.

Numbers have also been used in SMS texting: 4 sale. Go 2 the stor. g8 nwz.
This works in English but not in Spanish where the logograms have a different
pronunciation. e.g., quatro-saaley for 4-sale.

See Ianspel for an augmented alphabet that uses numbers.
www.foolswisdom.com/~sbett/u2-unifon.htm
> > PV: As far as the numeric digits go, I am sure it would be easy to
> > find a recognisable form of the digit 0 and digit 1, that does
> > not in any way resemble the letters tot <1> and Oak <O>. For example
> > a zero (0) with a dot in the centre and a straight up slash or
> > bar for a one. I have to consider this a red herring.

> DSHEP: Could be. As you point out , a naught with a slash is sometimes
used
> in serial numbers for example. In Germany, and perhaps other European
> countries, a one is written (at least in handwriting) as a two-stroke
> figure with a forward-leaning diagonal attached to the head of the
> vertical stroke. So, an alternative already exists for the numerals,
> and it really doesn't matter which of the two, letters or numerals,
> defers to the other to avoid confusion. It is only the widespread use
> > of the binary code, which will almost certainly not be changed, and
> the use of optical text-readers, which argues that it should be the
> letters instead that are altered, especially if one is introducing
> and encouraging the use of an entirely new alphabet upon the world.

Hugh,

You have recommended that we use Andagii as the standard in this discussion
group.
However, the link you provided no longer works.

I think we had a poll on member preferences for a group standard.
Lionspaw is no longer my favorite because it is not compact enough.
I prefer a condensed Shavian font. Recommendations?
Hugh B. wrote: Yes you’re able to do it now [use Unicode for Shavian].
Attached to this message is the latest version of my keyboard layout for Windows XP
that allows you to type in Shavian Unicode.
Here’s how to install it:
Before starting, make sure you have a font with Shavian characters at the
correct Unicode points installed – “Andagii” (www.alanwood.net/unicode/) or Ethan
’s “Shavian ESL Gothic Unicode”. [URL?]

And the so called l33t-speak would take on a whole new form!

Very well thought out, Paul. I have no aversion to either spelling out
letters eight becomes age+tot, seven s+egg+v+n or using the pound sign
(#) as a distinctive markers.

--Star

--- paul vandenbrink <pvandenbrink@...> wrote:

>
> Hi Star
> You are right. The only time I see numbers mixed with letters, is
> when you are counting. 1st, 2d, 3d, 4th 5th, 5th 7,th 8th, 9th,
> 10th, 11th, 12th, 13th, ..., 21st, 22d, 23d, 24th, etc.
> And typically, you write out numbers up to ten or twenty in T.O.
> That's just a convention to avoid in confusion in sentences like
> "I had 1 of my brothers over." or "I'd like to see you 2." (You dual)
> For large numbers, you do use Digits in T.O.
> For Shavian, I would expect because the aim is to minimize the
> number of letters and other characters we write, I would always use
> Digits. In T.O. they use the # sign or the Abbreviation "no." in
> front of an arbitrary sring of numbers used or a name or address.
> I don't know what conventions we should have for Shavian?
>
> Regards, Paul V.
>
>
> --- In shawalphabet@yahoogroups.com, Star Raven
> <celestraof12worlds@y...> wrote:
> > Here's my thought on this (from the silence of left field):
> >
> > h0w 0f73n d0 y0u s33 num834s a5 pa47 0f a w0rd?
> >
> > --Star
> >
> > --- dshepx <dshep@g...> wrote:
> >
> > >
> > > --- In shawalphabet@yahoogroups.com, "paul vandenbrink" wrote:
> > > >
> > > > Hi Shep
> > >
> > > > As far as the numeric digits go, I am sure it would be
> easy to
> > > > find a recognisable form of the digit 0 and digit 1, that
> does
> > > > not in any way resemble the letters tot and Oak. For
> example
> > > > a zero (0) with a dot in the centre and a straight up
> slash or
> > > > bar for a one. I have to consider this a red herring.
> > > > Regards, Paul V.
> > >
> > > Could be. As you point out , a naught with a slash is sometimes
> used
> > > in serial numbers for example. In Germany, and perhaps other
> European
> > >
> > > countries, a one is written (at least in handwriting) as a two-
> stroke
> > > figure with a forward-leaning diagonal attached to the head of
> the
> > > vertical stroke. So, an alternative already exists for the
> numerals,
> > > and it really doesn't matter which of the two, letters or
> numerals,
> > > defers to the other to avoid confusion. It is only the
> widespread use
> > >
> > > of the binary code, which will almost certainly not be changed,
> and
> > > the use of optical text-readers, which argues that it should be
> the
> > > letters instead that are altered, especially if one is
> introducing
> > > and
> > > encouraging the use of an entirely new alphabet upon the world.
> > >
> > > regards,
> > > dshep
> > >
> > >
> > >
> > >
> > >
> >
> >
> > ====> > http://www.livejournal.com/users/wodentoad
> >
> > Numfar! Do the Dance of Joy!
> >
> >
> >
> > __________________________________
> > Do you Yahoo!?
> > The all-new My Yahoo! - Get yours free!
> > http://my.yahoo.com
>
>
>
>


====http://www.livejournal.com/users/wodentoad

Numfar! Do the Dance of Joy!



__________________________________
Do you Yahoo!?
Send a seasonal email greeting and help others. Do good.
http://celebrity.mail.yahoo.com