Shawalphabet YahooGroup Archive Browser

A key to Kingsley Read's entry for the Shaw Alphabet competition may be
seen at
http://groups.yahoo.com/group/shawalphabet/photos/album/1025578948/pic/l\
ist

Philip complained:

> > Ha-ha and Hung have had their symbols swapped
> > ERR and AIR have had their symbols swapped

> I'd appreciate it if this old debate is not re-opened.


But the flaw remains...

On Wed, Nov 23, 2011 at 15:27, dshepx <david@...> wrote:
>
>
> Philip complained:
>
> > > Ha-ha and Hung have had their symbols swapped
> > > ERR and AIR have had their symbols swapped
>
> > I'd appreciate it if this old debate is not re-opened.
>
> But the flaw remains...

The situation remains. Calling it a "flaw" is begging the question.

For people who think it's a flaw, I would recommend they pick another
writing system and/or not complain or otherwise bring it up, because I
don't think a reform to the Shaw Alphabet would go over well with the
community of current users.

Cheers,
Philip
--
Philip Newton <philip.newton@...>

On 23/11/11 15:27, dshepx wrote:

>
>
> Philip complained:
>
>
> > > Ha-ha and Hung have had their symbols swapped
> > > ERR and AIR have had their symbols swapped
>
> > I'd appreciate it if this old debate is not re-opened.
>
>
> But the flaw remains...

I think the question is: “Is there really a flaw?”

Let’s see what the Shavian alphabet looks like.

First, let’s examine the 40 ”simple” letters.

There are 16 consonants organized in pairs: each voiceless consonant has
a voiced counterpart.
They are:
𐑐 /p/ — 𐑚 /b/
𐑑 /t/ — 𐑛 /d/
𐑒 /k/ — 𐑜 /ɡ/
𐑓 /f/ — 𐑝 /v/
𐑔 /θ/ — 𐑞 /ð/
𐑕 /s/ — 𐑟 /z/
𐑖 /ʃ/ — 𐑠 /ʒ/
𐑗 /tʃ/ — 𐑡 /dʒ/
For each pair, the voiced sound is represented by a deep letter, the
unvoiced sound, by a tall letter.

There are 16 vowels and diphtongs. They’re all written with small letters:

𐑦 /ɪ/ — 𐑰 /iː/
𐑧 /ɛ/ — 𐑱 /eɪ/
𐑨 /æ/ — 𐑲 /aɪ/
𐑩 /ə/ — 𐑳 /ʌ/
𐑪 /ɒ/ — 𐑴 /oʊ/
𐑫 /ʊ/ — 𐑵 /uː/
𐑬 /aʊ/ — 𐑶 /ɔɪ/
𐑭 /ɑː/ — 𐑷 /ɔː/

The remaining 8 letters are 2 semivowels and 6 consonants which can’t be
organized into voiced/voiceless pairs:
𐑘 /j/ and 𐑙 /ŋ/ are tall letters,
𐑢 /w/ and 𐑣 /h/ are deep letters,
𐑤 /l/, 𐑥 /m/, 𐑮 /ɹ/ and 𐑯 /n/ are small letters.
If the “voiced consonant = deep letter / voiceless consonant = tall
letter” rule was universal, 𐑘 and 𐑙 should be deep, 𐑣 should be tall,
𐑤, 𐑥, 𐑮 and 𐑯 should be deep. But they aren’t. That rule only
applies to letters which are members of a voiced/voiless pair where each
member of the pair shares the same point of articulation and the same
manner of articulation.
Therefore, I don’t consider the present situation (𐑙 = /ŋ/, 𐑣 = /h/)
as a flaw. It is perhaps a little misleading at the beginning but
nothing more.
Also, I’d add that the same glyph (𐑙) is used for /ŋ/ in the successor
of the Shavian alphabet: Qui(c)kscript
(http://en.wikipedia.org/wiki/File:Quickscript_alphabet_revised_names.png ,
letter 21).

Now, let’s look at the ligatures.
We can say for sure that:
𐑸 /ɑːʴ/ is 𐑭 + 𐑮,
𐑹 /ɔːʴ/ is 𐑷 + 𐑮,
𐑼 /əʴ/ is 𐑩 + 𐑮,
𐑽 /ɪəʴ/ is 𐑦 + 𐑩 + 𐑮,
𐑾 /ɪə/ is 𐑦 + 𐑩,
𐑿 /juː/ is 𐑘 + 𐑵.
But what’s the left part of 𐑺 /ɛəʴ/ and 𐑻 /ɜːʴ/ made of?
If 𐑺 = 𐑩𐑩𐑮 and 𐑻 = 𐑧𐑧𐑮, it’s clear there’s a flaw. But are we
sure this condition is verified? The drawing of 𐑩 and 𐑧 begins
vertically. That’s not true in the ligatures.

Jean-François Colson

> On Wed, Nov 23, 2011 at 15:27, dshep wrote:
> >
> > Philip complains:
> >
> > > > Ha-ha and Hung have had their symbols swapped
> > > > ERR and AIR have had their symbols swapped
> >
> > > I'd appreciate it if this old debate is not re-opened.
> >
> > But the flaw remains...

and Philip responded
> The situation remains.

Yes, it does, and shall be until remedied--if not by this group then
some later generation.
> Calling it a "flaw" is begging the question.
Petitio principii? I don't think so--what must I prove that isn't
self-evident? It is more difficult to insist upon the opposite--that
there is some, unclear what, but some reason nonetheless for the
existence of blatant exceptions to an otherwise rational structure. From
a pedagogical point of view, why erect a useful system of
arrangement--one that unlike other schemes of spelling and alphabet
reform actually offers a clear, visual identification of its functional
parts and then willingly contradict it? Mere eccentricity? Somewhere an
oversight occurred.
> For people who think it's a flaw, ...
I don't think it's a flaw; it is a flaw.
> I would recommend they pick another writing system and/or not >
complain or otherwise bring it up, because I don't think a reform > to
the Shaw Alphabet would go over well with the community of > current
users.
Oh dear! Mustn't create discord amongst the faithful.
dshep

On 11/23/11, Philip Newton wrote:
> On Wed, Nov 23, 2011 at 12:11, Michael Everson wrote:
>> On 23 Nov 2011, at 10:07, Philip Newton wrote:

>>> For that reason, I do not believe that _specific_ fonts are necessary.

>> Keyboard drivers are, though.

> True.

> Or at least, they're very useful -- you could hunt and peck together
> the characters from a "character map" application, or copy and paste
> them together from the Wikipedia article, or the like.

> But ideally, you'd want to just type and have Shavian come out, for
> which you'd need a keyboard driver that supports that.

Tonight I've written an Emacs input method (see at the end of the
message), and tried to test it by the letter names. It works in
Emacs 23.1, although I am not quite sure about my transcriptions.
I would appreciate if somebody checks my exercise in
the RP transcription and the Shavian tradition (especially about use
of Roar in vocalized syllables):
-----------------------------

𐑞 ·𐑖𐑱𐑝𐑾𐑯 𐑨𐑤𐑓𐑩𐑚𐑧𐑑

* 𐑑𐑷𐑤 𐑯 𐑛𐑰𐑐 𐑤𐑧𐑑𐑼𐑕:

𐑐𐑰𐑐 𐑚𐑦𐑚
𐑑𐑷𐑑 𐑛𐑧𐑛
𐑒𐑦𐑒 𐑜𐑨𐑜
𐑓𐑰 𐑝𐑬
𐑔𐑲 𐑞𐑱
𐑕𐑴 𐑟𐑫
𐑖𐑵𐑼 (?) 𐑥𐑧𐑠𐑼 (?)
𐑗𐑼𐑗 𐑡𐑳𐑡
𐑘𐑱 𐑢𐑴
𐑣𐑳𐑙 𐑣𐑭-𐑣𐑭

* 𐑖𐑹𐑑 𐑤𐑧𐑑𐑼𐑕:

𐑤𐑷𐑤 𐑮𐑷𐑼
𐑥𐑲𐑥 𐑯𐑳𐑯
𐑦𐑓 𐑰𐑑
𐑧𐑜 𐑱𐑛𐑡 𐑨𐑖 𐑲𐑕
𐑩𐑛𐑵 𐑳𐑐 𐑪𐑯
𐑴𐑒 𐑢𐑫𐑤 𐑫𐑟 𐑬𐑑 𐑶𐑤 𐑭 𐑷

* 𐑤𐑦𐑜𐑩𐑗𐑼𐑟 (𐑤𐑦𐑜𐑩𐑑𐑿𐑼𐑟 ??):

𐑸 𐑹 𐑺 𐑻 𐑼𐑱 𐑽 𐑾 𐑿

----------------------------

My Emacs input method:

(quail-define-package
"shavian-ucs" "utf-8" "𐑖8" t
"A SAMPA-like notation to input Shavian letters
in UTF-8 encoding.
" nil t nil nil nil nil nil nil nil nil t)

(quail-define-rules
("p" ?𐑐) ("b" ?𐑚)
("t" ?𐑑) ("d" ?𐑛)
("k" ?𐑒) ("g" ?𐑜)
("f" ?𐑓) ("v" ?𐑝)
("T" ?𐑔) ("D" ?𐑞)
("s" ?𐑕) ("z" ?𐑟)
("S" ?𐑖) ("Z" ?𐑠)
("ts" ?𐑗) ("j" ?𐑡)
("y" ?𐑘) ; [j], yea (SAMPA [j])
("w" ?𐑢)
("N" ?𐑙) ("h" ?𐑣)
;; --- Short letters:
("l" ?𐑤) ("r" ?𐑮)
("m" ?𐑥) ("n" ?𐑯)
("i" ?𐑦) ; [ɪ], if (SAMPA [I])
("ee" ?𐑰) ; [i:], eat (SAMPA [i:])
("E" ?𐑧) ; egg [ɛ]
("ei" ?𐑱) ; Age
("{" ?𐑨) ; [æ] Ash
("ai" ?𐑲) ; Ice
("@" ?𐑩) ; [ə]
("V" ?𐑳) ; Up [ʌ]
("q" ?𐑪) ; On [ɒ]
("ou" ?𐑴) ; OAk [oʊ]
("oo" ?𐑫) ; [ʊ], wOOl (SAMPA [U])
("u" ?𐑵) ; [u:], ooze
("au" ?𐑬) ; [aʊ], out
("oi" ?𐑶) ; [ɔɪ], oil
("A" ?𐑭) ; [ɑː], ah
("O" ?𐑷) ; [ɔː], awe

;; --- Ligatures et al.:
(";;" ?\;)
(";." ?·) ; naming dot
("Ar" ?𐑸) ; [ɑr] - are
("Or" ?𐑹) ; [ɔr], [ɔər] - or
(";E" ?𐑺) ; [ɛər] - air
(";3" ?𐑻) ; [ɜr] - err
("@r" ?𐑼) ; [ər] - ARRay
("i@r" ?𐑽) ; [ɪər] - ear
("i@" ?𐑾) ; [i.ə] - Ian
("yu" ?𐑿) ; [ju:] - yew
)

--
Sergio

Jean-François comments:
> On 23/11/11 15:27, dshepx wrote:
>
> > Philip complains:
>
> > > > Ha-ha and Hung have had their symbols swapped
> > > > ERR and AIR have had their symbols swapped
>
> > > I'd appreciate it if this old debate is not re-opened.
>
> > But the flaw remains...>> I think the question is: 'Is there really
a flaw'?>
> Let's see what the Shavian alphabet looks like.> ...
> There are 16 consonants organized in pairs: each voiceless consonant
> has a voiced counterpart.
>
> For each pair, the voiced sound is represented by a deep letter, the
> unvoiced sound, by a tall letter.

So far so good.

> The remaining 8 letters are 2 semivowels and 6 consonants which can't
> be organized into voiced/voiceless pairs:> ...> If the 'voiced
consonant = deep letter / voiceless consonant = tall letter'> rule was
universal, [y] and [ng] should be deep, h should be tall, l, m, r, > and
n should be deep. But they aren't.
First some simplification. The so-called short-level consonants (as
written in Shavian): l, m, n, and r, are sonorants--that is, in their
production they share more characteristics with vowels than with
consonants, which require closure or friction of the air stream as the
sound is made. It's just that we use them as consonants, so they are
considered consonants. Whether keeping them short was in fact a
deliberate design feature or came about fortuitously I wish I could say.
If necessary all of them could be given descenders and this can in fact
sometimes help reduce clutter and improve legibility when writing by
hand.
That leaves the tall letters for [y] and [ng]. Both [w] and [y] were
in Read's first draft short-level letters as would be appropriate but,
it appears to me, in order to obtain a more distinctive character for
the 'you' sound (as in Europe), the [y] character was lengthened in the
same direction it already pointed, which was upwards. and the [w] was
then extended downwards for symmetry's sake, or so I would think. Thus
the two semi-vowels, the ones most like vowels, are not short. They
could be left short however, and the [w] detached to pair with its
natural partner, voiceless [hw], a letter which also appeared in Read's
first draft, was lost, then reappeared again in Quickscript as well as
his later ReadSpel.
This sort of speculation infuriates some of the members of this
group, for whom Holy Writ exists in the form of 'Androcles' (though they
are ready to discard the schwa of unstressed final syllables, as in
buttn, so it isn't entirely Holy), but I'm too old to care about such
things and so shall continue. The [ng] could then be made deep
(ð`£) to contrast with a tall letter (ð`™) for [ngk] (thing
vs think), something some modern dictionaries have begun with.
In addition Read also included in his first draft a pair of
lettersthat correspond to the [x] of traditional orthography, as in
'expect' and 'exert', one voiced, one not--thus eliminating the one
instance where words spelled the traditional way are actually shorter
thanthe Shavian equivalant. Finally, an [h] could be borrowed from
Quickscript, ( h), tall as a voiceless letter should be, which could
pair up with the voiced [y]. That's no worse than the present situation,
and they do share at least one characteristic, neither sound can end a
word.
Note that I have introduced nothing that Read had not alsoconsidered
save the [ngk] borrowed from the Oxford Pocket Dictionary--the other
letters existed in one or other of Read's proposals. To judge from his
continued efforts and willingness to adjust and move things about as
needed it would seem that he was not overcome with horror at the notion
of change itself,but rather inspired I believe by a sustained interest
and never-ending quest to get it right.

As for 'err' and 'air'. Look at the combinations
> But what’s the left part of ð`º /ɛəʴ/ and
ð`» /ɜːʴ/ made of?> If ð`º = ð`©ð`©ð`®
and ð`» = ð`§ð`§ð`®, it’s clear there’s
a flaw. Exactly! Doesn't that seem more reasonable than the
reverse?
In anticipation of fulsome fan-mail,dshep

On Fri, Nov 25, 2011 at 02:36, Sergio Pokrovskij
<sergio.pokrovskij@gmail.com> wrote:
> I would appreciate if somebody checks my exercise in
> the RP transcription and the Shavian tradition (especially about use
> of Roar in vocalized syllables):
> -----------------------------
>
> 𐑞 ·𐑖𐑱𐑝𐑾𐑯 𐑨𐑤𐑓𐑩𐑚𐑧𐑑
>
> * 𐑑𐑷𐑤 𐑯 𐑛𐑰𐑐 𐑤𐑧𐑑𐑼𐑕:
>
> 𐑐𐑰𐑐 𐑚𐑦𐑚
> 𐑑𐑷𐑑 𐑛𐑧𐑛

𐑑𐑪𐑑, I would say; 𐑑𐑷𐑑 would be "taut" or "taught".

> 𐑒𐑦𐑒 𐑜𐑨𐑜
> 𐑓𐑰 𐑝𐑬
> 𐑔𐑲 𐑞𐑱
> 𐑕𐑴 𐑟𐑫

𐑟𐑵

> 𐑖𐑵𐑼 (?) 𐑥𐑧𐑠𐑼 (?)

I would say 𐑖𐑫𐑼. And I agree with your 𐑥𐑧𐑠𐑼.

> 𐑗𐑼𐑗 𐑡𐑳𐑡

𐑗𐑻𐑗 (since it's stressed).

> 𐑘𐑱 𐑢𐑴
> 𐑣𐑳𐑙 𐑣𐑭-𐑣𐑭
>
> * 𐑖𐑹𐑑 𐑤𐑧𐑑𐑼𐑕:
>
> 𐑤𐑷𐑤 𐑮𐑷𐑼

I would say 𐑤𐑪𐑤, and 𐑮𐑹.

> 𐑥𐑲𐑥 𐑯𐑳𐑯
> 𐑦𐑓 𐑰𐑑
> 𐑧𐑜 𐑱𐑛𐑡 𐑨𐑖 𐑲𐑕

Just 𐑱𐑡.

> 𐑩𐑛𐑵 𐑳𐑐 𐑪𐑯
> 𐑴𐑒 𐑢𐑫𐑤 𐑫𐑟 𐑬𐑑 𐑶𐑤 𐑭 𐑷

𐑵𐑟

> * 𐑤𐑦𐑜𐑩𐑗𐑼𐑟 (𐑤𐑦𐑜𐑩𐑑𐑿𐑼𐑟 ??):

Good question. Or even 𐑤𐑦𐑜𐑩𐑑𐑘𐑫𐑼𐑟.

> 𐑸 𐑹 𐑺 𐑻 𐑼𐑱 𐑽 𐑾 𐑿

𐑾𐑯

I agree with the rest of your spellings.

Cheers,
Philip
--
Philip Newton <philip.newton@gmail.com>

On Fri, Nov 25, 2011 at 01:34, dshepx <david@...> wrote:
>
> Oh dear! Mustn't create discord amongst the faithful.

In my opinion, the use of a shared standard is more valuable than
having that standard be perfect.

Also, having two systems which are very similar but for a few changes
is more confusing, in my opinion, than having two clearly different
systems.

Both of those would speak against reforming Shavian, I think. (Unless
you can get everyone on board and somehow completely discard the old
system.)

Cheers,
Philip
--
Philip Newton <philip.newton@...>

--- In shawalphabet@yahoogroups.com, Philip Newton <philip.newton@...> wrote:
>
> On Fri, Nov 25, 2011 at 02:36, Sergio Pokrovskij
> <sergio.pokrovskij@...> wrote:
> > I would appreciate if somebody checks my exercise in
> > the RP transcription and the Shavian tradition (especially about use
> > of Roar in vocalized syllables):
> > -----------------------------
> >
> > ð`ž ·ð`–ð`±ð`ð`¾ð`¯ ð`¨ð`¤ð`"ð`©ð`šð`§ð``
> >
> > * ð``ð`·ð`¤ ð`¯ ð`›ð`°ð` ð`¤ð`§ð``ð`¼ð`•:
> >
> > ð`ð`°ð` ð`šð`¦ð`š
> > ð``ð`·ð`` ð`›ð`§ð`›
>
> ð``ð`ªð``, I would say; ð``ð`·ð`` would be "taut" or "taught".
>
> > ð`'ð`¦ð`' ð`œð`¨ð`œ
> > ð`"ð`° ð`ð`¬
> > ð`"ð`² ð`žð`±
> > ð`•ð`´ ð`Ÿð`«
>
> ð`Ÿð`µ
>
> > ð`–ð`µð`¼ (?) ð`¥ð`§ð` ð`¼ (?)
>
> I would say ð`–ð`«ð`¼. And I agree with your ð`¥ð`§ð` ð`¼.
>
> > ð`—ð`¼ð`— ð`¡ð`³ð`¡
>
> ð`—ð`»ð`— (since it's stressed).
>
> > ð`˜ð`± ð`¢ð`´
> > ð`£ð`³ð`™ ð`£ð`­-ð`£ð`­
> >
> > * ð`–ð`¹ð`` ð`¤ð`§ð``ð`¼ð`•:
> >
> > ð`¤ð`·ð`¤ ð`®ð`·ð`¼
>
> I would say ð`¤ð`ªð`¤, and ð`®ð`¹.
>
> > ð`¥ð`²ð`¥ ð`¯ð`³ð`¯
> > ð`¦ð`" ð`°ð``
> > ð`§ð`œ ð`±ð`›ð`¡ ð`¨ð`– ð`²ð`•
>
> Just ð`±ð`¡.
>
> > ð`©ð`›ð`µ ð`³ð` ð`ªð`¯
> > ð`´ð`' ð`¢ð`«ð`¤ ð`«ð`Ÿ ð`¬ð`` ð`¶ð`¤ ð`­ ð`·
>
> ð`µð`Ÿ
>
> > * ð`¤ð`¦ð`œð`©ð`—ð`¼ð`Ÿ (ð`¤ð`¦ð`œð`©ð``ð`¿ð`¼ð`Ÿ ??):
>
> Good question. Or even ð`¤ð`¦ð`œð`©ð``ð`˜ð`«ð`¼ð`Ÿ.
>
> > ð`¸ ð`¹ ð`º ð`» ð`¼ð`± ð`½ ð`¾ ð`¿
>
> ð`¾ð`¯
>
> I agree with the rest of your spellings.
>
> Cheers,
> Philip
> --
> Philip Newton <philip.newton@...>
>

Or if the operating system you use is Windows or Linux, there are keyboard layouts available in the files section.