[YG Conlang Archives] > [westasianconlangs group] > messages [Date Index] [Thread Index] >
> Anyway, Arabic script is the only official script for Ajami.
> To use smth else produces the same effect
> as e.g. Russian written in katakana.
Katakana and Cyrillic are entirelly different scripts,
but Arabic, Hebrew, Greek, Latin, Glagolic and Cyrillic
are sister or cousin descendants of common ancestor script.
So I imagine another branch of this family of scripts,
in which the shapes of the letters
would changed as in Greek and Latin,
but without the change of their phonetic values.
Thus for example <H> is read still as [X\],
whereas [h] is written as <E>.
Waw has two forms, <F> and <Y>;
Greeks left <F> for consonant [w]
and the variant <Y> added to the end of the alphabet
for the vowel [u].
In my Gedeh script,
the variant <Y> is used for both [w] and [u],
whereas the variant <F> is used in digraphs for labiovelars
in the languages and dialects which have them:
<KF> [k_w], <QF> [k_?_w] (PIE gW), <CF> [g_w] (PIE gWH).
There are five semivowels in the wide sense;
their basic phonetic value is consonantal:
<A> [?], <E> [h], <Y> [w], <I> [j], <O> [?\]
As vowels, these letters are marked with accents:
Short unstressed vowels (low tone)
are marked with grave accent `
<À> [e_L], <È> [a_L], <Y`> [u_L], <Ì> [i_L], <Ò> [o_L]
Long unstressed vowels (low tone)
are marked with double grave accent ``
or with its shortened form, double dot above ¨ (dieresis, umlaut)
<Ä> [e:_L], <Ë> [a:_L], <ÿ> [u:_L], <Ï> [i:_L], <Ö> [o:_L]
short stressed vowels (high tone):
<Á> [e_H], <É> [a_H], <Ý> [u_H], <Í> [i_H], <Ó> [o_H]
Long stressed vowels (falling tone):
<Â> [e:_F], <Ê> [a:_F], <Y^> [u:_F], <Î> [i:_F], <Ô> [o:_F]
Long stressed vowels (rising tone):
<À´> [e:_R], <È´> [a:_R], <Y`´> [u:_R], <Ì´> [i:_R], <Ò´> [o:_R]
Grave accent above (or after) a stop consonant
changes it to a fricative:
B [b] B` [v]
C [g] C` [G]
D [d] D` [D]
Þ [t_?] Þ` [T_?]
K [k] K` [x]
P [p] P` [f]
T [b] T` [b]
Circumflex accent above a fricative marks lateralization
(originally it was lambda in upper index)
S [s] S^ [K]
U [s_?] Û [K_?]
Unfortunately, the Greek alphabet lacks cade.
There were some old alphabets retaining cade,
but it has the form of our <M>.
So I think that in the Gedeh alphabet,
cade originally looked like inverted <N>, |/|
(Cyrillic <i> can be used in Unicode documents)
and then it lost another leg and was further simplified to <V>
and finally the shape was rounded to <U>.
(Compare with Cyrillic <i> which is in handwriting
also equal to Latin <u>).
The Latin alphabet lacks also samek (east Greek ksi).
So let us suppose that in the Gedeh alphabet,
the samek was replaced by sigma (which developed later to <S>),
whereas at the place between <R> and <T>
was reintroduced the old shape of shin,
which was later sharpened to the shape similar to Latin <W>.
A B C D E Y Z H Þ I K L M N S O P U Q R W T F
Of course, although this system is logical,
it is nevertheless very weird and difficult to read.
So I suppose that the modern Gedeh texts are written
with some reformed orthography,
probably something like this:
E B G D A U Z X Þ I K L M N S O P C Q R S` T W
> I need to make an unofficial but standard romanization
> for several reasons:
>
> 1) very few people have reliable Arabic support on their computers:
> Unicoded/non-Unicoded, with vocalizations, additional Farsi, Urdu et
> sim. letters etc.; thus, to produce readable emails, I need a
> translit/transcript scheme;
>
> 2) I'm not fluent in reading Arabics,
> so for my convenience I need a kind of transcription aid
> for better remembering words, quicker search in a word list,
> quicker estimation of this or that grammar phenomenon,
For now I am using this transliteration of Arabic letters:
1569 hamza [?] '
1570 alef with madda above [] |~
1571 alef with hamza above [?] |'
1572 waw with hamza above [?] w'
1573 alef with hamza below [?] |,
1574 yeh with hamza above [?] J'
1575 alef [] |
1576 beh [b] b
1577 teh marbuta [t] +
1578 teh [t] t
1579 theh [T] t` (or þ)
1580 jeem [g] [dZ] [Z] g
1581 hah [X\] c
1582 khah [x] c` (or k` or x)
1583 dal [d] d
1584 thal [D] d` (or ð)
1585 reh [r] r
1586 zain [z] z
1587 seen [s] s
1588 sheen [S] < [K] s` < s^
1588 sad [s_X\] s¸ (or c)
1590 dad [d_?\] < [K_?] d¸ < s¸^ (or c^)
1591 tah [t_X\] t¸
1592 zah [z_?\] < [T_?] z¸ < t`¸ (or þ¸)
1593 ain [?\] 6
1594 ghain [G] 6` (or g`)
1601 feh [f] < [p] f (or p`)
1602 qaf [k_?] q
1603 kaf [k] k
1604 lam [l] l
1605 meem [m] m
1606 noon [n] n
1607 heh [h] h
1608 noon [w] w
1609 alef maksura [J]
1610 yeh [j]
1611 fathatan [a:] AN
1612 dammatan [u:] UN
1613 kasratan [in] IN
1614 fatha [a] a
1615 damma [u] u
1616 kasra [i] i
1617 shadda [] :
1618 sukun [] °
I am not sure yet how transliterate
1649 alef wasla
> - being simple and email-friendly.
I think that today we may freely use
not only the characters from us-ascii,
but also from iso-8859-1.
There are mainly accented vowels there,
but the thorn þ, edh ð
and the spacing marks ¨°´¸
could be useful.
> I think I came closely to a neat solution,
> but the situation is aggravated by the fact
> I'm not sure about precise phonetic value of
> some letters in Arabic borrowings.
> Mostly I stick to the pattern
> I find in Sefaradi pronunciation of Hebrew.
> So, I make both "kaf" and "qaf" to sound as [k],
> "s.ad" becomes [ts)] etc.
> But in this case "thaa" would turn into [t],
> and, by analogy, "dhaal" into [d],
> and I'm not quite happy with this,
> especially since both "taa" and "t.aa" already sound as [t].
> I'm looking for alternatives
> (as e.g. in Farsi, that made "thaa" > [s], "dhaal" > [z],
> but how shall I justify it?)
> or better arguements in favor of the [t]&[d] variant.
IMHO both possibilities,
[T] > [t], [D] > [d]
(as in Aram.)
and
[T] > [s], [D] > [z]
(as in Akk., Eth. and Hebr.)
are plausible.
Or maybe you could leave [T], [D]
as separate phonemes without merging
with [t], [d] nor with [s], [z]?
Or maybe let them separate only in written language.
> My current version (consonants) is this:
> Hamza - in most cases silent; [?] between a cons. and a vowel,
> rendered as _7_ in Latin Translit Scheme (LTS);
> alif - vowel or Hamza carrier
> baa - [b] word initial, [v] inside
> paa (baa with three dots below) - [p]
In Unicode, it is
&1662; peh (Persian, Urdu, ...)
Gedeh texts written with Arabic script use for [p]
&1697; 06A1 dotless feh
I do not know what phonetic value it had *here* in Adighe,
but I chose it by analogy with following pairs,
where dot above changes stop consonant to fricatives
1578 teh [t] - 1579 theh [T] t`
1583 dal [d] - 1584 thal [D] d`
1591 tah [t_X\] - 1592 zah [z_?\] < [T_?]
Similarly, [v] should be written
as beh with one dot below and one dot above,
but unfortunately there is no such combination *here*.
Instead of 1582 khah,
[x] could be written in some Gedeh texts as
&1708; kaf with dot above;
I do not know what phonetic value it had *here*
in old Malay and Pashto.
Unfortunately there is *here* no
jeem with one dot below and one dot above
as an allograph for 1594 ghain [G].
> taa - [t]
> thaa - [s]
> jiim - [dZ)], LTS _dj_
> chiim (jiim with three dots below) - [tS)], LTS _tx_
Unicode &1670; tcheh (Persian, Urdu, ...)
The pairs
&1576; beh [b] - &1662; peh [p]
&1580; jeem [dZ)] - &1670; tcheh [tS)]
sugest that three dots below instead of one
means "voiceless".
> Haa and xaa - [X], LTS _hh_
> daal - [d]
> dhaal - [z], occasionaly [d] or silent (in Ibero-Romance words)
> raa - [r] word initial, [4] inside; both are _r_ in LTS
> zain - [z]
> jain (zain with three dots above) - [Z], LTS _j_
Unicode &1688; jeh (Persian, Urdu, ...)
> siin - [s]
> shiin - [S], LTS _x_
> s.ad - [ts)], LTS _tz_
Why not _ts_? Because of sorting?
> d.ad - [dz)], LTS _dz_
> t.aa - [t]
> z.aa - [dz)], LTS _dz_
> `ain - in most cases silent;
> [?] between a cons. and a vowel, LTS _3_
> ghain - [g] word initial, [R] inside; both are _gh_ in LTS
> faa - [f]
> qaaf and kaaf - [k]
> gaaf (kaaf with a stroke) - [g]
&1711; gaf (Persian, Urdu, ...)
> laam - [l]
> miim - [m]
> nuun - [n]
> haa - [h] or mater lectionis
> waaw - [w] or mater lectionis
> yaa - [j] or mater lectionis, _y_ in LTS.
Should not the LTS differentiate all letters
which are differentiated in the official Arabic script?
P.A.