Arabic Orthography

From Lojban
Jump to navigation Jump to search
  • .filip.:
    • I thought of using Arabic letters to write Lojban.
    • The consonants are pretty straightforward, especially if you use Persian rather than Arabic as a base (which adds g, j, p, and tc, of which we need the first three).
    • I thought of using the existing vowel signs for the Lojban vowels. Since Arabic has only three signs compared to Lojban's six vowels, I thought of using the "doubled" signs (-n endings) for the "real" vowels and "single" signs for "close" vowels - that is, vowels which aren't the same as the standard Arabic vowel but which are sometimes represented with the same vowel sign anyway.
    • This gives kasra/kasratan for e/i, damma/dammatan for o/u, and fathatan for a. This leaves y to go with fatha.
    • I propose using alif for . since it's a glottal stop, and heh for ' because of the pronunciation. I couldn't think of anything particularly good for , so I picked yeh with hamza. I also need a "null" consonant for diphthongs to carry the second vowel sign; I picked `ayn.
    • So here's my proposal. What do you think?
    • lojban-xrabo.png
      • .aulun.:
        • Doesn't look bad - and it seems you are using front, middle, end and single type variants of Arabic (Farsi) characters. Yet, it also seems you didn't do this by normal typing but by composing the samples by means of a graphic tool ;-) I therefore think that - different from Yiddish mode - it is a hard job to 'really use' Arabic for Lojban writing. Nevertheless, nice idea.
          • .filip.:
            • I did indeed use a tool SC UniPad which automatically shapes Arabic letters (and does RTL automatically for Hebrew, Arabic, and Syriac, too). The screenshot is from UniPad.
              • .aulun.:
                • Great! Without need of extra fonts. Yet, Macintosh again excluded :[
ni'o la .dorotis. cu xabju lo midju be lo ganra sastu'a pe la .kanzas. fi'o kansa la nakfamti .xenris. ku noi te cange ku'o .e la fetfamti .em. ku noi speni lo te cange .i lo zdani cu cmalu ki'u lo nu lo mudri poi zy ke'a zilzba cu jai sarcu fai lo ka lo carce cu bevri ce'u ve'a lo minli be li so'i

  • la gustek.:
    • With all due respect, I think rablermorna has multiple problems, first of all, the use of symbols are linguistically and historically incoherent: they don't quite reflect the sound they represent and the vowel system entirely lacks some form of precedent in existing languages using a Perso-Arabic script; then, the exclusive use of ḥarakāt to represent vowels has three effects: 1) making it difficult to type texts 2) making it difficult to read texts given the size of ḥarakāt in a standard font and 3) it breaks diphthongs both visually and orthographically, and uses a totally unrelated symbol to represent the latter. This is why I suggest the following specification:
      Arabic Orthography for Lojban.png
    • As you can see, this proposal uses exclusively long vowels, no ḥarakāt. This allows to identify vowels more easily and to represent diphthongs in a way akin to the Latin orthography for Lojban. Most consonants are directly taken from Arabic, apart from ه for /ɛ/ and ۆ for /ɔ/ that come from Kurdish. I've also decided to use the alif maqṣūrah (ى) for the /ə/ sound, because of its similarity to the ي, thus being linked to the graphic similarity between ‹i› and ‹y›. Then, for consonants, I've given multiple possibilities for cases in which two look-alike consonants would be placed next to each other, or simply for aesthetics' sake. I've proposed both the ra and the ʿayn for the ‹r› character, given that the (Un)CLL specifies that any rhotic consonant is accepted for this character. I imported one Maghrebi character ‹ڤ› and three Persian characters ‹پ›, ‹گ› and ‹ژ› to respectively represent the /v/, /g/, /p/ and /ʒ/ sounds, although the latter is not really needed. The hamzah represents the glottal stop ‹.› as it normally represents the glottal stop in Arabic and that it always being isolated allows for better visual delimitation of names and single vowels. /a/, /i/ and /u/ all exist in a “hamzah-ed” form as it allows representing some commonly used sequences (such as ‹.i›) in a single character. As in rablermorna, vowel elimination is perfectly possible in words containing more than one consonant.
p پ
t ت ,ط
k ك
f ف
s ص ,س
c ش
b ب
d ض ,د
g گ
v ڤ
z ظ ,ز
j ژ ,ج
m م
l ل
n ن
r غ ,ر
x خ
' ح
. ء
a ا
.a أ
e ه
i ي
.i ئ
o ۆ
u و
.u ؤ
y ى

tcidu ji'a