Re: VMs: [VMS] Arabic again

[Jacques Guy <jguy@xxxxxxxxxxxxxxxx>]

21/10/2004 11:04:45 PM, Jorge Stolfi <stolfi@xxxxxxxxxxxxx> wrote:

>Indeed the word-length distribution (WLD) of the Quran, in vowelless 
>Arabic and a "native" encoding, is surprisingly similar to that of 
>Voynichese.

It has to do with most roots consisting of three consonants
and only long vowels being represented. Most derived forms are
prefixing or suffixing a single consonant, so that the distribution
of word lengths is quite... can't think of the word.. "narrow"
will do.

>On the contrary,
>it seems that some people have looked hard at Arabic, and rejected it.
>Comments, anyone?

A suggestion was "dotless Arabic". Remove the dots from the
consonants and you end up with:

r = z
d = dh (soft th)
t = n = b = y = th
s = sh
S = D (emphatic s and d)

Much guessing ensues (is this byt "house" or 
nyb "playing card"?)

But even dotless Arabic does not match Voynichese
at all. I have posted here about it several times.

For instance, if we assume that <qo> = "al-"
(alif, lam) we expect to see <q> (alif) word-finally. 
We don't.

We also expect to see it followed by just about
any other letter. We don't. 

If we say <qo> = "wa" (and, spelt "w") then we are 
saying that this is Arabic with the vowels written.
So <q> = w and <o> = a. 

But then again, we she should see <q> occurring 
in any position, followed by any consonant or vowel. 
We don't.

Same for whatever values we might assign to <qo>.



______________________________________________________________________
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
unsubscribe vms-list