--- In qalam@yahoogroups.com, "Richard Wordingham"
<richard.wordingham@...> wrote:

>

I got the Vai Unicode Proposal printed off today.

> Are these known to be loan words, or are this just a guess? The
> symbols for /zo/ (ZOO), /zO/ (ZO?) and /jO/ (YO) had a theme of two
> vertical circles, so they dont seem to be aftwerthoughs in the

system.

Not sure.

> And I get the impression that /mgb-/ does not occur in the text of
>the
> book of Ndole, which stacks up with the the lack of *mgbi and

*mgbu in

> the modern language.

According to Singler prenazalized are not necessary.

>
> Collapse /w/, /h/ and /0/ (nothing) with oral vowels, and /w/, /h/,
> /0/ and /N/ with nasal vowels.
>

>
> Lack of distinction of /l/, /d_</ and /nd_</ is well documented,

even

> when there were three distinct symbols!

Right.

>
> > 5. Collapse voiced consnants into unvoiced. minus 4 rows.
> >
> > 6. Possibly nasal syllables were marked with the related

implosive.

>
> Is this pure speculation?

In the VUP page 5 "implosive consonants are far more frequent than
non-implosive ones" *suggests* that the voiced v. unvoiced
distiniction for the non-implosive consonants is not important.
However, it is still there for some vowels fairly early.

I know this is pure speculation but we can use it to frame the
questions. First, how did Singler get 40 - 60 symbols? I would guess
that is because only a few symbols from each row are used, rather
than because so few consonants were used.

Next, would one want to use a completely systematic approach with
complete rows or not? If it were possible to find the original
consonants, (16 are used in the text that Singler provides) and then
complete the rows there would be a total of 7 x 17 = 119.

Another approach would be to use the "N'dole syllabary" and actually
use the consonants that N'dole used but fill out the rows. Call it
the "N'dole - Vai Syllabary". The rows from page 11 would be p, b,
b_, kp. gb. f, t, d, l/d_, s, j, k, w, and five nasal rows. 13 x 7
= + 18 nasals = 109.

Without collecting handwritten text, I think that is about as far as
one can go.

Is it just me or does it seem easier to provide full rows for each
consonant?

I fully realize how speculative this is, but it is a good exercise
none the less.

Suzanne

>Looking at the early glyphs, I can persuade
> myself that there were the following lacks of consonant
>distinctions:
>
> 1. h v. w v. 0
>
> 2. N v. h~ v. w~ v. 0~
>
> 3. p v. b v. m (but only p v. b look like a late distinction in the
> standard syllabary - 4 cases out of 7)
>
> 4. b_< v. mb_< (6 out of 7 look like late distinctions)
>
> 5. kp v. gb v. mgb (All the mgb characters are modifications of
> others; at least 3 of 7 kp v. gb distinctions look late)
>
> 6. f v. v (7 out of 7 look like late distinctions)
>
> 7. t v. d. v. n (but only t v. d looks liks a late distinction - 3

or

> 4 out of 7)
>
> 8. l v. d_< v. nd_<
>
> 9. s v. z (5 out of 7 look look late distinctions)
>
> 10. c v. J\ v. ñJ\ v. j (In 5 of 7 cases, 3 or more look like late
> distinctions)
>
> 11. k v. g v. Ng v. ñ
>
> 12. Syllabic nasal.
>
> > That gives only 63 symbols left.
>
> This comes roughly to 11 lots of 7 - you didn't count for the
> palatals, and I think at least one set of nasals is needed.
>
> Although my idea looks like 78 symbols, there do seem to be gaps

(or

> at least, very infrequently needed combinations) in the language's
> combinations:
>
> No /pe,be,me/, no /pO,bO/ and MO seems to be related to POO and

BOO,

> not PO and BO. Strike 2.
>
> No /hu~,wu~,u~,Nu/, no /he~,we~,e~,Ne/, no /ho~,wo~,o~,No/. Strike
> another 3.
>
> That still leaves 73 symbols!
>
> This discussion is rather speculative - I suspect we need texts

(e.g.

> private letters or accounts) to make further progress, and without

a

> knowledge of Vai we'd probably have a hard time analysing them.
>
> Richard.