Date: Sun Feb 14 2010 - 10:56:46 CST
I am often in North Africa and work a lot with different Berber language
Because of the frustrations with different writing systems and the keyboards
and fonts, I ended up systematically trying to improving them, while
remaining on the fence about whether to use Latin letters or Tifinagh.
I ended up with a unified proposal for writing Berber language varieties
(including Tuareg) using a common approach, in Latin letters or Tifinagh,
while allowing for calligraphic, regional and personal variation. My
proposal is backwards-compatible with existing fonts, keyboards, Unicode
and regional literacy work.
To summarize my approach, I looked across all the language variants by
letter sound. Even though the signs used for particular sounds vary
slightly from place to place and through time, which has clouded the
Unicode issue so far, the overall repertoire of phonemes is more or
less the same.
A common Tifinagh alphabet can be achieved that also covers modern Tuareg,
using the IRCAM code points (and their visual reference forms) as a
base (they are even named by sound). The visual form for different regions
would be done by choice of font.
For scientific use where the same explicit letter form should always
be rendered in text regardless of the font used, the existing approach
of collecting attested letters could be continued, including historic
The common alphabet coding would be useful for existing Berber variants
with their existing orthography, as well as for writing the same variants
in a converged style, and also any future converged literary forms.
What is needed as extra in Unicode is 2 Tifinagh letters for normal text,
and 1 more arguable letter for phonetic dialect writing.
The 3 are:
- Tuareg 'e' (long e, not schwa)
- Tuareg 'o'
- Tuareg palatalization modifier letter y, written as a zigzag - equivalent
to the Tifinagh labialization marker for Moroccan gw, kw.
Other features are achieved by use of combining diacritics.
My informal proposal is available here:
It begins by discussing Latin script, then moves to the equivalent Tifinagh.
My keyboard drivers and demonstration fonts are also on the site.
I'm in discussion with specialists to check my approach for Tuareg as
I'm not so familiar with it, but I have a lot of written material in
Northern Berber variants.
Should I eventually put my proposal in as a Unicode standard proposal
pdf? If so, could someone help me to formulate it? What are people's
comments, and thoughts on Tifinagh? Perhaps someone has already made
a similar proposal?
A complete common alphabet encoding would be a huge leap forward for
Berber, and further font and keyboard work could begin. Openoffice 3.2
already supports SIL graphite, which would allow even the ideal font
scenarios in my proposal to work.
This archive was generated by hypermail 2.1.5 : Sun Feb 14 2010 - 15:31:16 CST