[A12n-forum] Re: LOP & UNESCO stats

["Don Osborn" <dzo@xxxxxxxxxxxx>]

Thanks for keeping me in the loop on this. I'll cc to the A12n-forum list
(NB to readers - this topic originated on this list as "Presentation on LOP
African language survey at UNESCO," and was forwarded offlist; it now comes
back with some interesting observations from Gilles-Maurice de Schryver, to
which I reply).

A lot goes back to definitions, IMO. "Endangered" is a word thrown around a
lot. For instance I've read more than once in Nigerian press claims that
Igbo is "endangered" - a language with perhaps 20 million speakers! A useful
concept for such cases is that of "contraction" or "impoverishment."
Endangered languages are an important concern, but the contraction of
African languages is at least as urgent, and is directly relevant to the
purposes of projects like LOP (Language Observatory Project) and PAL
(PanAfriL10n / PanAfrican Localisation project). Not sure if UNESCO is
addressing this - and there is no "Red Book" for languages that are
"contracting" in terms of skilled use across the range of social settings.

On "language," Ethnologue is a unique and essential resource, but definitely
a "splitter" when it comes to defining what is a language. It seems that any
broad statement re extinctions glosses over relationships among languages -
this is a point that Gilles-Maurice already alluded to (dialects &
languages). For example are these two situations comparable?:
* when a "language" so closely related to another that it might be
considered a dialect gets "absorbed" into the other (in the manner of
dialect-leveling but with unequal partners - I understand from Ethnologue
itself that something like this is happening with Khassonke in relation to
Bambara; another example perhaps is Runyakitara in Uganda - if the 4
constituent languages cease to be identifiable as separate, has humanity
lost a count of 3?)
* when a language with no near relations, linguistically or geographically,
ceases to be spoken

Counts of languages in Africa vary, of course. On the other extreme of the
spectrum, perhaps, are ideas espoused by Prof. Prah of CASAS. In between
there is room for discussion and need for analysis (CASAS has done some
research into interintelligibility; one hopes that the BASAL project of
Prof. Tadadjeu and NACALCO have similar aims). If you look at what I've
tried to do in the wiki at
http://www.panafril10n.org/wikidoc/pmwiki.php/PanAfrLoc/MajorLanguages you
will note that in many cases listings in Ethnologue are regrouped, as a
basis for discussion relating primarily to localization and localization
policy (there's a longish discussion of the evolution of this list, which is
by no means in a final form, and references to other lists).  

Making clear statements about what has an orthography or not is tricky. My
impression is that most African languages have been written in some way,
sometimes in more than one way. But it is also true that many do not have a
fixed orthography. It might be said that the majority do not have an
established orthography, as in one that is officially standardized and used.


I had hoped to develop on the PanAfriL10n.org site a database on
orthographies of African languages that would draw on work such as Christian
Chanard's dbase at http://sumale.vjf.cnrs.fr/phono/index.htm (itself based
on Rhonda Hartell's 1993 book, some of the data of which was earlier
summarized at http://www.bisharat.net/A12N/#countrytables ) and an extensive
research of printed sources done by John Hudson for Microsoft (which
apparently covered usages and systems that are obsolete as well as current).
However the data overall is uneven and incomplete. It made more sense to
reference other works in the major language profiles on the PanAfriL10n.org
wiki (for examply under #6, orthography, in each language profile) - one of
the many tasks on the list for the wiki.

Another related area - tangential to this discussion but relevant to the
extent of classifying languages in ICT - is ISO-639. Part 3, based on
Ethnologue's language categories, does not seem to match practice and needs
in all kinds of localization. The ISO-639 system including part 3, however,
is in a state of development and to the extent that African language
specialists and standards groups in Africa can have an input, the result may
be flexible enough to accommodate both text that is specific to particular
African language varieties and localized interfaces (software) that is
usable across a range of closely related "languages" in one or more
countries.

Lots more to say on all this but I will stop here.

Don

> -----Original Message-----
> From: Gilles-Maurice de Schryver
> [mailto:gillesmaurice.deschryver@xxxxxxxxxxxxxx]
> Sent: Sunday, February 04, 2007 8:30 AM
> To: 'David Joffe'
> Cc: 'Don Osborn'; 'De Pauw Guy'; asamass@xxxxxxxx; 'Michael Meeuwis'
> Subject: LOP & UNESCO stats
> 
> 
> Hi,
> 
> Thx for this!
> 
> I cannot find 'the LOP tool' itself though -- was hoping to add it to
> aflat.org
> 
> Clicking through, I arrived here:
>
http://portal.unesco.org/culture/en/ev.php-URL_ID=8270&URL_DO=DO_TOPIC&URL_S
ECTION=201.html
> 
> where I cannot agree with most statements:
> 
>     * Over 50% of the world's 6000 languages are endangered.
> 
> -> Possible, depending on how one defines 'endangered'. With around
> 2000
> languages in Africa, and fewer than 100 endangered, that 50% definitely
> seems too high though. Cf.
> http://portal.unesco.org/ci/maps/babel/output/
> african_endangered_languages.php
> 
>     * 96% of the world's 6000 languages are spoken by 4% of the world's
> population.
> 
> -> Likely true; easy to check and the distribution is indeed likely
> Zipfian.
> 
>     * 90% of the world's languages are not represented on the Internet.
> 
> That would mean only 600 languages are. Most definitely wrong, as all
> studies in computational linguistics that deal with minority languages
> prove.
> 
>     * One language disappears on average every two weeks.
> 
> -> Likely true, if 'language' includes 'dialects'. One needs to be
> careful
> with this figure though, as that means around 25 languages disappear
> every
> year, 250 every decade, or thus over 1000 by the year 2050. Hmmmm ...
> 
>     * 80% of the African languages have no orthography.
> 
> -> Most definitely wrong. All Bantu languages, of which there are about
> 500
> (out of 2000 African languages), have an orthography. So, that's
> already 25%
> that _have_ an orthography. Unless one means a 'stable/standard'
> orthography, with which everyone agrees?
> 
> One would expect better stats from UNESCO ...
> 
> Gilles-Maurice.
> 
> 
> -----Original Message-----
> From: David Joffe [mailto:david.joffe@xxxxxxxxxxxxxx]
> Sent: 02 February 2007 11:04
> To: gillesmaurice.deschryver@xxxxxxxxxxxxxx
> Subject: (Fwd) [A12n-forum] Presentation on LOP African language survey
> 
> fyi
> 
> ------- Forwarded message follows -------
> From:                 "Don Osborn" <dzo@xxxxxxxxxxxx>
> To:                   <A12n-collaboration@xxxxxxxxxxxx>,
>       <A12n-forum@xxxxxxxxxxxx>
> Date sent:            Thu, 1 Feb 2007 13:28:41 -0500
> Copies to:
> Subject:              [A12n-forum] Presentation on LOP
> African language survey at UNESCO
> 
> [ Double-click this line for list subscription options ]
> 
> The Language Observatory Project (LOP) has developed a
> tool for surveying
> African language content on the web. There will be a
> presentation of it at
> the UNESCO headquarters in Paris on 22 February, as part
> of a 2-day
> colloquium in observance of International Mother
> Language Day (21 Feb.).
> 
> For more information see:
> *
> http://portal.unesco.org/ci/en/ev.php-
> URL_ID=23957&URL_DO=DO_TOPIC&URL_SECTI
> ON=201.html
> *
> http://portal.unesco.org/culture/es/ev.php-
> URL_ID=32773&URL_DO=DO_PRINTPAGE&
> URL_SECTION=201.html
> 
> A page on the LOP site with more info on aspects of the
> African language
> survey (meetings & presentations) is at:
> *
> http://gii2.nagaokaut.ac.jp/gii/blog/lopdiary.php?catid=
> 154&blogid=8
> 
> Don Osborn
> Bisharat.net
> PanAfrican Localisation project
> 
> 
> _______________________________________________
> A12n-forum mailing list
> A12n-forum@xxxxxxxxxxxx
> http://lists.kabissa.org/mailman/listinfo/a12n-forum
> 
> ------- End of forwarded message ----------
> http://tshwanedje.com/
> TshwaneDJe Human Language Technology
> 
> 
> 
> ------=neXtPaRt_1170595567
> Content-Type: text/plain;
> 
> All Email originating from UWC is covered by disclaimer
> http://www.uwc.ac.za/portal/uwc2006/content/mail_disclaimer/index.htm
> 
> ------=neXtPaRt_1170595567--