[A12n-forum] Between the lines (Re: NYTimes article on African languages & ICT)

["Donald Z. Osborn" <dzo@xxxxxxxxxxxx>]

This NYTimes article has gotten a lot of attention among people interested in
aspects of localization and African languages, and deservedly so. It highlights
some of the increasing level of activity on the continent and abroad to use its
indigenous languages in computing and on the internet. Nevertheless, there is
as much left out as included and I wanted to take a moment to mention some of
what wasn't said...

Don Osborn
Bisharat.net


> Using a New Language in Africa to Save Dying Ones
> 
> November 12, 2004
>  By MARC LACEY 
> 
> NAIROBI, Kenya, Nov. 11 - Swahili speakers wishing to use a "kompyuta" - as
> computer is rendered in Swahili - have been out of luck when it comes to
> communicating in their tongue. Computers, no matter how bulky their hard
> drives
> or sophisticated their software packages, have not yet mastered Swahili or
> hundreds of other indigenous African languages. 

True, but a fair number of people have been using languages like Swahili in
e-mail and some web content. Quantifying that is another matter, but some
research is aiming to get a clearer picture of the level of African language
use in ICT.

> But that may soon change. Across the continent, linguists are working with
> experts in information technology to make computers more accessible to
> Africans
> who happen not to know English, French or the other major languages that
> have
> been programmed into the world's desktops. 

True, but it would be easy to overstate the level of collaboration among
linguists on the one hand and ICT experts on the other. In fact, on the ground
level in many countries, from what I've heard and been able to tell,
collaboration between language and technical experts is still limited.
 
> There are economic reasons for the outreach. Microsoft, which is working to
> incorporate Swahili into Microsoft Windows, Microsoft Office and other
> popular
> programs, sees a market for its software among the roughly 100 million
> Swahili
> speakers in East Africa. The same goes for Google, which last month launched
> www.google.co.ke, offering a Kenyan version in Swahili of the popular search
> engine. 

The good news in this is that Africa counts, though not a few are disquieted by
the interest the software giant in particular is taking in African languages.
Purely from the perspective of advancing use of the languages, however, I think
that the competition with OSS is good and will lead to higher quality and
quantity of localized products. It might not be too popular to say this, but
there are some good people in MS working on African language localization, just
as there are in the OSS movement.

> But the campaign to Africanize cyberspace is not all about the bottom line.
> There are hundreds of languages in Africa - some spoken only by a few dozen
> elders - and they are dying out at an alarming rate. The continent's
> linguists
> see the computer as one important way of saving them. Unesco estimates that
> 90
> percent of the world's 6,000 languages are not represented on the Internet,
> and
> that one language is disappearing somewhere around the world every two weeks.

THis particular point was discussed a bit on the ILAT list - if this is to imply
that software localization for endangered languages is a preferred strategy, it
may be missing the main ways that ICT can help those languages (and hence
priorities).
 
> "Technology can overrun these languages and entrench Anglophone
> imperialism,"
> said Tunde Adegbola, a Nigerian computer scientist and linguist who is
> working
> to preserve Yoruba, a West African language spoken by millions of people in
> western Nigeria as well as in Cameroon and Niger. "But if we act, we can use
> technology to preserve these so-called minority languages." 

Yoruba is not a language of Cameroon or Niger, but it is spoken (as Ife/Ana) to
the west in Benin and into Togo. (Tunde, of course knows this, but the
reference is not his.)

> Experts say that putting local languages on the screen will also lure more
> Africans to information technology, narrowing the digital divide between the
> world's rich and poor. 

It will also make it more relevant generally.

> As it is now, Internet cafes are becoming more and more common in even the
> smallest of African towns, but most of the people at the keyboards are the
> educated elite. Wireless computer networks are appearing - there is one at
> the
> Nairobi airport and another at the Intercontinental Hotel in Kigali,
> Rwanda's
> capital - but they are geared for the wealthy not the working class. 

As I've suggested elsewhere, the so-called "bridging" of the "digital divide"
often looks more like its replication on ever more local levels, along
socio-economic and linguistic lines.

> Extending the computer era to the remote reaches of Africa requires more
> than
> just wiring the villages. Experts say that software must be developed and
> computer keyboards adapted so that Swahili speakers and those who
> communicate
> in Amharic, Yoruba, Hausa, Sesotho and many other languages spoken in Africa
> feel at home. 

This is in other words an expanded and nuanced view of "access" that goes beyond
just letting someone sit in front of a connected computer. This discussion has
been percolating for some time, but as obvious as it may seem, the notion of
localizing anything ICT to African languages has still nonplussed many industry
and development organization people who wonder why it is necessary (I won't
give examples here but there are many).

> Mr. Adegbola, executive director of the African Languages Technology
> Initiative,
> has developed a keyboard able to deal with the complexities of Yoruba, a
> tonal
> language. Different Yoruba words are written the same way using the Latin
> alphabet - the tones that differentiate them are indicated by extra
> punctuation. It can take many different keystrokes to complete a Yoruba word.

It also has some dot-under diacritical characters. Combination of those with
tine marks (combining diacritics) is the challenge - not only for input and
rendering, but also for standards.

> To accomplish the same result with fewer, more comfortable keystrokes, Mr.
> Adegbola made a keyboard without the letters Q, Z, X, C and V, which Yoruba
> does not use. He repositioned the vowels, which are high-frequency, to more
> prominent spots and added accent marks and other symbols, creating what he
> calls Africa's first indigenous language keyboard. Now, Mr. Adegbola is at
> work
> on voice recognition software that can convert spoken Yoruba into text. 

This is one of several strategies for input of extended characters (key
substitution; the others are use of deadkeys and key combinations). There are
other Yoruba keyboards with other approaches (this has been discussed on
several fora including A12n-forum).

Alt-I's keyboard actually won an award from IICD. Tunde, in recent
correspondence, is humble about his accomplishment, and also committed to the
longer term aim of promoting ICT in Yoruba. How this all shakes out will be an
interesting process in itself and also because there are questions of
localization for other languages in Nigeria and the region. (There is actually
a keyboard layout for all Nigerian languages called NITDA, but there is an
issue with one of the diacritics it uses.)

> Related research is under way in Ethiopia. Amharic, the official language,
> has
> 345 letters and letter variations, which has made developing a coherent
> keyboard difficult. Further complicating the project, the country also has
> its
> own system of time and its own calendar. 

Different issues here. Staying with keyboards, one innovation has been
adaptation of a graphic tablet as a keyboard for this syllabary (the "letters"
actually represent syllables).

> Still, computer experts at Addis Ababa University are making headway.
> Recently,
> they came up with a system that will allow Amharic speakers to send text
> messages, a relatively new phenomenon in the country. 

Like to know more about this.

> The researchers involved in the project envision it as more than a way for
> Amharic-speaking teenagers to gossip among themselves. Text messaging could
> be
> a development tool, they say, if farmers in remote areas of the country can
> get
> instant access to coffee prices or weather reports. 
> 
> The Ethiopian researchers hope a cellphone maker will see the country's
> millions
> of Amharic speakers as a big enough market to turn their concept into a
> commercial Amharic handset. 

There of course is language localization for cellphones in southern Africa -
less challenging with respect to displays since they use ASCII characters
there.

> Mr. Adegbola has similar dreams. He is distributing his keyboard free to
> influential Yoruba speakers, hoping to attract some deep-pocketed
> entrepreneur
> who could turn it into a business venture. 
> 
> In South Africa, researchers at the Unit for Language Facilitation and
> Empowerment at the University of the Free State are working on a
> computerized
> translation system between English and two local languages, Afrikaans and
> Southern Sotho. Cobus Snyman, who heads the project, said the goal is to
> extend
> the system to Xhosa, Venda, Tsonga and other South African languages. 

Machine translation could be really exciting in Africa, and potentially used in
ways unlike what folks in the North are used to thinking in terms of.  There
are several efforts on small scale. However, with this and other high-tech
applications like TTS & STT there is a need to be sure that orthographies etc.
are standardized & stable.

> One of Microsoft's motivations in localizing its software is to try to head
> off
> the movement toward open-source operating systems like Linux, which are
> increasingly popular. South Africa has already adopted Linux, which it
> considers more cost efficient and more likely to stimulate local software
> development. 

OSS has some special advantages and is also more likely to address needs for
less widely spoken languages. It certainly has a strong local movement -
including in the Francophone countries. The only fault is that except for a few
efforts like Translate.org.za (which is a big omission from this article) there
does not seem to be much software language localization underway (and the topic
has had few echoes so far when brought up in French language fora - at least
that I've noted).

> Patrick Opiyo, the Microsoft official in charge of the Swahili program,
> portrays
> the effort as more about community outreach than business development.
> Besides
> Swahili, the company is looking at making its products more available to
> those
> who speak Amharic, Zulu and Yoruba and the other two widely used languages
> in
> Nigeria - Hausa and Igbo. 

MS is a big an complex organization. A discussion on ILAT mentioned that some
people from MS were unimpressed by the idea of localizing anything to certain
indigenous North American languages (one of those "but eveyone speaks English"
moments).

> In Kenya, Microsoft has rounded up some of the region's top Swahili scholars
> to
> come up with a glossary of 3,000 technical terms - the first step in the
> company's effort to make Microsoft products accessible to Swahili speakers. 

It's actually a great advantage to have these resources going into solving some
localization issues (which is not to say one favors MS over OSS - I still
regret what MS did to WordPerfect). Again my perspective is first how to best
advance the potential to use African languages on computers and the internet.

> 
...
> 
>
http://www.nytimes.com/2004/11/12/international/africa/12africa.html?ex=1101289393&ei=1&en=472d9074c49fb681
> 
> 
> ---------------------------------
> 
> Copyright 2004 The New York Times Company
> 
> ----- End forwarded message -----
> 
> 
> _______________________________________________
> A12n-forum mailing list
> A12n-forum@xxxxxxxxxxxx
> http://lists.kabissa.org/mailman/listinfo/a12n-forum
>