a thought on microsoft...

Donald Z. Osborn dzo at BISHARAT.NET
Sat Nov 13 09:35:52 UTC 2004

Phil, These are good points. It's hard to imagine that we'd see a full-blown
software suite for every language including those with very few speakers, but
how things play out in the vast middle ground of languages with, say, 50,000 to
1,000,000 speakers is the interesting question.

One thing that I've been pushing, without a whole lot of success I'm afraid, is
the idea of lesser localizations, meaning accommodation of needs for using
various languages even if the operating environment is still in a more widely
used language (a convenient acronym for which is LWC - language[s] of wider
communication). These include the means for composition in minority languages -
fonts and keyboards where orthographies require it, content (not necessarily
limited to text), and better provision for use of audio (e.g., audio e-mail).
In addition to the problem of attitudes in industry and development agencies
that "everyone uses [insert LWC name]" there also seems to be an all or nothing
mentality - why do anything if you can't do everything (so no lite or
intermediate measures are seriously entertained). Such lesser localizations can
be applied fairly quickly and cheaply in ICT4D and education projects. I think
there is a need to expand thinking about localization and recognize that there
may be various solutions for different situations, and perhaps even a vision of
a sequence of progressively more technically complex solutions.

Beyond that, the issues you raise about literacy and orthographies are centrally
important. First-language literacy and pluriliteracy (multi/bilingual literacy)
issues need to be addressed in more comprehensive plans for expanding access to
ICT, and this in turn brings in language and educational policy considerations.

At the same time it would seem necessary not to get trapped into thinking that
literacy has to come first, in which case nothing will happen. Localized
content and even software will (depending on the language situation) likely
find people along a spectrum between those just learning to read anything for
the first time, to those who are literate in an additional language but never
learned to read their first one, to those who are literate in the first
language and perhaps others.

The issue of stable standard orthographies is important for progress in the
written form of a language, of course, but it's pretty much essential for full
utilization of the potential of basic software features like sorts and spell
checks and advanced tools such as machine translation or text-to-speech. This
is an issue that requires communication between linguists, educators, and

Standardizing orthography is a concern that seems particularly urgent in Africa,
even though there has already been significant work on this in past decades.
We're not only talking about languages with few speakers, but even for some of
the major ones there is still discussion on fundamentals of what are fairly
well established orthographies. And then there is the issue of different
countries' conventions for the many languages that cross borders.

Nevertheless, it would seem advantageous to forge ahead however imperfectly (and
a lot of the African language material that there is, from web content to
localized Googles, not to mention e-mails, does not follow any strict rules).
However, if it's worth doing (and it is), it's worth doing wrong ... at least
at this point and for things that can be fixed.

The last point may bring up a note of caution re software localization. While
messy localized content can be fixed, the existence of language localized
software that is somehow deficient, erroneous, or idiosyncratic in its usage of
the language - for reasons of orthography or other - might be more problematic
in the long term than to wait a little while in order to get it right.

Sorry if these musings ramble a bit. I'm working on something in this topic area
and this lets me run out some ideas that may be of interest to others.

Don Osborn

Quoting phil cash cash <pasxapu at DAKOTACOM.NET>:

> I wonder if endangered language communities with small populations
> warrant the kinds of software localization like Quechua and Swahili.
> It would seem, in any given situation, one would need a strong
> developing literacy and stable orthography to support this kind of
> development, not to mention some long term heritage language benefits.
> Just curious, and thanks for all your messages as I am learning a lot
> from your discussions,
> phil cash cash
> UofA

More information about the Ilat mailing list