[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-i18n-doc
Subject:    Re: new language
From:       Kevin Donnelly <kevin () dotmon ! com>
Date:       2004-01-29 11:44:24
Message-ID: 200401291144.24247.kevin () dotmon ! com
[Download RAW message or body]

The following is a reflection on what Rinse and Heiko have said - it's not a 
comment on Frisian itself, but on the trials and tribulations of doing a 
minority translation.

On Wednesday 28 January 2004 10:22 pm, Rinse de Vries wrote:
> We had about 20 volunteers, and about 5 actual translators.
> i guess we had about 45% of kdelibs (ppptranslations.po not counted)
> translated when the team broke up.

There's no doubt it's difficult, especially for a language where the actual 
vocabulary may not exist yet, and people are not entirely sure what is 
"correct" and may be scared that their version is "dialectal".

Rightly or wrongly, in KDE Welsh I have taken the view that anything goes.  I 
will correct obviously wrong things to the best of my ability, but if one 
person uses periphrastic tenses and another uses the impersonal, or one uses 
"rwan" for "now" and another uses "nawr", it all stands.  As we have built up 
our translation database (Omnivore - 
http://www.kyfieithu.co.uk/kartouche/omnivore/index.php?lg=en&), we can see 
which words are "by consensus" getting used for particular things, so most 
people will use these, and if they don't, again I change them (eg "plygell" 
instead of "ffolder" for "folder").

At some point (once the translations is 100%), we will perhaps have to sit 
down and do some standardisation, but that can wait.  (We'll have an app 
shortly to allow end-users to look up a desktop translation they think is 
wrong, and write a correction - I can then commit this if appropriate, and 
make a revised .mo file available for download.)  I think it's really 
important from the morale point of view to get something translated and 
available to show - even if people disagree with it and criticise it, that 
may spur others to get involved.  I think Heiko has a good point here - do 
something and let others critique it, but bear in mind that the best is the 
enemy of the good.  I am not actually a first-language speaker, but I began a 
recent talk on the translation effort by saying to my audience (who *were* 
first-language speakers) that they were welcome to criticise, but only if 
they did something about it themselves - in fact, they all agreed that this 
was a valid line to take, and were very supportive.

But I know it's difficult - things have progressed here in fits and starts, 
depending on who was available to translate.  A report in the paper can get 
you maybe 500 translations - people see the report, go to the website, and 
add 20, or in some cases a lot more.  About 5% of our 26,455 strings (about 
1300) have been translated by what I would call "passers-by"; 62% was 
translated by two people; another 33% by 6 people, 5 of whom seem to have 
dropped out, with the other one just started contributing (see the Hall of 
Fame at http://www.kyfieithu.co.uk/kartouche/hallfame.php?lg=en).  Much of 
this input came after press reports, and if one is lucky one will get someone 
who really is committed (even if he/she can only do it for a limited period, 
if they submit a couple of thousand translations it can be a real boost).  So 
it's important to keep on at the publicity angle, even if it's very 
time-consuming.  Heiko is lucky to have someone prepared to do that.

> The problem with Frisian is that I can't write Frisian good enough to
> translate it myself. In our team there were only 4 people with Linux
> installed. All others used Windows.

We have exactly the same problem here - *very* few people are aware of Linux 
at all.  That was the reason I did Kartouche, and it might be useful in a 
situation where it is desired to make things as simple as possible for 
"passers-by" to contribute.

> During the time our team was active, quite a lot of Frisian Linux-users
> contacted us wanting to download the Frisian modules. But none of them
> wanted to help with the project.
> After three months most of the members left the team. We tried for about a
> year to bring new life into the project, but failed.

Yes, it's sad but true - in most "community" projects (eg building a new 
village hall, raising money for the local school, translating KDE), only 
about 5% of the people concerned will actually get involved.  About another 
20% will be willing to help in a vague sort of way, with no commitments, 
another 40% will contribute money or moral support if specifically asked, and 
the other 35% will do nothing.  

With a minority language, you will also get lots of people (even some minority 
language speakers themselves) asking "what's the point?".  I think this is 
because they assume there is only *one* possible OS, aren't aware of free 
software, and can't see any way of persuading the foreign multinational that 
makes that one OS to support their titchy little language.  So to a certain 
extent there is a big education process involved here as well - believe it or 
not, when I am on the phone now explaining this to people, I can actually 
*hear* their eyes glaze over!  This will have to be done perhaps a thousand 
times before the message begins to percolate. 

But as Heiko says, it is an important message.  Without the minority language 
available for the most important tool of our age, whole generations of 
children will grow up ceding that portion of the language sphere to the 
majority language.  So the minority language will be cut off from a huge 
chunk of daily life.  This is particularly important in schools, where people 
form their first impressions of where it is "proper" or "appropriate" to use 
particular linguistic constructs.  If people realise that there is a viable 
alternative that is built around their language, they may be willing to go to 
sometimes significant lengths to use it.  Someone contacted me after our last 
press report, and it is obvious he knows very little about Linux, and had 
only just heard about it, but he is buying a boxed set purely on the basis 
that GNOME 2.4 supports Welsh, and that the Mandrake install has been 
translated into Welsh by someone else.  (KDE will be in Welsh in the next 
Mandrake, Fedora is being translated as we speak, and SUSE may be ready by 
v10, so things are progressing.)

> There are a lot of Frisian organisations that want to maintain the Frisian
> language for the future (next to Dutch, Frisian is the second official
> language in The Netherlands), we even have a Frisian Academy!! But none of
> them were interested in supporting the project.

Again, this is very true.  Frisian seems to be in a very similar situation to 
Welsh (same number of speakers - 600,000, same role in public life and rural 
areas, same sort of cultural support).  I'm sure the "official" support 
structures are similar.  But I believe we can expect very little help from 
official bodies, for a number of reasons:
- it is my experience that the "public sector" is not really very appreciative 
of the "voluntary sector", even though they cannot afford to say so; we can't 
do much about this except present them with a fait accompli;
- official bodies tend to be conservative, in the sense that they are averse 
to taking risks, and will tend to stick with whatever "everyone else" is 
using - ideally, we need local pilot studies to show that minority language 
support is viable;
- there may be a significant unwillingness to accept that something put 
together locally in people's spare time and given away freely could actually 
be of more long-term value than something bought in from a large foreign 
company with marketing budgets in the millions - nothing can be done here 
except engage in a long education process, which is in fact getting easier as 
Linux makes its way up the application stack.

I therefore believe that the first necessity is to have something deliverable, 
ie KDE in the minority language on a running system.  The second is to try to 
plug in to the cultural bodies relating to the minority language, which is 
again, unfortunately, a long-term process.  But if you can demo something 
interesting to enough people, they will themselves start asking why this is 
not being supported.  In Wales, we are at the start of this process now, and 
over the next 4 months we want to create a liveCD using all the Welsh 
software translated so far, so that we can distribute this to show to as many 
schools and businesses as possible (I'm not sure yet where we'll get the 
money!).

A longer-term necessity is to see whether the existing legislative framework 
relating to the minority language gives it any special status.  For example, 
if any citizen is supposed to be able to use either language in their 
dealings with official bodies, should that not also apply to information 
terminals, school PCs used by children, or PCs used by minority language 
speakers in government offices?  If so, there may be a legal requirement to 
support systems that will provide an interface in either the minority or 
majority language, as opposed to systems that provide only the majority 
language.  You may have a language campaign group in Frisia which would be 
interested in following this up.

Of course, others may wake up to this too.  Microsoft, playing catch-up, has 
just announced that it will now provide a Welsh "skin" for WinXP and MSOffice 
:-)  It is entirely coincidental that this happens after significant progress 
on a Welsh desktop, as happened in other similar language areas (Nynorsk, 
Catalan, Maltese) ...  But of course it's good for the language!

-- 

Best wishes

Kevin Donnelly

www.kyfieithu.co.uk - Meddalwedd Rhydd yn Gymraeg
[prev in list] [next in list] [prev in thread] [next in thread]