[Foundation-l] WMF 2015 strategic plan and multilingualism

List overview All Threads
Download

newer

older

Re: [Foundation-l] Remarks on...

[Foundation-l] Raising funds...

Teofilo

5 Mar 2011 5 Mar '11

12:19 p.m.

http://wikimediafoundation.org/wiki/Template:BLPLang is not currently used at http://wikimediafoundation.org/wiki/Wikimedia_Movement_Strategic_Plan_Summa… This can be construed as the WMF wanting to reach the people of the world to provide educational contents AND English-dominate them. The fact that http://meta.wikimedia.org/wiki/Communications_subcommittees/Trans#Core_set_… is now marked as "obsolete" disappoints me. It seems to mean that multilingualism has been rejected. Can the notion that a key document like a strategic plan is ready for release when it exists in only one language be discussed ? Or is it already too late ? Has multilingualism definitely lost the game ? For example because most of the supporters of multilingualism have left the management sphere of WMF. If you look at Jay Walsh's user page on meta : http://meta.wikimedia.org/wiki/User:Jaywalsh you can find an indirect acknowledgement that Canada is a multilingual country. Is multilingualism worse off or better off in the Wikimedia Foundation than it is in Canada ? Should http://blog.wikimedia.org/ remain 100% English ? Why not have 1 or 2% of non-English with English translation ? 5 or 10% of English-with-some-translation ? Which degree of openness to non-English language should be shown on http://blog.wikimedia.org/ ? What is the purpose of linking to the blog from non-English main pages such as http://wikimediafoundation.org/wiki/Portada anyway ? Would it not be fairer to tell people "we have nothing pertinent in your language on this website. Please learn English first and come back. See you again" ? Shouldn't a number of English-only contents be moved to the USA, UK, Australia, etc. chapter websites ?

Show replies by date

Amir E. Aharoni

5 Mar 5 Mar

8:27 p.m.

2011/3/5 Teofilo <teofilowiki(a)gmail.com>om>:

...

The fact that http://meta.wikimedia.org/wiki/Communications_subcommittees/Trans#Core_set_… is now marked as "obsolete" disappoints me. It seems to mean that multilingualism has been rejected.

This is an interesting idea that should be revived. Put mildly, Wikimedia really shouldn't be focused on English. There are a lot of people who don't know English and don't plan to learn it; to paraphrase Larry Wall, it is not occasionally forgotten, but occasionally remembered. A UN-like model, with several major languages, into which important Foundation releases *must* be translated, is a realistic solution that will enable more people to read them. This, however, also poses the danger of perpetuating current linguistic conflicts. For example, translating the WMF blog into Chinese will allow a lot of people who know Chinese, but not English, read it, but it will yet again put Chinese above the regional languages of China; the same can be said about Russian, Spanish, French, Indonesian and other major languages. Nevertheless, done properly, it's better than staying English-only.

M. Williamson

9 p.m.

2011/3/5 Amir E. Aharoni <amir.aharoni(a)mail.huji.ac.il>il>:

...

will enable more people to read them. This, however, also poses the danger of perpetuating current linguistic conflicts. For example, translating the WMF blog into Chinese will allow a lot of people who know Chinese, but not English, read it, but it will yet again put Chinese above the regional languages of China; the same can be said about Russian, Spanish, French, Indonesian and other major languages.

Of course the same could be said of any option that does not translate everything into every language of the world which has monolingual speakers. This doesn't mean we should translate all WMF documents into Hopi, Kunama, Irish and Pirahã; that would be practically impossible, in fact even the most determined of organizations, missionary groups, have not achieved translation of anything into all languages with monolingual speakers. Yes, translating documents only into (for example) English, Spanish and French would leave out monolingual Quechua speakers, monolingual Basque speakers and monolingual Yoruba speakers, but it's simply not possible to reach every language, so decisions have to be made about which ones we are going to include. It is my hope that these decisions are data-driven - http://www.ethnologue.com/ethno_docs/distribution.asp?by=size and the size of Wikimedia communities speaking a language as well as the (in)frequency of bilingualism in those communities are a good place to start.

Casey Brown

9:27 p.m.

On Sat, Mar 5, 2011 at 4:00 PM, M. Williamson <node.ue(a)gmail.com> wrote:

...

It is my hope that these decisions are data-driven - http://www.ethnologue.com/ethno_docs/distribution.asp?by=size and the size of Wikimedia communities speaking a language as well as the (in)frequency of bilingualism in those communities are a good place to start.

Exactly. As far as I understand it, the language priority list *was* created based on these metrics. However, it hadn't been updated for years, and /that's/ why it was marked obsolete. That particular list was obsolete; no one was rejecting multilingualism or anything like that. It's actually the opposite: we didn't try to figure out how we should prioritize languages because we looked at them all as equal. All translation work is done by volunteers, and who were we to say "your language isn't as important, we'd rather you translate into X", especially if we hadn't really researched how to make those priority lists? If you translate something into Hopi, Kunama, Irish, or Pirahã, it'll get published just as quickly as if you translate something into French, Spanish, German, or Chinese. -- Casey Brown Cbrown1023

Amir E. Aharoni

9:47 p.m.

2011/3/5 Casey Brown <lists(a)caseybrown.org>rg>:

...

All translation work is done by volunteers, and who were we to say "your language isn't as important, we'd rather you translate into X", especially if we hadn't really researched how to make those priority lists? If you translate something into Hopi, Kunama, Irish, or Pirahã, it'll get published just as quickly as if you translate something into French, Spanish, German, or Chinese.

Yes, but in practice this creates strange situations, when a very active linguistic community, such as Catalan or Macedonian, has the messages translated, but major languages such as French or Russian are left untreated. This happened quite a lot during the latest Fundraising, for example. Of course this means that translators into Russian and French should be more active, but maybe some other model can be considered. This can be yet another topic for the discussion in the Language committee meeting in May :)

Gerard Meijssen

9:48 p.m.

Hoi, The notion that everyone working on Wikipedia and MediaWiki is a volunteer is a fallacy. The one thing I have been advocating is that the different languages and scripts are performing technically on a level playing field. This is not the case and there is a lot that can be achieved with modest investments. At this stage we do not want to invest in specific languages to create content. If a language is viable and can operate on a level playing field the communities will do their thing in the way that fits for them. Thanks, GerardM On 5 March 2011 22:27, Casey Brown <lists(a)caseybrown.org> wrote:

...

On Sat, Mar 5, 2011 at 4:00 PM, M. Williamson <node.ue(a)gmail.com> wrote:

David Gerard

9:58 p.m.

On 5 March 2011 21:48, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote:

...

The notion that everyone working on Wikipedia and MediaWiki is a volunteer is a fallacy. The one thing I have been advocating is that the different languages and scripts are performing technically on a level playing field. This is not the case and there is a lot that can be achieved with modest investments. At this stage we do not want to invest in specific languages to create content. If a language is viable and can operate on a level playing field the communities will do their thing in the way that fits for them.

Yes. The advocates of a minimal Foundation are missing the point that "in their own language" is an extremely good reason to spend money on the necessary translations and so forth. The Chapter structure is a brilliant way to get this sort of thing locally self-organising and not be run from San Francisco. But as Amir points out, this results in very patchy coverage. Really. Take the sentence: "Imagine a world in which every single human being can freely share in the sum of all knowledge. That's our commitment." - and think of how to fund everything that implies, or to fund its encouragement, or to fund encouraging the funding of its encouragement. The WMF as it stands is *tiny* for such a goal. - d.

Gerard Meijssen

10:06 p.m.

Hoi, The Wikimedia Foundation is a five hundred pound gorilla in the field of building language resources in the languages that have a smaller footprint on the Internet. I am convinced that for a million Euros we can make sure that all languages have technically a level playing field. I have proposed to spend 100,000,- Euro and this will make major improvements for the scripts, the fonts and the standards for the languages we have a Wikipedia for. This is given the current budget chicken feed. Thanks, GerardM On 5 March 2011 22:58, David Gerard <dgerard(a)gmail.com> wrote:

...

On 5 March 2011 21:48, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote:

The notion that everyone working on Wikipedia and MediaWiki is a

volunteer

is a fallacy. The one thing I have been advocating is that the different languages and scripts are performing technically on a level playing field. This is not

the

case and there is a lot that can be achieved with modest investments. At this stage we do not want to invest in specific languages to create

content.

If a language is viable and can operate on a level playing field the communities will do their thing in the way that fits for them.

M. Williamson

7 Mar 7 Mar

12:22 a.m.

Gerard, that seems like quite a bit of money, I'd be curious to know what exactly that would be spent on, in detail. 2011/3/5 Gerard Meijssen <gerard.meijssen(a)gmail.com>om>:

...

On 5 March 2011 21:48, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote:

The notion that everyone working on Wikipedia and MediaWiki is a

volunteer

is a fallacy. The one thing I have been advocating is that the different languages and scripts are performing technically on a level playing field. This is not

the

case and there is a lot that can be achieved with modest investments. At this stage we do not want to invest in specific languages to create

content.

If a language is viable and can operate on a level playing field the communities will do their thing in the way that fits for them.

_______________________________________________ foundation-l mailing list foundation-l(a)lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/foundation-l

Casey Brown

5 Mar 5 Mar

9:44 p.m.

On Sat, Mar 5, 2011 at 3:27 PM, Amir E. Aharoni <amir.aharoni(a)mail.huji.ac.il> wrote:

...

A UN-like model, with several major languages, into which important Foundation releases *must* be translated, is a realistic solution that will enable more people to read them. This, however, also poses the danger of perpetuating current linguistic conflicts. For example, translating the WMF blog into Chinese will allow a lot of people who know Chinese, but not English, read it, but it will yet again put Chinese above the regional languages of China; the same can be said about Russian, Spanish, French, Indonesian and other major languages. Nevertheless, done properly, it's better than staying English-only.

This is a very interesting idea. (My last post was mostly directed at the original poster's comments about us rejecting multilingualism, not your proposal. Sorry if I didn't make that clear!) I don't know if we should necessarily say that they "*must* be translated" into those languages, since we are volunteer-driven, but this could be more of a coordinator-thing... like we need to make sure that we keep volunteers on-hand and supported who speak these languages. If we see a gap in a certain language, we'll try to replace that person ASAP. We might want to try to revive translation teams, and make sure that we always have a well-staffed translation team with members ready to translate into these big languages. Anyway, I would love to have people working to make an updated priority list that they think we should use. :-) The metrics that Mark suggests are a great idea. Number of speakers, number of monolingual (or native) speakers, and size of the editing community would be great things to consider. -- Casey Brown Cbrown1023

M. Williamson

7 Mar 7 Mar

12:21 a.m.

...

priority list that they think we should use. :-) The metrics that Mark suggests are a great idea. Number of speakers, number of monolingual (or native) speakers, and size of the editing community would be great things to consider.

This may seem like nitpicking but I think it's an important consideration. Number of native speakers is not a good metric if we are trying to be practical, since the vast majority of Yoruba speakers on the Internet are fluent in English. English bilingualism is much more frequent in, say, the Indian internet community than it is in the Chinese internet community. I think we should take the following approach towards "crucial" languages: 1) Start with one language that all official documents must be in; due to current structure of WMF, this would be English. 2) To choose the subsequent language, search for the language with the greatest number of non-English speakers; for the sake of argument let's say it's Spanish (although it may be Chinese in reality). 3) For the third language, search for the language with the greatest number of people who do not speak English or language 2 (in this case, Spanish). 4) And so on and so forth. This cuts down on potential redundancy; of course if somebody wants to translate all documents into Irish or Basque, we will be happy to receive such translations. I do agree with the suggestion that we require documents be translated into a certain (small) group of languages. We do rely on volunteer translators, but we are not saying "you must translate this" but rather "we will not release this to the public until it has been translated to this language".

Samuel Klein

5:43 a.m.

Hi Teofilo, Thanks for raising all of these points. I agree that we should have our important essays and communication in multiple languages. And I was wishing myself that we had a multilingual blog planet that combined all of the monolingual ones... sometimes we segment ourselves by language in a way that just doesn't let us connect with one another. http://meta.wikimedia.org/wiki/Planet_Wikimedia#Other_languages

...

http://wikimediafoundation.org/wiki/Template:BLPLang is not currently used at http://wikimediafoundation.org/wiki/Wikimedia_Movement_Strategic_Plan_Summa…

In this particular case: the language templates on that wiki are added once multiple language versions of a doc are in place. "BLPLang" or "StaffLang" templates are custom to the BLP or Staff page. So the lack of a template just means no other language translations have been posted yet. On Sun, Mar 6, 2011 at 7:21 PM, M. Williamson <node.ue(a)gmail.com> wrote:

...

on the Internet are fluent in English. English bilingualism is much more frequent in, say, the Indian internet community than it is in the Chinese internet community. I think we should take the following approach towards "crucial" languages: 1) Start with one language that all official documents must be in; due to current structure of WMF, this would be English. 2) To choose the subsequent language, search for the language with the greatest number of non-English speakers; for the sake of argument let's say it's Spanish (although it may be Chinese in reality). 3) For the third language, search for the language with the greatest number of people who do not speak English or language 2 (in this case, Spanish). 4) And so on and so forth.

This was part of the thinking behind the original 'core languages' list. We also considered * how many different cultures/countries/regions could be reached with a language (not just raw # of primary- or secondary-language speakers) -- there are many low-population parts of the world that speak Spanish or French but not English or Chinese or Arabic, and * how many active Wikipedians there are, both readers and contributors, who prefer to read or write in that language. (Even if you and your friends can read English, if you are much more likely to forward a link to a Hindi-language essay, that is relevant to our community growth and balance.) We have slightly better data for some of this now, and perhaps different thoughts on how to balance the different criteria. .

...

This cuts down on potential redundancy; of course if somebody wants to translate all documents into Irish or Basque, we will be happy to receive such translations. I do agree with the suggestion that we require documents be translated into a certain (small) group of languages. We do rely on volunteer translators, but we are not saying "you must translate this" but rather "we will not release this to the public until it has been translated to this language".

Yes. es/fr/zh may still be a fair place to start - in terms of covering the world's regions and population by secondary-language. SJ

Jay Walsh

6:35 a.m.

On Mar 5, 2011, at 4:19 AM, Teofilo wrote:

...

Although we haven't discussed this much here, I'd like to add that one of the document/design production factors I've been thinking about a great deal is how to prepare information for broad localization in other areas. At a minimum we wanted to be sure the summary report could be localized on the wiki, which is why we simultaneously released it in PDF and back on the strategy wiki. I don't believe any localization has happened, but I know an open field when I see it :) Beyond just localization, I'd also like to add that we spent quite a bit of time simplifying the text in the summary carefully to remove un-necessary words/ideas or complex elements so it could be translated with more ease than a more complicated document. In a similar vein, I'd like to point out that the design concept proposed for Wikipedia 10 was very much developed to widely include all the languages of the world: http://ten.wikipedia.org/wiki/Design And over 100 logos were developed in scripts and languages from all over the world. Multilingualism is at the core of our design thinking, as is the concept that everyone be able to take simple graphic forms and turn them into their own works.

...

Should http://blog.wikimedia.org/ remain 100% English ? Why not have 1 or 2% of non-English with English translation ? 5 or 10% of English-with-some-translation ? Which degree of openness to non-English language should be shown on http://blog.wikimedia.org/ ? What is the purpose of linking to the blog from non-English main pages such as http://wikimediafoundation.org/wiki/Portada anyway ?

I'm very much open to a multilingual blogging space in time. Right now we're working on a system to expand the wikimedia blogs into a few channels, and to open the door for a space with potentially many different channels using different languages - including perhaps reports and info from chapters in their own language and hopefully also in english.

...

Would it not be fairer to tell people "we have nothing pertinent in your language on this website. Please learn English first and come back. See you again" ?

There's no simple answer for how to accommodate many languages. Do we start with 3? Which ones? All? I don't know - but at a minimum I'm glad to say that we factor into our production and writing the ideas of simple, easy to comprehend text that can be understood by those for whom English isn't a first language, and text that lends itself to easier translation. It doesn't overcome the challenge, but it's an important step. -- Jay Walsh Head of Communications WikimediaFoundation.org blog.wikimedia.org +1 (415) 839 6885 x 6609, @jansonw

Samuel Klein

5:28 p.m.

On Mon, Mar 7, 2011 at 1:35 AM, Jay Walsh <jwalsh(a)wikimedia.org> wrote:

...

Multilingualism is at the core of our design thinking, as is the concept that everyone be able to take simple graphic forms and turn them into their own works.

That's a nice way of putting it. Has anyone thought of making icons for the core menu items and interface strings? I know that some of the main page redesign discussions have looked into simple icons for each of the top-level categories or sections. That would both help with people who aren't very literate yet and with people who are trying to navigate MediaWiki (whether on our projets or elsewhere) in a language not their own.

...

+1 to this, and Jay's whole post. SJ

4813

days inactive

4815

days old

wikimedia-l@lists.wikimedia.org

Manage subscription

13 comments

8 participants

tags (0)

participants (8)

Amir E. Aharoni
Casey Brown
David Gerard
Gerard Meijssen
Jay Walsh
M. Williamson
Samuel Klein
Teofilo