Meta talk:Language proposal policy/Archives/2007

From Meta, a Wikimedia project coordination wiki

Conflicting invented codes

The following text, relevant to invented language codes for languages without a standard ISO code, was added to the main page by at 01:30, 9 January 2007.

Since RFC4646 was published, the extension code after a general 3-letter code will be wrong as it conflicts with "extlang" codes which will be normalized along with the final release of ISO 639-3. In other words "fiu-vro" is an incorrect example, because "vro" will conflict violate ISO 639-3 (and future standardization of "extlang" codes in the next revision of RFC4646). So only "fiu-voroo" shuold have been used. All language code subtags that are not conforming to an existing ISO 639 language code (3 letters) or ISO 3166 country/region code (2 letters), or ISO 16924 script code (4 letters), MUST be at least 5 characters long to qualify as a valid "variant" code!

So only "fiu-voroo" should have been created, or "x-voro" (using an unrestricted private code after the "x-" prefix). Please reread RFC 4646 (and its accompanying RFC 4647 for parsing and handling locale codes) !

That's an interesting problem, but the language subcommittee will probably soon decide that languages must have a standard ISO code before being proposed (information will be provided on having one created). —{admin} Pathoschild 17:19, 9 January 2007 (UTC)
What will happen to those language editions which have already been created despite lacking an ISO codes, in case there are any? —Nightstallion (?) 19:20, 9 January 2007 (UTC)
This has been addressed in the new draft, which now states:
2. The language should have a valid ISO-639 (search) or BCP 47 (list) code.
If there is no valid ISO-639 or RFC 4646 code, it should be a natural language or a well-established constructed language. The Wikimedia Foundation does not seek to develop new linguistic entities; there must be an extensive body of works in that language. The information that distinguishes this language from another should be sufficient to convince standards organizations to create an ISO-639 or BCP 47 code.
What is done with existing projects having a non-standard code is beyond the scope of the language subcommittee. —{admin} Pathoschild 23:02:43, 26 February 2007 (UTC)


If a guidance how (or under which condition) to set up a test wiki on incubator wiki, it would be useful. Perhaps it could be a part of FAQ. --Aphaia 07:51, 10 January 2007 (UTC)

I've added the information in the Frequently asked questions section. —{admin} Pathoschild 00:27, 11 January 2007 (UTC)

Multiple written forms for same language

Different written forms of the same basic language should be included on the same wiki, possibly on different pages.

I question whether this is really good policy, and suggest that in some cases each written form would be better off in its own wiki, and even be better able to collaborate that way.

In each of the two cases I am familiar with, namely Ladino & Kurdish, the default written form of the wiki (system messages, page direction) is totally dominant on the wiki. In each case, the less dominant written form (on the wiki, not less dominant in real life) has very few pages and almost no activity.

It may be that interwiki links between parallel articles in different scripts are far more efficient (and better common sense) than other manual systems of linking back and forth on the same wiki.

Also, in the examples above (Kurdish & Ladino) we don't even have the technical ability to uphold this policy and have them on the same wiki, because MediaWiki only supports LTR or RTL pages on the same wiki, not both. In both cases, only articles in the LTR script have taken hold on the LTR wikis.

In these examples, it is also important to note that the two different alphabets are so unlike each other that there is no possibility of automatic conversion between them (unlike what was successfully done for Chinese). A different example, however would probably belong on a single wiki, namely Aramaic. Here we have a single language with a single alphabet for all varieties (all use the same 22 letters), and is always RTL, but different groups use different scripts for those same 22 letters. This is an obvious candidate for automatic conversion, and should certainly be on a single wiki.

I instead suggest that we be agnostic about this, and let the users themselves decide whether such languages should have their multiple alphabets sharing the same wiki or in parallel wikis.Dovi 05:23, 15 February 2007 (UTC)

This was addressed in the new draft of the policy, which states:
3. The language must be sufficiently unique that it could not coexist on a more general wiki. In most cases, this excludes regional dialects and different written forms of the same language.
The degree of difference required is considered on a case-by-case basis. The subcommittee does not consider political differences, since the Wikimedia Foundation's goal is to give every single person free, unbiased access to the sum of all human knowledge, rather than information from the viewpoint of individual political communities.
{admin} Pathoschild 23:02:28, 26 February 2007 (UTC)

GerardM-Pathoschild draft

The following discussions were posted to the GerardM-Pathoschild draft discussion page (permanent link to this draft).

Sounds fine

This version sounds quite fine to me. —Nightstallion (?) 16:05, 20 January 2007 (UTC)

Me too... Toira 00:48, 26 January 2007 (UTC)


I assume that this will apply to all language-divided projects? Just confirming: at the moment Wikisource and Wikinews seem to have their own methods. Also, would it not make more sense to locate the approval discussions on Incubator rather than Meta, in the interests of keeping it centralised? Other than that I have no objections. Dbmag9 13:10, 21 February 2007 (UTC)

Hello Dbmag2. This should theoretically apply to all projects, which means that the individual processes will be merged into this one when we've ironed out the glitches. It will be much easier to do after we demonstrate the efficiency of this process and modify the policy as needed to integrate local policies' ideas.
I'm not sure whether or not we should move the process to the Incubator, since it falls within both wikis' scopes. That's something we should discuss with the Incubator community, particularly if they've made any progress in drafting a policy. —Pathoschild 21:02:25, 21 February 2007 (UTC)
Draft Incubator policy at incubator:Incubator:Policy/Draft. That's just something I've knocked up in a few minutes based on this proposal. Hopefully it will be a useful starting point for discussion. Dbmag9 19:52, 22 February 2007 (UTC)
I've tweaked and expanded it a bit (unofficially). :) —Pathoschild 02:02:39, 23 February 2007 (UTC)

Hmmm... Local project policies for new languages are one thing, and there may be reasons for or against having special project policies. But a totally separate question is location: My own feeling is that the local "project-specific" incubators new languages at Wikisource and Wikiversity are better environments than the generic incubator wiki (for all new projects and all new languages). Much more intuitive, user friendly, and easier to keep track of would be new languages for existing projects at a project-specific wiki, and new projects (in all of their languages) at Dovi 18:24, 27 February 2007 (UTC)

I have to support Dovi. Sure, it is possible to discuss improvements, to unify the policy etc., but it is quite different to prepare some people to work in a Wikipedia and to prepare them to work in a Wikisource. And only in the environment of a general Wikisource project (i.e. oldwikisource) it will be possible to show this. But, see the new thread on this question at oldwikisource: Wikisource:Scriptorium#Wikisource:Language domain requests versus m:Meta:Language proposal policy. Thx, -jkb- 09:42, 3 April 2007 (UTC)


Hi, this proposal looks good to me - it's obviously ambiguous in parts (eg. in relation to what constitutes a dialect etc), but it's probably hard to be definitive in a general sense with such matters. I've made a minor change to language - nothing substantial.

My main query would be in relation to new language incubation. In Wikiversity, new languages are currently incubated at the beta Wikiversity project instead of the incubator wiki - Wikisource also has its own similar procedure. This policy requires that a test project be set up on the incubator wiki before or immediately after conditional approval for the language domain. I'm wondering if this policy should reflect the different procedures that these projects operate under - or does this policy specify that these procedures be standardised and on the incubator wiki? Thanks. Cormaggio 18:48, 23 February 2007 (UTC)

The procedures will eventually be standardized, but they'll coexist temporarily while we're ironing out the standardized procedure and streamlining the process. —Pathoschild 21:02:44, 23 February 2007 (UTC)

Already-Existant Languages

Someone raised this on the mailing list (forgive me for being too lazy to find out who). I assume that languages which already have a project go through the same procedure, but with acceptance for the language itself implicit: i.e. they still make a test at Incubator, and the langcom still checks if there are enough people, but they do not need to discuss whether the language is suitable. Dbmag9 19:52, 25 February 2007 (UTC)

Linguistic considerations aren't as important if there is already a flourishing project in the language. However, they would regain some importance if existing projects in that language are underdeveloped, controversial, or redundant (such as some projects in dialects). —Pathoschild 06:02:59, 26 February 2007 (UTC)

List by creation

Do we have a timeline of projects (particularly Wikipedias) by date of creation? Thanks.--Pharos 23:34, 6 June 2007 (UTC)

For the Wikipedias, you can see en:Wikipedia:Multilingual coordination. It isn't a timeline, but you can find the dates there. SPQRobin 23:26, 8 July 2007 (UTC)

Question: Classical Mongolian

How should one go about to open a Classical Mongolian Wikipedia? This is a different writing system and is not really a separate language (may be a dialect), but the script is used in Inner Mongolia where almost a majority of Mongolians live. The current version of Mongolian Wikipedia is in Cyrillic, which is used in Mongolia. These two scrpits are completely different and there are very few people who know both.

Just follow the procedure on the other side of this page. And be sure to link to here because it shows the form of Mongolian used in Inner Mongolia has a separate ISO code (it's "mvf"), which lends it some easily-recognizable status as a separate entity. Also be sure you're ready to discuss the extent of differences in these two standards of the w:Mongolian language, and how significant these differences are apart from the writing system. Because it might be possible to develop an automatic conversion system as has been done with w:Traditional Chinese characters and w:Simplified Chinese characters. Good luck.--Pharos 06:11, 16 June 2007 (UTC)

Thank you vey much. That automatic conversion idea is very nice one. I think the software complexity would have to be considerably high, as it should transliterate at least word by word. There are also other subtle differences in the languages (or dialects). But if there is such a conversion system, that would be great! Temur 08:32, 16 June 2007 (UTC)

That is right. It is like arabic script turned 90 degrees anticlockwise. Temur 22:25, 2 July 2007 (UTC)

For the Mongolian used in Inner Mongolia, it's considered as "Peripheral" Mongolian(mvf) by westerners. But it's not peripheral at all, so we should use the address rather than Using the "Peripheral" Mongolian (Hohhot dialect). Otherwise, move to --虞海 (Yú Hǎi) 06:54, 7 March 2009 (UTC)

Related projects

test wiki

"Anyone can create a test project at any time after opening a new request." but in the propsal summary example, there is a link "development wiki project". So anyone can create a test project at any time? MF-Warburg(de) 14:00, 19 July 2007 (UTC)

I corrected the page, since the Incubator doesn't require an open request. Note that the summary box is routinely updated after the request is opened, though. —{admin} Pathoschild 18:46:20, 19 July 2007 (UTC)

Does a test project in the incubator require a permission by langcom?

At Requests for new languages/Wikipedia Aeres, User:GerardM writes:

To go to the incubator stage you need a go ahead from the language committee. You need an ISO 639 code. There is none. GerardM 19:21, 13 July 2007 (UTC)[1].

I fail to find such a provision in the Language proposal policy. Quite the contrary:

  • When and how are test projects created?
    Anyone can create a test project at any time. For more information, see the Incubator wiki's main page.

Did I miss something? There are test projects created almost daily. I haven't seen any prior "go ahead" at Requests for new languages. --Johannes Rohr 10:35, 1 August 2007 (UTC)

No, you didn't miss something. Such a provision doesn't exist, see also #test wiki and incubator:Incubator:Community_Portal#Starting_tests. --MF-Warburg(de) 11:14, 1 August 2007 (UTC)
Agreed, no approval is needed for a test project. This was discussed extensively by the subcommittee without consensus, with GerardM arguing in favour of subcommittee control of the Incubator. He tends to put forth his own opinions as policy. However, the test project will not lead to a new Wikipedia until the requirements (like an ISO 639 code) are met. There are no exceptions to this. —{admin} Pathoschild 19:52:13, 01 August 2007 (UTC)

Can the langcom take a look on this proposal? This is a group effort (example) with at least 400 pages (and growing...), more active than a dozen of Wikisource wikis. 555 21:31, 15 August 2007 (UTC)

It's been conditionally approved. —{admin} Pathoschild 02:24:19, 12 October 2007 (UTC)
Why only conditionally? Localisation is present, try . What else is required? --Johannes Rohr 17:32, 12 October 2007 (UTC)
There are 815 untranslated interface messages. Some of them may be optional (as described on the request's talk page), but many aren't. —{admin} Pathoschild 18:29:35, 12 October 2007 (UTC)


There is a discussion at en:Wikipedia:User_categories_for_discussion#Category:User_als (stable link) about als:, the Alemannic German Wikipedia. This project is an exceptional case in that it does not concern a language that has no official code, but a group of mutually intelligible dialects, which individually have ISO codes (wae, gsw, swg[2]). Unfortunately, the "als" subdomain chosen (originally for "Alsatian", a subdialect of gsw) has been assigned to Tosk Albanian by ISO 639-3 in 2007, so that this wiki now isn't at an undefined subdomain, but at a well-defined one that designates a completely different language. What can be done about this? Suggestions include relocation of als: to "de-x-als" or "de-x-alemannisch" or "gsw-swg-wae" or similar: is this a possibility? (the als: community would have to be asked to choose their preference, of course). --Dbachmann 12:24, 11 October 2007 (UTC)

Final approval for bcl, now

Thank you.

Latin Wikipedia

See also: Latina wikipedia closing and hellenic wikipedia opening

According to this page, Wikipedia projects in ancient languages are not accepted, as 'there is no living native community to use the resources'. Why, then, does the Latin Wikipedia exist? Is there a native Latin-speaking community somewhere I'm not aware of? Terraxos 20:32, 21 November 2007 (UTC)

Requests predating the language subcommittee were determined by community vote, which were then submitted to Wikimedia developers who arbitrarily created wikis in batches. This caused several problems, such as political repression by out-voting, attempts to falsify votes by creating many accounts per person, empty wikis that even today attract vandalism, spam, and bias, takeover of a small wiki by another language group or a group of friends, and so forth. The language subcommittee was formed to prevent these problem in future requests; dealing with existing wikis is beyond their current authority (that is up to the community and Board of Trustees). Thus far, no approved wiki has had such issues. —{admin} Pathoschild 04:04:19, 22 November 2007 (UTC)
Also Latin is used by the Vatican. Consequently it is definitely a classical language but it is not a dead language. It is also not indicated as such by Ethnologue. GerardM 09:11, 22 November 2007 (UTC)
I think Gerard has hit the nail on the head with the difference between a 'dead' language and a 'classical' language. We might consider a specific exception for classical languages (even without "living native communities"), though this probably isn't necessary as every major classical language it seems already has a Wikipedia. Really, the issues facing a Latin Wikipedia are -very similar- to the issues facing an Esperanto Wikipedia.--Pharos 20:13, 23 November 2007 (UTC)
There is a difference. The issue is that the wikipedias that do exist are not considered really by the language committee. New requests for languages are. When a classical language is a dead language, I do not give it much chance. GerardM 22:30, 23 November 2007 (UTC)

Thanks for the explanations, guys. I agree that there is value in having a Latin Wikipedia - I was just confused by the fact it appeared to flatly contradict this language policy.

Is there a list somewhere of which Wikipedias were created before the language proposal policy, and which after? I have some suspicions - e.g., I find it very hard to see how Volapuk could have possibly passed these criteria - but I would be interested to see the list. Terraxos 03:51, 1 December 2007 (UTC)

Volapuk is not a dead language. For a constructed language it is reasonable to expect that modern concepts can dealt with. So I wonder why you think that a language like Volapuk would not qualify. Thanks, GerardM 23:16, 1 December 2007 (UTC)
Actually, I take back my comment about Volapuk above. I was unconvinced it deserved a Wikipedia because I was unable to find its ISO code, but I've since discovered it does have one (code:vol) and I just wasn't searching the list properly. I apologise for my unfair criticisms above. Terraxos 19:51, 12 December 2007 (UTC)

Is there any actual reason behind the "no ancient languages" rule? I mean, there's Old English, Gothic, Old Church Slavonic and Latin, but no Ancient Greek or Sanskrit. It just seems wholly arbitrary and without reason. 01:00, 25 March 2008 (UTC)

The one reason why no dead languages is because of the need to change the language in order to write a modern encyclopaedia. By writing an encyclopaedia it is no longer the language it is said to be. Thanks, GerardM 11:14, 26 March 2008 (UTC)

There is a Sanskrit wikipedia (besides, some people would argue that there are still some native speakers of Sanskrit left). The only major classical language left out is Ancient Greek. One would suppose that the success of the Latin Wikipedia would refute the arguments offered by the 'subcommittee' on why classical languages should not have their own wikipedias. But alas, a narrow doctrine, which I find hard to reconcile with the creative, liberal, and democratic spirit in which Wikipedia was founded, has prevailed.-- 17:36, 29 March 2008 (UTC)

I agree that the cancellation of the Ancient Greek Wikipedia, just shortly after it was tentatively approved and after people had been working on translating the interface for months, was wrong. Notice: Had the interface translation not been a prerequisite (also a mistake by the language committee in my opinion) it would already be up and running!

Ancient Greek is studied and taught by thousands of academics worldwide, and by tens of thousands (if not more) of enthusiasts. It is an essential part of a classical education, and one of the basic underpinnings of western culture. It can support a Wikipedia the same way as Latin and Sanskrit, and equally deserves one.

GerardM wrote above:

The one reason why no dead languages is because of the need to change the language in order to write a modern encyclopaedia.

What is wrong with adapting a language to meet the needs of a Wikipedia? Nothing at all. All languages do so (even languages like English and French), and certainly languages like Latin and Sanskrit. Actually, in the past I read some of the wonderful archived discussions about how to do this for Ancient Greek, all of which were written in a spirit of high-level cooperation that included some people who obviously had academic backgrounds.

The cancellation of the Ancient Greek Wikipedia will be a loss to Wikimedia in general, and I hope the decision will be changed. In general, the process of supporting new languages seems to have become far more restrictive. I wonder how some new voices and differing perspectives can be brought into the language committee. Is there a way to apply for membership? Dovi 18:26, 29 March 2008 (UTC)

Latin and Esperanto are not unreasonable, as it's a language that's actually used for communication of modern concepts in the real world. There's a case for Sanskrit. But I fail to see what the point of an Ancient Greek Wikipedia is, and why Wikimedia should support that point. It's not a language that any two people would actually use to communicate with in the real world, so having a Wikipedia in it is just a game for people who want to play around with Ancient Greek.--Prosfilaes 22:35, 27 April 2008 (UTC)
there is no difference among ancient greek, latin or sanskrit. all they are extinct languages with some modern uses. sometimes ancient greek is more useful than latin, and the contrary.
I agree. There is a practical difference between the three; Latin and Sanskrit Wikipedia existed before the start of the language committee and are therefore outside of its remit. Ancient Greek is not. GerardM 07:06, 30 April 2008 (UTC)
This was discussed extensively on the Foundation mailing list. As one of the foundational languages of Western culture it is nearly inconceivable that Old Greek would cease being used for communication and concepts. And indeed it never did (though it does so to a lesser degree than Latin). Furthermore, even if it had does not justify denying it a wiki. Latin and Greek are not, and never have been "extinct" languages. It is more justifiable to call them "classical" languages, i.e. languages that continue to have influence and lives of their own.
It would be a disgrace for WP to have a solid place for languages like Esperanto and Volapuk, but none for classical languages like Greek and Latin. (Not that I have anything against the Esperanto and Volapuk wikis, and I wish them success too.)
That Latin and Sanskrit were approved before the committee is a very poor excuse for keeping them. Basically you are saying they shouldn't be here, but we already goofed... Whoops... That doesn't speak too well for their future success, and for their being an accepted part of the WM family. What happens when there is a new WM project relevant to these languages? Will the LC then say "no," or will it say that the language was already (wrongly) legitimized before the LC began?
Also note that Greek WP was already approved before the language committee suddenly changed its policy, and after people had been working for months under the impression of that very approval by the language committee.
Nothing I wrote above is new; all has been stated at length by many people on the mailing list this past month. I am sorry that the LC has taken such an extraordinarily narrow view on classical languages. Dovi 12:59, 1 May 2008 (UTC)
Latin is a moot point. I fail to see why it's a disgrace to have an article in a language that is actually used to communicate in the real world (see, for example, Monato) and not in a language that's not so used. It would be impressive if the people who want an Ancient Greek wiki could show any evidence that it was actually used in the real world, unlike Latin and Esperanto.--Prosfilaes 23:12, 2 May 2008 (UTC)
See the Wikimedia foundation forum. there is a large discussion about this. latin is closely to ancient greek. and esperanto to another artificial languages like interlingua. but nothing of this is comparable with any modern languages, linguist can say that any modern languages are better to communicate in the real world than latin or esperanto. classics and artificial languages have the same justifications to exist. if we are extremists with your viewpoint (Prosfilaes) we should close all the wikipedias in any classic or artificial languages. all of them are inevitably linked. their reason for having a wikipedia.
Obviously I failed to make clear that I consider evidence that the language is actually used to communicate real-life issues, as shown for Esperanto by the magazine w:Monato, to be sufficient. Esperanto clearly passes the test, as I believe does Latin. In this manner, Latin and Ancient Greek aren't very closely related at all.--Prosfilaes 00:43, 5 May 2008 (UTC)
In what form latin pass the test? according to your?
  • Well the main problem with old languages is that they have no words for modern inventions, political terms etc, so they should be invented. This is not the case of Latin, this language has all words one could need for modern Wikipedia, Latin dictionarier contain words even for vacuum cleaners and microwave ovens. Latin radio stations discuss modern political topics.
  • Contrary, small living languages with native speakers can experience the same problems. One can even find than it is difficult sometimes explain some complicated mathematical topic in Ukrainian, not to say for example Abkhaz, Chuvash or Udmurt languages, so the terms sometimes inventer 'on air'. --Nxx 19:16, 17 August 2008 (UTC)