What is the smallest community that can support a wikipedia?

List overview All Threads
Download

newer

older

Suggestion to reject the proposal...

Creole Spanish rejected

Asaf Bartov

26 Jan 2017 26 Jan '17

11:41 p.m.

Hello. The question in the subject line, asked by the Aramaic expert on the Assyrian thread, has been floating in space (and making me curious) for years. To my knowledge, we have never had a good answer. So I'm taking the opportunity to attempt discussion of it. It seems to me it would be good to get at some approximation of an answer. For example, Milos just mentioned Cora, an indigenous Mexican language with about ~10,000 speakers. Thinking of a *Wikipedia* in that language seems to me a complete waste of time. Statistically, it would seem it could never recruit more than a handful of volunteers, and it would not have a reader base, nor ever offer even a modest genuinely-useful corpus of up-to-date, encyclopedic knowledge. It makes absolute sense to document the Cora lexicon (on major Wiktionary projects, i.e. in other languages), to curate any extant literature (on the multilingual Wikisource), to record and document live speakers (and any folklore) on Commons, etc. But I think this language won't ever achieve an encyclopedia, and I think it is unhelpful to pretend otherwise. You may disagree, perhaps. What I am interested in hearing the committee's opinion about is the general question: can we identify the criteria for a minimally-viable Wikipedia? I will take a shot at a very rough, partly arbitrary definition of "minimally-viable Wikipedia": a wiki community commanding sustained participation from at least 5 very active editors and at least 20 active editors, and able to reach 20,000 non-stub articles in under 10 years. (many other definitions can be offered.) It seems clear, for example, that 1 million literate speakers of high average education level, stable orthography, and available secondary sources and higher education in that language (e.g. Estonian) are definitely enough to sustain such a community. But there's still a lot of room to ponder -- would 500,000 speakers also be enough, provided the other characteristics are in place? Would 10 million speakers be enough, if there's no higher education or secondary sources in a given language? Etc. etc. Langcom is probably the densest concentration of expertise able to approach this question. Is the committee interested in thinking about it and maybe working towards some working recommendation/guideline? (I don't think it necessarily has to result in any policy change for LangCom. It may just be a useful guideline for interested volunteers/communities to compare themselves with, for example.) Cheers, A.

Attachments:

attachment.htm (text/html — 2.8 KB)

Show replies by date

Milos Rancic

27 Jan 27 Jan

12:31 a.m.

There are a couple of issues here... First and most important, I think that Language committee should maintain eligibility for a language as the rule, provided that there is at least one native speaker interested in working on Wikimedia projects. We shouldn't demotivate people because of the size of the population speaking their native language. ("On hold" is mostly about such languages.) The costs of supporting a project are proportional with its usefulness: less useful, less traffic, less CPU, less RAM; more useful, more traffic, more CPU, more RAM... So, it's not about if Wikimedia could or couldn't support it. Imagine a tribe of 20 people in the rainforests of New Guinea, but close enough to be able to get a computer and internet connection. They likely speak their own language. And one person there is willing to use Wikipedia as a tool to make children literate in their own language. That person has to pass a lot of obstacles: making her/himself literate likely in Tok Pisin or Indonesian, Internet savvy, to invent the way how to write their native language and to convince others that literacy is a good thing. We shouldn't make obstacles to such person. It is not likely that something exactly like that would happen -- at least not soon --, but it's about our principle. The other very important thing is that our main brand is Wikipedia, not Wiktionary, not Wikisource. People want to have Wikipedia in their languages, not other projects. In the cases like Estonian is, we know that we'll find there a lot of useful materials. In the cases of any non-first-world-country we will find tons of quite problematic materials, no matter even of the size of the population. I think we shouldn't be strict when the native population is very small; and that we should use our main brand to gather a little bit more knowledge, written in a language spoken today, but not in 50 years. Wikipedia is influencing cultures. As language is spoken with smaller number of people, as more Wikipedia influences the language and the culture, both. It could turn out that Wikipedia actually made that language to survive; actually, I think Wikipedia is the main tool for small languages to survive. The number of speakers limits are very questionable. It could be about a large number of speakers (in millions, maybe even more) who don't have positive attitude towards their own language. The languages like those are not going to survive and it's not likely that they would even ask for Wikipedia in their language. At the other side, it could be about much smaller number of speakers, with population having strong positive attitude and willing to work on it (Scottish Gaelic has less than 100,000 speakers). It could be even about so called "shifting" languages, which just 30 years ago didn't have good chances to survive 21st century, but they experienced revival (Welsh). Ethnologue says that there are more than 5000 languages up to 6a "vigorous" status. All of those languages will survive 21st century and a lot of them are below 10,000 speakers (more than 2000). I simply don't think that we should be giving any suggestion from the position of power. Our suggestion "It would be better if you'd use English Wiktionary or Multilingual Wikisource" would be interpreted as an order. I think we should speak with them after we see they started working on the projects of their choice. On Thu, Jan 26, 2017 at 11:41 PM, Asaf Bartov <abartov(a)wikimedia.org> wrote:

...

Asaf Bartov

12:48 a.m.

I had hoped to head off the arguments you bring in your first few paragraphs, Milos, by explicitly saying I'm not even suggesting any change in policy. So yes, language will remain *eligible* etc. according to the committee's policy, and the vanishingly-unlikely scenario you describe can still take place. I think it is *separately* worthwhile to try and have some approximation of an answer to the question about rough minimal conditions for *a useful encyclopedia* in a language, *as distinct* from *a language preservation vehicle*. FWIW, I agree Wikipedia is attractive for that, and if I were trying to preserve my endangered ancestral language, I would certainly build a wiki, and probably try to build a Wikipedia (or a Wikinews-for-children, perhaps). But again, leaving aside the usefulness of Wikipedia for language preservation, and leaving aside *any* thought of changing policy, I am interested in whether the committee is interested in thinking about the question in my middle paragraph, above. A. On Thu, Jan 26, 2017 at 3:32 PM Milos Rancic <millosh(a)gmail.com> wrote:

...

the

opportunity to attempt discussion of it. It seems to me it would be good to get at some approximation of an

answer.

For example, Milos just mentioned Cora, an indigenous Mexican language

with

about ~10,000 speakers. Thinking of a *Wikipedia* in that language

seems to

me a complete waste of time. Statistically, it would seem it could never recruit more than a handful of volunteers, and it would not have a reader base, nor ever offer even a modest genuinely-useful corpus of up-to-date, encyclopedic knowledge. It makes absolute sense to document the Cora lexicon (on major Wiktionary projects, i.e. in other languages), to curate any extant literature (on

the

multilingual Wikisource), to record and document live speakers (and any folklore) on Commons, etc. But I think this language won't ever achieve

encyclopedia, and I think it is unhelpful to pretend otherwise. You may disagree, perhaps. What I am interested in hearing the

committee's

opinion about is the general question: can we identify the criteria for a minimally-viable Wikipedia? I will take a shot at a very rough, partly arbitrary definition of "minimally-viable Wikipedia": a wiki community commanding sustained participation from at least 5 very active editors and at least 20 active editors, and able to reach 20,000 non-stub articles in under 10 years. (many other definitions can be offered.) It seems clear, for example, that 1 million literate speakers of high average education level, stable orthography, and available secondary

sources

and higher education in that language (e.g. Estonian) are definitely

enough

to sustain such a community. But there's still a lot of room to ponder -- would 500,000 speakers also

enough, provided the other characteristics are in place? Would 10

million

speakers be enough, if there's no higher education or secondary sources

in a

given language? Etc. etc. Langcom is probably the densest concentration of expertise able to

approach

this question. Is the committee interested in thinking about it and

maybe

working towards some working recommendation/guideline? (I don't think it necessarily has to result in any policy change for LangCom. It may just be a useful guideline for interested volunteers/communities to compare themselves with, for example.) Cheers, A. _______________________________________________ Langcom mailing list Langcom(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/langcom

_______________________________________________ Langcom mailing list Langcom(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/langcom

Gerard Meijssen

7:40 a.m.

Hoi, What is the purpose of answering that question when it makes no difference in the way we work. Why would we consider it, what is the point. Asking for academic reasons? The reason why we have a language policy is to prevent the random creation of new projects. It is a blunt instrument in the way we practice it. In its practice we have two levels of prevention. Not so much for the first project and a lot for subsequent projects. In effect it is why there so few new Wikis for other projects. In the practice of the language policy we hardly use our discretion. We could allow for new Wikisources for instance when an organisation indicates that it is important for them and that they will see to specific named aspects of its operation. I take Wikisource as an example because it is mostly an environment for editors and not so much for readers. Discovery of fully developed books are not easy within the Wikimedia environment; it is used externally using an API... My point is, we have a good reason for what we do and by opening up this can of worms, we will open up for being more restrictive in the creation of new projects in the future. There is no need nor reason to do so. Quite the contrary, personally I feel we should be more open to Wikisource and really support its processes and its finished works. Thanks, GerardM On 27 January 2017 at 00:48, Asaf Bartov <abartov(a)wikimedia.org> wrote:

...

taking the

opportunity to attempt discussion of it. It seems to me it would be good to get at some approximation of an

answer.

For example, Milos just mentioned Cora, an indigenous Mexican language

with

about ~10,000 speakers. Thinking of a *Wikipedia* in that language

seems to

me a complete waste of time. Statistically, it would seem it could

never

recruit more than a handful of volunteers, and it would not have a

reader

base, nor ever offer even a modest genuinely-useful corpus of

up-to-date,

encyclopedic knowledge. It makes absolute sense to document the Cora lexicon (on major

Wiktionary

projects, i.e. in other languages), to curate any extant literature (on

the

multilingual Wikisource), to record and document live speakers (and any folklore) on Commons, etc. But I think this language won't ever

achieve an

encyclopedia, and I think it is unhelpful to pretend otherwise. You may disagree, perhaps. What I am interested in hearing the

committee's

opinion about is the general question: can we identify the criteria for

minimally-viable Wikipedia? I will take a shot at a very rough, partly arbitrary definition of "minimally-viable Wikipedia": a wiki community commanding sustained participation from at least 5 very active editors and at least 20 active editors, and able to reach 20,000 non-stub articles in under 10 years. (many other definitions can be offered.) It seems clear, for example, that 1 million literate speakers of high average education level, stable orthography, and available secondary

sources

and higher education in that language (e.g. Estonian) are definitely

enough

to sustain such a community. But there's still a lot of room to ponder -- would 500,000 speakers

also be

enough, provided the other characteristics are in place? Would 10

million

speakers be enough, if there's no higher education or secondary sources

in a

given language? Etc. etc. Langcom is probably the densest concentration of expertise able to

approach

this question. Is the committee interested in thinking about it and

maybe

_______________________________________________ Langcom mailing list Langcom(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/langcom

Milos Rancic

4:51 p.m.

On Fri, Jan 27, 2017 at 7:40 AM, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote:

...

What is the purpose of answering that question when it makes no difference in the way we work. Why would we consider it, what is the point. Asking for academic reasons?

Asaf, I agree with Gerard's position in relation to the purpose of the question/answer. May you clarify which practical implications you want to achieve with this question? If it's about the set of suggestions for people trying to preserve their languages, I suggest a very cautious approach just after they get what they want. Yes, it seems more rational to push them to make a good bilingual dictionary, but we have to be careful not to alienate them. If you had in your mind something else, please let us know.

Gerard Meijssen

5 p.m.

Hoi, There is little point at this time in a dictionary. With work going on for a Wikidatification of Wiktionary it is a real bad investment. There is much more to be gained by having books digitised. Thanks, GerardM On 27 January 2017 at 16:51, Milos Rancic <millosh(a)gmail.com> wrote:

...

On Fri, Jan 27, 2017 at 7:40 AM, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote:

What is the purpose of answering that question when it makes no

difference

in the way we work. Why would we consider it, what is the point. Asking

for

academic reasons?

Milos Rancic

6:53 p.m.

On Fri, Jan 27, 2017 at 5:00 PM, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote:

...

There is little point at this time in a dictionary. With work going on for a Wikidatification of Wiktionary it is a real bad investment. There is much more to be gained by having books digitised.

Antony Green

6:56 p.m.

New subject: What is the smallest community that can support a wikipedia?

And those Bible translations will be copyrighted, so it wouldn't be possible to put them up on Wikisource either. Am 2017-01-27 um 18:53 schrieb Milos Rancic:

...

On Fri, Jan 27, 2017 at 5:00 PM, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote:

There is little point at this time in a dictionary. With work going on for a Wikidatification of Wiktionary it is a real bad investment. There is much more to be gained by having books digitised.

In a lot of cases, all of the books in particular languages would be consisted just of portions of Bible. Thousands of languages belong to the category where original materials are needed (along with translations). _______________________________________________ Langcom mailing list Langcom(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/langcom

-- Dr. Antony Green Rudolf-Seiffert-Str. 31 WE 1703 10369 Berlin, Germany E-Mail: toniogreen(a)web.de Mobile: +49-176-82295920

Antony Green

5:04 p.m.

New subject: What is the smallest community that can support a wikipedia?

I remember saying a long time ago – perhaps at the meeting in Berlin – that if there's a request for a new Wikipedia in an endangered language, we should certainly comply with it once it meets the requirements, but in addition, we should also inform the requesters of the possibility of recording their language's lexicon in a major-language Wiktionary (typically whatever major language speakers of the endangered language are most likely to be bilingual in, e.g. Spanish in the case of an endangered language in Mexico), since they may not be aware of that possibility. Wiktionary is not as well known as Wikipedia, and not everyone realizes that Wiktionaries are intended to include words in all languages, not just their nominal administrative language. There's much less bureaucracy involved in adding Cora words to es.wiktionary.org than in starting up a Cora Wikipedia, so the former is something people can do to record their language while they're waiting for the latter to happen. Regards, Antony Am 2017-01-27 um 16:51 schrieb Milos Rancic:

...

On Fri, Jan 27, 2017 at 7:40 AM, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote:

What is the purpose of answering that question when it makes no difference in the way we work. Why would we consider it, what is the point. Asking for academic reasons?

-- Dr. Antony Green Rudolf-Seiffert-Str. 31 WE 1703 10369 Berlin, Germany E-Mail: toniogreen(a)web.de Mobile: +49-176-82295920

Milos Rancic

5:07 p.m.

On Fri, Jan 27, 2017 at 5:04 PM, Antony Green <toniogreen(a)web.de> wrote:

...

Yes. Also recorded materials in spoken language would be great to have. Although, we would need here to be cautious and follow contemporary norms of anthropology and field work.

Gerard Meijssen

5:09 p.m.

...

On Fri, Jan 27, 2017 at 5:04 PM, Antony Green <toniogreen(a)web.de> wrote:

I remember saying a long time ago – perhaps at the meeting in Berlin –

that

if there's a request for a new Wikipedia in an endangered language, we should certainly comply with it once it meets the requirements, but in addition, we should also inform the requesters of the possibility of recording their language's lexicon in a major-language Wiktionary

(typically

whatever major language speakers of the endangered language are most

likely

to be bilingual in, e.g. Spanish in the case of an endangered language in Mexico), since they may not be aware of that possibility. Wiktionary is

not

as well known as Wikipedia, and not everyone realizes that Wiktionaries

are

intended to include words in all languages, not just their nominal administrative language. There's much less bureaucracy involved in adding Cora words to es.wiktionary.org than in starting up a Cora Wikipedia,

so the

former is something people can do to record their language while they're waiting for the latter to happen.

Antony Green

5:17 p.m.

New subject: What is the smallest community that can support a wikipedia?

Practically, it just means we remind them of the existence of Wiktionary and that existing Wiktionaries will accept entries in their language. On another note, you mentioned the "Wikidatification" of Wiktionary, and I wonder what you mean. People at en-wikt often complain that Wikidata seems to be completely uninterested in coordinating Wiktionary entries the way it coordinates Wikipedia articles. Am 2017-01-27 um 17:09 schrieb Gerard Meijssen:

...

Hoi, What does that practically mean? It is volunteers who do this for their own purposes. What is it that we have to care about except that it is them doing it for their own reasons ? Thanks, GerardM On 27 January 2017 at 17:07, Milos Rancic <millosh(a)gmail.com <mailto:millosh@gmail.com>> wrote: On Fri, Jan 27, 2017 at 5:04 PM, Antony Green <toniogreen(a)web.de <mailto:toniogreen@web.de>> wrote:

I remember saying a long time ago – perhaps at the meeting in

Berlin – that

if there's a request for a new Wikipedia in an endangered

language, we

should certainly comply with it once it meets the requirements,

but in

addition, we should also inform the requesters of the possibility of recording their language's lexicon in a major-language

Wiktionary (typically

whatever major language speakers of the endangered language are

most likely

to be bilingual in, e.g. Spanish in the case of an endangered

language in

Mexico), since they may not be aware of that possibility.

Wiktionary is not

as well known as Wikipedia, and not everyone realizes that

Wiktionaries are

intended to include words in all languages, not just their nominal administrative language. There's much less bureaucracy involved

in adding

Cora words to es.wiktionary.org <http://es.wiktionary.org> than

in starting up a Cora Wikipedia, so the

former is something people can do to record their language while

they're

waiting for the latter to happen.

Yes. Also recorded materials in spoken language would be great to have. Although, we would need here to be cautious and follow contemporary norms of anthropology and field work. _______________________________________________ Langcom mailing list Langcom(a)lists.wikimedia.org <mailto:Langcom@lists.wikimedia.org> https://lists.wikimedia.org/mailman/listinfo/langcom <https://lists.wikimedia.org/mailman/listinfo/langcom> _______________________________________________ Langcom mailing list Langcom(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/langcom

-- Dr. Antony Green Rudolf-Seiffert-Str. 31 WE 1703 10369 Berlin, Germany E-Mail: toniogreen(a)web.de Mobile: +49-176-82295920

Milos Rancic

5:50 p.m.

On Fri, Jan 27, 2017 at 5:09 PM, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote:

...

What does that practically mean? It is volunteers who do this for their own purposes. What is it that we have to care about except that it is them doing it for their own reasons ?

This is quite delicate issue. There could be various taboos which we should respect. For example, there are information that can't be heard by foreigners, by women, by men, by people with blue eyes... All of those information could be gotten by a person close to a particular group. Note that I am talking here about very small groups, likely up to few hundreds of speakers. If it's about the groups with at least tens of thousands of speakers, we don't need to be that cautious. I am even thinking that we should create a kind of "staging" website, where people would be able to upload such materials to be checked, first. But that's a long shot. In reality, we should just take care about what's going on and react when needed.

2682

days inactive

2683

days old

langcom@lists.wikimedia.org

Manage subscription

12 comments

4 participants

tags (0)

participants (4)

Antony Green
Asaf Bartov
Gerard Meijssen
Milos Rancic