![]() | This page is currently inactive and is retained for
historical reference. Either the page is no longer relevant or consensus on its purpose has become unclear. To revive discussion, seek broader input via a forum such as the village pump. |
A conlang is a constructed language.
The purpose of this page is to form a consensus. Since the foundation of Wikipedia, a number of articles (almost 200) about constructed languages have been written, but not everyone agrees on how notable a conlang must be in order to be kept. On one camp, there are people who think all conlang articles should be deleted (although a generous few of them would let us keep an article on Esperanto to merge all conlang articles to); on the other extreme, there are those who think how 'notable' an artlang is should not be criteria for exclusion, as opposed to its 'quality'. We need to set a bar somewhere, or else conlang articles (and conlang VfDs) are going to continue on without rhyme or reason.
The discussion originally ran from July 28 until August 28, 2005, with a vote on proposed criteria beginning August 28. At that point more people got involved, and some objected to the whole procedure. Renewed discussion has been taking place at Wikipedia talk:Conlangs/Straw poll and Wikipedia:Conlangs/Alternative proposal since about August 30.
Note: This page is getting too large! If you can, please move comments or sections to subpages (e.g., Wikipedia:Conlangs/Only Esperanto, if you think only Esperanto should be mentioned on Wikipedia), with a sentence to describe the subpage's content.
Most of the criteria you propose are applicable only to Auxlangs. Actual use, by anyone other than a fictional character, is irrelevant for artlangs. Personally, I would say that Kélen and Wenedyk, (both of which I've had the pleasure to translate) are finer examples of the art than Klingon. However, your criteria would include Klingon because some people have learnt to speak it, even though this is more due to the fact that it's part of the Star Trek cult than its inherent artistic merit, and exclude (what I consider to be) better conlangs whose creators are happy for their language's speakers to remain fictional.
Pete Bleackley
I agree that the current conditions are a little auxlang-oriented. But I am glad that Almafeta is at least trying to establish some objective criteria. Currently everybody can just shout "notable" or "non-notable" at will, and therefore I genuinely think we nééd those criteria badly.
A few remarks:
Number of speakers: like the others said, this can apply only to languages that were designed with the purpose of being spoken. In other words, IALs. True, there are some artistic languages that have some speakers, too. But IMO the number of Klingon speakers says as much about the importance of Klingon as the number of sold Star Trek T-shirts about the importance of Star Trek. Nobody can of course be sure, but a huge number of those people have learnt the language nót because they think it is important, or because they want to communicate in it, but because they are Star Trek fans. Another thing is that the "number of speakers" is terribly hard to establish: where do you get the info, and how well must a person speak the language for being qualified as a "speaker"?
Books with ISBN numbers: that's an dangerous one. There has been a period when the world was rather susceptible to the idea of a world language. All kinds of authors started building something Esperanto-like and could easily find a publisher to publish there ideas. Nowadays, I think such a thing is virtually impossible, especially when the subject is not an IAL but an artlang. Why would a publisher publish such a think if he knows in advance that he won't sell a huge number of copies? Even Tolkien couldn't manage such a thing; only now books are written about his conlangs.
Furthermore, nowadays we have the Internet. Why bother going through hell by looking for a publishing house if you can easily self-publish your stuff in that way. Guaranteed no income, but conlanging is no way to become rich anyway.
Older than usual: I can live with that. Although, see my comments to the previous point.
Langmaker Top 100: perhaps Jeffrey Henning, the Big Boss of Langmaker, can tell us in how far this would be a reliable criterion. I'm not asking him to disclose the algorithms he uses, but I would like to know how useful in his opinion the result is; how easily one could manipulate the outcome; and in how far the outcome is influenced by people who might be looking for something completely different (I noticed for example that languages like "Gothic", "New English" and my own "Hattic" always score well, and I suspect that it because their names attract a lot of googlers). If you are reading this, Jeffrey, tell us this: if someone would, purely theorically, have a conlang called "Vagina". Would it become part of the Top-10, and if so, how quickly?
Mistaken for a real language: By whom? When? Where? I have a website about Hattic, a naturalistic conlang with an entire quasi-history; it is never mentioned that the whole thing is fictional, and indeed I have stumbled upon people who thought the whole thing was real. But that is merely the result of how a language is presented, not how significant it is.
Caused controversy: Frankly, I think the whole current set of VfDs regarding Brithenig, Verdurian, Wenedyk and others can be considered controversy too, isn't it?
ISO code: I don't really know the mechanisms behind ISO 639-3. But I have seen the list of conlangs listed there, and my impression is that someone just picked a few conlangs from the Internet during a spare few minutes. The list hasn't changed at all since a number of years, and I also noticed that the list contains both Ceqli and Tceqli, which as far as I know is exactly the same language (or perhaps the same language in two different stages of its development).
Wikipedia: Question is: how does a language get a Wikipedia? Again, it seems to me that the number of users is decisive here. And even then, the whole thing seems to be rather prone to politicising (see the story of the Toki Pona wikipedia).
My point of view: I think the most of the conditions Almafeta mentions can indeed contribute to a language's notability. But I can't accept that a language that does not fulfill them is therefore to be considered non-notable. Note for example that a language like Parseltongue wouldn't survive these points either, and undoubtedly many more languages wouldn't. What I miss is therefore some criteria that might apply specifically to conlangs of the artistic genre, like:
Food for thought, I would say!
-- IJzeren Jan 12:53, 28 July 2005 (UTC)
I shall comment Almafeta's proposition in italics, adapting it to a particular case, Verdurian. Cctoide 15:38, 28 July 2005 (UTC)
If at least two of these minor conditions can be met, a conlang is notable:
If any of the following major conditions can be met, a conlang is notable:
-- Cctoide 15:38, 28 July 2005 (UTC)
Over on WP:COMIC, the suggested criteria include at least 33 weeks of 'active production' of strips, and at least 100 comics. This was seen as being sufficiently selective to prevent each and every brand new webcomic from making its own Wikipedia article.
Should the criteria for conlangs be similar -- i.e., excluding all the innumerous conlangs that were created in one day's or one week's time, but allowing for efforts that have had a significant effort in development and 'finishing'? Almafeta 14:28, 28 July 2005 (UTC)
Moved to Wikipedia:Conlangs/Artlangs, ficlangs, and loglangs
Just about how the vote will go (that's a long way off, but it's good to get this set straight now): Each proposed rule will have its own level 2 heading, with the following level four headings: Minor, Major, and Oppose. If you vote Minor, you think it the supports notability but does not merit inclusion in Wikipedia on its lonesome; if you vote Major, you think it supports notability and merits inclusion in Wikipedia reguardless of any other factors; and if you vote Oppose, you think that it is not applicable or misleading. If nobody has any objections... Almafeta 22:22, 28 July 2005 (UTC)
What about notability? Would a conlang that's trying to do something interesting and unique be notable despite lack of completion? (Though notoriety outside its environment might cover this.) DenisMoskowitz 17:21, 2005 July 28 (UTC)
We could set some criteria about completeness of the language's description - lexicon size, whether a certain list of basic questions about its grammar and phonology are answered, etc. I would also propose a minimum corpus size criterion: are there multiple texts written in the language, with some significant total wordage. A language with several thousand words in its lexicon may not really be very complete if the only text available is a few sample sentences in the grammar document, and maybe the Tower of Babel story or the Lord's Prayer. On the other hand, a language designed for lexical parsimoniousness should not be penalized for its small lexicon, if the amount of text written in the language proves that this language is more powerful than its small lexicon would suggest. Corpus size counts extra if some significant part of was written by people other than the creator of the language.
I would also suggest that the minimum number of speakers in Almafeta's criteria should be reduced. 500 speakers are certainly plenty to prove notability, but I think 50 are plenty even without one of the other "minor" criteria; having even 10 or 20 speakers is pretty notable for a naturalistic conlang that was not designed (like most auxlangs) to be easy to learn. -- Jim Henry | Talk 17:37, 28 July 2005 (UTC)
I would like to note that #almea and the KutjaraWiki are not managed by or linked by any way to zompist. #almea hardly ever discusses Verdurian, and the Wiki is not run by Mark. Cctoide 18:23, 28 July 2005 (UTC)
FWIW, "notability" is an ever-elusive thing. Given the various motives people have for creating languages, from auxiliary languages, artistic languages, and cryptographic languages, I am not sure that we need a policy here that can be applied like an algorithm. I would prefer that conlang articles speak for themselves: if a language can be described in terms that make it "interesting" enough, and if the article that is written about it shows some sophistication and is otherwise encyclopedic, I would generally vote to allow such articles to remain. Smerdis of Tlön 14:51, 2 August 2005 (UTC)
Moved to Wikipedia talk:Conlangs/Criteria
Moved to Wikipedia:Conlangs/Notability, verifiability, merit, completeness
(moved to /Suggestions by Thomas Winwood)
Why not have a look at the List of constructed languages, decide which ones each of us consider "worthy" of inclusion and then discuss why or why not just these languages should be in WP, thereby discovering inclusion-criteria that have already been used and whether these criteria are necessary or sufficient? Note that the list is divided into auxlangs, artlangs and languages mentioned or used in a verified work of fiction already.
I don't care much for auxlangs at all, so to make an example pour encourager les autres:
Furthermore, many of the listed conlangs, especially the fictional langs, are so incomplete as to not be usable for communication at all (exceptions: Kesh, Klingon, Láadan, Quenya) because they only consist of word lists or are described, with few or no examples shown, while the "masterpieces" can be so complete so as to be perfectly learnable by the sufficiently interested. This means the representation/layout of the sketched langs and the more complete langs need to be different. Compare the entry for Babel-17 and Quenya. -- Kaleissin 21:08:54, 2005-08-16 (UTC)
Moved to Wikipedia:Conlangs/Differential
Moved to Wikipedia:Conlangs/Notability, verifiability, merit, completeness
I am starting in the Spanish language Wikipedia in a WikiProject for constructed langauges:
The idea with the project is to both define the policies regarding inclussion of artificial languages, and how these languages should be presented.
Those interested can sign in the WikiProject at Spanish lanaguage Wikipedia. You are also wellcome to start a similar WikiProject here.
There are a few wikis on constructed languages. I am running one at Wikicities (which uses the same GFDL as Wikipedia).
There is room there to any constructed language and any detail of publications (from simple one paragraph article to complete grammars, translations and original works). OR is accepted.
Note that the GFDL applies to the articles, not to the languages. Anyhow any language that was nominated to be in wikipedia would fit in Wikicities.
— Carlos Th (talk) 21:52, 3 August 2005 (UTC)
I've posted to http://conlang.wikicities.com/wiki/Talk:Main_Page a proposal for a peer-reviewed journal along the lines discussed above. -- Jim Henry
/Sai's two cents on what verifiability means for conlang articles
![]() | This page is currently inactive and is retained for
historical reference. Either the page is no longer relevant or consensus on its purpose has become unclear. To revive discussion, seek broader input via a forum such as the village pump. |
A conlang is a constructed language.
The purpose of this page is to form a consensus. Since the foundation of Wikipedia, a number of articles (almost 200) about constructed languages have been written, but not everyone agrees on how notable a conlang must be in order to be kept. On one camp, there are people who think all conlang articles should be deleted (although a generous few of them would let us keep an article on Esperanto to merge all conlang articles to); on the other extreme, there are those who think how 'notable' an artlang is should not be criteria for exclusion, as opposed to its 'quality'. We need to set a bar somewhere, or else conlang articles (and conlang VfDs) are going to continue on without rhyme or reason.
The discussion originally ran from July 28 until August 28, 2005, with a vote on proposed criteria beginning August 28. At that point more people got involved, and some objected to the whole procedure. Renewed discussion has been taking place at Wikipedia talk:Conlangs/Straw poll and Wikipedia:Conlangs/Alternative proposal since about August 30.
Note: This page is getting too large! If you can, please move comments or sections to subpages (e.g., Wikipedia:Conlangs/Only Esperanto, if you think only Esperanto should be mentioned on Wikipedia), with a sentence to describe the subpage's content.
Most of the criteria you propose are applicable only to Auxlangs. Actual use, by anyone other than a fictional character, is irrelevant for artlangs. Personally, I would say that Kélen and Wenedyk, (both of which I've had the pleasure to translate) are finer examples of the art than Klingon. However, your criteria would include Klingon because some people have learnt to speak it, even though this is more due to the fact that it's part of the Star Trek cult than its inherent artistic merit, and exclude (what I consider to be) better conlangs whose creators are happy for their language's speakers to remain fictional.
Pete Bleackley
I agree that the current conditions are a little auxlang-oriented. But I am glad that Almafeta is at least trying to establish some objective criteria. Currently everybody can just shout "notable" or "non-notable" at will, and therefore I genuinely think we nééd those criteria badly.
A few remarks:
Number of speakers: like the others said, this can apply only to languages that were designed with the purpose of being spoken. In other words, IALs. True, there are some artistic languages that have some speakers, too. But IMO the number of Klingon speakers says as much about the importance of Klingon as the number of sold Star Trek T-shirts about the importance of Star Trek. Nobody can of course be sure, but a huge number of those people have learnt the language nót because they think it is important, or because they want to communicate in it, but because they are Star Trek fans. Another thing is that the "number of speakers" is terribly hard to establish: where do you get the info, and how well must a person speak the language for being qualified as a "speaker"?
Books with ISBN numbers: that's an dangerous one. There has been a period when the world was rather susceptible to the idea of a world language. All kinds of authors started building something Esperanto-like and could easily find a publisher to publish there ideas. Nowadays, I think such a thing is virtually impossible, especially when the subject is not an IAL but an artlang. Why would a publisher publish such a think if he knows in advance that he won't sell a huge number of copies? Even Tolkien couldn't manage such a thing; only now books are written about his conlangs.
Furthermore, nowadays we have the Internet. Why bother going through hell by looking for a publishing house if you can easily self-publish your stuff in that way. Guaranteed no income, but conlanging is no way to become rich anyway.
Older than usual: I can live with that. Although, see my comments to the previous point.
Langmaker Top 100: perhaps Jeffrey Henning, the Big Boss of Langmaker, can tell us in how far this would be a reliable criterion. I'm not asking him to disclose the algorithms he uses, but I would like to know how useful in his opinion the result is; how easily one could manipulate the outcome; and in how far the outcome is influenced by people who might be looking for something completely different (I noticed for example that languages like "Gothic", "New English" and my own "Hattic" always score well, and I suspect that it because their names attract a lot of googlers). If you are reading this, Jeffrey, tell us this: if someone would, purely theorically, have a conlang called "Vagina". Would it become part of the Top-10, and if so, how quickly?
Mistaken for a real language: By whom? When? Where? I have a website about Hattic, a naturalistic conlang with an entire quasi-history; it is never mentioned that the whole thing is fictional, and indeed I have stumbled upon people who thought the whole thing was real. But that is merely the result of how a language is presented, not how significant it is.
Caused controversy: Frankly, I think the whole current set of VfDs regarding Brithenig, Verdurian, Wenedyk and others can be considered controversy too, isn't it?
ISO code: I don't really know the mechanisms behind ISO 639-3. But I have seen the list of conlangs listed there, and my impression is that someone just picked a few conlangs from the Internet during a spare few minutes. The list hasn't changed at all since a number of years, and I also noticed that the list contains both Ceqli and Tceqli, which as far as I know is exactly the same language (or perhaps the same language in two different stages of its development).
Wikipedia: Question is: how does a language get a Wikipedia? Again, it seems to me that the number of users is decisive here. And even then, the whole thing seems to be rather prone to politicising (see the story of the Toki Pona wikipedia).
My point of view: I think the most of the conditions Almafeta mentions can indeed contribute to a language's notability. But I can't accept that a language that does not fulfill them is therefore to be considered non-notable. Note for example that a language like Parseltongue wouldn't survive these points either, and undoubtedly many more languages wouldn't. What I miss is therefore some criteria that might apply specifically to conlangs of the artistic genre, like:
Food for thought, I would say!
-- IJzeren Jan 12:53, 28 July 2005 (UTC)
I shall comment Almafeta's proposition in italics, adapting it to a particular case, Verdurian. Cctoide 15:38, 28 July 2005 (UTC)
If at least two of these minor conditions can be met, a conlang is notable:
If any of the following major conditions can be met, a conlang is notable:
-- Cctoide 15:38, 28 July 2005 (UTC)
Over on WP:COMIC, the suggested criteria include at least 33 weeks of 'active production' of strips, and at least 100 comics. This was seen as being sufficiently selective to prevent each and every brand new webcomic from making its own Wikipedia article.
Should the criteria for conlangs be similar -- i.e., excluding all the innumerous conlangs that were created in one day's or one week's time, but allowing for efforts that have had a significant effort in development and 'finishing'? Almafeta 14:28, 28 July 2005 (UTC)
Moved to Wikipedia:Conlangs/Artlangs, ficlangs, and loglangs
Just about how the vote will go (that's a long way off, but it's good to get this set straight now): Each proposed rule will have its own level 2 heading, with the following level four headings: Minor, Major, and Oppose. If you vote Minor, you think it the supports notability but does not merit inclusion in Wikipedia on its lonesome; if you vote Major, you think it supports notability and merits inclusion in Wikipedia reguardless of any other factors; and if you vote Oppose, you think that it is not applicable or misleading. If nobody has any objections... Almafeta 22:22, 28 July 2005 (UTC)
What about notability? Would a conlang that's trying to do something interesting and unique be notable despite lack of completion? (Though notoriety outside its environment might cover this.) DenisMoskowitz 17:21, 2005 July 28 (UTC)
We could set some criteria about completeness of the language's description - lexicon size, whether a certain list of basic questions about its grammar and phonology are answered, etc. I would also propose a minimum corpus size criterion: are there multiple texts written in the language, with some significant total wordage. A language with several thousand words in its lexicon may not really be very complete if the only text available is a few sample sentences in the grammar document, and maybe the Tower of Babel story or the Lord's Prayer. On the other hand, a language designed for lexical parsimoniousness should not be penalized for its small lexicon, if the amount of text written in the language proves that this language is more powerful than its small lexicon would suggest. Corpus size counts extra if some significant part of was written by people other than the creator of the language.
I would also suggest that the minimum number of speakers in Almafeta's criteria should be reduced. 500 speakers are certainly plenty to prove notability, but I think 50 are plenty even without one of the other "minor" criteria; having even 10 or 20 speakers is pretty notable for a naturalistic conlang that was not designed (like most auxlangs) to be easy to learn. -- Jim Henry | Talk 17:37, 28 July 2005 (UTC)
I would like to note that #almea and the KutjaraWiki are not managed by or linked by any way to zompist. #almea hardly ever discusses Verdurian, and the Wiki is not run by Mark. Cctoide 18:23, 28 July 2005 (UTC)
FWIW, "notability" is an ever-elusive thing. Given the various motives people have for creating languages, from auxiliary languages, artistic languages, and cryptographic languages, I am not sure that we need a policy here that can be applied like an algorithm. I would prefer that conlang articles speak for themselves: if a language can be described in terms that make it "interesting" enough, and if the article that is written about it shows some sophistication and is otherwise encyclopedic, I would generally vote to allow such articles to remain. Smerdis of Tlön 14:51, 2 August 2005 (UTC)
Moved to Wikipedia talk:Conlangs/Criteria
Moved to Wikipedia:Conlangs/Notability, verifiability, merit, completeness
(moved to /Suggestions by Thomas Winwood)
Why not have a look at the List of constructed languages, decide which ones each of us consider "worthy" of inclusion and then discuss why or why not just these languages should be in WP, thereby discovering inclusion-criteria that have already been used and whether these criteria are necessary or sufficient? Note that the list is divided into auxlangs, artlangs and languages mentioned or used in a verified work of fiction already.
I don't care much for auxlangs at all, so to make an example pour encourager les autres:
Furthermore, many of the listed conlangs, especially the fictional langs, are so incomplete as to not be usable for communication at all (exceptions: Kesh, Klingon, Láadan, Quenya) because they only consist of word lists or are described, with few or no examples shown, while the "masterpieces" can be so complete so as to be perfectly learnable by the sufficiently interested. This means the representation/layout of the sketched langs and the more complete langs need to be different. Compare the entry for Babel-17 and Quenya. -- Kaleissin 21:08:54, 2005-08-16 (UTC)
Moved to Wikipedia:Conlangs/Differential
Moved to Wikipedia:Conlangs/Notability, verifiability, merit, completeness
I am starting in the Spanish language Wikipedia in a WikiProject for constructed langauges:
The idea with the project is to both define the policies regarding inclussion of artificial languages, and how these languages should be presented.
Those interested can sign in the WikiProject at Spanish lanaguage Wikipedia. You are also wellcome to start a similar WikiProject here.
There are a few wikis on constructed languages. I am running one at Wikicities (which uses the same GFDL as Wikipedia).
There is room there to any constructed language and any detail of publications (from simple one paragraph article to complete grammars, translations and original works). OR is accepted.
Note that the GFDL applies to the articles, not to the languages. Anyhow any language that was nominated to be in wikipedia would fit in Wikicities.
— Carlos Th (talk) 21:52, 3 August 2005 (UTC)
I've posted to http://conlang.wikicities.com/wiki/Talk:Main_Page a proposal for a peer-reviewed journal along the lines discussed above. -- Jim Henry
/Sai's two cents on what verifiability means for conlang articles