Wikipedia:Conlangs/Artlangs, ficlangs, and loglangs

Main page: Wikipedia:Conlangs

What sort of provisions should be made for conlangs that are primarially artistic (like Eklektu), fictional (like Quenya), logical (like Loglan), or philosophical (like Toki Pona)? Right now, my guidelines aren't good enough for those sorts of languages. (For example, if we made a notability criteria having a certain number of words -- such as the often-quoted 2,500 words needed to achieve conversational fluency in English -- we'd ignore languages like Basic English, Toki Pona, and Baza, all of which are designed to work around an extremely restricted set of words.)

Does anyone have a good idea what would make one of the four above types of conlang notable? Almafeta 18:26, 28 July 2005 (UTC)

Well, Loglan is notable for being a test of Sapir-Whorf and for being the most agressively logical conlang, and Quenya is notable because of the notability of Lord of the Rings. I'm not sure whether Eklektu or Toki Pona are notable. DenisMoskowitz 18:36, 2005 July 28 (UTC)

I would argue that any language with as many speakers as Loglan, Lojban or Toki Pona is notable regardless of its other characteristics. As I mentioned elsewhere, I think the minimum number of speakers should be set far lower than 500, arguably lower than 50, especially for languages that aren't designed to be regular and/or easy. Further down the page there is some discussion about "completeness" criteria (how thoroughly the phonology, grammar, semantics are defined; how large a corpus exists in the language; etc) that would apply to artlangs (and engelangs like Ithkuil) that may be too difficult to attract learners. Influence on other conlangs is a criteria that applies well to artlangs and engelangs as well. --Jim Henry | Talk 21:09, 28 July 2005 (UTC)

Agreed. But I maintain that in the case of artistic languages the number of speakers shouldn't matter at all. IJzeren Jan 21:38, 28 July 2005 (UTC)

I agree, in the sense that an artlang (or engelang) may merit inclusion in Wikipedia even if it has no fluent speakers at all (even the creator) and has not been used in professionally published/broadcast fiction/media -- if it is complete enough (see discussion below) and there is some evidence that it is well known or influential on other conlangers. However, I think having some speakers besides the creator is a plus; an artlang that just barely merits inclusion based on its completeness, fame, influence, etc., would be pushed over the edge into definite notability by having some speakers besides the author (even if only 10 or 20). We might do this with a structure similar to Almafeta's minor criteria: at least four of these criteria for expressivity/completeness, or at least three of those criteria plus at least 10 speakers other than the creator and friends. --Jim Henry | Talk 22:00, 28 July 2005 (UTC)

Agreed on all accounts! It looks like our points of view are not that different after all. With one small reservation: before establishing a number of required additional criteria, I'd first like to have established a definitive list. --IJzeren Jan 13:35, 29 July 2005 (UTC)

OK, that makes sense. But do you agree that the two-tier system (any one major criterion, any two minor criteria) in Almafeta's original proposal is too simple? --Jim Henry | Talk 15:51, 29 July 2005 (UTC)

To be honest, my feelings about that system are mixed. Yes, the way it looks now it is too simple (and too auxlang-oriented) and definitely deserves refining. But on the other hand, I don't think we do our "cause" much good if we turn it into an algorithm that is so complicated that hardly anyone can crack it. --IJzeren Jan 09:24, 30 July 2005 (UTC)

Besides, how do you verify a language's number of speakers? Whom do you ask? And furthermore: how proficient must a person be in the language to be called a "speaker"? Usually it won't be much more than writing a few sentences, mind. IJzeren Jan 21:45, 28 July 2005 (UTC)

We would probably have to have a weaker definition of "speaker" than (say) the "Foreign Service Level 3" that was reported used in Dr. Culbert's Esperanto speaker census, simply for want of means to measure spoken fluency. Ability to read and write at asynchronous speeds would suffice, because it's easy to verify online. For Toki Pona, for instance, we can point to a mailing list where a fair amount of the traffic is in the conlang itself (not just about it) and the number of people posting in the language is significant. --Jim Henry | Talk 22:00, 28 July 2005 (UTC)

I definately think traffic on a message board by a number of speakers in different locales would establish a certain amount of notability. Also, organizations or clubs for speakers of the language existing outside of the location of the creator would be a good sign. -Satori 23:23, 28 July 2005 (UTC)

The membership count of an active mailing list (at least 1 article per day on average) would also be a good yardstick. Almafeta 17:29, 29 July 2005 (UTC)

Supplementary criteria proposed by Pete Bleackley

Here are some proposed ideas for things that may be used to recommend the inclusion of an artlang - I'll give the justification for them here, along with whether I think they're major or minor reasons for inclusion, and (just for fun) how my own creations measure up. I'll also copy them to the votes page.

====Is associated with a well developed conculture==== Major Goes hand in hand with a really good artlang. An example is Verdurian Drawback is that it may be difficult to assess. For my own languages, I's say cannot be established from published material.

====Sets itself challenging artistic goals, and achieves them==== Major Any artwork that does this is noteworthy. An example is Kélen, which is an intelligible and naturalistic language, despite having no verbs. For my own conlangs no.

====Has been used in at least 5 translation relays==== Minor Shows that a language is usable, and intelligible to third parties. For my conlangs I've only done 3 so far.

====At least 10 original texts are published in the language==== Minor Shows that the author has thought about the cultural background and use of the language, not just the grammar. For my languages no.

====At least 20 original texts published in the language==== Major Now that's impressive. For my languages nowhere near.

====Demonstates plausible historical derivation from a natural language==== Minor A worthwhile and interesting pursuit. See Wenedyk or Brithenig for examples. Only rated minor since it's a fairly common excercise. For my conlangs no.

====Demonstrates plausible historical derivation from another conlang==== Major Rated higher than the above, because the author first needs to create a language to derive it from. For one of my languages, I'd say not independantly verifiable.

====Contributes to the notability of another conlang==== Minor This is interesting, but probably needs other supporting factors. For my conlangs, no

--132.185.132.12 09:07, 12 August 2005 (UTC) Pete Bleackley

Reply to Mr. Bleackley's criteria

Most of the criteria Pete Bleackley propose are good for establishing completeness and expressivity of conlangs, and therefore, to some extent, notability. In some cases I would want them to be more specific (quantified) or to have the bar set higher. E.g., the minima of 10 or 20 "texts" — of what length, kind, complexity? During the discussion archived in Wikipedia:Conlangs/Notability, verifiability, merit, completeness I proposed minimum corpus sizes — a total number of words published (or anyway publicly available) in the language.

However, these criteria don't address the basic issues of verifiability and original research. If a conlang is complete enough to qualify under the criteria proposed above, it will probably eventually attract "active discussion" (in Almafeta's words) or independent review (my preferred terminology), and then will be verifiable without original research, thus suitable for Wikipedia. A conlang with a corpus of twenty texts totaling 100,000 words is remarkable and notable, but if no one has written about it yet but the conlang's creator, can we write a verifiable, non-OR article about it at Wikipedia? --Jim Henry | Talk 16:57, 12 August 2005 (UTC)

I felt that originality was a more noteworthy criterion than mere length - somebody could fairly easily create a large corpus size by translating some of the "and A begat B" passages of Genesis. While a text translated from a known source (e.g. the Babel text, or a relay text) is of interest, I'd think it more noteworthy that a conlanger is writing original material in his language. Pete Bleackley --132.185.132.12 13:19, 15 August 2005 (UTC)

OK, it makes sense to have a separate criterion for original corpus vs. overall corpus including translations, and the first could have a lower limit than the second (maybe 2000 words original or 10,000 words total as minor criterion, 20,000 words original or 100,000 words total as a major criterion). But I still think the criteria ought to be based on total word count, not total number of distinct texts. And -- again -- having a large corpus, whether original, translated, or both, makes a conlang notable but doesn't make it verifiable without original research if the conlang's creator is the only person who has written about it. --Jim Henry | Talk 13:50, 15 August 2005 (UTC)

Also, the ability of the language to translate set texts from other languages is an important additional measure of its "completeness". My own conlang gjâ-zym-byn has an original corpus of >10,000 words (most of which (diary entries) will never be public and therefore is irrelevant for Wikipedia notability, though), and translating things from other languages is generally more challenging for me than writing more original text. Writing original stuff, you can subconsciously work within the limitations of the language; translating, you're forced to confront them. I coin new words and clarify grammatical issues at a much higher rate when working on a translation than when simply writing diary entries day to day. --Jim Henry | Talk 14:04, 15 August 2005 (UTC)

Well, I had said "original texts published". Several criteria for translated texts had already been proposed, so I proposed ones for original texts. Pete Bleackley --132.185.132.12 09:24, 16 August 2005 (UTC)