Wikipedia talk:Conlangs/Alternative proposal

Start talking already :) --Kaleissin

No algorithms, please

Ultimately, any article on a conlang has to fail or succeed on the article itself. No doubt, many conlang articles will have some features of vanity articles, and many may be obscure creations posted by the language's creator herself. Even these articles arguably ought to be allowed to remain IMO if they show some evidence of sophistication and are, for lack of a better word, "interesting."

A checklist of features will not discourage vanity creations. It will instead lead to lawyering the checklist. Sufficiently determined conlangers will seek to make sure that enough of the criteria that can be met with a one-person effort will be satisfied.

Ultimately, the best test of noteworthiness for an article on a conlang is whether the article is informative and interesting to people who take an interest in conlangs. No algorithm is going to be able to replace this subjective criterion. Conlang articles will always have to be considered on a case by case basis. I don't believe it is desirable or feasible to have a policy, at least not one that takes the form of a checklist. Smerdis of Tlön 14:49, 30 August 2005 (UTC)[reply]

I agree we should keep this possibility in mind for each criterion we discuss and vote on -- can it be gamed? Can a solo conlang creator meet this criterion by spending enough money, time, and effort, even if everyone else ignores the language or at most puts a brief mention of it in their conlang links page? If so, we should avoid such criteria. But I think we do need some more specific notability/verifiability policy to apply to conlangs, to counter those who automatically vote Delete on all conlang VfDs just because they are conlangs, without considering the notability and verifiability of each on its merits. Criteria such as Has at least N speakers, Has generated at least N independent discussions, etc., are not so gameable and would hopefully help guide future VfDs in a rational way. --Jim Henry | Talk 15:28, 30 August 2005 (UTC)[reply]

Let's not exaggerate here. What we making here is nothing set in stone. We are merely offering a set of guidelines that may or may not be used for determining whether a conlang is notable or not. I don't have the illusion that it will discourage people from nominating even a language that meets all criteria for deletion, nor from voting for deletion in such a VfD.

Some factors can be gamed easier than others, I'll grant you that. But would anybody seriously spend years of his life manipulating these criteria in such a way that his conlang warrants a wikipedia article? Frankly, I happen to be the author of a conlang that survived a VfD, but do you really think that makes me much happier? I think a careful combination of factors should make it not too easy to game the system. And even if a conlang meet the criteria and there are good reasons to assume that somebody has been gaming the system, we can always make an exception. --IJzeren Jan 16:21, 30 August 2005 (UTC)[reply]

Conlang article discussion group/studio?

I'm bothered by the idea of a checklist, since it will be gameable and possibly not accepted by the majority of wikipedians. But I'm also bothered by not having any ammunition to say, "this conlang is notable" in a VfD discussion.

What if we had a wikipedia group, possibly the same group that is discussing this project, to subjectively discuss new conlang articles before they're even added to the main wikipedia namespace? Such a group would discuss the notability of conlangs outside of the VfD process. It could even attempt to discover notable conlangs that are missing from the wikipedia.

I'm concerned that this could be seen as a "conlang cabal" but it seems better than the alternatives. DenisMoskowitz 16:13, 2005 August 30 (UTC)

This is not exactly what you're suggesting, but see [1] for a related proposal. --Jim Henry | Talk 16:29, 30 August 2005 (UTC)[reply]

The same suggestion was also made recently by Thomas Winwood (here). Like I said then, I like the idea of such a conlang panel very much, and I'd certainly volunteer to be part of it. But we shouldn't overestimate its role. Its decisions can never been final and/or decisive. And it cannot serve as a "filter" between wikipedia namespace and the rest of the world; anybody can write articles about anything, including conlangs, and gatekeepers are simply not in the spirit of the project. No, in my opinion the role of such a panel should be the following:

It provides the project with a reliable source of information, available on request.
It is committed to keeping the entire conlang section clean, maintaining its high quality, and where possible improving it.
It monitors new articles, and when necessary nominates them for deletion.
It may start new articles about conlangs that are not in the wikipedia but ought to be.

I think such a panel should include some knowledgeable people who are active in other wikipedias; for example, I know that the German wikipedia has an excellent Portal for constructed languages, maintained by a small, dedicated group. Furthermore, all "currents" within conlanging (artlangs, auxlangs, loglangs, etc.) should be properly represented.

--IJzeren Jan 20:15, 31 August 2005 (UTC)[reply]

Those roles sound worthy to me. It would be nice to have a panel putting in good articles on notable conlangs, to give conlangs a good name on Wikipedia. If the only conlang articles are positive summaries added by self-interested creators, it gives the hobby in general a bad reputation among wikipedians. (I've held off on adding my own conlang so as not to be part of this problem.) DenisMoskowitz 15:56, 2005 September 1 (UTC)

Rikchik, right? A truly fascinating language! I'd say, if there's one language around where uniqueness can warrant an article, It'd be Rikchik! --IJzeren Jan 20:06, 1 September 2005 (UTC)[reply]

In the short to medium term, I think we need to focus on expanding stubs (like Teonaht) into substantial articles rather than creating new articles. We could also discuss a list of conlangs that arguably deserve articles; probably Talk:Constructed language would be a good place, or more specific pages like Talk:Artistic language. And conlangs that aren't (yet) notable or verifiable enough for Wikipedia can be discussed in an original research, not-necessarily-NPOV manner at the Conlang Wikicity. --Jim Henry | Talk 16:25, 1 September 2005 (UTC)[reply]

I think we should do both things: expanding stubs and starting new articles. And monitoring all conlang-related stuff. I propose that we make a list of ALL conlang-related articles (actually, my watchlist is pretty complete in that respect; I'll post it somewhere tomorrow). Then we should find out who should belong to the panel. There are a few very prominent conlangers around here (Mark Rosenfelder, Jeffrey Henning, Pablo Flores, Carlos Thompson and others); I'd like to invite them in, too.

As for my own contributions: I will of course keep an eye on everything here and where possible contribute. I am ready to assist with all the knowledge I have. However, my main point of focus is, and remains, the Dutch wikipedia. I've been responsible for almost the entire conlang section there, and frankly, I'm quite proud of it. I haven't done much with Lingua Ignota yet, but I have prepared a Dutch version! :) --IJzeren Jan 20:06, 1 September 2005 (UTC)[reply]

Maybe it would make sense to organize this as a WikiProject? Calling it a "panel" seems to imply an authority that such a self-selected group of people would not have.

My watchlist contains probably all of the general conlang articles and many of those on specific languages; I'd welcome a list such as you propose, though. --Jim Henry | Talk 20:18, 1 September 2005 (UTC)[reply]

Re:WikiProject - Good idea! If you know how to do that, please go on with it.

Re:Links page - I just made on here: User:IJzeren Jan/List Of Conlang-Related Articles. Do with it whatever you like. I'd say, please add everything relevant that you have on your watchlist, so that we can combine our forces better! --IJzeren Jan 20:42, 1 September 2005 (UTC)[reply]

Wikipedia:WikiProject suggests that the coordination overhead is not worth the trouble unless at least 5-10 people are involved. Do we have that many? Probably. Anyway, y'all go read it and see what you think, whether it fits what we're talking about doing. --Jim Henry | Talk 21:24, 1 September 2005 (UTC)[reply]

From what I can see, I think a WikiProject would be a good idea. I support it, and I'm ready to invest time and work in it. I can think of several tasks for such a project:

Making sure that all current conlang articles meet certain criteria. We expand stubs. For example, it should IMO be made clear in the article WHY the language is notable/interesting/unique. And, for example, we could add a sample of each conlang to the corresponding article. And a template. And ...
Applying the criteria that turn out from the current vote/poll. If we find conlangs that would fit the criteria, we start new articles about them. If we find articles about conlang that do not fit, we nominate them for deletion.
We keep an eye on all conlang-related stuff, especially on new articles.

If we decide to go on with it, I think it's important that all "factions" are properly represented. That is, artlangers enough, but I would also like to have at least one auxlang expert in our midst. And an engelang expert.

Anyway, who would be interested in participating? Shall we invite a few people?

--IJzeren Jan 20:29, 3 September 2005 (UTC)[reply]

Let's use them all

In the recent course of discussions, I've been infulenced back in the direction of the point I made earlier that there are as many ways for a conlang to be notable as there are conlangs. This, to some extent explains why we've got so many suggested criteria. To a large extent, we're chasing our own tails by trying to enumerate a hard and fast set of criteria. On the other hand, we do need something we can use to put a stop to the recent wave of rampant deletionism. I'm suggesting that a conlang should have an article if there are verifiable reasons why it is potentially of interest to persons other than its creator. These reasons should be elaborated in the article, with reference to sources. Our suggested criteria should be given as a list examples of reasons why a conlang might be of interest, with the caveat that this list is not considered exhaustive, merely illustrative. If an article fails to verifiably demonstrate that the conlang is of interest, the first recourse should be to attempt to improve it or to find somebody who can. Only if this is found to be impossible should deletion be considered.

If somebody can draw up a full draft on the lines of what I've writtten above, we can all hack at it until we're happy with the wording, and then release it as the policy.

While I'm at it, can anybody add anything to the articles I've creted for Teonaht, Kélen and Lingua Ignota? All worthy languages.

Pete Bleackley --132.185.132.12 09:34, 31 August 2005 (UTC)[reply]

I like this idea a lot. DenisMoskowitz 16:06, 2005 August 31 (UTC)

I pretty much sympathise with what you write, Pete, but I don't think it is workable. Paradoxically, it is essentially the same thing as the current status quo. Don't get me wrong: I'm not at all that happy with the whole concept of conlang notability. But we are building and maintaining an encyclopedia here, we want to have a good, healthy conlang section, and therefore we need to draw a line somewhere. And we better make that line objectively verifiable, otherwise we keep having the situation that anybody can yell "non-notable, conlang", and no matter what counterarguments are brought in, people who don't care for conlangs will join the choir. Mind, "notable" means in this case: suitable for inclusion in wikipedia; it says nothing about the quality of a language.

Furthermore, I can't agree with your proposal to allow anything that MIGHT be interesting to somebody else but the conlanger him/herself. That brings us in conflict with several of the WP pillars (like "WP is not an art gallery", "WP is not a soapbox", etc.). Any auxlanger who just reinvented Esperanto will argue that his project is going to save the world; and any artlanger who just created the weirdest of all weird languages will argue that his project could be interesting to people who like weird languages.

Another thing is that I disagree with several of the criteria mentioned in the list. For example, the "has a unique script" criterion IMO has no influence at all on the importance of a language. The list was basically an inventory of criteria proposed by several people.

I do however agree with your remark that an article should always mention WHY the language is important or interesting. In all conlang articles that I wrote for the Dutch wikipedia, that's what I have tried to do anyway.

BTW, Pete, why don't you take a user name? Now your votes and arguments are constantly ignored by anyone who doesn't like what you are saying. Frankly, with your state of service you deserve better!

--IJzeren Jan 20:15, 31 August 2005 (UTC)[reply]

I'm fairly new to wikipedia, so I wanted to find my feet a bit before creating an account. But since people seem to be valuing my contributions ... --PeteBleackley 10:33, 1 September 2005 (UTC)[reply]

Gaming the criteria

Okay, I looked at the various questions in my proposal, specifically from the point of view in how far they can be gamed by the creator of a conlang. This is what I think:

Number of speakers. Not very easy to manipulate. But it's pretty hard to get reliable data about the number of speakers of a language. Not only because we'd first need to define what level of proficiency of required, but also because it is virtually impossible to check it (esp'lly without original research). All in all, if someone claims more than 10 speakers, I think he or she better come up with some evidence.

Completeness. Well, if you build a conlang, you create words and expand the grammar. That's normal. I can't see how this would be "gaming the system".

For the non-conlangers out there: creating a complete grammar that isn't a clone of the grammar of your native tongue is very hard work! As for words, it is rather easy to spot a dictionary made with a random generator and a wordlist. Such conlangs even have a name (that I personally consider rather pejorative in fact): relex. This is why the "haha only serious" made up words that have occured in comments and in various RfDs are so ridiculous, only serving to prove that the commenters don't know what they are talking about. Words in a language never exist in a vacuum. If they aren't part of a system, it's not a language. In fact, if anyone managed to create a language whose words did exist in a vacuum, that in itself would be so notable that it alone would be a good enough reason for inclusion! Oh, and we should arrange for some way to send flowers to the asylum the inventor was locked up in, or ask him/her to repeal the third law of termodynamics while he/she was still on a roll. --Kaleissin 14:15:57, 2005-09-01 (UTC)

Uniqueness. All these things belong to the character of the language. I honestly cannot see how one would create a unique script, solely in order to score here. Besides, I personally have some doubts in how far uniqueness can really contribute to notability. To me personally, only the "exponent of a particular genre" is a valid criterion. And that's not something that can be gamed.

There are however people that only create scripts... and for some funny reason they usually have one set of 26 letters or two sets of 26 letters each... not copying your native system is hard. --Kaleissin 14:15:57, 2005-09-01 (UTC)

Publication. Yes, this can be manipulated by the creator himself. We have both editions without an ISBN, and editions with an ISBM that belong to the so-called "vanity press". But frankly, even those DO in my opinion add up to a language's notability. If you think that's not enough to warrant inclusion, then simply vote "MINOR".

Reputation.

The Top-200 at Langmaker.com - I don't know precisely how this works, and I understand why Jeffrey doesn't want to give details. All in all, I dó think it can be gamed, although I'm not sure to which degree; I have the impression that it requires quite some work for a new conlang to get it into the Top-100.
Caused controversy - Even if the creator dóes manage to cause controvery, I think this adds up to notability.
Googlehits, Independent discussions - Very easy to manipulate. But a little research can unmask quite of lot of creator-inspired discussion, if necessary.
Attention in the popular/scientific media - Hard to manipulate. Of course, an artist who does nothing to make his work known will probably have less success than someone who does. But that's normal and not restricted to conlanging.
Inspired other conlangs - As a means of gaming the system? No.
Established notability inside and/or outside of the conlanger community - To a certain degree perhaps. But this might just be where the Conlang Panel we discussed comes in.

Year of creation. No, impossible.

ISO code. I know far too little about ISO 639-3. The shortish list of conlangs that have an ISO 639-3 code looks pretty haphazard to me. And apparently it hasn't been worked on for years. The way it looks now, there is no way of manipulating this. Once the list gets some form, we might reconsider this point. But for now, I doesn't hurt.

Wikipedia. At present, it is so difficult to create a new wikipedia even in a natlang with few speakers, that I don't think there's anything to fear here. Should that change, so that anybody can have a wikipedia for any language, we could of course change the rules. But in the meantime, gaming this is completely impossible.

Notability by proxy. I honestly can't see how the creator of a conlang can manipulate this.

All in all, I think it's not at all that bad. Only Publication and Google Hits can really be manipulated by a conlang creator into some kind of fake notability. In all other cases I believe an "increase notability" is real. --IJzeren Jan 12:07, 1 September 2005 (UTC)[reply]