Invented Language Database

[elemtilas]

Hi. A group of us are starting a new Conlang Database, to be housed on the Language Creation Society's web space. We will progressively add basic data about the world's invented languages to a database which will be publicly searchable and able to produce reports involving stats or lists according to people's searches. If you'd like to be involved, come and join the Discord server: https://discord.gg/w4pAf3P

[bradrn]

I clicked on the invite link, and got a message: ‘Invite invalid’. Can you double-check if that link is correct?

As for the conlang database, a good place to start might be Ser’s ZBB Conlang Index: https://www.verduria.org/viewtopic.php?f=3&t=23

[Moose-tache]

Is this "database" a list of links, or will the data itself be stored on the LCS website? If it's the former, then why isn't this just a wikipedia page called "list of conlangs?" At least there it might get enough attention to prevent it from going to ruin like every other conlang list. If it's the latter, who owns the data? If the LCS makes ad revenue off of other people's copyrighted data, they'll need release forms that no one is going to sign.

I think Ser's list is a good example of what can go wrong (no offence, Ser). First, it's self-reported, so only people who see it can be on the list, and only people who want to be are on the list. This means that projects barely above the level of a scratch pad make the list, while complete conlangs frequently do not. Second, for several months after the fall of kneequickie the list was basically a museum of broken links. Does the LCS have any plan to prevent these problems?

[Man in Space]

If it's the latter, who owns the data? If the LCS makes ad revenue off of other people's copyrighted data, they'll need release forms that no one is going to sign.

Point of order: Languages are not copyrightable.

[Moose-tache]

If it's the latter, who owns the data? If the LCS makes ad revenue off of other people's copyrighted data, they'll need release forms that no one is going to sign.

Point of order: Languages are not copyrightable.

The language in an abstract way, no. But a written document about a language absolutely is. Wheelock's Latin, for example, goes for about $11 on Kindle. I doubt they would be happy if you pirated it and told them they can't copyright Latin. If you download someone's work to your servers and post to on your monetized website, you need their permission.

[elemtilas]

I clicked on the invite link, and got a message: ‘Invite invalid’. Can you double-check if that link is correct?

Link corrected in OP.

[elemtilas]

Is this "database" a list of links, or will the data itself be stored on the LCS website? If it's the former, then why isn't this just a wikipedia page called "list of conlangs?" At least there it might get enough attention to prevent it from going to ruin like every other conlang list. If it's the latter, who owns the data? If the LCS makes ad revenue off of other people's copyrighted data, they'll need release forms that no one is going to sign.

I think Ser's list is a good example of what can go wrong (no offence, Ser). First, it's self-reported, so only people who see it can be on the list, and only people who want to be are on the list. This means that projects barely above the level of a scratch pad make the list, while complete conlangs frequently do not. Second, for several months after the fall of kneequickie the list was basically a museum of broken links. Does the LCS have any plan to prevent these problems?

From what I understand:

1. the DB consists of several pieces of metadata: name, creator, date of creation, relevant link(s), writing system, Gnoli designation, source type, lexico-semantics, and development level achieved
2. dunno about data ownership -- will f/u
3. the DB is a mixture of self report and other report; the idea is to document as many invented languages as possible with the relevant metadata
4. this is not an LCS project per se; I believe the project's initiator plans to attract a core of dedicated / long term data aggregators and also a cadre of other interested persons; there is also a plan to regularly (but not overly frequently) advertise the DB on as many language invention forums as are known or yet to be revealed to the broader community. At the very least, associating with the LCS & storing the DB within their webspace will all but insure a permanent existence of the DB project.

[bradrn]

I clicked on the invite link, and got a message: ‘Invite invalid’. Can you double-check if that link is correct?

Link corrected in OP.

Yep, it works now!

Gnoli designation

What’s that?

source type, lexico-semantics

And what exactly are these supposed to mean?

[Kuchigakatai]

Is this "database" a list of links, or will the data itself be stored on the LCS website? If it's the former, then why isn't this just a wikipedia page called "list of conlangs?" At least there it might get enough attention to prevent it from going to ruin like every other conlang list. If it's the latter, who owns the data? If the LCS makes ad revenue off of other people's copyrighted data, they'll need release forms that no one is going to sign.

I think Ser's list is a good example of what can go wrong (no offence, Ser). First, it's self-reported, so only people who see it can be on the list, and only people who want to be are on the list. This means that projects barely above the level of a scratch pad make the list, while complete conlangs frequently do not. Second, for several months after the fall of kneequickie the list was basically a museum of broken links. Does the LCS have any plan to prevent these problems?

As mentioned above the plan seems to be it'll have a mix of data provided by the conlangers themselves and volunteers.

The problem I'd have with doing this on Wikipedia is their notability requirements, as I believe they'd insist on deleting such an article from existence, suggesting doing it on a Fandom.org site instead, an alternative that isn't that much better than an LCS-hosted site. I don't know if they plan to host it with ad revenue, but at least I've noticed they tend not to do that at least... Something worth asking.

Also, no offence taken! I've discussed the weaknesses of my little index before in fact. Part of the reason why it is the way it is is that I basically took up maintaining it after the original maintainer (some Israeli guy whose username I don't remember, not Astraios) stopped coming to the ZBB, and partly I've never bothered to improve it with metadata myself because people don't seem to use it much either. (I reckon right now that people would perhaps use it more if I did though...) By the way, one fun thing is that usernames with A- appear first and are therefore more prominent. In one of these discussions about my index, Pabappa joked he was going to change his username to Aaaaaa, or maybe just A, in order to appear first at the very top...

[Moose-tache]

Feel free to disregard my grumpy old lady attitude, but I think a list of metadata is pretty pointless. If I want to learn about a language, its date of creation isn't what I'm looking for. The dilemma is simply this: if you link to personal websites with more information, then you're just making a Richard Kenneway page, and it will be a graveyard of broken links by the end of the year. If you host the documents themselves, you will have issues with the creators, not getting permission from many of them or losing permission for no reason, and of course you'll have to constantly update the latest version of each language. The actual thing that is needed is exactly what is hardest to build: a downloadable collection of up-to-date reference grammars for a large fraction of the completed conlangs of the world.

[Kuchigakatai]

"Metadata" can be many things, I was thinking of features and characteristics there, which would still be useful in a website like the one you describe.

Your comment is very interesting though! That sounds like something that'd need to be done by asking people to upload their reference docs... after being told what will happen to them exactly... In a way you're talking about something like FrathWiki (maybe with stricter copyright... I recall masako was once bitten by its default use of CC-BY-SA), except accepting a wider variety of formats.

[elemtilas]

Gnoli designation

What’s that?

The Gnoli Triangle, a spacial way of describing invented languages. Essentially a three-point continuum between Artistic, Engineered, and Auxiliary Languages.

source type, lexico-semantics

And what exactly are these supposed to mean?

Those are basically how relatively naturalistic & priori/posteriori the language is.

[elemtilas]

Feel free to disregard my grumpy old lady attitude, but I think a list of metadata is pretty pointless. If I want to learn about a language, its date of creation isn't what I'm looking for. The dilemma is simply this: if you link to personal websites with more information, then you're just making a Richard Kenneway page, and it will be a graveyard of broken links by the end of the year. If you host the documents themselves, you will have issues with the creators, not getting permission from many of them or losing permission for no reason, and of course you'll have to constantly update the latest version of each language. The actual thing that is needed is exactly what is hardest to build: a downloadable collection of up-to-date reference grammars for a large fraction of the completed conlangs of the world.

At the very least, if you want to learn about a language, and you don't even know it exists, a resource like this one can at the very least enlighten you (or tantalize you mercilessly) with the existence of and some interesting information about the language. As for collections of up to date grammars, best bet will be Frath Wiki itself. Other than that, as you say, you'd be at the whims of linkrot.

As for the current project ending up like Langmaker, I think there is a distinct advantage now. Whenever anyone makes some wonderful resource like that, the risk is that this one person will somehow disappear or otherwise stop work on it. It's a lot for one person to maintain over time. In some way associated with or attached to the LCS, at least we don't have to worry about the DB itself disappearing, as it will be maintained as long as there an LCS in business. There's also a greater likelihood that someone down the line can be attracted to the project to take it over.

[Vardelm]

Seems to me that there needs to be a website that allows people to create a profile and then create an entry of one or more conlangs associated with that profile. Give them a series of options they can choose from that generally describe the language (apriori vs posteriori, art vs eng vs aux, alignment, etc.; think like a simplified WALS) and allow links to whatever webpage/doc they want, but it won't be hosted there. It also shouldn't be so much information that it equates to a grammar. This distributes the maintenance & places the onus on the author to update, and if it's never updated and links go defunct, maybe admins can just flag it as a potentially dead conlang, but you still have a record that the conlang was a WIP at one point in time.

[bradrn]

Seems to me that there needs to be a website that allows people to create a profile and then create an entry of one or more conlangs associated with that profile. Give them a series of options they can choose from that generally describe the language (apriori vs posteriori, art vs eng vs aux, alignment, etc.; think like a simplified WALS) and allow links to whatever webpage/doc they want, but it won't be hosted there. It also shouldn't be so much information that it equates to a grammar. This distributes the maintenance & places the onus on the author to update, and if it's never updated and links go defunct, maybe admins can just flag it as a potentially dead conlang, but you still have a record that the conlang was a WIP at one point in time.

Like this? https://cals.info/

[Vardelm]

Like this? https://cals.info/

Actually, yeah, I guess so!

The UI needs an intervention, and personally I find the WALS categories a bit wonky. Since this exists, though, I don't see a huge need for another database, unless said database has features that make it more useful and user friendly than CALS. I'd focus on upgrading the CALS sight, if the admin is open to such. Creating another database sight just fragments the community & sort of defeats having a database int the 1st place.

EDIT: Or you work with the CALS people to make something to replace CALS.

[Moose-tache]

I'll stick to my current method: mailing a letter to Jeffrey Henning asking him to list all the conlangs he can think of.

[Qwynegold]

date of creation

I never know what to make of this. No one makes a conlang in one day, so what are you supposed to enter? The date you began the project, or the time frame during which you have been working on it?

[Vardelm]

I never know what to make of this. No one makes a conlang in one day, so what are you supposed to enter? The date you began the project, or the time frame during which you have been working on it?

We should have:

• date of initial concept/idea
• date of initial commencement of development
• date of initial major revamping due to disatisfaction
• date of most recent major revamping due to disatisfaction
• date of primary moment of inspiration that made the lang "gel"
• date of initial public release
• date of initial public feedback stating "this doesn't suck" or similar
• date of completion of initial, "full-ish" grammar documentation
• date of most recent, "full-ish" grammar documentation
• date you realized you would probably never get a smiley award for the language

[Qwynegold]

lol

What is gel though?