[Tagging] Which languages are admissible for name:xx tags?

[Jeremiah Rose]

Hello, sorry I'm late to this discussion. I added a bunch of Klingon language country names to OSM a few days ago, which were reverted.

Regardless about whether the OSM community think names in {tlhIngan Hol} should get the {HISlaH} "yea" or {ghobe'} "nay", I think it's important that any rule about multilingual names should be written in such a way to not exclude indigenous and minority language communities. What counts as a "full" or "real" language is actually a hard problem even among professional linguists. Ultimately whether a language is a language doesn't have to do with how it came to exist, it has to do with whether it has a lexicon and grammar that can be used to communicate, and whether it is used among a community of people. Making rules about what kinds of communities or domains of use should count as valid quickly becomes a thorny political exercise. There are a number of indigenous language communities who do not have living native speakers due to historical efforts to suppress their languages, but have active language development and language revival efforts, and could want to make use of maps for them. Excluding minority languages from official recognition and contemporary media is exactly the kind of thing that has been used to suppress them in the past. It can be difficult to write sensible rules excluding conlangs that wouldn't impact such communities. I would object to rules that require large numbers of native speakers or first-language speakers for "notable" languages, for reasons that have nothing to do with constructed languages.

Less seriously, while nations, major cities, and famous places do have a lot of language tags, I don't think a flood of conlang names is actually a problem. To my knowledge Sindarin doesn't have names for Earth places because it is a language of Middle-Earth and the elves were more concerned with places like Thangorodrim than ones like Barcelona. This is often the case with alien or fantasy conlangs—Dothraki don't need directions in Denver. Auxlangs like Esperanto conceive of themselves as international projects and thus may have a large number of names for countries. However, as a practical matter, there are only a few historically important conlangs that have a ISO 639-3 code at all. While there are ongoing revisions to those standards, unfortunately in recent years the body has consistently rejected new codes for conlangs. Specifically, a number were proposed in 2017 but rejected under new criteria adopted at that time. Esperanto has first-language speakers and would meet that criteria, but Klingon has been alive only for decades, not for generations yet. So it's very unlikely that aside from a few grandfathered exceptions, there will be new ISO 639-3 codes for conlangs for people to worry about, at least for a few decades. A more urgent problem might be how to support names in conlangs and other languages that don't yet have ISO 639-3 codes.
https://iso639-3.sil.org/code_changes/change_request_index/data/2017
https://iso639-3.sil.org/sites/iso639-3/files/criteria_for_coded_languages.pdf

While the toponymy of {Qo'noS} is well outside the scope of OSM, certainly some Klingon country, city, and state names for places here on {tera'} are verifiable within the contemporary human Klingon community. The Klingon language community runs the Klingon Language Institute, there is a very active Learn Klingon Facebook group, klingonwiki.net, and two major annual conventions: {qep'a'} "Great Meeting" in Indianapolis in July, and {qepHom} "Little Meeting" in Saarbrücken in November, as well as occasional smaller {qepHommey} "Little Meetings". Klingonists are active on social media platforms such as Twitter and YouTube as well as producing podcasts, zines, books, operas, and plays. (You may not have seen or participated in fluent conversations conducted in Klingon, but I certainly have, and I'd be happy provide examples.) One could easily approach Klingonists in any of these venues about these names and get the same answers. Maintainability is a problem with a variety of data in OSM, but the major conlangs have very active online communities in general, and I wouldn't have put the Klingon names in if I didn't intend to keep them updated. The reason I added Klingon country names to OSM is because I've been an active OSM editor for years with thousands of edits, am also a Klingonist, and thought that adding Klingon-langage names would enable the Klingon-language community to make use of OSM in cool ways. {'etlh QorghHa'lu'chugh ragh 'etlh nIvqu' 'ej jejHa'choH.} [1]

In Klingon, we don't have more than a few dozen agreed-upon names for countries as well as a handful of major cities and regions. The usual process for new vocabulary, including toponyms, in Klingon is to submit suggestions for new words to the {chabal tetlh} in spring for discussion and clarification, and an official list of {mu'mey chu'} "new words" is released at the {qep'a'} convention in July, so that new words can be standardized in the community. I've suggested some, because it's awkward when people drop into English just to say a place name. This is a slow process, and anyone going rogue to make up words for random places in the world would have them disputed by the Klingon community as a whole.  

As far as the copyright issue goes, languages are not copyrightable, although dictionaries and fixed works describing them are. This has not been definitively settled by litigation for conlangs; the Loglan/Lojban dispute and the Axanar lawsuit is the closest it has come in recent decades, but the issue was rendered moot ( https://conlang.org/axanar/ ). I included a link to the KLI New Klingon Words List to document their verifiability within the Klingon community, not imagining the names of countries would be taken as copyrightable material. Most skilled Klingonists make their own lists of official words including toponyms, and it would be easy to add Klingon names without relying on that source.   

If an item must have a Wikidata entry to have a name:XX=* tag, would it need to have a Wikidata entry to have a name=* tag at all? It would represent a very large change to how multilingual information is represented in OSM. This is a different issue to whether languages should be excluded from OSM.

>* "serious" constructed languages (Esperanto)
{tulwI' Hol Sagh law' tlhIngan Hol Sagh puS net jalchugh, not SuvwI' Hol jatlhbogh ghot qIHlu'pu'!} [2]

As far as the other examples go, FredLang-you-just-made-up doesn't have a language community with which it could be verified. I can think of no reason why languages like Esperanto and Latin should be excluded. Latin is used in very restricted domains these days, but lots of people still learn and use it; there are many Latin experts to verify names with. They are otherwise tagged appropriately within the existing standards of OSM name tagging. I'm more curious about the use of name:etymology=* than name:la=*. What criteria would be used to exclude languages like Esperanto and Latin, and what is the rationale for that criteria?

Jeremiah

[1] "Even the best blade will rust and grow dull unless it is cared for."
[2] "If someone thinks Esperanto is more serious than Klingon, they have never encountered a person who speaks the warrior's language!" ;)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20200328/61ec2ae2/attachment.htm>