An deasbaireachd aig Uicipeid:Ainmean chànan

This came up offline but we better put it here for the record. The issue of language names, sort of following on from the above.

General Observations

I'd like to propose an approach broadly based on the conventions used by various software developments, including Microsoft, which is a "moderate" approach to translating language names. In a nutshell, Gaelic forms are used for:

• Main official state languages (Spàinntis, Sìnis, Amthara...)
• Most other European living languages (Basgais, Catalanais...)
• Languages particularly relevant/known in the Gaelic world (Laideann, Eabhra...)

Beyond that, the native language name or the English form is favoured for a variety of reasons. "Luwian" just doesn't come up often enough for people to recognise or remember it and most are night impossible to accommodate withing Gaelic orthography and phonology that there's no point really.

• Native form, especially if used in English (Xhosa, Zulu, Kinyarwanda, Hmong...)
• Otherwise English form (Guaraní, Luwian, Udmurt, Dyirbal...)

Thoughts? Akerbeltz 15:56, 18 dhen Ghearrain 2012 (UTC)

Sounds good to me. Maybe it could go to another page, Uicipeid:Ainmean chànan, under Uicipeid:Poileasaidhean na h-Uicipeid. On second thoughts, it could include names of countries too. Anyway, two questions:

What if there are more forms? I think I've seen both Spàinnis and Spàinntis, Seacais/Seicis/Teacais or Mongòlais/Mongoilis - can you think of a general rule or would we treat them case by case?

Will we use English even where it has the typically English -an or -ese suffix, when the native form isn't in Latin script, eg Chittagonian and Assamese? --Thrissel 17:03, 18 dhen Ghearrain 2012 (UTC)

Ya, we'll move it to a page like that at one point. I kinda like keeping debates here because more (relatively speaking) people watch this page.

Ignoring the ghastly Wallmap, the closes thing to a standard list is the one Microsoft collated. Redirects rule. That aside, my *only* red line is that I don't think forms which break Gaelic phonology/text to speech rules are good. It messes with the language. I usually go for Spàinntis because Spàinnis is a regular cluster simplification, like coltas > collas; so speakers can derive both the form with and without t from the same spelling but not the other way round.

I think for the most part, we'll go with the English form even in -ese cases (Asamais is covered by relevant/known but you could persuade me otherwise). -an forms we should debate case by case, usually these are better formed using the country name in Gaelic e.g. Cànain Chittagon.

Another option, which I could live with is suffixing -(a)is and -(e)ach but onto unchanged roots i.e. Luwis, Chittagonach, Assamais, Samoyedais etc). Akerbeltz 17:58, 18 dhen Ghearrain 2012 (UTC)

Hi, preparing the [list] took a lot longer than I am prepared to confess to. Here are some observations:

1. I tried for the majority of the languages to locate existing versions of the names. My preferences in order typically werefor (i) an existing and established Gaelic name where I knew or could locate one, (ii) a Galicised version of a name in wide international usage, comforming to Gaelic orthographical conventions (of which I don't profess to have any great knowledge, (iii) the English version of the name, where there were no suffixes, or the suffixes were local.
2. It is work in progress, and therefore even aggressive editing - and particularly the addition of missing information - is very much welcome. I had reached a stage in compiling the data for this list where I felt that it would be good to get feedback before going further.
3. Consequently, I have been very reluctant to create dead links, particularly where there is no settled name for the language.
4. We should leave the door open to the discovery of Gaelic forms and pay particular heed to early attested forms, for example, for the Indo-Ayran languages, from the Gaelic diaries, letter, or memoirs of colonial administrators, traders, missionaries, philologists and so on, if such documents exist.
5. I think we should avoid English suffixes as much as possible. It would seem that it would make more sense linguistically to borrow from the language itself, e.g. Bagla, or from other Celtic languages. Daibhidh mac Uisdean 21:08, 18 dhen Ghearrain 2012 (UTC)

Okay, just my two cents:

Most other European living languages (Basgais, Catalanais...): I wouldn’t keep it to living European languages only, East Germanic languages could easily be tranlated, altought they are not longer spoken. And Sòrbais Ìosal is understandable as well, altought it is not found in a dictionary. In the infobox for languages we might run into some other languages like "Central Ripuarian": Ripuarian Meadhanach or English version?

Native form, especially if used in English (Xhosa, Zulu, Kinyarwanda, Hmong...), okay, Kölsch, Esperanto....

Another point would be: Languages with or without article? Gàidhlig had been moved to A' Ghàidhlig and back againn quite often. I would prefer without article. --Sionnach 21:40, 18 dhen Ghearrain 2012 (UTC)

Without article, with extreme prejudice. It serves no purpose except making the use of search machine less user-friendly. Just as most people wouldn't expect to find en:Sun or de:Elbe under en:The Sun or de:Die Elbe. In fact we have far too many pages starting with an article which I would like to see moved, e.g. names of diseases. --Thrissel 22:09, 18 dhen Ghearrain 2012 (UTC)

Diseases are likely a special case as the article is often the single distinguishing element vs a common noun, e.g. a' bhreac (smallpox) vs breac (trout). So we have to tread carefully. Akerbeltz 01:23, 26 dhen Ghearrain 2012 (UTC)

Policy Discussion

Moving a draft to below the debate as before, please post above it.

• Established names, yes, assuming you mean names which have been around 50+ years like Eadailtis, Arabais etc
• The door will be open but I've done a lot of work in this area, especially when it came to compiling lists for Microsoft and I doubt there's any big surprises waiting for us on a dusty shelf. In many cases, such old forms are ad-hoc, idiosyncratic and without much reference to phonology etc - bad transliterations for the most part. We need to be wary of those.
• I'm taking a loose view of "European" but as East Germanic is dead and apart from Gothic (Gothais seems fine to me) fairly unknown. The key issue is that Gaelicized forms of little known languages are not going to gain currency this century, if ever. For the most part, even German linguists have given up on adding -isch to all but the most widely known languages - Dyirbalisch just makes no sense in German and we're talking a bit language here. But I also take Daibhidh's point about the suffixes, which is why I mentioned the possibility of using non-Gaelic roots plus a suffix. I've added a proposal below, tell me what you think.
• Article... redirects rule. I tend to agree with Sionnach that it makes more sense to use the name without the article but we should list both forms in the lede (e.g. 'S e Gàidhlig no A' Ghàidhlig...) Akerbeltz 22:25, 18 dhen Ghearrain 2012 (UTC)

I guess we can summarise Akerbeltz's rule hierarchy as follows: first, established Gaelic names; second Gaelicised English forms, where English adds a suffix; and finally the native form as represented in other European languages. I basically agree with this but have a proposal and question.

1. Firstly, where we have any early attested form coined or used by a native speaker - unless absurd - we should encode this in the language, in order to preserve important assumptions about naming conventions, the speaker's own pronunciation, position on issues of standardness/nonstandardness and so, essential linguistic data. (Before I labour the point does anyone know of any such material in Gaelic from colonial India, Afganastan etc.? This whole debate may be academic.)
2. Clearly older names restrict themselves to the eighteen letters of the Gaelic alphabet. Why are newer names not subject to the same restriction? Daibhidh mac Uisdean 03:29, 19 dhen Ghearrain 2012 (UTC)

My proposed rule therefore - in simple form - is:

1. Established Gaelic, then
2. Any attested early form, then
3. Gaelicised English, then
4. Native form. Daibhidh mac Uisdean 03:29, 19 dhen Ghearrain 2012 (UTC)

Just because a language is dead, doesn't mean that we can't write an article about it. Would East Germanic be Gearmanach an Ear no Gearmanach Earach or would it be in English because it is not an established form in Gàidhlig?

There is a need for two separate things, articles for extinct or proto- languages, e.g. proto-West Germanic, and another for the current West Germanic language family. Daibhidh 12:42, 19 dhen Ghearrain 2012 (UTC)

Any attested early form: Hm, I would suggest to add this into the article, just in case it ever happens that there is more than one source, we might be in trouble to decide with one to use as article name. --Sionnach 09:02, 19 dhen Ghearrain 2012 (UTC)

Moved Daibhidh's comment (sorry, was unclear, I meant "post below last comment and above Dreach nan riaghailtean)

Sionnach has a good point about language families. I think that can be incorporated along the lines of "linguistic family names will be based on existing Gaelic root forms (e.g. Gearmanach, Sìona-Thibeiteach...) but non-Gaelic root forms in most other cases (Vasconiceach, Siouanach, Inuiteach...)

The problem with just using forms that some author at some point coined is that just because a writer comes up with something, doesn't necessarily mean it's a good idea. When faced with non-native names, the first people who create written forms are often ill-equipped to do so. Just think of the way English-speaking map-makers mangled Gaelic placenames. Or the way colonial administrators mangles Cantonese. "Hong Kong" (coined on the spot and by people with no linguistic knowledge) is a very long way from "Heung Gong" (coined much later when people started developing "proper" romanization schemes). Some of the old names may indeed be useful but need to be handled with care, we don't want to break Gaelic any more. In any case, as I mentioned it's extremely unlikely that we'll suddenly unearth a Gaelic form for Menominee or Tarahumara. But I'll stick something in. Akerbeltz 10:21, 19 dhen Ghearrain 2012 (UTC)

Ok, tweaked. Also added something about Biblical forms which are a surprisingly useful source. Akerbeltz 10:53, 19 dhen Ghearrain 2012 (UTC)

You helped me out with Hitis! Daibhidh 12:42, 19 dhen Ghearrain 2012 (UTC)

Sorry, maybe this is nit-picking or maybe I have an off-day, but the way rule 2) is written implies to me we use any kind of original writing system, and only if there are two forms we use a transcription to Latin alphabet by ISO-9 - so rule 3) applies to what "others"? I mean, couldn't Gujarati be just as well under 3a) and Udmurt or Dyirbal under 2)? --Thrissel 17:13, 19 dhen Ghearrain 2012 (UTC)

Good point. Tweaked. Sometimes the forms used in English is identical to the native form, e.g. Lakhóta, Sesotho etc, so yes, there will be overlaps. Now, deep breath, a sip of slivo... shlibo oh that stuff you make with plums. Well, mostly plums - and you'll feel better :) Akerbeltz 17:36, 19 dhen Ghearrain 2012 (UTC)

Actually the guy next door did bring some slivovica, but that was three and a half months ago and I doubt there's any left. Anyway, it's better now but even with the help of... the barley stuff I'd probably still miss in the rules a reason for having Guaraní under 3) rather than Avañe'ẽ under 2) - provided en-wiki's got the the endonym right. --Thrissel 19:24, 19 dhen Ghearrain 2012 (UTC)

It meant to say under 2) that native names will be given priority if (amongst other) they are close to the English form e.g. Lakhóta over Lakota/Dakota but Avañe'ẽ is so distant from the English form that Guaraní should be used. I'm getting nowhere with work today either, maybe I should take something distilled myself. Akerbeltz 19:38, 19 dhen Ghearrain 2012 (UTC)

I added a proposal about initial bh, ch, dh, fh, gh, mh, ph, sh, th. Thoughts? We seemed to have reached a lull so I'll probably move this to a new policy page soon. Akerbeltz 10:37, 25 dhen Ghearrain 2012 (UTC)

Just a very minor point under 3 b): It would be nice to add examples of the the use of -ish, and -(e)ach as well, because it would make it easier to spot the right one. So far there are Gilbertese -> Gilbertais and Ripuarian -> Ripuarais. The rest is fine to me.--Sionnach 07:51, 26 dhen Ghearrain 2012 (UTC)