WT:LTD redirects to this talk page, to ease using the aWa tool to archive discussions here. For more archived discussions, see the corresponding Wiktionary:-space page, Wiktionary:Language treatment/Discussions.
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
The term "Old East Slavic" must have been created to avoid offending Ukrainians and Belarusians and to please nationalists who use "Old Ukrainian" and "Old Belarusian", since "Old Russian" may imply to some that Belarusian/Ukrainian languages are derivations of Russian. However, "Old East Slavic" (Russian: древнеру́сский язы́к(drevnerússkij jazýk)) translates into Belarusian as старажытнару́ская мо́ва(staražytnarúskaja móva) and into Ukrainian as давньору́ська мо́ва(davnʹorúsʹka móva) (alternative names also exist, including "Old Ukrainian"). All three terms are literally translated as "Old/Ancient Russian/Rusian", not "Old East Slavic". Note that in Ukrainian ру́ський(rúsʹkyj) refers more specifically to Rus rather than modern Russia. (Unlike Russian and Belarusian, the Ukrainian term росі́йський(rosíjsʹkyj) means "Russian" both referring to ethnicity and the country. Cf. Russian ру́сский(rússkij) / росси́йский(rossíjskij) and Belarusian ру́скі(rúski) / расі́йскі(rasíjski)). Despite the current tensions between Russia and Ukraine, linguistically it makes much more sense to rename "Old East Slavic" to "Old Russian". Call me biased, whatever, but respectable East Slavic linguists all use "Old Russian", it's of the Ancient Rus, not modern Russia, the centre of which was in Kiev, modern Ukraine. :) --Anatoli(обсудить/вклад)02:47, 12 March 2014 (UTC)Reply
I don't think the names that modern speakers use should really count. They're bound to have some kind of national pride in them, like discussions on the Wikipedia article show. (And I'm hoping to avoid a straw man, but by the same logic we'd call Dutch "Netherlandish", German "Dutch", Greek "Hellenic", Armenian "Hayeric", Hittite "Neshite", and so on...) I think we should focus purely on what modern English-language scholarship calls the language. —CodeCat03:23, 12 March 2014 (UTC)Reply
Re: English language usage. Ngram, Ngram 2 can't even find "Old East Slavic (language)", not sure if I interpret this correctly or if it is possible to check the English usage via Google.
Being half-Russian, half-Ukrainian, I take pride in belonging to both but pride can take various ways. E.g. ру́ський(rúsʹkyj) is sometimes translated by Ukrainians as давньоукраї́нський(davnʹoukrajínsʹkyj, “Old Ukrainian”) and some Ukrainian nationalists claim they have nothing to do with Russian but others take pride in being the cradle of all Slavs. We all know the Serbo-Croatian language story and Israeli-Palestinian conflict doesn't make Hebrew and Arabic belong to different language families. Modern Ukrainian, Russian, Belarusian are distinct languages but they are all derived from one, so your comparison is not valid.
In any case, Belarusians (educated), considering the name of the language, are not even trying to deny that Belarusian is derived from Old Russian (=Rusian, Old East Slavic), the language of Rus, in Ukrainian, the term ру́ська мо́ва(rúsʹka móva) is also used, without "Old", since ру́ський(rúsʹkyj) has this meaning ((modern) Russian language is called росі́йська мо́ва(rosíjsʹka móva) in Ukrainian). Ukrainians also know that Украї́на(Ukrajína) is quite a new term, although there are some ridiculous claims about so-called "Ukr" tribes, derived directly from Proto-Slavic, bypassing Old Russian stage. I wonder what discussion you're referring to?
BTW, I disapprove the policy of Russian government and respect the right of Ukrainians to decide their fate but we are talking about language names and linguistics here. Bad timing for this discussion, considering the Ukrainian and Crimean crisis. --Anatoli(обсудить/вклад)04:22, 12 March 2014 (UTC)Reply
I think this should be decided solely by the most commonly used name in English. Thus, I tentatively support "Old Russian" based on the Ngrams above, and based on a hunch that English speakers are less likely to have to look up the language if it is called "Old Russian" than if it is called "Old East Slavic". However, there needs to be a more thorough investigation into the evidence, since Ngrams are often inaccurate and/or skewed by unexpected factors. --WikiTiki8904:55, 12 March 2014 (UTC)Reply
Agreed but Google books count of "Old Russian language" is also significantly larger than "Old East Slavic language" (even though "Old Russian language" may not always have the same sense as "древнерусский язык"). I'm happy if someone else can analyse it better using Google or other sources. --Anatoli(обсудить/вклад)05:02, 12 March 2014 (UTC)Reply
Re: solely by the most commonly used name in English. We should also take into consideration the literal names in the affected languages. I think "Old East Slavic" is a result of some kind of political correctness or fairness, ignoring the fact that Ukrainian and Belarusian don't use this fairness in the name, also not note French "vieux russe", Estonian "vanavene keel", Dutch "Oudrussisch" and many Slavic names. The rest of names must have followed English Wikipedia in their namings. --Anatoli(обсудить/вклад)05:21, 12 March 2014 (UTC)Reply
I disagree, whatever the origin of its usage, if "Old East Slavic" became were shown to be used more commonly in English, then that is the name we should use. The only other factors we should consider are the ambiguity of the name (for example, if "That Language There" were the most commonly used term for some language, we probably still shouldn't use it), or possibly some other minor issues that I can't think of at the moment. --WikiTiki8905:30, 12 March 2014 (UTC)Reply
I meant it seems it was coined out of some considerations but I agree that if it were more common we should use it. I can't think of another example but in Japanese ハングル語 (Hanguru-go) "Hangeul language" was specifically coined, out of political correctness, to avoid giving preferences to either Korea or their languages (North and South Koreas have different names in East Asian languages, see 韓国 and 朝鮮). --Anatoli(обсудить/вклад)05:38, 12 March 2014 (UTC)Reply
If you agree that if it were more common then we should use it, then isn't that the same thing as only relying on which is more common? --WikiTiki8905:43, 12 March 2014 (UTC)Reply
It is not easy to get a sense of which name is most common in English. Chaff like "an old Russian car" skews counts of "Old East Slavic" vs "Old Russian", but if "language" is added, the number of hits drops so low that it is statistically insignificant / unreliable. However, based on manual review of a search with "in", I think "Old Russian" is the more common of the two names. - -sche(discuss)06:25, 12 March 2014 (UTC)Reply
-sche's data
On BGC, "Old Russian language" gets 4 relevant hits, 7 chaff hits of things like "thirty-three-year-old Russian-language specialist" or "the old [=venerable] Russian language", and 2 hits alongside "Old East Slavic language" in Library of Congress subject heading catalogues.
"Old East Slavic language" gets 3 relevant hits, 6 hits in LOC subject heading catalogues or printed editions of Wikipedia, and 1 hit that actually uses both terms: "the purely vernacular, Old Russian — or, more precisely, Old East Slavic — language".
"in Old East Slavic" gets 13 relevant hits, 2 in books that also use "Old Russian".
"in Old Russian" gets 32 clearly relevant hits, 25 clearly irrelevant hits (like "dressed in old Russian uniforms") and 13 hits where its not clear whether it means Old Russian or just Russian that is old, and that's just the first 7 pages: I stopped counting after that because it was clear at that point that "Old Russian" was more common than "Old East Slavic".
I support the rename to Old Russian per the following Google Scholar Data, showing that "Old Russian" is more than 10 times as common as all other names I could think of, combined. —Aɴɢʀ (talk) 19:20, 12 March 2014 (UTC)Reply
Google Scholar data
"Old Russian" language – 13,200 hits
"Old Ukrainian" language – 847 hits
"Old East Slavic" language – 167 hits
"Old Belarusian" language – 139 hits
"Old Ruthenian" language – 96 hits
"Old Byelorussian" language – 38 hits
"Old East Slavonic" language – 26 hits
"Old Eastern Slavic" language – 9 hits
"Old Eastern Slavonic" language – 4 hits
“Old Russian” reflects systemic prejudices that go back to times of the empires, when Encyclopedia Britannica defined wrote about White Russians and Little Russians, some of whom were Ruthenians. These prejudices were still felt in academics and journalism when the Soviet Union broke up, and are only going away now. I don’t think Wiktionary is part of an establishment that tries to wilfully reinforces these backward practices.
Furthermore, the moment when the Russian regime is exploiting such ideas in its anti-Ukrainian propaganda is the shittiest possible time to start this discussion. —MichaelZ. 2014-03-12 19:18 z
I feel kind of the same about it. I can't motivate the choice with widespread usage, but I feel that "Old East Slavic" just is a better, more descriptive name for the language. So I oppose. —CodeCat19:42, 12 March 2014 (UTC)Reply
@Mzajac I agree about bad timing and with the criticism of the Russian regime and I'm very sad about the latest events - Putin, in a short period, has created a situation, which was unimaginable for hundred years - Russians fighting Ukrainians. I wish we didn't mix politics in here, though. The political split of Yugoslavia caused the artificial language split but we all know that Serbo-Croatian is one language, despite the tensions. Jewish and Arabic are still semitic languages and Ukrainian and Russian are still derived from the same source (after a big injection of Old Church Slavonic into Russian and Polish injection into Ukrainian). I'm not stating any Russian supremacy or support any name-calling. Regarding the names, I disagree that "White Russians" and "Little Russians" were a result of prejudice, just as "Белару́сь" (Бе́лая Русь) and "Малоро́ссия" are just historical words, names among others like "Great Rus", "Red Rus", "Black Rus", "Carpathian Rus". Nikolay Gogol, native of modern Ukraine, also used "Малороссия" when referring to his homeland. The "offensive" meaning (if "Малоро́ссия" ("Little Rus(sia)" sounds offensive to Ukrainians) was acquired mistakenly in the modern times. In any case, Русь(Rusĭ) originated on the territory of modern Ukraine by people who lived there and this is where all East Slavs originate from. I don't see or Ukrainians don't see anything offensive in terms "давньору́ська мо́ва" and "ру́ська мо́ва" when referring to Old East Slavic language but I won't insist on continuing this discussion, if it hurts someone's feelings. --Anatoli(обсудить/вклад)02:10, 14 March 2014 (UTC)Reply
Guys, call it Old Bulgarian and that's it. Though modern Bulgarian belongs to the southern branch, the early Bulgarian and Russian texts do show their being one language. Alexdubr (talk) 14:31, 17 March 2014 (UTC)Reply
You are confusing two different languages. Old Bulgarian, which we call Old Church Slavonic, was also a South Slavic language, but it was used even in Russia/Ukraine/etc. as the main written and liturgical language. The main spoken language, however was Old Russian or Old East Slavic, which was an East Slavic language and the ancestor of modern Russian/Ukrainian/etc. --WikiTiki8916:14, 17 March 2014 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Merge Komi-Zyrian (kpv), Komi-Permyak (koi), Komi-Yodzyak (no code) with Komi (kv). Komi-Zyrian is dominating over others. Currently, they all use the same alphabet but there were differences in the past. There's very little information on the grammar differences but they are mostly considered dialects, not separate languages. Merging kv and kpv should be straightforward, Zyrian is the language of Komi Republic.-Anatoli(обсудить/вклад)10:39, 6 April 2014 (UTC)Reply
Actually, it has been planned as far as Komi-Zyrian and Komi-Permyak are concerned, see WT:LANGTREAT: "kv and kpv refer to the same lect; one will eventually be deleted.". Apparently, there are differences with Komi-Permyak but those can be labeled as "Permyak", similar Serbo-Croatian or Albanian varieties. --Anatoli(обсудить/вклад)01:54, 7 April 2014 (UTC)Reply
Komi-Permyak has a distinct written tradition and should not and cannot be merged. Your statement that Komi-Zyrian is dominating shows your pro-Russian and anti-minority POV which should never be a basis for consensus on this wiki. Komi-Zyrian can be handled under {{kv}}, yes. -- Liliana•12:34, 7 April 2014 (UTC)Reply
Do you have paranoia or something? I'm not Putin and not suppressing any minorities. Komi-Zyrian is a language of majority of Komi people in Komi Republic, has a much larger number of speakers and much more materials written in this variety of Komi. Komi-Permyak is used by a minority in Perm Krai and has only 63,000 speakers. By all means, I'm not forcing anyone to merge, this is only a suggestion. They are mutually comprehensible and each of them has dialects. It's possible to merge Komi varieties like it's possible to have one L2 header for Albanian, Norwegian or Serbo-Croatian, marking varieties accordingly: тӧлысь (Zyryan)/тӧлісь (Permyak), выль (Zyryan)/виль (Permyak). --Anatoli(обсудить/вклад)12:58, 7 April 2014 (UTC)Reply
That was completely rediculous and has no place on this wiki. It shows your anti-Russian bias more than Anatoli ever showed any pro-Russian bias. —CodeCat13:00, 7 April 2014 (UTC)Reply
Thanks. I can't help to have some Russian bias, though. This is my language, my culture. I can help with the Russian language and Russia-related topics, even if Russian may now be interesting perhaps as a "language of enemy" for some. I don't blame people for criticizing Russian politicians looking at Russia with suspicion. I don't support Russian politics and propaganda but I don't have to apologize for being Russian either. :) Certainly I shouldn't be blamed for being against minorities. Why would I add contents for minority languages, if I wanted to supress them? --Anatoli(обсудить/вклад)13:19, 7 April 2014 (UTC)Reply
There is one problem: as WT:DATACHECK reveals, "Komi-Zyrian language (kpv) has a canonical name that is not unique, it is also used by the code kv." Presumably one of the two needs to be retired (or, failing that, at least renamed). - -sche(discuss)18:52, 2 May 2014 (UTC)Reply
Komi-Zyrian is less ambiguous and clear. Although I favoured "Komi", perhaps we should retire it and leave Komi-Zyrian for kpv and kv and Komi-Permyak for koi. Komi-Zyrian is implied if Komi is used. --Anatoli(обсудить/вклад)00:12, 3 May 2014 (UTC)Reply
I support a merger of Campidanese (sro) and Logudorese (src) into Sardinian (sc), for reasons I outlined on my talk page and repeat here for others' benefit: those two lects differ from each other, quoth WP, "mostly in phonetics, which does not hamper intelligibility among the speakers". They are perhaps comparable to the dialects of Irish, with standard Sardinian (sc) existing as a unification of the dialects (again, comparable to Irish). Notably, we already include standard Sardinian (sc) — which means the additional inclusion of sro and src is quite schizophrenic.
There is some disagreement over whether Gallurese (sdn) and Sassarese (sdc) are dialects of Sardinian, dialects of Corsican, or languages separate from both Sardinian and Corsican. I would not mergethem at this time. - -sche(discuss)03:19, 12 May 2014 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
@-sche (and anyone else who loves language-name conflicts), we currently call aum "Abewa" and asa "Asu". Firstly, it seems that practically nobody calls aum by that name; Wikipedia and Ethnologue both titles their entries for it "Asu", and Roger Blench, who may be the only person ever to study it, calls it "Asu" as well. As for asa, which currently occupies that name, Wikipedia and most hits on Google Books call it "Pare" (the remainder call it "Asu" or "Chasu", for the most part). As you may have guessed, there is already a language that we call "Pare", namely ppt, but this New Guinean language is also called "Akium-Pare" and (Wikipedia's choice) "Pa", which thankfully appears to be untaken. The chain of changing language names does seem rather silly, but the overall purpose of this is to move the only language out of these three that actually has any literature on it (and thus the one I just added a translation in), namely asa, to its most commonly used name. —Μετάknowledgediscuss/deeds05:40, 9 September 2015 (UTC)Reply
I'm down with renaming aum away from Abewa.
On a balance, I'm also OK with renaming asa away from Asu. I can find a decent amount of references to "Asu": it's hard to say whether more or less than "Pare" because both terms turn up so much chaff. There is some documentation of its vocabulary available, incidentally; The Making of a Mixed Language: The Case of Ma'a/Mbugu by Maarten Mous mentions "Pare (Chasu)" muruke "sweat" and tika "lift" (in the context of their having been borrowed without change into Normal Mbugu, and then glottalized into Inner Mbugu muru'u and ti'i); Mbugu also borrowed Pare ku-kasha "hunt", Zigua ku-kala "to hunt", and Shambaa u-kalá "hunting" and ngwilizi "eagle" (source of Normal + Inner Mbugu ngwirizi, variation of l/r being dialectal in Shambaa; contrast Pare ngwirini). Isaria N. Kimambo's Political history of the Pare of Tanzania, c. 1500-1900 (1969) implies Pare and Asu are different: "Other naming procedures include the use of u- for territorial names, e.g. Upare for the Pare country; and ki- for language, e.g. Kipare for the Pare language. The only exception here is Chasu which refers to the Asu language." John D. Kesby's Rangi of Tanzania: an introduction to their culture (1981) seems to clarify, however: "in the northeast of Tanzania, [...] the people called Pare in Swahili refer to themselves as Asu". (And Maarten Mous provides a bit more detail, that Pare/Asu has at least two dialects, north and south, with the north one apparently also being called Vudee and several spelling variations thereof.)
As for ppt, some works refer to it as "Pari", e.g. The Abandoned Narcotic: Kava and Cultural Instability refers to "the Pari (Pa) language" (not to be confused with lkr Päri - Lokoro). I guess we can rename it "Pa" for now, and switch to "Pari" later if something else called Pa comes up. - -sche(discuss)05:48, 10 September 2015 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming dig
From "Chidigo" to just "Digo", partly because we should try to purge prefixes from our language names where appropriate and partly because the latter name is vastly more used by linguists. —Μετάknowledgediscuss/deeds19:35, 5 September 2015 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
We currently call pmm "Pomo", which makes it sound like the Pomoan "language" which some people hypothesize exists, and for this reason I almost added a translation (with nested translations for Northern and Central Pomo) using the code in this way. I propose it be renamed "Pol" or "Pol Pomo", the name Wikipedia and the International Encyclopedia of Linguistics use. - -sche(discuss)13:57, 19 June 2015 (UTC)Reply
Latest comment: 8 years ago1 comment1 person in discussion
The ISO merged aam "Aramanik" into aas "Aasax", saying here "Aramanik [aam] is listed as a Southern Nilotic language of the Nandi group, presumably because the Aramanik people assimilated to the Nandi. The original Aramanik language was a Cushitic language (or a non-Nilotic language with heavy Cushitic overlay) usually called Aasax (Fleming 1969) and is already included in a separate Aasax [aas] entry. Maarten Mous, in A Grammar of Iraqw, also gives them as synonyms." - -sche(discuss)22:11, 2 November 2015 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
These are currently named "Gallurese Sardinian" and "Sassarese Sardinian", which has led to them sometimes being nested under Sardinian in translations tables, but this is erroneous because they are (transitional) dialects of Corsican spoken on Sardinia, not dialects of Sardinian. I propose to drop "Sardinian" from their names. See also WT:RFM#Sardinian_templates. - -sche(discuss)17:52, 17 August 2015 (UTC)Reply
We could. They're subject to the same LDL CFI whether we keep them independent or merge them into Corsican (as contrasted with dialects of Italian, for example, which would find themselves subject to much higher CFI if merged into it), and a merger would reduce duplication while we could still note differences with {{label}}s and {{qualifier}}s... so perhaps we even should merge them. But they occupy grey areas. Gallurese is transitional between Corsican and Sardinian; Sassarese is transitional between Corsican, Sardinian and Tuscan (which I guess we consider it?). Ah, dialect continua... - -sche(discuss)01:45, 18 August 2015 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Pretty much every source I could find referred to Zamboangueño et al. as varieties or dialects, not languages. Two of them in particular make it very clear:
“The result of the study showed that while there are observable differences in certain language features beween and among the four variants, they are nonetheless, mutually intelligible with each other even among native speakers who do not have any special language training. Thus, for the purpose of this pilot study, al four variants were identified as dialects of PCS.” Sister María Isabelita O. Riego de Dios, A Pilot Study on the Dialects of Philippine Creole Spanish
“The two variants of PCS share enough distinctive differences from regular Spanish or regular Philippine usage that they must be considered historicaly related dialects of the same language” Charles O. Frake, Lexical Origins and Semantic Structure in Philippine Creole Spanish
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
There are two problems here: First, we currently have the code yaf referred to a language we call Kiyaka, but which Wikipedia calls Yaka (without the noun class prefix) and which authors on Google Books seem to agree to call Yaka as well. There are a couple other languages sometimes called Yaka, but fortunately they all have other names that are more common and therefore there is no conflict, and the principal name of yaf should therefore be modified; there are very few categories associated with this one, so it should be easy to change.
Secondly, the Wikipedia article states that the codes noq, ppp, and lnz refer to its dialects, but Glottolog seems to consider them separate languages (possibly just following the ISO rather than actually making a judgement). Therefore we should consider merging these codes, if in fact there are not enough differences (some data would be helpful). @-sche —Μετάknowledgediscuss/deeds02:16, 18 June 2015 (UTC)Reply
axk currently goes by "Yaka"; if yaf were renamed "Yaka", yaf would need to use a disambiguator or axk would need to be renamed, but to what? axk's alt name of "Aka" is taken (by soh), and I haven't offhand found evidence that anyone calls it by its alt name "Beka".
Ethnologue, although it grants Ngoongo a separate code (noq), labels it a dialect of Yaka in its entry on Yaka. I can find a German reference stating "Eng mit den Yaka verwandt sind die Lonzo, Pelende und Suku." ("Closely related to the Yaka are the Lonzo, Pelende and Suku", the last of which WP and Ethnologue consider to speak a separate language.) A French reference says "Les Pelende ont un accent linguistique propre, mais ils s'entendent avec les Yaka, Suku, Lonzo, Luwa, Hungana, Tsamba, Ngongo, Mbala et Kongo." ("The Pelendes have their own linguistic accent, but they get along with / can understand the Yaka, Suku, Lonzo, Luwa, Hungana, Tsamba, Ngongo, Mbala and Kongo.") Another says "Le kipelênde comme le kiyaka est un dialecte du kikongo commun. Plus répandu, «Le kiyaka comprend quelques neuf dialectes distincts, présentant parfois des variantes assez considérables.»" ("The Kipelênde like Kiyaka is a dialect of the common Kikongo. More widespread, "The Kiyaka includes some nine separate dialects, sometimes with quite considerable variations.")
I can find a small Yaka corpus, but not any comparison of the different [might-be-]dialects.
Hmm, re names: soh is also called (Jebel) Silak, or (Jebel) Sillok; Wikipedia uses the name Sillok, but I'm not finding many resources to assess how common that is (and some refer it by a hyphenated string of dialects). If we can move soh (and perhaps should anyway), then we could move the rest down without disambiguating (so axk would be Aka, and yaf would be Yaka). —Μετάknowledgediscuss/deeds23:22, 18 June 2015 (UTC)Reply
I sometimes wonder if we should start preferring disambiguators to alt names: they'd be annoying to type, but I think the mere fact that we're discussing a three-link chain of renames (language A takes B's name, B takes C's name, C takes D's name) shows how much clearer they'd be. I pity the new user who e.g. adds aja content under an ==Adja== header, and I pity the veteran user who has to notice that that has happened. In this case, I can't find evidence of soh being called Sillok, but the people who speak it and the place they live are called Sillok, so at least it wouldn't be unclear. Perhaps we could just rename axk and soh to have disambiguators, though: "Aka (Congo)" and "Aka (Sudan)". - -sche(discuss)02:02, 19 June 2015 (UTC)Reply
@-sche: I've generally followed the guideline that we avoid such parenthetical geographic locators; were we to use them in general, it would change a great deal of our names. I know few others care, but perhaps we ought to put this to the community at large in the BP? —Μετάknowledgediscuss/deeds19:56, 28 July 2015 (UTC)Reply
In this specific case, I think there is compelling reason to deviate from the general practice/guideline of preferring alt names to parentheticals even without changing that guideline: in order to use alt names here, we'd have to chain-rename ≥3 languages such that the name each one most often went by was assigned to a different one, and one of them would end up with an unattested name, which would all be extremely confusing. As for whether/how to change the general guideline: I'll think the matter through more thoroughly before I post anything in the BP. I don't think I'd propose switching to parentheticals in all cases (I think, for instance, that Pyu/Tircul and Riang/Reang use different scripts and so are unlikely to be mixed up). I would only prefer parentheticals where people would be likely to mix up which language was meant by a given name, and where the mix-up would be likely to go unnoticed (e.g. because the script was the same). - -sche(discuss)21:23, 28 July 2015 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
There are far more Tonga languages than anyone would want to have to deal with, but I am excluding all but two of them in this discussion for the sake of ease. We currently call toi and tog "Tonga" and "Chitonga", but both languages are called by both names, and use of alternative names seems to be vanishingly rare. Our current system is leaving me (and at least one person who tried to give a translation in one of the languages) thoroughly confused, so as much as I find them ungainly, I'd much rather we use parenthetical geographic identifiers than have to go through this madness. (Pinging @-sche as usual (and you ought to take a look at the other ones I've posted recently on this page when you get a chance, if you are so inclined.) —Μετάknowledgediscuss/deeds05:28, 20 September 2015 (UTC)Reply
I'm all in favor of parenthetical disambiguators, but what should they be? Wikipedia calls toiTonga language (Zambia and Zimbabwe) and togTonga (Nyasa) language, but that seems suboptimal to me since the parentheticals aren't parallel. Ethnologue suggests the majority of toi speakers are in Zambia and all tog speakers are in Malawi, so how about "Tonga (Zambia)" and "Tonga (Malawi)"? —Aɴɢʀ (talk) 13:25, 20 September 2015 (UTC)Reply
Support a rename of toi to "Tonga (Zambia)" and of tog to "Tonga (Malawi)". While we're at it, I think we prefer (do we?) to drop "ki-", "chi-", "gi-" and such African language-name prefixes, so toh could be renamed from "Gitonga" to "Tonga (Mozambique)". Happily, to is distinct as Tongan, and we don't have tnz yet, but it seems to be consistently called "Ten'edn" or "Maniq" (the latter being properly an ethnonym) by its speakers, who SIL says are totally unfamiliar with "Tonga" as the name of a language. - -sche(discuss)23:50, 26 September 2015 (UTC)Reply
@-sche: I don't know if we've talked about it before, but I think in general we should avoid language-name prefixes. However, there are some exceptions; I prefer "Luganda" to "Ganda", for example, because it's far more commonly used. I'm fine with renaming Gitonga as you suggest. By the way, thanks for dealing with some of these language issues; Kikuyu and Rwanda-Rundi are still lingering on this page, so please give them some love/research when you have a chance. —Μετάknowledgediscuss/deeds04:07, 27 September 2015 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Gambian Wolof (wof) should be merged into Wolof (wo), IMO. Ethnologue says "Senegalese Wolof [wol] intelligible by speakers of Gambian Wolof but with significant enough differences to require adaptation of materials", which seems to have been their motive in splitting these lects (like so many others). - -sche(discuss)06:51, 25 February 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming skc
Currently called "Sauk", but Wikipedia, SIL publications on the language, and work by Alexandra Aikhenvald all call it "Ma Manda". A happy side effect of the move would be that we could add "Sauk" as an alias for sac, as it is probably the second most common name for that language. —Μετάknowledgediscuss/deeds06:34, 3 November 2015 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Considering war and wrz
Our current situation is to call war "Waray-Waray" and wrz "Waray"; this is not necessarily an optimal solution. Wikipedia chooses to call war "Waray" and wrz "Warray"; although "Warray" is less common than "Waray" to refer to wrz (as far as I can tell), this gives the commonest name of war to that language, which probably deserves priority due to being much more studied. At Template talk:war, you can see that the idea to rename war to "Winaray" was rejected and Liliana's choice of "Waray-Waray" won out. However, it's clear that our current situation has caused some confusion (User:DTLHS/cleanup/mismatched translation codes shows a lot of misuse of wrz when war was intended). Basically, what we have now isn't bad, but the fact is that it's resulted in mismatched codes, so we might want to try a different approach. —Μετάknowledgediscuss/deeds20:46, 14 September 2015 (UTC)Reply
I'd prefer minimizing ambiguity by calling war "Waray-Waray" and wrz "Warray" so that no language at all is called by the ambiguous name "Waray". —Aɴɢʀ (talk) 14:56, 15 September 2015 (UTC)Reply
To be fair I think most of the mistakes were caused when the language was renamed but the translations weren't edited, not by whoever added them in the first place. DTLHS (talk) 23:49, 26 September 2015 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
In 2011, the ISO retired the code for Tingal [tie], merging it into Tegali [ras]. I think we should follow suit. Wikipedia notes that there is dialectal variation in Tegali, but it's not between Tegali proper and Tingal, it is rather between Tegali proper and Rashad (but even those dialects are "nearly identical"). - -sche(discuss)06:15, 11 August 2015 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Shuadit (sdt) aka Judeo-Occitan or Judeo-Provençal is most definitely not an independent language. The literature seems to be quite sure on this point: Banitt refers to it as a "langue fantôme", Vouland as a "non-langue", and Alessio as a "langue imaginaire", to quote three especially scathing francophone scholars. One main scholar is Szajkowski, who seems to have made up a great deal about it (including the name Shuadit, of which there is no evidence of use) and who "was no linguist, and his knowledge of Occitan was quite poor" (Strich and Jochnowitz). Moreover, the so-called last speaker, and a chief primary source, Armand Lunel, was evidently a semi-speaker who was not actually fluent in the "language". Glottolog sums all this up by saying: "This entry [Shuadit] is spurious. This means either that the language denoted cannot be asserted to be/have been a language distinct from all others, or that the language denoted is covered in another entry." To the extent that anyone wants to enter the paltry Hebrew-script text, it can be done as Old Provençal or Occitan, depending on how old it is. —Μετάknowledgediscuss/deeds06:33, 14 April 2016 (UTC)Reply
Pretty much just a script variant. That was a popular thing to do, because using the Latin script was seen as too associated with the Church and distant from the Jewish educational tradition. There have been many other isolated incidences of languages like Urdu and Samogitian being written in Hebrew script as well. —Μετάknowledgediscuss/deeds06:45, 14 April 2016 (UTC)Reply
@Metaknowledge: do you think that much of Judaeo‐Romance contains only superficial differences? Most of them are considered ‘extinct’ by Wikipedia, bearing Ladino and Judeo‐Italian. Judaeo‐Italian notwithstanding, Ladino appears to be the language of Latin Jews around the world. --Romanophile♞ (contributions) 07:13, 14 April 2016 (UTC)Reply
I think so, but I'd like to read through the chapter on each in the Handbook and consult some other sources before deciding. I obviously like Jewish languages, but I think that the line between language and dialect is being abused, and that we should find better ways to document these. —Μετάknowledgediscuss/deeds07:20, 14 April 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
This name is used rather uncommonly, and really only by Jewish philologists, not mainstream linguists. It should be changed to the much more common "Judeo-French".
Unrelated to the name change, I'm not fully sure that we should have this as a separate language. Kiwitt and Dörr (2015) say: "It should be noted, however, that the major part of linguistic data attested in Judeo-French sources is simply common Old French written in Hebrew script, with some texts showing little to no register variation in comparison with Christian Old French sources." They go on to discuss one extensive text, a biblical glossary, where only 6% of words were not attested in Christian Old French texts. Basically, this is similar to Hindi and Urdu — is it worth keeping separate? (And if you think it'd be strange to have Hebrew-script entries under ==Old French==, remember that we have Arabic-script entries under ==Afrikaans==.) @-sche, Renard Migrant, Wikitiki89 —Μετάknowledgediscuss/deeds04:52, 13 April 2016 (UTC)Reply
Interesting. I’d be fine with a merge or a rename. Script variants and dialectisms can simply be marked with Judeo-French. I would love to work on this dialect, but I have no idea where to find any texts. --Romanophile♞ (contributions) 05:08, 13 April 2016 (UTC)Reply
Before commenting here, I was planning to read the Judeo-French chapter of my new copy of the Handbook of Jewish Languages, but I now realize that Metaknowledge also recently acquired this book and has probably already this chapter and probably only started this discussion because of that. So I'll just assume that his conclusion is the same that I would have drawn and say that I agree that this should be merged with Old French. --WikiTiki8921:10, 13 April 2016 (UTC)Reply
Rename per nom; see also ngrams. Also merge per nom; perusing the references that turn up if I just search for references that mention both lects (google books:"Judeo-French" "Old French"), I find that they agree:
Raphael Patai, Encyclopedia of Jewish Folklore and Traditions (2015, →ISBN, page 316: "Judeo-French is Medieval (Old) French as spoken and written by French and Rhenish Jews. It differs from the other “Judeo” languages in that there were no dialectal differences between it and the Old French spoken by the non-Jews[.]"
Aaron D. Rubin, Lily Kahn, Handbook of Jewish Languages (2015, →ISBN, page 139: "However, this term does not imply the existence of a set of linguistic features common to these sources that would allow identifying a 'Judeo-French' language or dialect distinct from the varieties of Old French encountered in Christian sources."
If I remember correctly, a number of Old French words are attested first in Rashi's writings, and those writings been used in recent years by scholars of Old French to fill in gaps in knowledge of other aspects of the language. I'm sure some of our etymologies already include those Zarphatic words as Old French, so we might as well make it official. I agree we should both rename the lect and merge it with Old French. Chuck Entz (talk) 02:29, 14 April 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Hello everyone. I'm curious as to how one would go about requesting an exceptional code for the standard Kichwa language? There are several SIL/Ethnologue codes for various Kichwa dialects (Imbabura (qvi), Chimborazo (qug), Cañar Highland (qxr), etc), but not one is for the standard Kichwa language that's taught in schools and used by the government in Ecuador. There is a common Quechua code (qu) is currently used for Quechua Wiktionary and Wikipedia, but both projects are exclusively written in standard Southern Quechua. Both Kichwa and Southern Quechua are part of the Quechua II branch of the Quechuan languages, but they are different dialects with different standardized grammars and different standardized writing conventions. --Dijan (talk) 18:28, 12 May 2014 (UTC)Reply
Hmm... how mutually intelligible are Kichwa and Quechua? It wasn't that long ago that the various Quechua dialects' codes were removed from Module:languages, though I'm having trouble locating the discussion(s) that led to that.
They are mutually intelligible for the most part (similar to differences between Danish and Swedish), but they are also significantly different from each other to be considered separate. Linguistically, Kichwa belongs to a different subgroup of Quechua. The grammar of Kichwa is more simplified (loss of possissive suffixes, loss of the voiceless uvular fricative, etc) and the vocabulary is affected by native languages spoken before the Incan conquest of the territory of today's Ecuador and Colombia (meaning, Quechua was imposed as a foreign language, whereas it is a native language in the regions where Southern Quechua is spoken - in southern Peru and Bolivia).
I was referring to designing our own code and using that as an umbrella for all the Kichwa varieties - which now all use a standardized alphabet different from the Peruvian varieties (such as Southern Quechua), but I couldn't find the procedure for it.
There was an attempt to create a separate Kichwa Wikipedia, but apparently no one got around to it and it got complicated as somone pointed out that an official ISO code must be requested specifically for the standard variety. And for some the problem was that it was trying to use the Chimborazo (qug) code (which is one of the most widely spoken varieties of Kichwa). Apparently the issuing of codes is very strict on Wikipedia. --Dijan (talk) 20:57, 27 May 2014 (UTC)Reply
If none of the ISO codes cover this we should create our own code. The languages seem different enough that they warrant separate treatment, at least. -- Liliana•21:43, 29 February 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Although LANGTREAT notes that they should be subsumed into fa, the following codes still exist in Module:languages:
pes, for "Western Persian", the most common variety of Persian. We should merge it into fa because it *is* fa.
prs, Eastern Persian / Dari, the variety of Persian spoken in Iran Afghanistan. Its status as a separate language, and its very name 'Dari', were products of Afghanistani politics. Not even Afghani speakers of the language call it 'Dari' or consider it separate from Persian; we shouldn't consider it separate, either.
aiq, Aimaq, a variety spoken by nomads in Afghanistan and Iran. It is sometimes considered specifically a variety of Dari (which is itself little more than another name for Persian, as explained). It differs from standard Persian mainly in matters of pronunciation, something we usually handle with {{a}} rather than separate L2s.
haz, Hazaragi, another Afghan variety. WP summarizes scholarly opinion (with citations, for which look here): "The primary differences between Standard Persian and Hazaragi are the accent and Hazaragi's greater array of Mongolic loanwords. Despite these differences, Hazaragi is mutually intelligible with other regional Persian dialects."
deh and phv, Dehwari and Pahlavani, which it is hard to find information on because even WP simply redirects the words to "Persian".
As indicated above, my opinion is that we should merge all of those codes into fa.
Incidentally, LANGTREAT originally also banned Tajik (tg), but this was not supported by scholarship or by our own practice (we had hundreds of Tajik entries), so after two discussions, I updated the page to note that Tajik is allowed. LANGTREAT made no mention of Judeo-Persian (jpr), Bukhari (bhh), Judeo-Tat (jdt) or Tat (ttt), and past discussions of them have assumed they were separate languages, so I also updated the page to reflect that. - -sche(discuss)05:16, 20 November 2013 (UTC) (fixed think-o)Reply
The Persian lects are an interesting issue; they are on the whole pretty similar, but Persian and Tajik have separate literary and cultural traditions, and I believe Dari does too. I think it is best to keep them separate. The Jewish varieties are often written in the Hebrew script and have a separate cultural tradition, so I think it would probably be handy to keep them separate as well. All the rest probably ought to be merged into their macrolanguages, unless there are script conflicts I'm unaware of (it is, of course, easier to keep with one script per language). —Μετάknowledgediscuss/deeds05:23, 20 November 2013 (UTC)Reply
The difference between Dari and Persian is not great AFAIK, there are some references on Wikipedia. Tajik should stay separate, not just because it's in Cyrillic. It's very different from Persian and has many Russian and Turkic loanwords. There is also a significant difference in phonology (vowels). Persian e, o and â are usually i, u and o in Tajik. ZxxZxxZ (talk • contribs) may be able to say a bit more. --Anatoli(обсудить/вклад)05:36, 20 November 2013 (UTC)Reply
Regarding Dari, to merge Dari into Persian, we would have to figure out what to do with transliterations. In standard Persian, ē merged with ī into what we transliterate as i (no diacritic) and ō merged with ū into what we transliterate as u. In Dari, ē and ī and ō and ū are still differentiated. Furthing complicating the problem, standard Persian e, o, ey, and ow are pronounced i, o, ay and aw in Dari, although these differences are not phonemic. The other differences are not as much of a problem, but see a brief description at w:Dari#Phonology. --WikiTiki8914:37, 4 December 2013 (UTC)Reply
That has been the practice on Wiktionary. Dari should be under the Persian heading with {{context|Dari|lang=fa}} label. That is what we do already. Take a look at فاکولته, for example. Regarding the transliteration of long vowels, ZxxZxxZ (talk • contribs) has been trying to implement a classical Persian transliteration (which is pretty much used for Dari by scholars) as a standard for all Persian entries. So far, it's been a slow and selective process. I'm not opposed to it and it can easily be indicated as the standard Wiktionary practice in the Appendix:Persian transliteration. --Dijan (talk) 06:35, 5 December 2013 (UTC)Reply
I support keeping Dari (and Western Persian) under Persian heading, but I'm not sure what to do about transliteration. By the way, Eastern Persian / Dari (prs) is the variety of Persian spoken in Afghanistan, not Iran, that of Iran is the first one, Western Persian (pes). --Z17:48, 12 December 2013 (UTC)Reply
James Minahan says "The Hazara language, called Hazaragi, is a Farsi dialect, although the Hazaras are physically Mongol. The intermixture of the Indo-European and Mongol linguistic groups resulted in a dialect of Dari Persian that contains extensive words and forms from Farsi, Turkic, and Mongol. [...] Most Hazaras also speak Dari Persian [...] as a second language. In Iran most are bilingual, speaking both Hazaragi and Farsi. [...] Until the 1980s educated Hazaras used Farsi or Arabic as the literary language, but a movement to create a Hazaragi literary language has gained momentum." This suggests that Hazaragi is a separate language; we also already have a few translations into it. Therefore, I've kept it separate and updated WT:LT accordingly.
Barbara West, in the Encyclopedia of the Peoples of Asia and Oceania, writes that the Aimaq speak "Aimaq, a dialect of Dari or Afghan Persian"; Brian Williams writes that "Aimaqs also have a strong Persian admixture, and their language is Dari or Farsi", Jake Kircher writes that "Once there was a generally used, common Aimaq language but now, few seem to speak it anymore. Dialects spoken today resemble Dari (Afghan eastern Farsi) admixed with words of Mongolian and Turkic origin." In general it seems that it should be handled the same way as Dari, via {{lb}} and {{a}}.
Several references consider Pahlavani an alternative term for Pahlavi, but the 2003 International Encyclopedia of Linguistics says Pahlavani is still spoken today in a village in Afghanistan, Haji Hamza Khan, where it is "similar to Dari Persian but still distinct" (earlier editions of the IEL are explicit that this is the only village where the language is spoken). I suppose it can be left unresolved for now.
Dehwari is spoken by Persians in Baluchistan; several references seem to consider it to be Persian, but Denys Bray's 1934 The Brāhūī problem writes as if it and Persian are separate languages; I suppose it too can be left unresolved for now.
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
We currently have separate codes for three lects considered part of the Lega macrolanguage: lea, lgm, and khx. These are reasonable to separate into two languages, lgm (which should be called Lega-Ntara) and lea (which should be called Lega-Malinga) as there is 67% mutual intelligibility, but khx is clearly a variety of lea. All this is per the treatment of the Beya dialect of lea in The Bantu Languages. —Μετάknowledgediscuss/deeds04:23, 22 November 2015 (UTC)Reply
Support; merge khx into lea, with lgm remaining separate, and name everything per nom. I'm not sure why Wikipedia names the two main dialects using placenames. The full array of alternate names I encountered in (cursorily) researching the matter:
Mwenga Lega = Lega-Ntara / Lega Ntara (variously translated in refs as "Lower Lega", "Upper Lega" or "Eastern/Northern Lega") = Isile, Ishile, Kisile; Mwenda-Liga
Shabunda Lega = Lega-Malinga / Lega Malinga (variously translated in refs as "Upper Lega", "Lower Lega" or "Forest Lega" or "Western/Southern Lega") = Lega (Kilega) / Liga (Kiliga) proper; dialects: Kanu (Kikanu), Gala (Kigala), Yoma (Kiyoma), Sede (Kisede), Gonzabale, Beya (Beia), and possibly (Ki)Nyamunsange and Banagabo and Kabango and Bene
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming ki
This is a hard one, and I'm not advocating one way or the other, just wishing to raise the issue. Ngrams reveal that in 1992, "speaking Gikuyu" and "Gikuyu language" became more common than "speaking Kikuyu" and "Kikuyu language" (as well as becoming the linguistic standard), yet overall the spelling "Kikuyu" is still more common, presumably in speaking of the people, who are less obscure than their language. We follow Wikipedia in using the spelling "Kikuyu", but this spelling is clearly no longer favoured for the language (if you're curious, the "g" is to reflect the etymon, KikuyuGĩkũyũ). Should we change it? —Μετάknowledgediscuss/deeds19:55, 5 September 2015 (UTC)Reply
I really don't have a strong opinion on this. Personally, I still think of the language as Kikuyu, which makes it difficult for me to come out and say "Yes, we should rename it", but the reasons you mention make it difficult for me to come out and say "No, we shouldn't rename it". So I abstain. I'll be happy if we continue to call it Kikuyu, but I won't be unhappy if we start calling it Gikuyu. (I will be unhappy if we start calling it Gĩkũyũ, though, since that really isn't an English word.) —Aɴɢʀ (talk) 08:19, 27 February 2016 (UTC)Reply
I am very familiar with the English name Kikuyu for both the language and the people (especially for the language), but I have never seen Gikuyu used in English. I think Gikuyu is the native name (more properly Gĩkũyũ). —Stephen(Talk)22:50, 27 February 2016 (UTC)Reply
As you note, there are phrases where ngrams suggets Gikuyu is now more common as a name for the language — but there are also phrases (1, 2) where Kikuyu is still more common even as a language name. I'd stick with the current spelling. - -sche(discuss)05:00, 28 February 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Judging by Ngrams, "Malaccan Creole Portuguese" is the least common name for this language; more common is "Malacca Creole Portuguese" with no n, and most common is "Kristang". In Glottolog's list of materials on it, I note that most of the modern material on it (by Baxter and Marbeck) calls it Kristang. I suggest renaming. That entails updating several entries and moving several categories. - -sche(discuss)19:01, 20 March 2016 (UTC)Reply
Kristang gets a misleadingly high number of hits because it’s also the name of the people that speaks it, and part of the synonym Papia/Papiah/Papiá Kristang. — Ungoliant(falai)01:05, 2 April 2016 (UTC)Reply
Kristang, but Papia Kristang or Malacca Creole Portuguese are also good options. The most important works about this language use Kristang more prominently. — Ungoliant(falai)19:30, 3 April 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
We currently have a code spq for this dialect of Spanish; Wikipedia has an article for it at Amazonic Spanish which states that "Ethnologue's reasons for doing this [making a separate code for it] are poorly documented." Although it has some mild differences, it is clearly a dialect of es and should be merged into it. (There are no entries, but we should record the merger.) —Μετάknowledgediscuss/deeds17:35, 3 April 2016 (UTC)Reply
Merge, IMO. (The only thing that gives me pause is the difference in attestation requirements: if Amazonian Spanish isn't well documented, then treating it as a separate lect allows its entries to be held to a lower attestation requirement. But the same could be said of a lot of varieties that don't have their own codes, e.g. New Mexico and Southern Colorado Spanish, for which there are several published references.) - -sche(discuss)00:50, 4 April 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
The Senoufo lects are a mess, and we have no consistency in naming them (some are Senoufo, some Sénoufo, some without the Senoufo at all). This one, seb seems not even to be a separate lect at all, but instead what Supyire spp is called in Côte d'Ivoire. —Μετάknowledgediscuss/deeds20:25, 3 April 2016 (UTC)Reply
Yes, Wikipedia and Omniglot agree they are the same. (The old 2003 International Encyclopedia of Linguistics said their "Relationship [was] undetermined" at that time.) As for the other Senoufo lects: I noticed one while checking translations at water, and removed "Senoufo" from its name before I added entries in it because I saw how rarely it was actually referred to with "Senoufo" in the name. Supyire too seems to be mostly referred to without "Senoufo". - -sche(discuss)00:45, 4 April 2016 (UTC)Reply
@CodeCat This is pretty funny. I'm fairly certain I used to use "Brittonic" and then you corrected me to "Brythonic" (rightfully, since it is the one we use currently), but it's humorous to me that we're now suggesting the change. I definitely think we should change it to "Brittonic". —JohnC518:21, 12 April 2016 (UTC)Reply
Google Books Ngrams shows what looks to me like virtually a statistical tie since 1950, though Brythonic has been more common since the turn of the century. I really don't have a strong preference either way. —Aɴɢʀ (talk) 09:28, 14 April 2016 (UTC)Reply
I reckon that we should add exceptional codes for the ones that have words directly recorded from them, even though it's precious few for each.
Palta (separate article) could be qfa-jiv-pal, in imitation of Linguist list's jiv-pal, but given that its family assignment is not one hundred percent certain and a three-part code is annoying and abnormal, we could also settle for sai-pal. I don't know why Jivaroan's language family uses qfa rather than sai, which seems to be our default for unsorted South American languages.
Re "why Jivaroan's language family uses qfa rather than sai": presumably human error. It could be updated to "sai-jiv". Re whether to use "sai-pal" or "sai-jiv-pal": for consistency with other codes ("nai-yuc-yav", etc) and with the schema described in WT:LANG, we should use "sai-jiv-pal" if we accept that it was a Jivaroan language. But as you say, the family identification is speculative (although the evidence which does exist is consistent with it). I suppose we could use "sai-pal" to be 'safe' / 'conservative' about the family identification and get a shorter code... this also lets us add the others as "sai-" codes without worrying we ought to reassign them if we later create a family code for the families they belong to. (Indeed, we already have a code for the Cariban family which Campbell and Grondona say Patagón belonged to.) - -sche(discuss)22:27, 26 May 2016 (UTC)Reply
I've added Palta as sai-pal, added Rabona, Patagón, Bagua, Copallén Tabancale, Chirino and Sácata, and also renamed the Jivaroan code to sai-jiv and changed Esmeralda's code from qfa-und-esm to sai-esm to fit the usual naming scheme. - -sche(discuss)15:42, 29 June 2016 (UTC)Reply
I have added Puruhá as sai-prh. There was at one point a grammar of it (though it has been lost), so we know it existed as a discrete lect. And based on personal- and place-names, words in it have been reconstructed by scholars. Even if the only words in it we can add are words in the Reconstruction: namespace, that does seem worth having a code for (a code also lets us reference it when giving the etymologies of those personal- and place-names). The other languages could probably be given codes on the same basis (if words in them have been reconstructed). - -sche(discuss)00:56, 1 July 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
We currently have a language called Saliba (sbe) and one called Sáliba (slc). In my mind, having two languages' names only differ in a diacritic is not acceptable. It confuses automated programs as much as human editors. I'm not sure there are any acceptable alternative names, so I propose using geographic disambiguation for sbe as "Saliba (Papua New Guinea)". @-sche —Μετάknowledgediscuss/deeds04:15, 26 June 2016 (UTC)Reply
Alternately we could call slc "Saliva". This seems to get about 10× more Ghits for both the ethnic group and the language (though interference from saliva is possible). --Tropylium (talk) 04:34, 27 June 2016 (UTC)Reply
Indeed the current names are so confusing that I gotthembackwards when I added the only three entries we have in the two languages. I think that disambiguating them both with parentheticals is clearer than allowing one to keep the ambiguous name (either while renaming the other to "Saliva" or while disambiguating the other). Current practice suggests that we should rename slc to "Saliva" or "Sáliva", but I wouldn't mind if we started making more frequent use of disambiguators instread. - -sche(discuss)08:50, 27 June 2016 (UTC)Reply
Done, except that per previous discussion on using geographic disambiguators rather than national ones, I used "New Guinea" rather than "Papua New Guinea". (Other languages also use "New Guinea" as a parenthetical disambiguator in their canonical names, and only mention "Papua New Guinea" in alt names because SIL uses national disambiguators like that.) - -sche(discuss)05:24, 28 June 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Ethnologue encoded this language twice: once as lrk "Loarki" (the name its 20,000 Pakistani speakers call it), and a second time as gda "Gade Lohar", the name its ~500 speakers on the Indian side of the border call it. (The International Encyclopedia of Linguistics entry on Gade Lohar conservatively only says the languages "may be the same" as Loarki, and notes its long list of other names: Gaduliya Lohar, Lohpitta Rajput Lohar, Bagri Lohar, Bhubaliya Lohar, Lohari, Gara, Domba, Dombiali, Chitodi Lohar, Panchal Lohar, Belani, and Dhunkuria Kanwar Khati. The IEL entry on Loarki is more explicit, breaking down the population by country and countain Gade Lohar's 500 speakers as Loarki speakers, because Loarki is "probably the same as Gade Lohar in Rajasthan, India, a Rajasthani language.") I propose to merge gda into lrk. - -sche(discuss)20:50, 11 August 2015 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
The ISO has retired the code duj and replaced it with dwu (in the process of splitting off dwy "Dhuwaya"). If we want to follow the ISO, all our Dhuwal entries and categories need to be switched from duj to dwu, which seems like a lot of unnecessary bother. (Whoever does this should also add dwy to the module.) - -sche(discuss)05:54, 24 February 2016 (UTC)Reply
Latest comment: 8 years ago4 comments2 people in discussion
The following discussion has been moved from the page User talk:-sche.
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
What do you think about adding a code for this language, and under what name? Wikipedia describes it at Katembri language; they cite Fabre for the claim that this language is only preserved in a single brief wordlist, where it is called Kiriri (the wordlist is on page 22 (section 3.4) of this pdf). Regardless, that document does seem to be a good place to find more words for water. —Μετάknowledgediscuss/deeds03:24, 2 March 2016 (UTC)Reply
There's a Kiriri attested in a single wordlist collected from an elder from the 1960s that's a Katembri language, and another Kiriri attested in a single wordlist collected from an elder from the 1960s that's a Xukuru language? Well, that's confusing.
The fact that there's only a limited number of words is no reason not to include the language, but it would be good to avoid the ambiguous name Kiriri. How about calling them Katembri and Xukuru, like Wikipedia does? Wait, (as a minor point of curiosity,) if the wordlist is labelled Kiriri, where'd the alternative name come from?
Given the naming issues, I'm suddenly confused about whether I have correctly identified the wordlist being referred to. I don't know anything about any of these languages, so I feel lost (it's so much better in Austronesian, for example, where I at least feel like I have a hold on what goes where). Anyway, that naming scheme makes sense; we can use qfa codes and not worry about the families, no? —Μετάknowledgediscuss/deeds05:01, 2 March 2016 (UTC)Reply
qfa is the prefix for exceptional family codes. All of our exceptional language codes which start with qfa do so because they start with a family code that starts with qfa, like qfa-ctc-col. There's been at least one case where we've created a family code for an accepted family (qfa-len, the Lencan languages) in order to use it in constructing a language code (qfa-len-slv for Salvadoran Lencan), but Wikipedia notes that scholars aren't certain what family either Kiriri belonged to, so we couldn't do that here because we couldn't accurately, confidently assign either one a family code (even an exceptional family code). I suppose we could construct codes starting with qfa-und, like qfa-und-ktm for Katembri. I wouldn't want to use bare qfa-___ (e.g. qfa-ktm for Katembri) because it would look like a family code. - -sche(discuss)08:21, 3 March 2016 (UTC)Reply
Burgundian language (Germanic) (gme-bur) — the name "Burgundian" is unoccupied but could cause confusionDone as gem-bur due to uncertainy over whether or not it is East Germanic
Ch'olti' language (myn-cho) — I wasn't 100% sure that this lect is a separate languageDone (notes below) as myn-chl for maximal distinction from Ch'orti'
Amarizana: add per nom. Alexandra Y. Aikhenvald in Languages of the Amazon cautions that "many languages in Amazonia have 'namesakes' [and] more than one group may hide behind the same name", and more than one language has been called Amarizana: "one of the clans of the Piapoco is the Amarizanes (and the name is sometimes applied to the whole group). A now extinct language, also called Amarizana, and from the same Arawak family, used to be spoken in the Meta territory of modern Colombia." Nonetheless, the only references I find unambiguously mean this language. Čestmír Loukotka, Johannes Wilbert, Classification of South American Indian Languages (1968), page 131, lists some Amarizana words alongside and hence obviously distinct from Piapoco, including nuita "head", notuy "eye", nukagi "hand", kaxü "house", sietai "water", eriepi "fire" and keybin "sun". Julian Granberry's A Grammar and Dictionary of the Timucua Language even provides some etymology, connecting Amarizana eri(-...) "fire" to Achagua eri "sun, day", Arekena ale "sun".
Amasi: this happens to highlight what a mess our African language family codes are. Several codes use the prefix nic- even though their most immediate superfamily is alv, e.g. nic-vco should be alv-vco. Fortunately, fixing the nic- codes should not require updating very many pages. One that is done, precedent would have us use alv-bco-... rather than alv-... (compare nai-yuc-tip, qfa-ctc-cat), although the argument in favor of a shorter code is obvious. :-/ Some words are listed in a 1973 article in Africana Marburgensia ('AM') and in a pre-draft working paper cited by WP ('B'), including bú (AM) / bu (B) "dog", ázɔ́lí (AM) / azɔle (B) "tree", ɣà-nēm (AM) / ɣanim (B) "man", ɣà-zhyī (-zhyì?) (AM) / ɣaʒɛ (B) "woman", mwɔ̄ (AM) / muɔ (B) "water".
Anauyá: add per nom. Also called Anauya, but the version with diacritic is more common. [3] has uni "water" and ahiri "sun", the latter confirmed by I Simposio Antonio Tovar sobre Lenguas Amerindias: Tordesillas... (Emilio Ridruejo Alonso, Mara Fuertes, Carlos González-Espresati; 2003) and both are seemingly in the aforementioned Classification of South American Indian Languages, although I can't see the exact snippet.
Atanque(s): out of the various names WP mentions, namely "Atanque (Atanques) or Cancuamo (Kankuamo), also known as Kankwe and Kankuí", plus others I ran across (Atanke), "Atanques" seems to be most common, at least as the name of the language. ("Kankuamo" is quite common as a placename(?) that forms part of the designation of a tribe.) A 1962 article in Anthropological Linguistics has some words, including jo̱ke "gourd cup", cognate to cho̱kue(“gourd cup”), and mo̱ga "two", cognate to mo̱ga(“two”), and the 1981 Comparative Chibchan Phonology has more words (and may drop the underline from the os of those words; it is hard to see, because all words are underlined), including ji "worm", jinua "six".
Ayomán (rarely also Ayoman): I've added a code for Jirajaran, sai-jir, so this language's code should be sai-jir-ayo.
@-sche: Can't we just do two-part codes so we don't have to feel obligated to create these horribly long ones? It wouldn't clash with all of our preëxisting practice, despite there being some precedent. Also, I'm worried that your careful work on this is going to make this RFM section far too long, and also cause you to burn out. Perhaps this should be a user page that this section links to? —Μετάknowledgediscuss/deeds07:57, 4 July 2016 (UTC)Reply
Where an existing ISO family code like alv exists, I suppose we could go with two-part codes, but then what should be done with languages that have no ISO family code but instead belong to families for which we've had to create qfa- codes? I suppose they can be treated the same as they are now. But if we accept nai and sai as family codes for this purpose, I suppose that means some qfa- things like Salvadoran Lenca and Catacao can be re-coded. I will update the existing three-part non-proto-language codes if we go that route. I've started Wiktionary:Beer_parlour/2016/July#Shortening_some_.27exceptional.27_language_codes. Yes, I considered as probably should start storing long comments and information on addable vocabulary in userspace. I wouldn't worry too much about burnout; we can take time; the only reason there would be a rush to add these codes ASAP is if we wanted to add words in particular ones of them, and if we wanted to add words, we'd need to do some research to find words to add. - -sche(discuss)17:40, 4 July 2016 (UTC)Reply
I mean, is it separate from fr? I don't know. In any case, I guess on further consideration it doesn't stop us from adding the Germanic language as "Burgundian", because if the Romance language needs to be added, it can be Bourguignon. (And since they're from different (sub)families, it should be easy to tell which one was meant if someone enters a word from one incorrectly as the other.) I'll collect information about them at User:-sche/Burgundian (Wiktionary:About Bourguignon). - -sche(discuss)16:50, 5 July 2016 (UTC)Reply
Oh, there are tons more. This is only the low-hanging fruit; a lot more languages with paltry data are waiting to be dealt with. I'm avoiding the Bantu ones for now, because pretty much all of them are in dialect continua and probably should be left alone unless good scholarship on their mutual intelligibility can be found (which I suppose I should go about finding). There's a pile of Australian ones that I'll get around to listing at some point (I thought maybe I'd give you time to digest all this first), and then even more messier ones from South America. —Μετάknowledgediscuss/deeds09:01, 5 July 2016 (UTC)Reply
FWIW, re 'time to digest', I'd feel free to post any others you have the data to post (except the ones you mention are parts of dialect continua ... might as well leave them alone, as long as some part of the continuum has a code, although if no part does, then we should probably rectify that). There's no harm in it sitting around on the site unattended-to for a while, whereas letting it sit on one's computer sometimes (at least for me) means forgetting where one put it. (I can no longer find the information I thought I had collected on the separability vs mergeability of Haida dialects.) - -sche(discuss)16:50, 5 July 2016 (UTC)Reply
Regarding Ch'olti': The oldest stage of the language (written in so-called hieroglyphs), sometimes confusingly just called "Ch'olti'" but more often called by other names, has its own code emy. The Colonial- and post-Colonial-era stage is considered distinct from emy, and also considered distinct from the more recent stages of Ch'orti', e.g. one reference says "In the Mayan classificatory tradition, the Ch'olti' language, as recorded in the 1695 grammar of Pedro Moran, is generally held to be related to but separate from the modern language of Ch'orti' (see Kaufman's 1976 classification, for example)." Post-Epigraphic-era Ch'olti' and modern-era Ch'orti' (caa) are theoretically distinguished from each other as different branches of Eastern Ch'olan (and from ctu, as it is a Western Ch'olan language), but the size of the difference between Ch'olti' and Ch'orti' is hard to ascertain, especially because, quoth WP, "the post-colonial stage of the language is only known from a single manuscript written between 1685 and 1695" (as afore-mentioned). For that matter, the size of the difference between Epigraphic Ch'olti'an and Colonial Ch'olti' is not obvious to me; Søren Wichmann, The Linguistics of Maya Writing (2004), page 271, says "In this section we show how Classic Ch'olti'an became seventeenth-century Ch'olti'. The chief grammatical difference between the grammars of Classic Ch'olti'an and Ch'olti' is the difference between straight- and split-ergativity." As an example, mi "father" is used in Classic and Colonial Ch'olti'(an) and in Ch'orti'. Nonetheless, given that the corpus of post-emy Ch'olti' is small and well-defined, it shouldn't be that hard to include it separately from emy and caa. - -sche(discuss)21:42, 5 July 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
We seem to have this as a single macrolanguage, yok, despite the fact that the constituent lects seem to constitute at least a few languages. -sche added some entries, but it's all tagged by (dia)lect, so it will be easy to separate them. I think we should retire yok and replace it with exceptional codes to reduce confusion, but I am not sure what those divisions should be. —Μετάknowledgediscuss/deeds20:17, 3 July 2016 (UTC)Reply
Yes, the difficulty is in deciding how to divide it. Christopher Loether cautions that although e.g. "Gayton (1948) listed 26 named groups or 'tribelets'[,] many of these named groups speak dialects which are nearly identical phonologically, lexically and syntactically, while others speak varieties which are indeed quite distinct from their neighbours' speech. Kroeber [...] divided [what he considered a single] language into two main divisions: Valley and Foothill Yokuts. He further divided the Valley Branch into Northern and Southern, and the Foothill branch into Kings, Tule-Kaweah, Poso, and Buena Vista. Newman (1944: 5-3) agreed with Kroeber's analysis of a single Yokuts language and stated that his data corroborated Kroeber's dialect divisions."
Kroeber lists 20+ dialects, of which 21 are named in [[ˀilik']]. Wikipedia has a tree/bush diagram from Whistler and Golla of 23-28 dialects, including all of those 21 plus Koyeti, Merced "(?)", Noptinte (Nopchinche(s), Nopthrinthre(s), Nopṭinṭe, Nopthrinte, Noptinci), Yachikumne a.k.a. Chulamni, Lower San Joaquin Yokuts, and Lakisamni "(?)", and Tawalimni. (Several have multiple names, e.g. Ayticha is also called Kocheyali as well as Ayitcha; Palewyami is also Altinin and Poso Creek Yokuts in addition to Paleuyami. And Hometwoli is also Taneshach?)
Kroeber's and Newman's division; Whistler and Golla's division
Kroeber's and Newman's division is:
Valley
1. Northern Valley
2. Southern Valley
Foothill
3. Kings River (including Gashowu)
4. Tule-Kaweah
5. Poso [Creek]
6. Buena Vista
Whereas, Whistler and Golla's division is:
1. Poso [Creek] (Palewyami)
General Yokuts
2. Buena Vista (Tulamni, Hometwoli)
Nim-Yokuts
3. Tule-Kaweah (Wikchamni, Yawdanchi)
Northern Yokuts
4. Kings River (Chukaimina, Michahay, Ayticha, Choynimni)
5. Gashowu
Valley Yokuts
6. Southern Valley (Yawelmani, Tachi, etc)
7. Northern Valley (Chukchansi, Kechayi, etc)
8. Far Northern Valley (misc dialects)
For Yawelmani and Chukchansi, decent resources exist; in addition to those two, Wikchamni and Tachi are also being taught according to WP, and in addition to those four, Choinimni and Kechayi also have at least some speakers according to WP.
WP says the Yokutsan family consists of "half a dozen" languages, but evidently not the six just named, because those six leave out several major branches that Kroeber, Newman, and Whistler and Golla all agree on.
I suggest we create a family code nai-yok for the Yokutsan languages, and then distinguish the following branches which Kroeber, Newman, and Whistler and Golla all consider distinguishable, without splitting them further at this time:
Palewyami (nai-ply — or putting the y at the start so the codes sort together and are more apparently connected — on further thought, nah) a.k.a. Poso a.k.a. Poso Creek
Buena Vista Yokuts (nai-bvy) a.k.a. Tulamni-Hometwoli
Kings River Yokuts (nai-kry) a.k.a. Choinimni, etc
Gashowu (nai-gsy), which Kroeber and Whislter/Golla agree is intermediate between Kings River and Northern Valley, though Kroeber considers it ultimately/genetically Kings
Southern Valley Yokuts (nai-svy) a.k.a. Yawelmani, Tachi, etc
Northern Valley Yokuts (nai-nvy) a.k.a. Chukchansi, Kechayi, etc
Delta Yokuts (nai-dly) a.k.a. Far Northern Valley Yokuts
A more conservative approach would keep Southern Valley Yokuts, Northern Valley Yokuts, and Delta Yokuts together as "Valley Yokuts", but Delta Yokuts is relatively divergent. - -sche(discuss)22:16, 3 July 2016 (UTC)Reply
Not much, I'm afraid. A couple of decades ago I read everything I could find on the Wikchamni, who used to live in the area where my brother lives now. I read all of the sources you mentioned above, but I was more interested in the ethnobotany of the Yokuts than their languages, per se, and that was a long time ago.
On a side note, I remember one of my professors at UCLA back in the 80s saying that Yawelmani was one of the best-understood languages in the world at the time from a theoretical perspective, because so many linguists had been publishing papers on it- it was sort of the linguistic equivalent of a model organism. Chuck Entz (talk) 04:25, 5 July 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Module:languages includes codes from the individual members of several language/dialect groups which WT:LANGTREAT says, without citing any discussion, should be merged. Something needs to change: either WT:LANGTREAT should be updated to note that the individual varieties are allowed, or their codes should be removed from the module. The following language/dialect groups are affected: (Note 1: whenever the merging of a particular dialect group had been discussed and that discussion was cited by LANGTREAT, I simply updated the module.) (Note 2: Haida, Cree and Kalenjin face the same issue; I expect to write about them later.) - -sche(discuss)05:16, 20 November 2013 (UTC)Reply
Stephen Tyler (not the singer), in his oft-cited works on Gondi, states "Though I have no real evidence, the general pattern seems to be for geographically adjacent Koya and Gondi populations to speak different, but mutually intelligible Gondi dialects. Where these populations are geographically non-contiguous, the dialects are not mutually intelligible. This same pattern probably prevails among all Gondi dialects." WP says "The more important dialects are Dorla, Koya, Maria, Muria, and Raj Gond." Ethnologue, meanwhile, as encoded only two varieties, ggo (Southern Gondi) and gno (Northern Gondi). Should we deprecate those two codes? Or deprecate the macro-code gon and recognise those dialects? Or allow all three? - -sche(discuss)05:16, 20 November 2013 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Malay lects
I propose to merge the following Malay lects into Malay [ms]:
Jakun [jak]
Orang Kanaq [orn]
Orang Seletar [ors]
Temuan [tmw]
These are all mere dialects of Malay with no written tradition and perfectly mutually intelligible. Even Ethnologue says they should be considered dialects of Malay rather than separate languages. -- Liliana•23:39, 28 February 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming nxg
This language is far more commonly called "Ngadha" than "Ngad'a"; the latter spelling is so rare that when I was trying to verify our "Ngad'a" translation of water using that spelling for the language name, I couldn't find any references at all (they all spell it "Ngadha"). This rename entails moving a few categories and updating a handful of entries. - -sche(discuss)02:32, 29 February 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
We currently include both the macrolanguage Bontoc (code: bnc) and its dialects, particularly Eastern Bontok (not even using the same spelling, you notice! code: ebk) which we have about ~35 translations in. IMO, it rarely makes sense to include both a macrolanguage and also all of its dialects; we should usually have one or the other but not both. Ethnologue says the dialects are "reportedly similar", as if they split bnc into dialects in 2010 without without knowing enough about them to tell whether they were similar or distinct. The International Encyclopedia of Linguistics considers Central Bontoc to be only 56% intelligible with Eastern Bontoc, which is only a few percentage points better than the intelligibility of the various Bontocs with Ilocano, suggesting that at least Central and Eastern Bontocs, if not the others, are different languages. Our ~15 "Bontoc" (bnc) entries seem to be Central (Igorot) Bontoc and could be relabelled accordingly if we deprecated bnc. - -sche(discuss)07:48, 29 February 2016 (UTC)Reply
@-sche: Relabelled "Central Bontoc" or "Igorot Bontoc"? And is it "Bontoc" or "Bontok"? Whatever the details, I support the idea of reducing the macrolanguage Bontoc (bnc) to an etymology-only language in favour or having translations and entries for the various Bontoc languages. — I.S.M.E.T.A.01:30, 4 March 2016 (UTC)Reply
"Central Bonto(c|k)" is more common than "Igorot Bonto(c|k)" or "Bontok Igorot", and "Bontoc" is more common than "Bontok". I've tweaked the canonical spelling of Central Bontoc (lbk) accordingly; I suppose the other Bontocs which are currently spelled with k should also be updated. - -sche(discuss)01:58, 4 March 2016 (UTC)Reply
A number of works refer to "the Bontoc language" without specifying which of the Bontoc languages they mean, and we couldn't easily include words from these works if we deprecated bnc; there are even books like Clapp's Vocabulary of the Igorot Language as Spoken by the Bontok Igorots which conflate all the languages of the Igorot people (perhaps understandably, given the point above that the Bontocs and e.g. Ilocano are equally different from each other). However, if we accept that being barely halfway mutually intelligible makes Central vs Eastern Bontoc separate languages, then we're not losing anything of quality by not following (and not being able to easily add content from) books that fail to distinguish such different lects. - -sche(discuss)03:01, 4 March 2016 (UTC)Reply
@-sche: Quite. Though, there will be occasions when it will not possible to work out easily from which of the Bontoc languages a given term in a borrowing language will have been derived. — I.S.M.E.T.A.03:28, 4 March 2016 (UTC)Reply
Btw, Native South Americans: Ethnology of the Least Known Continent lays out the case from Nimuendaju (who documented Mura) that Rivet, at least, if not others, was too hasty in grouping this with Mura. Nimuendaju considers Mat. and Mur to be isolates. - -sche(discuss)05:12, 15 August 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
We currently separate {{rw}}, {{rn}}, {{haq}}, {{suj}}, {{han}}, and {{vin}}, which can all be treated as a single language and often are; their separation seems largely political. The Wikipedia page Rwanda-Rundi language catalogues the differences between rw and rn: they are pretty minor, and a lot seem to have to do with regular spelling differences and tones, which are not even reflected in the standard orthography and could thus be relegated to pronunciation sections. To quote Zorc and Nibagwire's Kinyarwanda and Kirundi comparative grammar (2007):
The terms dialect and language are used loosely in everyday communication. In linguistic terms, the two are bound together in the same definition: a language consists of all the dialects that are connected by a chain of mutual intelligibility. Thus, if a person from Bronx, New York can speak with someone from Mobile, Alabama, and these two can converse with someone from Sydney, Australia without significant misunderstandings, then they all form part of the English language. Kigali [the capital of Rwanda] and Bujumbura [the capital of Burundi] are similarly connected within a chain of dialects that collectively make up the Rwanda-Rundi language.
Kimenyi's A Relational Grammar of Kinyarwanda (1980) explains that:
This language [Kinyarwanda, rw] is very close to both Kirundi [rn], the national language of Burundi, and Giha [haq], a language spoken in western Tanzania. The three languages are really dialects of a single language, since they are mutually intelligible to their respective speakers.
I got the ping, thanks. :) I've just been busy. I'll look into this more closely soon, but on the face of it, it does seem like we could merge them. (And that reminds me that en.Wikt really needs to have a discussion about merging Nynorsk and Bokmal. It's bizarre that we manage to accept that Drents and Twents — and, to use the example above, the English of Alabama and the English of Australia — are not separate languages, but haven't managed to accept that Nynorsk and Bokmal aren't. But that's for another discussion...) - -sche(discuss)06:04, 19 December 2013 (UTC)Reply
I was just talking to a native speaker today to get their perspective on this. They said that the vocabulary varies a lot dialectally, but not along national lines, and it's still easy to understand people on the other side. They claimed that the biggest differences were in the tones, but that's not even marked in the orthography. I think that's a pretty strong ase for merger. —Μετάknowledgediscuss/deeds19:22, 28 December 2015 (UTC)Reply
R. David Zorc and Louise Nibagwire have an entire Comparative Grammar (2007, →ISBN devoted to the differences between the two, which Google unfortunately only shows snippets of, including the TOC which lays out spelling differences, noun class differences, "word pairs, one matched, the other completely different", and false friends, as well as dialect-specific tonal marking. However, vocabulary differences and false friends exist even between English dialects (luck out), and tonal marking and other pronunciation differences which aren't reflected in the orthography can be handled in pronunciation sections, as with Iranian Persian vs Dari. The only thing that gives me pause is the point that dialects on the extremes of the continuum "may not" be intelligible with each other per WP, but then, if the dialects aren't split up along national lines / along the same lines as the codes, then that's not a good argument against merger. - -sche(discuss)00:14, 7 March 2016 (UTC)Reply
Elena Zinovʹevna Dubnova, in The Rwanda Language (1984), page 15, writes: "A. Coupez maintains that "as a matter of fact, Rwanda, Rundi and Ha ( and maybe other languages spoken east of the latter two areas) are so close that they can be regarded as dialects of one language" (17. p. 11). According to the other view they are closely related but still different languages (11, p. 26)." Other sources concur, "In linguistic terms, Kirundi and Kinyarwanda are mutually intelligible dialects of the Rwanda-Rundi language." Merged under the code rw. - -sche(discuss)19:17, 14 August 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
These are three extinct languages of Argentina that lack ISO codes, but two of them have recorded material (the third, Puntano, seems pointless to add). The only problem is that some linguists consider these to be dialects of the same language, although that is debated and cannot be satisfactorily resolved with the limited preserved lexica from each. I would prefer we follow es.wiktionary's lead in adding separate codes for Allentiac (sai-all) and Millcayac (sai-mil). —Μετάknowledgediscuss/deeds05:36, 28 June 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
I suggest renaming
ato from Atong to Atong (Africa) or Atong (Cameroon)
aot from A'tong to Atong (Asia) or Atong (India) if we accept the non-inclusion in the latter name of the border areas of Bangladesh where it is also spoken
References seem to prefer the spelling "Atong" to "A'tong" for aot, and Wikipedia says "The correct spelling Atong is based on the way the speakers themselves pronounce the name of their language. There is no glottal stop in the name and it is not a tonal language." - -sche(discuss)04:37, 29 July 2016 (UTC)Reply
@-sche: I support renaming the languages to "Atong (Cameroon)" and "Atong (India)", respectively; in the latter, "India" can be taken to refer to the subcontinent, rather than the country. — I.S.M.E.T.A.18:55, 31 July 2016 (UTC)Reply
Glottolog mentions a dialect Chánguena (Changuina); Linguist List gives separate codes to Chumulu and Gualaca. Do we know how distinct these are? - -sche(discuss)07:31, 6 July 2016 (UTC)Reply
Do we need a code for the fourth of the languages Oramas covers (in the same work where he covers Gayon, Ayaman, and Jirajira), Ajagua (Axagua, Jagua)? Loukotka says: "once spoken on the Tocuyo River near Carera, state of Lara, Venezuela. [A. Espinosa (Vazquez de Espinosa) 1948, p. 35, only two words: Oramas 1916, pp. 49-57, only patronyms.]" - -sche(discuss)23:02, 10 August 2016 (UTC)Reply
Cf User talk:-sche#Kiriri, where a few other needed additions are mentioned. This language is also known as Catrimbi or Kariri de Mirandela, or Kiriri, and is described by Loukotka as the "lost language of the ancient mission of Saco dos Morcegos, now the city of Mirandela". AFAICT, we should add this language (which is also known as Kariri), we already have Xukuru (which is also known as Kirirí and Kirirí-Xokó), and we should possibly add Xukuru-Kariri (postscript: Done), which is also known as Xocó. Keeping them all straight is going to be difficult and is complicated by the fact that each is known only from short words elicted from elders in 1961. - -sche(discuss)20:21, 14 July 2016 (UTC)Reply
Done but without the accent, since English-language sources seem to drop it at least as often as they retain it. Cabrera (1929) is said to record a few words and Serrano (1945) five more. I've also added a code for Comechingón / Comechingon / Comechingona. - -sche(discuss)17:51, 15 August 2016 (UTC)Reply
Also called Shebaye, Shebayo, per Campbell, American Indian Languages: The Historical Linguistics of Native America; he says David Payne "adduces persuasive evidence from the scant fifteen words recorded in extinct Shebayo (Shebaye) of Trinidad to show that it belongs with the Caribbean group (for example, it appears to have da- ‘my’, and these languages are the only ones which have an alveolar stop and not a nasal for 'first person singular')". - -sche(discuss)05:04, 14 August 2016 (UTC)Reply
Done as "Shebayo", the most common spelling. Douglas MacRae Taylor, Languages of the West Indies (1977), page 15, has: The Shebayo list, taken from De Laet's Novus Orbis, is as follows (divergent spellings found in different editions are shown in parentheses): heia (heja) 'pater', hamma 'mater', wackewijrrij 'caput', wackenoely (wackenoey) 'auris', noeyerri (noeyerii) 'oculus', wassibaly (wassi) 'nasus', darrymaily 'os', wadacoely 'dentes', watabaye 'crura', wackehyrry 'pedes', ataly 'arbor', hoerapallii 'arcus', hewerry 'sagittae', kyrizyrre 'luna', and wecoelije 'sol'. - -sche(discuss)19:22, 15 August 2016 (UTC)Reply
Loukotka says nothing of Voto is attested (which is ironic, because the subfamily is apparently named after it). The Indigenous Languages of South America: A Comprehensive Guide considers it a variety of Rama, possibly based on the fact that the Ramas were also called Votos. - -sche(discuss)05:04, 14 August 2016 (UTC)Reply
Glottolog doesn't list it or have any resources on it, either. Therefore, not added until content in it can be found, or at least determined to exist. - -sche(discuss)05:45, 15 August 2016 (UTC)Reply
As to how different the dialects of Dorasque are: A. L. Pinart's Vocabulario Castellano-Dorasque, Dialectos Chumulu, Gualaca y Changuina gives si (and ji) as the Chumulu form(s), and gives ti as the Gualaca form and ji as the Changuina form for "water". Overall, it's hard to tell how intelligible the dialects would be; some words are quite similar (Chumulu utká, Gualaca utkál "yellow"; Chumulu katuvá, Gualaca katavá "bow"), others are very different: Chumulu sagúsaña, Gualaca θake "blue"; Chumulu sérkala, Gualaca okiyigua "asiento" (OTOH, all three dialects use sérkala for "bench"). "Woman" is biá in Chumulu and Changuina, ωiá in Gualaca. I suppose we can follow Pinart in entering it as one language. - -sche(discuss)01:12, 10 August 2016 (UTC)Reply
Okwanuchu language (nai-okw): Berkeley.edu's Indian Languages project says some (<100) words were recorded in the early 20th century; the Handbook of North American Indians: California refers readers to Kroeber (1925) and Voegelin (1942); Glottolog cites Victor Golla California Indian languages (2011). Kroeber says "The dialect is peculiar. Many words are practically pure Shasta; others are distorted to the very verge of recognizability, or utterly different." Victor Golla, California Indian Languages, speculates at length that Okwanchu may have been "a bilingual mix of Shasta and some other language". There was a people "whom the McCloud River Wintu considered Wintu and called Waymaq ('north people') [whom] Du Bois believed [...] were closely related to, if not identical with, the Shastan Okwanuchu; the survivor of the group whom she interviewed gave her a short vocabulary that included words of Shasta origin (Du Bois 1935:8)." These words included atsa 'water', au-u 'wood', katisuk 'bring'. Golla also says "Okwanuchu speech may also be attested in [eighteen] words identified as 'Wailaki on McCloud' (cf. Wintu waylaki 'north people') that Jeremiah Curtin recorded" in 1884, namely: gü'ru 'man', ki'rikega 'woman', hänumaqa 'old man', apci 'old' (in ki'rikega apci 'old woman'), ä'toqe'äqa 'young man', kewatcaq 'young woman', tse'gwa 'one', hoka 'two', qätsi 'four', tseapka 'five' = 'one hand', hukaapka 'six', tsuwara 'sun', kapqu'[r]wara 'moon', kau 'snow', atsa 'water', gri'tuma 'thunder', itsa 'rock', tarak (Golla's note: "[terak?]") 'earth'. Golla notes: "Curtin employed the BAE transcription system, in which q represents a velar fricative, not a velar stop." "Of these forms, five (man, old man, four, thunder) are not attested in any other variety of Shasta."
Konomihu language (nai-knm): The Handbook of North American Indians: California says "An unpublished Konomihu word list was collected by Angulo (1928a)." Glottolog cites Shirley Silver Shasta and Konomihu (1980), Roland B. Dixon The Shasta-Achumawi: A New Linguistic Stock, with Four New Dialects (1905), Lars J. Larsson Who Were the Konomihu? (1987). Kroeber says "Kroeber says "it is still questionable whether their speech is more properly a highly specialized aberration of Shasta or of an ancient and independent but moribund branch of Hokan from which Karok and Chimariko are descended together with Shasta. [...] Konomihu is their own name." Silver's work documents a number of words, including kihínàpxī́k "woman".
New River Shasta language (nai-nrs): "the language is attested only in a few wordlists" per Berkeley.edu's Indian Languages project. Kroeber, who mentions the alternative exonym Amutahwe, says they were "perhaps rather nearest to the major group in speech, although at that their tongue as a whole must have been unintelligible to the Shasta proper. [...T]he tribe melted away without a survivor, leaving only a fragmentary vocabulary."
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming bvx
The di- in its current name, Dibole, is one of those language prefixes, but curiously enough, the most commonly used name for this language is actually Babole, with a different prefix. (Luckily, unprefixed Bole isn't used, because it's taken up already by bol.) We should rename this to Babole accordingly. —Μετάknowledgediscuss/deeds17:32, 29 June 2016 (UTC)Reply
If I know anything about Bantu languages, "Babole" refers to the people who speak it rather than the language itself. Is this really what they call it? —CodeCat18:24, 29 June 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming pgd
We currently have this as "Gāndhārī". This spelling is indeed used frequently in the literature, but we try to avoid difficult diacritics, and this spelling can be seen as using IAST to render the native name of the language, whereas "Gandhari" is the corresponding English. I think that switching to "Gandhari" would be the better choice. @Aryamanarora, -sche —Μετάknowledgediscuss/deeds08:52, 5 July 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
We follow the ISO in covering this as a single macrolanguage, ers. Following Yu (2012), we should keep ers as Ersu, but also create sit-tos for Tosu and sit-liz for Lizu. —Μετάknowledgediscuss/deeds20:17, 3 July 2016 (UTC)Reply
Done as sit- because there is apparently disagreement over whether Tibeto-Burman comprises a monophyletic group, and if we later remove it in favor of sit, the fewer things we have to rename, the better. - -sche(discuss)03:13, 29 August 2016 (UTC)Reply
Done, although this is another of those languages that is attested only in two (or in this case three) disparate wordlists, by Sapper, Ricke, and Johnston, all published by Lehmann, who considers them nonetheless the same language and presents them in two columns, "Johnston" and "Sapper-Ricke". - -sche(discuss)19:45, 18 August 2016 (UTC)Reply
Loukotka, and Campbell citing him (and Wikipedia citing Campbell), says only a single word is attested, but none of them say what that word is. Other sources, none of which strike me as reliable, provide different(!) words, of which the most common is quindio (one book has quindus, one has kindiyo, parallelling the spellings Quimbaya - Kimbaya), meaning "paradise", but other words mentioned include chascará, batatabatí, and fihisca "spirit" (the last of which may actually be from some other language). proel.org says eight words are attested, but doesn't say what they are. Several books provide placenames which may attest the language. - -sche(discuss)23:57, 20 August 2016 (UTC)Reply
An old source (Daniel Garrison Brinton, 1898) says "The Jupua and Curetu dialects are properly one and the same, the difference which appears in their vocabularies arising simply from inequality in the ears and the orthographies of observers. This is evident by the following comparison..." and then proceeds to offer a comparison which IMO doesn't actually conclusively demonstrate anything. In any case, Cueretú / Curetu itself does not currently have a code. - -sche(discuss)04:20, 14 July 2016 (UTC)Reply
I can find a number of sources attesting Yupua words, but none seem like reliable primary sources: Peruvia Scythica: The Quichua Language of Peru mentions "wui" as the word for house, but does so as part of trying to connect a huge number of unrelated languages based on chance sound correspondances. Brinton's 1898 Studies in South American Native Languages has a list of words which he sources to Martius (surely referring to the Wörtersammlung Brasilianischer Sprachen) and compares to Curetu words sourced to Wallace (surely referring to A Narrative of Travels on the Amazon and Rio Negro), but I haven't located the Yupua words in the originals; the words Brinton gives are: blood: thik (Yupua), dü (Wallace); bow: patopai, patueipei; earth: thitta, ditta; flesh, ga'hi', se'hea'; finger, moh-asoing, mu-etshu; fire, pieri, piure; flower, pagari, bagaria; foot, göaphoe, giapa; hair, poa, phoa; hand, moho, muhu; head, co'ëre, cuilri; house, wu'i, wee; mouth, thischüh, dishi; sun, hauvä, aoué; tongue, toro, dolo; tooth, gobâckaa', gophpecuh; water, thäco, deco; woman, nomöa, nomi; he also offers the additional Yupua words hóggoa "water" (sic), göaphae "foot", ga'hi "meat", jih "jaguar", ikama "deer", jocheo' "star". Ruhlen has manapẹ "husband / man", apara "we two", ti "this", -mai- "we", tsīngeē "boy", pilo "fire", poa "feather". - -sche(discuss)05:29, 14 July 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Berawan
It does not make sense to have both the macrolanguage (lod, which the ISO retired) and the dialects (zbc, zbe, zbw). We should either merge the dialects or retire lod. However, Berawan is composed of more than three varieties, which the literature often speaks of, though there are some works that use the same coarser division as Ethnologue. - -sche(discuss)09:57, 1 September 2016 (UTC)Reply
Being a "lumper" rather than a "splitter", I support merging the dialects into the macrolanguage. Dialect labels can be added with {{lb}} as needed. —Aɴɢʀ (talk) 11:24, 1 September 2016 (UTC)Reply
Oddly, Ethnologue does not provide a claim of how mutually intelligible vs unintelligible they are. Blust, in The Consonants of Long Terawan, asserts four dialects, all spoken within 25 miles of each other: Long Terawan (Ethnologue's zbw) and Batu Belah (part of zbc), and Long Teru (part of zbc) and Long Jegan (zbe); he considers e.g. Long Pata (which WP names, and which is noted by Ray's 1913 Languages of Borneo) to be a variant of one of these, though he is nonspecific as to which one. When I compare Blust's Terawan data and Burkhardt's The historic evolvement of true triphthong phonemes in Long Jegan Berawan, the differences are not large, especially considering the obvious differences in style of notation: e.g. Blust has Terawan gitoh "hundred", Burkhardt has Jegan getoʔ (Blust noted some uncertainty in his transcription of vowels); Terawan dimmeh "five", Jegan dimmiᵊy; T buluh "feathers or body hair", J bullǝw, T iciw "day", J iciᵊw, T iko "tail", J eko, T puté "white", J pote, T depeh "fathom", J dǝppiǝ̈. A larger difference is Terawan lebbih "two" vs Jegan duβiᵊy, T puʔ "head hair", J pǝuk, T manoʔ "bird", J manǎuk. The only entry we have in the dialects is pi "water", which is (per our entry) the same in all of them. It does seem that these could be merged. - -sche(discuss)18:19, 1 September 2016 (UTC)Reply
→ I would prefer qfa-und-capund-cap to avoid it appearing to be a family (qfa-) code, or ine-cap... or und-cap^? And are words in the language attested? WP says no texts survive. - -sche(discuss) 06:08, 11 July 2016 (UTC)- -sche(discuss)17:57, 14 August 2016 (UTC)Reply
I'm fine with three-part codes for a couple rare lects. As for attestation, I chose to put this language on the list chiefly based on The Expositor's Bible, vol. 35, p. 218, where it says: "Thus the ancient Cappadocian language is discussed and a lexicon of it compiled in a monograph which appeared in the Museum of the Evangelical school at Smyrna (1880-84), pp. 47—265." —Μετάknowledgediscuss/deeds06:28, 11 July 2016 (UTC)Reply
Let's not add this, at least until we find positive evidence that it needs to be added. In addition to the statement that no texts survive, I see scholars such as Holger Pedersen savaging claims (by Karolidis, Kretschmer, et al) to have found relics of the language in Cappadocian Greek. - -sche(discuss)18:13, 14 August 2016 (UTC)Reply
Done as sai-cno since -chn was taken. In addition to the catechism, Fitz-Roy (1839) has three words, and Jose Garcia (1889) has three words, per Loukotka. - -sche(discuss)06:18, 17 August 2016 (UTC)Reply
For a language that's known only from a single wordlist, this sure has a lot of names... Jeikó, Geicó, Jeicó, Jaikó, Geikó, Yeikó, all of those with unaccented 'o', and Eyco... - -sche(discuss)02:02, 20 August 2016 (UTC)Reply
Hmm, this is an interesting case. On the one hand, they seem to be mutually unintelligible, and speakers seem to refer to them as separate languages, and outside linguists seem to treat them as separate things (compare Eskayan)... and we do consider even such very similar, often-overlapping things as Norwegian and a slightly different orthography of Norwegian to be entirely separate languages. On the other hand, Damin is said — by researchers with access to its full vocabulary, as opposed to access to only alimitedsurvivingcorpus — to have a vocabulary of "perhaps no more than 250 words in all", which makes it hard to consider it a language, and does make it seem similar to pandanus avoidance registers or especially thick cant or jargon. According to Wikipedia, some markers are different, but others are the same, e.g. genitive -kan and future -ur. I suppose treating it as a dialect of sorts does make the most sense. - -sche(discuss)17:35, 14 August 2016 (UTC)Reply
Water is سو(su); the word also means 'tempering (of a sword)'. 'Ebçi' (apparently containing the occupational suffix '-çi'; ç = č) is 'woman', mentioned in one military treatise when it says that Indian swords are "useful only for hanging on the neck of a woman who cannot give birth to a son". 'sovuq' is 'cold' and 'sol' is 'left': andın songra sol egining ūzārā salgıl taqı boynuñ ūzārā tezgindūrgil "after that, place it over your left shoulder and make it rotate over your neck". See Munytu'l-Ghuzāt: a 14th-century Mamluk-Kipchak military treatise and Vocabulaire arabe-kiptchak de l'époque de l'État mamelouk. - -sche(discuss)00:47, 31 August 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Based on their limited mutual intelligibility, and following the references on these languages (see w:Rgyalrong languages' bibliography), we should split jya, and its category and any entries, into four languages:
Situ (sometimes called Eastern rGyalrong) (sit-sit?)
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Sonqor Turkic is an Oghuz lect spoken in northwestern Iran. The Turkic Languages by Fuchs Christian et al speaks of it, Qashqai/Kashkay (code qxq) and Afshar as "deviating" dialects of Azeri. The Oxford Handbook of Inflection, edited by Matthew Baerman, speaks of Sonqor as "a minority language of Iran heavily influenced by Kurdish". (The Handbook provides some Sonqor example words: ušaġ-ækæ-le' "child-SPEC-PL" "those children" (compare ušaġ to Azeri uşaq), mincuġ-ækæ-re "bead-SPEC-ABL" "of those pearls" (Azeri muncuq-), šéʕr-eke-sin-ne "poem-SPEC-POSS-ABL" "about that poem by him" (Azeri şeir-).) Other sources note that "Sonqor and others" "transitional dialects". How should Sonqor and Qashqai be handled? Ethnologue claims there are "significant differences [between North Azeri and] South Azerbaijani [azb] in phonology, lexicon, morphology, syntax, and loanwords", but we merged them (but not qxq) after this discussion, based on references saying "[the] dialects of Azerbaijani do not differ substantially. Speakers of various dialects normally do not have problems understanding each other" (heck, even Azeri and Turkish speakers do not have that much difficulty understanding one another). Gilles Authier, in New strategies for relative clauses in Azeri and Apsheron Tat, in Clause Linkage in Cross-Linguistic Perspective (2012, →ISBN, similarly says "There is an almost perfect mutual intelligibility between Azeri and Kashkai speakers. I tested this personally by submitting recordings to both audiences." It sounds to me as if Kashkay should be merged into Azeri, and as if we don't need a code for Sonqor. However, I'm pinging our only recently active editor who speaks some Azeri, User:123snake45. - -sche(discuss)03:33, 11 September 2016 (UTC)Reply
As you noted, only political/orthographic considerations compel us to keep Turkish and Azeri apart. I'd keep all these merged in Azeri. While we're at it, have we ever discussed whether the name "Azeri" should be changed to "Azerbaijani"? —Μετάknowledgediscuss/deeds03:46, 11 September 2016 (UTC)Reply
I do not speak anything Turkic; but I keep getting the general impression we make too little use of the "X is a dialect of Y" functionality. This would be well-suited for handling cases where one variety is mostly intelligible with a bigger standard language, but has its share of its own quirks (loanwords, occasional diverging phonetics, etc). --Tropylium (talk) 13:32, 12 September 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming bzj
The name we use, "Belize Kriol English", is certainly among the more uncommon names of this language. Much better would be just "Belize Kriol"; there might also be argument to be made for the more dated term "Belizean Creole" (which is the title of the Wikipedia entry). There is one entry which would have to be changed. —Μετάknowledgediscuss/deeds20:25, 3 April 2016 (UTC)Reply
On Ngrams and in the 'raw' Google Books results, and when I check academic sites on Google (site:.edi) and academic papers (filetype:pdf — most results for most non-famous languages' names are academic papers, dictionaries, etc), the most common name is "Belizean Creole" (which doesn't seem dated), then "Belize Creole", with "Belize Kriol" bringing up the rear. Approximately the same pattern holds when I page through the results looking only for primary reference works about this language in particular, as opposed to works that just mention it. There is some interference from the fact that "Belizean Creole" and "Belize Creole" are apparently also used to name the people who speak it and not just the language, but I suggest a rename to "Belizean Creole". - -sche(discuss)18:27, 20 August 2016 (UTC)Reply
@Angr, since you are knowledgeable of Celtic languages: do you agree that the Gallaecian language and Ivernic language should be added? Should "Ivernic" be called that, or another name? And do you know if any words in it are attested? (It's not clear to me if ond and fern, which Wikipedia mentions, are Ivernic words, or words borrowed by some other language from (potentially slightly different) Ivernic words.) - -sche(discuss)20:18, 12 August 2016 (UTC)Reply
@-sche:, AFAIK "Ivernic" is unattested; ond and fern are Old Irish words which Cormac mac Cuilennáin (who lived in the 9th century) believed to be of "Ivernic" origin. I don't think we need to add it. Gallaecian, on the other hand, is attested. If it were to be subsumed under anything else, it would be Celtiberian (xce), but I'd rather keep it separate. —Aɴɢʀ (talk) 20:38, 13 August 2016 (UTC)Reply
Light Warlpiri, a small mixed language (350 speakers, all also fluent in Warlpiri). I'm not sure if we should have this or not.
It's gotten a lot of press, but I lean toward saying it's not worth including — mixed languages are always a little hard to justify, and one that's just come into being is covered by Warlpiri and English sections well enough, I reckon. —Μετάknowledgediscuss/deeds05:30, 4 October 2016 (UTC)Reply
It apparently has its own orthography and is spoken on a different island, which might have motivated a non-merger, although I can find no discussion of it so it seems most likely that it simply went entirely unnoticed. Reconstruction:Proto-Austronesian/daNum and water make use of it. The only references I can find that mention it do speak of it as a dialect of Malagasy. Merged. - -sche(discuss)01:24, 4 October 2016 (UTC)Reply
On Google Books, "Puxian" and "Pu-Xian" seem to be the most common names, about equally common: but on Google Scholar, about 3.5 times as many works use "Puxian" as "Pu-Xian", so I will rename the language to "Puxian". "Pu Xian" does get a few hits. Incorporating "Min" ("Puxian Min") seems to be uncommon. Another attested spelling (not common) is "Putian". Wikipedia says Xinghua and Hinghwa are additional alt names. - -sche(discuss)13:50, 1 November 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Searches for "speak X" and "X language" on Google Books reveal that [tem], which we currently call "Timne", is referred to as "Temne" far more frequently than by any other spelling. The native spelling, Themnɛ (or the ASCII-friendly Themne), sees extremely limited use. See also w:Temne language. —Μετάknowledgediscuss/deeds05:40, 4 February 2017 (UTC)Reply
Coroado Puri language (sai-crd), of which Revista de la Universidad Nacional de Cordoba, volume 6, includes two quite disparate wordlists, with "sun" alternately given as Hope or Aaam, "moon" as pethara or kesha, "water" as something illegible (nhanan?) or gioi. The Viagem a Curitiba e Província de Santa Catarina also has some words, including goio "water". Greenberg, from who knows where, sources ope as "sun", oron as "new moon", and also e.g. čama "animal", puara "chest", teke "belly", bo, ambo "tree", ke "wood, fire", tong, ton "nape of the neck" (compare Greenberg's Puri thong), uka, høka "stone" (compare Puri ukhua), tima "love/want/enjoy" (compare Puri tamathi), baj "to live".
Menien (sai-men), attested in a Vocabulario Menien by Príncipe von Wied. He wrote in German, was translated with only a few alterations into French, and then was translated badly with a number of serious errors into Portuguese (says Loukotka, giving some examples from all three editions). Words (agreed on by the German and French editions) include hioi "wax", koinin [sic not h-] "child", and kohira [sic not c-] "red".
Curiously (and unhelpfully, but ultimately irrelevantly), the Real diccionario de la vulgar lengua guanaca (Joaquín Meza, 2008) is not a dictionary of this language, but (according to its description) of vulgar Salvadoran Spanish. - -sche(discuss)21:06, 12 August 2016 (UTC)Reply
Whether Corobici existed as a separate language, and whether words in it are attested, is unclear to me (and apparently to a lot of writers).
one unlikely idea
The 1911 Indian Languages of Mexico and Central America says:
"Gomara (1: 264; 2:457) and Herrera mention a tribe (the Corobici) which seems to be identical with the Mangue (or Chorotega). The latter author says (II, dec. 3, 121), "Hablaban en Nicaragua cinco Lenguas diferentes, Coribici, que lo hablan mucho en Chuloteca," etc. [= "They spoke in Nicaragua five different languages: Corobici, who speak a lot in Chuloteca,"] Nevertheless, Peralta thinks the Coribici were the ancestors of the Guatuso (see below). It would seem that Mangue is a comprehensive term precisely equivalent to Chorotega, properly used, that is, to include the Chiapanecan element in this region — Choluteca, Dirian, and Orottinan."
However, Mangue (code: mom) and Corobicí are in different language families, so an equation of the two is very suspect.
Other sources quote Gomara (sic, not Herrera as above) as saying "There are in Nicaragua five very different languages: Corobici, which is praised much;" etc.
Native-languages.org, based on who knows what, identifies Corobicí as bzd. They give four words, which I would not trust without other confirmation: tun "man", si "water", unsat "earth" and guá "house". (For comparison, Wikipedia gives the following Rama words: kiikna "man", kumaa "woman", nguu "house", sii "water".)
The 2013 Native Languages of the Americas, volume 2, citing Mason, envisions a "Rama-Corobici" family which includes the two languages "Rama" and "Guatuso (incl. Corobici)"; the Classification and Index of the World's Languages by Charles Frederick Voegelin and Florence Marie Voegelin also includes Corobici in Guatuso. Guatuso is also called Maléku Jaíka, and Ethnologue says it's specifically not intelligible with bzd.
However, the 1950 Handbook of South American Indians says "Guatuso, with its variety Corobici or Corbesi, and Rama with its dialect Melchora, are obviously very different from each other and from other Central American Chibchan languages, and Mason (1940) was evidently in error in making a Rama-Corobici subfamily." It goes on to place Guatuso and Rama in separate (sub)families.
Comparative Chibchan Phonology (1981), speaking about someone who has "gives a small sample of" Corobici, says "The 'Corobici' words are [actually] words from the dialect of Rama that was spoken in the region of Upala, Costa Rica, up to the 1920s. Really, not a single word of the language of the Corobicies (an extinct group) was recorded". An 1890s report by Daniel Brinton to the US Congress agrees, "Nothing remains of the Corobicies or Corvesies except the name Corobici or Curubici" used in placenames.
Other authors say the Corobici were the ancestors of both the Rama and Guatuso. Tozzer says "The modern Guatuso are probably descendants of, and synonymous with, the ancient Corobici." Lothrop says "It is generally assumed that the Rama were once a tribe identical in language and speech with the Corobici."
It's hard to find information on this Chapacuran language, firstly because it's not discussed much in the literature, and secondly because it has the same name, in the same range of spellings, as the Cariban language of the Cumanagotos (cuo). Some sources speak of it as a dialect of Abitana (Wanham), and Loukotka's wordlist is quite similar to his Abitana wordlist; indeed, the two are more similar than some wordlists of the same language (e.g., of Peba) are to each other. Glottolog only has resources on Kuyubi / Kaw Ta Yo, and although the equation of that with Cumana is non-obvious, it would at least be a less confusing name (any spelling of Cumana is likely to be misinterpreted as cuo). - -sche(discuss)03:36, 31 August 2016 (UTC)Reply
There may be two lects called Guachí. Wikipedia places a Guachí / Wachí, which it considers possibly Guaicuruan, in Argentina. Glottolog places a possibly-Guaicuruan Guachí in southwestern Brazil, in Mato Grosso do Sul, where Opaie is also spoken. The index to Čestmír Loukotka, Johannes Wilbert, Classification of South American Indian Languages (1968), says they describe Guachí is two places: pages 51-52, which I can't see, and page 66, which says: "Opaie or Ofaie-Chavante - spoken on the Ivinhema, Pardo and Nhandui Rivers in the Brazilian state of Mato Grosso, now by only a few individuals. [Nimuendaju in Ihering 1912, pp. 256-260; Nimuendaju 1932b, pp. 567-573; Ribeiro ms.a.] The so-called language 'Guachi' on the Vaccaria River in the same state is only a dialectal version of Opaie. [Nimuendaju 1932b, pp. 567-573.]" Opaie is not Guaicuruan. It is possible that pages 51-52 describe a different lect. It is separately possible that the more recent (2004) Guaicurú no, macro-Guaicurú sí: Una hipótesis sobre la clasificación de la lengua Guachí (Mato Grosso do Sul, Brasil) knows more than Nimuendaju in 1932 and Loukotka in 1968 about the separateness and familiar relationships of Guachí. [5] lists two Guachí words, [iava] ‘diente(s)’ and [iacté] ‘pierna’. - -sche(discuss)04:46, 11 July 2016 (UTC)Reply
I'm having trouble finding evidence that this lect was recorded. Greenberg says "In Paezan, we find Guambiana, Guanaca, and Totoro with the noun plurals -ele, -el, and -le, respectively", but existence of the Paezan family has been discredited and it's not clear where Greenberg gets his Guanaca data; Glottolog apparently doesn't list the lect and I have not yet found anyone other than Greenberg who mentions words in it. Loukotka cites Castillo y Orozco's 1877 Vocabulario páez-castellano as a source for the lect, but I haven't found it in that work; maybe it's under an alternative name or a spelling I couldn't think of to check? I've found the spellings Guanaca, Guanáca, Guanaco, Guanuco, Guanukco(?), Wuanaka mentioned or used in various sources. Huanca and Huanuco are names of a variety of Quechua (whereas WP says Guanaca is "perhaps Barbacoan"). - -sche(discuss)15:51, 18 September 2016 (UTC)Reply
I'm ambivalent about this. WP says it's known only from one wordlist "and even its exact location is unknown. [And] Smits and Voorhoeve (1998) assumed it to be equivalent to Kalabra". There are, as WP notes, lexical differences, but compare other languages attested (with more certainty) in multiple wordlists that show great differences, including Tapachultec and Coroado Puri. OTOH, we could include it and just merge it later if more information came out that it was Kalabra. - -sche(discuss)01:44, 20 August 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
This obscure lect does seem like a separate language, but wasn't described until Cobbinah (2013). The Wikipedia article follows his lead and calls it thus, but it seems odd for us to do so, given that all the other Bainouk languages are given names starting with Bainouk. We also need to decide on a code, perhaps alv-bgu. —Μετάknowledgediscuss/deeds00:19, 2 April 2016 (UTC)Reply
Added. Wikipedia says Gubëeher (which I can also find referred to as that single word, or with the alternative "prefixes"/first-elements Nyun Gubëeher or Nun Gubëeher) is merely closely related to the two Bainouks, so it's probably tolerable that the spelling is Baïnounk here and Bainouk there. Incidentally, WP citing Glottolog says bcb is a duplicate code for an alternative name of bab. Should they be merged? and should we drop the "Bainouk"/"Baïnounk" prefixes? - -sche(discuss)03:25, 13 September 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming aig
We currently call this "Antigua and Barbuda Creole English" — names containing and are suspect, in my opinion. Wikipedia calls it "Leeward Caribbean Creole English". The most common name in literature seems to be "Antiguan Creole", which is what I suggest renaming it. - -sche(discuss)03:22, 19 April 2016 (UTC)Reply
The country is called Antigua and Barbuda, so I don't see anything suspicious about the name per se. Nevertheless, I prefer Wikipedia's name, since it's also spoken in Montserrat, Saint Kitts and Nevis, and Anguilla. —Aɴɢʀ (talk) 17:53, 19 April 2016 (UTC)Reply
I concur with Angr, although I have hesitations about changing it. The use of "Antiguan Creole" seems to be mainly by scholars who are just studying the lect spoken on that island and not worrying about what is and isn't considered the same language. —Μετάknowledgediscuss/deeds04:12, 26 June 2016 (UTC)Reply
This language got encoded twice, once as Kela [kel] and once as Yela [yel], based on its two names in two different provinces of DR Congo. Wikipedia opts for "Kela-Yela" as the compromise name, and I can't find a better one, unfortunately. (This also opens up "Kela" for [kcl], which we currently call "Kala", though that name seems to be less commonly used.) —Μετάknowledgediscuss/deeds06:00, 4 October 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Sena is mutually intelligible with Chichewa, so it really shouldn't be considered a separate language, but that's probably not worth a merger, as the speakers seem to disagree. What the speakers do seem to agree with is that there is one Sena language, despite it having gotten two ISO codes, seh in Mozambique and swk in Malawi, which we call "Sena" and "Malawi Sena" respectively. We should probably just do away with swk. —Μετάknowledgediscuss/deeds23:48, 8 October 2016 (UTC)Reply
Renamed. It seems to be as common or perhaps more common to spell it with a y, although searchign is complicated by chaff from the SOP phrases "old Uighur", "old Uyghur". - -sche(discuss)01:53, 28 March 2017 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
This is pretty decidedly too distinct to be considered a dialect; I was just reading a grammatical description and it's heavily different from Swahili. Maho, Kisseberth, and others seem to agree that it is a separate language. As for the name, it seems that "Chimwiini" (sadly, with the language prefix) is a better option than "Bravanese", used at w:Bravanese dialect. I suppose the code could be bnt-mwi. @-sche, I noticed this because of muke; what other Chimwiini entries did you add? —Μετάknowledgediscuss/deeds03:28, 25 March 2017 (UTC)Reply
I know what your reasoning is for choosing bnt-mwi as the language code, but I think that it's better if the code matches the English name, thus bnt-chi or bnt-cmw. —CodeCat18:56, 25 March 2017 (UTC)Reply
I agree it merits separation; indeed, I see articles going back to at least the 1960s pointing out that "Certainly Bravanese and standard Swahili are not mutually intelligible" (Morris Goodman's 1967 Prosodic Features of Bravanese, J. of African Languages 6). And yes, for better or worse codes have tended to be named based on the English names. Did you search for "Mwini" with one i? I find roughly comparable numbers of Google Books hits for "Mwini" Swahili, "Chimwiini" Swahili and "Bravanese" Swahili. (Ngrams, catching I-don't-know how much chaff, finds Mwini to be the most common spelling recently, followed by Bravanese — the most common spelling until recently — and then Mwiini, with Chimwiini not common enough for it to plot.) It looks like we could avoid the prefix if you wanted, though I defer to you if you checked more literature than the cross-section Google has digitized. Google Scholar has ~120 articles on "Chimwiini" Swahili compared to ~90 for "Mwini" Swahili, some of which are the irrelevant last name Mwini, and also ~90 for "Bravanese" Swahili, but many are bibliographies mentioning the same few articles. Other spellings I see, besides Chimwiini, Bravanese, Mwiini, and Mwini, include Chimwini, Chimini, and Brava. - -sche(discuss)23:37, 25 March 2017 (UTC)Reply
In my own files, and likely elsewhere, the Chichewa word mwini gets in the way. But in any case, most of the works I see dedicated to the language that aren't on the older end use "Chimwiini". Anyway, bnt-cmw is fine. —Μετάknowledgediscuss/deeds23:43, 25 March 2017 (UTC)Reply
"Arc" should only be used for "Imperial Aramaic" (aka "Official Aramaic"), ideally written in the Old Aramaic script rather than Hebrew. Current usage does not reflect that though and there are a whole mix of dialects intertwined within the "arc" code, so that one at least should be split. "Syc" should stay as it is. --334a (talk) 08:04, 21 December 2012 (UTC)Reply
I think SIL has done a really bad job with the classification of Aramaic languages. ARC was an umbrella code that was used to describe all later Aramaic varieties in ISO-639-2 in ISO 639-3 they introduced SYC for classical Syriac, by far the most widespread form of literary Aramaic.--Rafy (talk) 17:38, 28 December 2012 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
This is an issue that needs to be resolved soon: there are 90 module errors related to it.
First let me lay out the linguistic details:
East Frisian, a language of Lower Saxony in Germany, is, along with West Frisian in the Netherlands and North Frisian in Germany, one of the w:Frisian languages. It has historically consisted of two dialect groups: one near the Ems River, and the other near the Weser River. The Weser dialects are all now extinct, with the w:Wursten Frisian dialect surviving into at least the 1700s, and the last speaker of w:Wangerooge Frisian dying in 1953. Although most of the Ems dialects died out by the 18th century, Saterland Frisian has survived to the present day. The extinct dialects were absorbed into Low German to become East Frisian Low Saxon.
The ISO has typically made a massive mess out of the Germanic languages, and they really screwed up in this case- until recently, it was impossible to tell if their code for "East Frisian", frs, referred to Frisian East Frisian or to East Frisian Low Saxon. If the first were true, it would overlap with Saterland Frisian, stq. If the second were true, it would be just another of the Low German lects that we decided earlier to treat as German Low German, nds-de (not Dutch Low Saxon, nds-nl, in spite of having "Saxon" in the name). Because of this ambiguity, the frs code was pressed into service for Frisian East Frisian in upwards of 140 etymologies and translations (there might have been entries, too, but I have no way to tell).
A few weeks ago, I mentioned to User:-sche that the online description of another lect had been updated. In the process of checking this out, he discovered that frs was now unambiguously described as East Frisian Low Saxon, and thus redundant to nds-de. After making the usual checks of the categories and changing uses of frs that he knew about, he removed frs from the data module. Unfortunately, East Frisian is mostly only mentioned in etymologies as cognates, and cognates don't show up in any of the categories, nor do redlinked translations. User:Leasnam (who added most of the frs references in the first place), -sche and I have been able to whittle it down from 137 entries in Category:Pages with module errors, to the present 90 just by getting rid of unnecessary ones and by changing recognizable instances of Saterland Frisian and East Frisian Low Saxon to the correct language codes.
As I see it, there are two halfway-decent options:
Merge all of Frisian East Frisian into Saterland Low Frisian, stq, since the latter is the only surviving dialect of the former.
Create an exception code for Frisian East Frisian, such as gmw-efr or gmw-fre.
There's also the possibilty of restoring frs as Frisian East Frisian, but that would put us in direct contradiction to the ISO standard, and leave things open for all kinds of confusion.
The first probably fits the linguistic facts best, but the second may be more practical, at least in the short run.
I've found no references on East Frisian Low Saxon, and very little on Saterland Frisian (there's a Saterland Frisian Wiktionary, but most of the remaining terms aren't mentioned there). There is, at least, one good dictionary of Frisian East Frisian available online. Those with better sources apparently don't have the time to work on this right now. Chuck Entz (talk) 02:41, 11 April 2015 (UTC)Reply
I don't think restoring frs is a good idea for the confusion you mentioned. I think option 1 is the best; we would then treat Saterland Frisian as the main dialect of East Frisian. Option 2 would just introduce more ambiguity, since one language would suddenly become a part of another. This is why we eliminated the Low German varieties in the first place. —CodeCat17:28, 11 April 2015 (UTC)Reply
Yes, it wouldn't make sense to have both "East Frisian" and "Saterland Frisian" (they are not sufficiently distinct), so option 2 would only make sense if we changed instances of stq to the new pan-East-Frisian code and {{label}}-ed them. But we do need to recognize that there are inclusible East Frisian words which are not Saterland Frisian (namely, all the words and forms that we know — from records — existed in the non-Saterland varieties of East Frisian). Precedent exists of us repurposing codes to refer to slightly more things than the ISO intended them to refer to, e.g. we use gcf to refer to both gcf and acf and we used en to refer to both en-proper and hwc (Hawai'ian Creole English) and pld (Polari). Hence, I would add "East Frisian", "Eastern Frisian" etc as alternative names of stq, and then change all remaining uses of frs to stq to solve the module errors. Then, at leisure, we can go back through the affected pages and specify, whenever possible, which precise dialect they are from. (It may look a bit ugly to have e.g. "Wursten Saterland Frisian foo" or "Saterland Frisian foo(Wursten dialect)", but it's probably the best we can do, since changing the canonical name of stq to "East Frisian" would just invite people to become confused about it and misuse it again.) - -sche(discuss)16:58, 13 April 2015 (UTC)Reply
By the way, with all these module errors on highly visited pages, we can't really wait for this to go through RFM procedure. I think whoever sees this next, if (s)he has the time, should implement -sche's temporary solution. (I myself will do it if I can get my work done in meatspace quickly enough.) —Μετάknowledgediscuss/deeds06:36, 16 April 2015 (UTC)Reply
This language's Wikipedia article helpfully explains that this extinct language of Ethiopia is unattested. However, since Amharic is an LDL and wordlists of the Weyto dialect of Amharic have been made that contain words purported to derive from it, we should make it an etymology-only language code. —Μετάknowledgediscuss/deeds02:23, 4 April 2017 (UTC)Reply
I wonder why the SIL/ISO assigned a full code in the first place to an unattested extinct language of unclear family association that is not even certain to have existed... - -sche(discuss)03:16, 4 April 2017 (UTC)Reply
The only reason though why I used Dupaninan Agta instead of Dupaningan Agta is because that's the spelling that came out of Wiktionary when I inputted duo. --Mar vin kaiser (talk) 10:18, 5 April 2017 (UTC)Reply
Which, in turn, is because that's the spelling the Ethnologue/SIL/ISO used when we copied codes over, years ago. But the spelling with g seems slightly more common. Should it be made the canonical spelling, with the other one retained as another name? The names without "Agta" (i.e. "Dupaninan", "Dupaningan") should also be otherNames, since they are sometimes encountered. - -sche(discuss)17:04, 5 April 2017 (UTC)Reply
@Mar vin kaiser, Stephen G. Brown I have updated things so that the more common spelling is the one used in all the entries and translations tables and categories now: Dupaningan Agta. I'm sorry if this causes you to have to relearn how to type the header, etc. The "joys" of muscle memory! :-p - -sche(discuss)07:27, 10 May 2017 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
I propose to merge [kjy] "South Kewa"/"Erave Kewa", [kjs] "East Kewa" and [kew] "West Kewa"/"Pasuma Kewa" into [kew] as "Kewa". AFAICT most literature treats Kewa as a single language, and the only effect having three codes has had upon us so far is that our Kewa content is duplicated under several headers (as in ipa and utyali). There seems to be a far more marked difference between normal Kewa and its pandanus avoidance register than between South, East and West. - -sche(discuss)06:00, 11 August 2015 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
[pqw-pro] → [poz-pro]
As discussed earlier this year. Western Malayo-Polynesian is a solely geographic group, it is not recognized by our language categorization system, and a proto-language appendix seems to be superfluous.
In addition to the mentioned durian issue, we have currently 13 Proto-WMP lemmas, with a breakdown as follows:
10 entries are fully identical to corresponding Proto-MP entries (e.g. *huaji = *huaji, *wada = *wada)
2 entries (*huaji-ŋ, *qari-mauŋ) are reconstructed from very scarce data, and the most likely situation is that they were just randomly lost in Central-Eastern MP.
1 entry (*azak) is, per the cited source (Blust's dictionary), probably a late Wanderwort originating in Malay(ic) and not inherited.
Even if there are regional differences, you don't have to resort to separate protolanguages to explain them: for one thing, the substratum languages encountered off of Southeast Asia had to have been vastly different from those farther east. As for animal (and to a lesser extent plant) names, there's the matter of the w:Wallace Line and other such boundaries: the farther east you go, the fewer non-marine Asian species you find. By the time you get to Polynesia, the only flightless land animals (New Zealand is an exception, or course) are human-transported creatures such as pigs, chickens, dogs and rats, and the only widely-distributed plants are those with seeds that can drift on the currents, or Polynesian canoe plants- if you don't have wild beasts, you're not likely to preserve inherited names for them. Chuck Entz (talk) 02:30, 30 June 2015 (UTC)Reply
Perusing Glottolog's large bibliography for this language, I note that Kxoe predominated and Kxoe was little used until about the year 2000, after which Khwe has predominated and Kxoe has been little used. Wikipedia uses Khwe and mentions that it is the preferred spelling. Renamed. - -sche(discuss)07:51, 22 May 2017 (UTC)Reply
Nakrehé language (sai-nak) — not entirely sure this is separate from Krenak
The Encyclopedia of the World's Endangered Languages and The Indigenous Languages of South America: A Comprehensive Guide treat this as an alternative name of Krenak or a dialect. Not added at this time. - -sche(discuss)22:47, 12 May 2017 (UTC)Reply
Also called Cenúfana, Zenúfana, Cinifaná, Sinufana, Sinú, Cenú, Zenú; dialect names Fincenú, Finzenú, Pancenú, and supposedly Sutagao. Added; see my comments below. - -sche(discuss)05:21, 11 May 2017 (UTC)Reply
Sörsörian language (map-sor) — I can't actually find any material on this one
The evidence for this language is quite sketchy — it's only known from one wordlist by someone who may not have spoken Tawasa at all — but it's extinct if it existed at all. I suppose it can be added and then its (the wordlist's) entries' etymologies and those of Timucua and Muskogean entries' etymologies can discuss each other. - -sche(discuss)15:51, 18 September 2016 (UTC)Reply
The variety of names this language has and their homography with the names of the people and the place they lived, has made it hard to search for information on this language. And some scholars say there is nothing to find: The Indigenous Languages of South America (edited by Lyle Campbell and Verónica Grondona) quotes "Adelaar and Muysken (2004: 624): [Sinu] 'cannot be classified for absence of data.' Loukotka (1968: 257) grouped Zenú (Senú) with the Chocó Stock, though nothing was known of the language." However, I've been poking around for quite a while now, and finally found a copy of A. Oyuela-Caycedo's full article in the Handbook of South American Archaeology (edited by Helaine Silverman, William Isbell), which I'd previously only been able to see a page of. On the page I'd seen, Oyuela-Caycedo says "the descendants of the Sinú have lost their language, making it impossible to classify them in terms of known linguistic families (Adelaar and Muysken 2004). However, taking toponyms into consideration the area seems to have been occupied by Chibchan speakers." On the next page, however, Oyuela-Caycedo goes on to discuss Spanish records and mentions that the "name or title" of the Finzenú chief was "Tota", with which clue I tracked down Jaime Alberto Castro Núñez's Historia de la medicina en Córdoba: notas preliminares (2002), mentioning a few other words and names: "A la llegada de los cappunia, como los indios llamaban en su lengua a los españoles, la cacica de Finzenú era Tota; el Zenúfana era Nutibara y el Panzenú era Yapé." "At the arrival of the cappunia, as the Indians called in their language the Spanish, the chief of Finzenu was Tota, that [of] Zenúfana was Nutibara and that [of] Panzenú was Yapé." (Other mentioned names are Anunaybe and Quinuchu.) The ambiguity over whether Tota was a proper name or title appears to be original; I find what seems to be a copy of Sebastián de Benalcázar's writings, saying "[...] Tota, nombre que no sabemos si era el de su cargo o propio de la última reina de este riquísimo pueblo". - -sche(discuss)05:21, 11 May 2017 (UTC)Reply
The Mongolic lect spoken by the long-bilingual Steppe Tungus. Juha Janhunen has described and documented it in several works. 'Water' is ůhu(n) (ůhů(n)); compare Classical Mongolian 'usun' vs modern standard Mongolian 'us', 'usu'. - -sche(discuss)23:55, 30 August 2016 (UTC)Reply
Alungul and Angkula are very similar (but, OTOH, so are other related languages), with Barry Alpher mentioning the following words: Alungul oth:o, Angkula otho "liver" (but e.g. Ikarranggal otho is also similar); Alungul mbolvm, Angkula ombolvm "mosquito" (but also e.g. Ogunyjan ombolvm); Alungul orormv, Angkula otil "nape" (the latter "likely a loan from Uw-Olkola odel", and other languages use different words); Alungul obmo(gng), Angkula obmu "nose" (but also e.g. Athima ubmu); Alungul atïl, Angkula atï "see" (Ikarranggal ara); Alungul amadhv, Angkula amadh "shin" (Athima amadhv); Alungul anggul, Angkula angkul "tooth" (Athima angkul, Ogunyjan enggul). Many words were even recorded from the same informant, with one scholar writing "West recorded considerable material in Ogh Alungul and Ogh Anggula from the now deceased Mr Jack Burton, but subsequently l was able to record and transcribe not only a few fragments more of these but a quantity of another dialect." It is difficult to say if they should be treated as one language, and what it would be called. They have separate codes for now. They could always be merged later. - -sche(discuss)18:35, 25 May 2017 (UTC)Reply
In general, we do try to avoid using "Standard" in language names (e.g. we have "Estonian", not "Standard Estonian"), so a rename would be good, and "Moroccan Amazigh" does seem to be more common than "Moroccan Tamazight". - -sche(discuss)18:00, 24 June 2017 (UTC)Reply
If giving variant names, indicate preferred form first
Reference where found:
The French name is used by Institut royal de la culture amazighe (IRCAM). Amazighe is the name found in the French version of the constitution approved by a referendum on the 1st of July 2011.
The qualificative "marocain"/"Moroccan" is used to make sure it does not refer to a specific Moroccan dialect of Berber [tzm] (Moroccan being broader than the Central Atlas qualitificative used for [tzm]) or to the Berber/Amazigh family as a whole [ber] (Moroccan being narrower than the whole family’s coverage area).
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Judging by Google Books and Scholar searches, and based on the references cited by Glottolog and Wikipedia, the most common name for this language in recent literature (since the 1990s?) seems to be Barngarla, while the most common name historically/overall (still found in some recent references) is Parnkalla. Some old references have been updated from Parnkalla to Barngarla, e.g. Mark Clendon's Clamor Schürmann’s Barngarla grammar: A commentary on the first section (where the referred-to original had Parnkalla), which argues based on recordings as well as etymology that the name is /parnkarla/, with the /ŋ/-form Banggarla as a northern dialect form "or even an exonymic pronunciation". Banggarla, and Barngala with no second 'r' and Parnkarla with 'P' and two 'r's, seem relatively uncommon, and many other spellings exist (see Glottolog). I suggest a rename to Barngarla, or Parnkalla (this entails moving the categories). (Fr.Wikt has "banggarla" and "barngala" as separate languages, but this seems to be an oversight and I will let them know.) - -sche(discuss)15:43, 26 June 2017 (UTC)Reply
@Crom daba, what would happen to words in older literature such as de Smedt / Mostaert 1933 or Todaeva 1973 which are not clearly marked as being either Minhe or Huzhu? Now they can be added under Monguor. Someone can later label them as Mangghuer or Mongghul using {{lb|mjg|}}. After splitting, a lot of useful stuff from older sources will hang in the air. --Vahag (talk) 14:49, 22 December 2016 (UTC)Reply
Most sources are identifiable as either Mangghuer or Mongghul, we could specify which resource contains what in the about page. A bigger problem for me is how do we even enter data from the old sources? Do we put in Todaeva's Cyrillic and Mostaert's pre-IPA phonetic symbols or do we transcribe it into Pinyin? I wasn't around long enough to see what the precedent here is. Crom daba (talk) 22:08, 22 December 2016 (UTC)Reply
If you can specify which resource contains what in the about page, then I support the split. Otherwise, I would like to have three codes, like we do with ku (Kurdish macrolanguage), kmr (Northern Kurdish variety), ckb (Central Kurdish variety).
As for entering words from older sources, you can normalize them into Pinyin, as long as the rules of normalization are clearly defined. Look at what I have done with Udi at WT:AUDI. --Vahag (talk) 06:56, 23 December 2016 (UTC)Reply
After some research, it appears that correspondence of Pinyin spellings (as written by Dpal-ldan-bkra-shis et al) and older attestations is non-trivial, so I will put off transcribing anything which isn't already in Pinyin, at least until I see an example of it supporting the full range of dialectal and historical Monguor variation. Crom daba (talk) 00:16, 24 December 2016 (UTC)Reply
@Crom daba, Vahagn Petrosyan, Angr, Metaknowledge: I added codes for Mangghuer (xgn-mgr) and Mongghul (xgn-mgl). I suggest that we move as much content as possible to the new codes and update the orthography as we go, and that will give us an idea of whether or not it is feasible to retire the macrolanguage code yet. - -sche(discuss)00:58, 2 June 2017 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Solon is a language spoken by a Tungusic people living in Manchuria and Inner Mongolia, they consider themselves Evenks, but their language is somewhat different and is not usually counted among Evenki dialects in literature. We follow Ethnologue in categorizing it as Evenki, but most Russian Tungusic literature (Tsintsius, Vasilevich, ...) counts it as a separate language and I too think this would be preferable.
Also worth mentioning is that we have the language of Oroqen as a separate language already for whom Janhunen claims are " in fact much closer to the "Ewenki proper" (i.e., the Evenks of Siberia) than the Solon are" (quote from Wikipedia). Crom daba (talk) 08:37, 7 August 2017 (UTC)Reply
Sadly (as the lack of response shows), I think you may be the only editor with knowledge of Tungusic. The English-language works I can find discussing the lects, while mostly general rather than specialist, also seem to mostly speak of them as separate languages, even though Wikipedia redirects "Solon language" to Evenki. I see that Solon currently has an etymology-only code ("evn-sol"); do you want it to have a "full" code and its own header and entries like дяви, @Crom daba? That seems reasonable, although the code should be "tuw-sol", I think, to fit the customary naming scheme describe in WT:Languages; if that's OK, ping me to add it or add it yourself to Module:languages/datax (and then update the entries that refer to it and remove the etymology-only code "evn-sol"). - -sche(discuss)06:44, 29 December 2017 (UTC)Reply
Yes, perusing those searches and Glottolog's list of reference works about it, I see that some books do still use "Angas" but "Ngas" does seem to have been more common for at least a couple decades. (And you have access to better resources on African languages than I do and I trust your judgment.) I recall that we renamed the related Mwaghavul (from "Sura") only a couple years ago. Renamed.- -sche(discuss)06:15, 29 December 2017 (UTC)Reply
Alazapa/Alasapa/Pinto (nai-ala), a fragmentarily attested language of northern Mexico, scarcely described in English and not that much better described in Spanish. Considered to be related to Quinigua and Cotoname by Norman A. McQuown's 1968 Handbook of Middle American Indians (volume 5: Linguistics, page 100), it is sometimes identified with or considered a dialect of Coahuilteco, apparently as part of the former belief that the "the Coahuiltecans belonged to a single language family and that the Coahuiltecan languages were related to the Hokan languages of California, Arizona, and Baja California. Most modern linguists, however, discount this theory for lack of evidence and believe that the Coahuiltecans were diverse in both culture and language. At least seven different languages are known to have been spoken[.]" Some of the scholars who responded to and followed up on del Hoyo's vocabulary of Quinigua provide a few Alazapa words, like axi "tobacco" (compared to Karuk úuh "tobacco", Esselen ka'a "tobacco"). - -sche(discuss)03:48, 6 June 2017 (UTC)Reply
I'm away from my books at the moment, but I seem to remember Yuman languages having something along the lines of /up/ for tobacco. Chuck Entz (talk) 04:01, 6 June 2017 (UTC)Reply
Yes, Cocopa has ˀu·p "tobacco", and ˀu·p xyay "smoke tobacco". Quechan/Yuma itself has axta/ak’sa’ for a tobacco pipe; in a short search, I didn't find the [Quechan] word for tobacco itself, but the list I found the Alazapa word in was comparing it to mostly words for "tobacco" but some words for "pipe". (I updated the spelling of the Karuk and Esselen words to the spellings given in dedicated works on those languages.) - -sche(discuss)04:30, 6 June 2017 (UTC)Reply
The Digor and Iron dialects of Ossetian seem quite different, and already many (most?) of our entries distinguish which is meant. It seems to me that there is a fair chance that the two are separate enough to deserve being called different languages here. —Μετάknowledgediscuss/deeds04:46, 19 December 2012 (UTC)Reply
I've seen them referred to as separate languages before, but there's still some debate over that. Doesn't matter to me. But would there still be plain Ossetian language entries or would all be sorted into the new languages? There are some that aren't labelled as either Iron or Digor.Word dewd544 (talk) 17:58, 20 December 2012 (UTC)Reply
Iron is by far the more common dialect, and the literary Ossetian language is based on Iron. However, Digor is different enough that it could be considered a separate language. The main things against it here are the relatively small number of speakers and that it does not yet have a written standard, as far as I know. But there is now a Digor dictionary out there, and it’s probably just a matter of time before Digor develops a literary standard of its own. I think it’s unlikely that we will get enough Digor contributions to make a difference, but it is always possible that someone will start entering words from a Digor dictionary. The Digor language code is oss-dig. We could use os for Ossetian proper (and Iron), and oss-dig for Digor. —Stephen(Talk)02:20, 21 December 2012 (UTC)Reply
A standard language based on one widely-spoken dialect, and another lect sometimes considered a dialect and sometimes considered its own language? This reminds me of Tosk vs Gheg Albanian: some references say they're mutually unintelligible separate languages, speakers say their differences present no impediment to communication. Unfortunately, we lack speakers of the Ossestian lects, and the dictionary of Digor is said to waffle, the author calling it a language and the editor calling it a dialect. Stephen is probably right that it's just a matter of time before Digor develops its own standard (and merits separation as much as Luxembourgish and Limburgish do from each other and from German); OTOH, Wiktionary, like Wikipedia, is not a crystal ball. My preference would be to wait and not split them for now. If we do split them, I agree with using {{os}} for Iron (compare {{lt}} and {{sgs}}), and we should devise an exceptional code for Digor that fits our usual naming scheme (ira-odg or ira-dig), rather than using Linguist List's ersatz "oss-dig". - -sche(discuss)03:42, 21 December 2012 (UTC)Reply
Not merged at this time; cf the WT:ES discussion; but especially if the developments with regard to resources in the lects over the past few years suggest they should, in fact, have separate codes, please feel free to reopen (or start a new, non-stale) discussion. - -sche(discuss)14:47, 12 January 2018 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming lua
This language is currently called "Tshiluba", which is a really awful choice. First of all, tshi- is that good ol' language prefix that we often try not to have in language names (which I think is ci- in modern orthography), and there are in fact two Luba languages (the other is lu "Luba-Katanga"). To avoid confusion, we rightfully give neither the name Luba, but this is not much better, and we should rename it to "Luba-Kasai", as Wikipedia does. —Μετάknowledgediscuss/deeds19:28, 27 November 2015 (UTC)Reply
On the one hand, we do try to avoid prefixes. On the other hand, "Luba-Kasai" seems to more often be a placename and an ethnonym than a language name, and "Tshiluba" seems to be about twice as common. - -sche(discuss)21:15, 8 January 2016 (UTC)Reply
The issue is that, AFAICT, "Tshiluba" is more commonly used because it refers to both Luba languages! This is not so much about prefixes so much as the issue of the name being exceedingly ambiguous in its referent. —Μετάknowledgediscuss/deeds01:28, 9 January 2016 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
@-sche I think Taishanese dialect of Cantonese and Teochew dialect of Min Nan both need language codes. They are covered by Chinese pronunciation modules but theire transliterations is very different from Cantonese and Min Nan accordingly. E.g. in 鉛筆/铅笔 (qiānbǐ) "yon3 bit2" is not standard Jyutping and "ing5 big4" is Teochew Peng'im, not POJ. @Kc_kennylau, Wyang, Suzukaze-c, Justinrleung. Perhaps these subdialects needs nesting in translations and numbered tone marks (also Gan, Jin, Xiang) also need superscripts, just like Cantonese Jyutping. --Anatoli T.(обсудить/вклад)11:32, 12 November 2016 (UTC)Reply
@Justinrleung Thanks but if I try add a translation using 'yue-tai' I get the error:
Lua error in Module:languages/templates at line 28: The language code 'yue-tai' is not valid.: Lua error in Module:parameters at line 110: The parameter "<strong class" is not used by this template. --Anatoli T.(обсудить/вклад)06:15, 14 November 2016 (UTC)Reply
If there's a need to enter them into translations tables (on account of their different transliteration and, according to Wikipedia, sometimes different vocabulary), they could be given full codes, which as Meta notes would be named a little differently (using the system described in WT:LANG): zhx-tai and zhx-teo. I await Wyang's input. As an aside, we should consider taking a Chinese approach to Arabic, i.e. not have separate headers for each dialect, but retain the option of listing each dialect's pronunciation and maybe having each dialect in translations tables. - -sche(discuss)16:36, 14 November 2016 (UTC)Reply
where {{zh-l}} extracts and displays the simplified form from the entry, as well as extracts all the readings of the word from the entry, and (if on a computer) displays the readings on hover-over or (if on a mobile device) something. Additional topolect-specific translations can be added as:
We'd need a full-blown {{zh-t}} that can serve the function of linking to zh.wikt if an entry exists, and you'd also need to run a bot to update those every now and then (or convince Ruakh to do it). This would imply some changes to the translation adder as well, and possibly other gadgets and bits of code lying around. In short, that's a big jump from what Anatoli proposed. It does sound like a good idea, though, if you want to put the requisite work into it. —Μετάknowledgediscuss/deeds00:10, 16 November 2016 (UTC)Reply
@Wyang: Thank you for pinging me but I'm not familiar with Cantonese and Banlamgu and their regional speeches and I currently have no idea about this... Dokurrat (talk) 13:35, 12 January 2018 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Diegueño
I thought I might dig out a few references I have and add a few entries in one of the lects that go by this name, but I'm a bit confused by the way we have the language codes set up.
Diegueño is the name that anthropologists have traditionally used for the language originally spoken around Mission San Diego in the southwestern corner of California. Older references referred to it as single language covering most of San Diego County, California and northern Baja California, Mexico, but I always understood it to be at least three languages:
Northern Diegueño, known to its speakers as 'Iipay 'aa, and generally called Ipai in the literature
Central Diegueño, now called Kumeyaay or Kumiai
Southern Digueño, now calley Tipai or Tipay
Just to confuse things, Kumeyaay is sometimes used to refer to all three, and there are some scholars who have merged Tipay and Kumeyaay into a single language that they call Tipai. Then there's Kamia, which has been used in older literature for a variety of groups who all seem to have been Diegueño of one sort or another. The ISO has a single code, dih (which we call Kumiai) and our lone entry using that code is the 'Iipay 'aa word for water. That would make sense if we treated all of Diegueño as one language, but we have an exception code for Tipai: nai-tip, and a single entry. As far as I know, no one currently considers Ipai to be part of Kumiai unless they consider Tipai to be part of it, too- hence my confusion.
I've only studied Ipai (perhaps I should say "dabbled in"), so I can't judge for myself how different the lects are from each other. Based on what little I have read, it would seem to me that we have just a few credible options:
Treat Diegueño as a single language, keeping dih and retiring nai-tip
Treat Ipai as separate, but merge the rest of Diegueño into Tipai
Treat each of the three lects as independent, preferably all with exception codes (nai-ipa, nai-kum and nai-tip, perhaps?).
I would recommend against using dih for anything but the single-language option- this isn't a macro-language with a standard or prestige lect, and it would probably just confuse things. Chuck Entz (talk) 06:12, 19 June 2017 (UTC)Reply
(After digging into the history of the codes, I've refreshed myself that) I added the code for Tipai in diff, so I could add diff, after seeing that we called dih "Kumiai" and taking that to mean that it referred to the Central dialect and the others needed codes. (Apparently, the ISO/Ethnologue's actual reason for calling it "Kumiai" is that some people use "Kumiai" instead of "Diegueño" as the cover term, as you note.) I probably didn't add a code for the Northern variety at that time because I didn't want to bother figuring out what to call it ("'Iipay 'aa" struck me as a suboptimal name; do we have other names with spaces in them where the part after the space isn't capitalized?) at a time when no-one was coming forward with content that needed to be added in it. :p
Victor Golla, California Indian Languages (2011, →ISBN, page 120, says: "While Kroeber (1925) and others treated the Kamia as a Diegueño subgroup, there is no firm evidence in support of this approach, although the name they are known by [Kamia] appears to be a variant of Kumeyaay (Langdon 1975a). With this possible exception, all of the groups definitely known to have spoken varieties of Diegueño were located west of the present San Diego-Imperial County line or in Baja California west of the Sierra de Juarez. [...] Although Ipai and Tipai are to some extent mutually intelligible, they show numerous differences in vocabulary and structure (for a comparison of Mesa Grande Ipai and Jamul Tipai see A. Miller 2001:359-363) and have sometimes been treated as separate languages. Winter [...] judged [Tipai] to be no closer to (Northern) Diegueño than to Cocopa. The most recent classification (Langdon 1991; Miller 2001:1-4) distinguishes [Ipai, Tipai and Kumeyaay]."
Amy Miller's referred-to work, A Grammar of Jamul Tipay, says "A comparison of descriptions of Mesa Grande [...] with the results of my own research reveals that differences between Mesa Grande and Jamul pervade the phonology, lexicon, derivational morphology, inflectional morphology, syntactic morphology, syntax, and discourse."
OTOH, Tipai Ethnographic Notes: A Baja California Indian Community (2001, →ISBN, edited by Langdon et al, says: "These Mexican Diegueno, who call themselves Ipai or Tipai 'people,' cannot be described as now forming a tribe; they are a group of Indian families speaking mutually intelligible dialects (Northern and Southern) of a language[.]"
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming az
I propose to rename the language name Azeri to Azerbaijani. The reasons are (1) to distinguish it from the Iranian language Azeri, (2) Ethnologue, Glottolog and Wikipedia call the language Azerbaijani, (3) Azerbaijani is closer the native form Azərbaycan dili, (4) Azeri has pejorative connotations in Armenian. @Allahverdi Verdizade, currently the most active Azerbaijani contributor, agrees. --Vahag (talk) 14:01, 24 January 2018 (UTC)Reply
Support. And to summon evidence more appropriate to our dictionarying efforts, Google Books searches for "speaking X" or "the X language" have invariably returned more hits with "Azerbaijani" than with "Azeri" when I have tried them. —Μετάknowledgediscuss/deeds17:22, 24 January 2018 (UTC)Reply
Support. Azeri is a little "vague" while Azerbaijani is kind of long and sounds only limited to Azerbaijan, but I support the change, "Azeri" is used for people rather than the language itself in Turkish as well. --Anylai (talk) 17:33, 25 January 2018 (UTC)Reply
Support for the reasons Vahag and Meta outlined. Someone with a bot will need to implement the rename, because at least three thousand entries will need to be updated, and that's just the ones where the L2 header needs to be changed; in other entries, translations tables and descendants lists and etymologies where the name is spelled out will need updating. - -sche(discuss)23:51, 25 January 2018 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Some spurious languages to merge or remove, 1
merge Dororo [drr] and Guliguli [gli] into [kzk]
In 1953, Lanyon-Orgill provided short wordlists of two languages he called Dororo and Guliguli. His lists are so similar to Kazukuru that subsequent scholars have suspected they are dialects, if not alternative transcriptions, if not jokes. Karen Davis, in A grammar of the Hoava language, Western Solomons, notes "there was no one in present day Kusaghe who had heard of [Dororo or Guliguli], and Lanyon-Orgill does not identify his informant. As one of the names of the dialects, Guliguli, can mean 'masturbate' in Hoava-Kusaghe, I have my doubts about the existence of this language." Michael Dunn and Malcolm Ross expand on Davis's point in their 2007 article Is Kazukuru really non-Austronesian, with the view (accepted also by e.g. Harald Hammarström and Sebastian Nordhoff, in Melanesian Languages on the Edge of Asia: Challenges for the 21st Century) that if the lects were real, they were the same language as Kazukuru, which I propose to merge them into. (Sample words in Kazukuru, Guliguli, and Dororo: pito, bito, bito "arrow"; vinovo, vino, bino "banana"; viniti, vini, vinitini "body"; minata, minate, minate "die"; meta, mata, mata / meta "eye"; rano, rano, rano "head"; muni, moni, muni / moni "night".) - -sche(discuss)06:44, 12 May 2017 (UTC)Reply
As noted by Hammarström, Melanesian Languages on the Edge of Asia: Challenges for the 21st Century, the existence of a Yarsun language seems have been based on the confusion of language names with place names which is not uncommon in Papua (Yarsun is near Anus Island, and that's not a joke); "no such language is attested". Likewise with Wares. - -sche(discuss)06:44, 12 May 2017 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Bunurong
This Pama-Nyungan lect seems to have been given an exceptional code with no discussion, for its use at one transwikied entry, ngargee. It's by no means clear that we should be giving it a separate code rather than treating it as a dialect of Woiwurrung wyi; Wikipedia claims they are 90% mutually intelligible, but doesn't cite that claim directly. —Μετάknowledgediscuss/deeds04:54, 6 July 2016 (UTC)Reply
I noticed this a while ago, but left it alone because I could not at the time find a reference (i.e. outside WP) that confirmed that they were mutually intelligible, probably because the huge variety of spellings made searching for information difficult. However, I can now find references that suggest we should be merging more than just these two. Leigh Boucher and Lynette Russell's Settler Colonial Governance in Nineteenth-Century Victoria (2015, →ISBN, page 8, speaks of "the Woiwurrung (Wurundjeri), Boonwurrung, Wathaurung, Taungurong and Dja Dja Wurrung [being] mutually intelligible languages that share up to 80 per cent of their terminology." A paper by Barry Blake and Julie Reid on Sound Change in Kulin, in the La Trobe Working Papers in Linguistics, v 6-8 (1993), speaks of a single Kulin language, with "material available on three dialects: Boonwurrung (B), Woiwurrung (W) and Thagungwurrung (T)" (emphasis mine). Dja Dja Wurrung = dja, and Taungurong / Thagungwurrung = dgw, and Wathaurung = wth, all of which we currently treat as separate. Kulin suggests that at least the four eastern ones, if not also Wathaurung, could be merged. - -sche(discuss)06:48, 6 July 2016 (UTC)Reply
Bunurong was merged at the time this discussion was current. The discussion has since gone stale, and I am going to leave the others alone (unmerged). - -sche(discuss)06:26, 29 January 2018 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
ISO code changes for 2017
Here is a document summarising the changes to ISO codes that have been made in 2017. I have gone through all the approved request documents and briefly summarised their findings and my conclusions about what I propose we should do. —Μετάknowledgediscuss/deeds10:36, 31 January 2018 (UTC)Reply
Only a clan name, confirmed with fieldwork. I concur with removal.
Ndaktup [ncp]
Merge into Kwaja [kdz], confirmed with fieldwork. I concur with removal.
Lyons Sign Language [lsg]
Apparently spurious. I concur with removal.
Mediak [mwx]
Only a clan name, confirmed with fieldwork. I concur with removal.
I also agree, and have removed them all, adding Mosiro and Mediak and several other things as alt names of oki (our spelling of the canonical name, Okiek, seems more common than Wikipedia's Ogiek), and adding Ndaktup and several other forms as alt names of kdz. - -sche(discuss)16:56, 31 January 2018 (UTC)Reply
Confirmed with fieldwork and dictionary creation is underway. I concur with creation.
Done, tentatively with Latn, Deva and Tibt scripts listed, based on the two scripts Ethnologue lists for broader Manang [which WP thinks it is a dialect of] and how I see linguists documenting it. - -sche(discuss)23:35, 7 February 2018 (UTC)Reply
But is this consistent with how we treat other Tibetan/Tibetic varieties? E.g. kte is subsumed into bo, somewhat like with Chinese varieties. This may need further discussion, maybe separate from this big list of codes, with Wyang. - -sche(discuss)23:58, 7 February 2018 (UTC)Reply
I realised this when reading your comment at Talk:གསར་པ. We decided on a Chinese-style merger, but this compounds the lack of interest and expertise that has plagued documenting anything other than Lhasa Tibetan (which is admittedly the only one I have formally studied either). —Μετάknowledgediscuss/deeds00:35, 8 February 2018 (UTC)Reply
Magɨ (Madang Province) [gkd]
Request document has no content; the case seems to be based on Daniels (2016), which is the only documentation. The name should presumably be simply "Magɨ". I concur with creation.
Or even "Magi" (although this is also an alt name of Mailu). And, aha, so this is a separate language from gmg Magiyi/Magɨyi; I wondered about that back when I was looking at gmg, since they're both sparsely-documented Sogeram languages of Madang. - -sche(discuss)17:43, 31 January 2018 (UTC)Reply
Confirmed with a Swadesh list as worthy to be split from Beng [nhb]. I cannot find any documentation after a quick search, nor do I find confirming fieldwork. I abstain on creation.
Glottlog, citing Anna Maloletnyaya's 2014 Brief presentation of Ngen language, agrees that "Ngen, a Southeastern Mande language of the Ben-Gban group not intelligible to Beng". Therefore, added. - -sche(discuss)03:55, 7 March 2018 (UTC)Reply
Western Armenian [hyw]
Request document is chiefly concerned with social needs (e.g. a new Wikipedia). There are significant dialectal differences, but Western and Eastern Armenian are known to be mutually intelligible and have been successfully treated as [hy] at Wiktionary (though there has been little to no prior discussion). There is currently a parallel discussion of this issue in the Beer parlour. I disagree with creation.
On the subject of languages we could include: Jedek, just documented (too recently to have been included in this page of ISO codes). (Our own incomparable Stephen G Brown has already created a Wikipedia article for it, Jedek language.) - -sche(discuss)22:21, 8 February 2018 (UTC)Reply
It is one of many very closely related lects (although geographically displaced from the closest ones to itself). I skimmed the paper, but it really wasn't easy to make a quick determination about whether we should include it. —Μετάknowledgediscuss/deeds06:32, 10 February 2018 (UTC)Reply
Changed to Hyolmo, based on the current name being a misnomer (the speakers are not Sherpas and the language is not closely related to Sherpa) and preference of native speakers. I find that Yolmo is in fact the most-used spelling on Google Books and Google Scholar, so I suggest we use that name instead.
Changed to Lidzonka, based on preference of native speakers. This name has almost never been used in the linguistic literature, whereas Dzodinka has and the request document itself admits that "both are correct". I disagree with the name change.
Changed to Shuhi, based on local usage; the requester actually uses the name Xumi when documenting the language. All other use in the linguistic literature appears to follow the original name, Shixing. I disagree with the name change.
Changed to Rigwe, based on the current name being erroneous. What little linguistic literature there is seems to have fully shifted over. I support the name change.
Changed to Paloor, based on the language's new orthographic norms. There is very little linguistic literature on it, most of it in French, and it is difficult to tell what name is or will be more common. I abstain on the name change.
The new spelling may well take hold, but it hasn't yet (and we have no contributors in the language who might clamor to use the new spelling when adding entries), so, not renamed at this time. We should reexamine this in a few years. - -sche(discuss)16:50, 22 February 2018 (UTC)Reply
Australian Sign Language [asf]
Changed to Auslan, based on popular and linguistic usage. This does seem to be more common as the standard name, and is preferred by native speakers. I support the name change.
Has been done, apparently.
The name of [khm] was also changed from "Khmer, Central" to "Khmer", but we have chosen to exclude this code, so it does not affect us.
Oh, no, the more people who know how to do this kind of thing, the better! :) You just have to be sure all the requests have been handled (done/accepted, or rejected). But yes, WT:LTD is the catch-all for anything that doesn't have a more logical place to be archived; and then, if there's a change in how we treat a language (e.g. a language has been split between two codes), update WT:LT and link to the discussion. If a language code has been removed (e.g. merged into something else), I find that it's helpful (and prevents accidental re-creation) to leave a --placeholder comment in the module (as I did for e.g. "frc" and am slowly doing for other codes). - -sche(discuss)16:42, 22 February 2018 (UTC)Reply
OK, let me try this again. I'd like to have Alanicxln renamed to Sarmato-Alanic and have entries use xln-pro. Alanic and Sarmatian (which has no language code otherwise) occupy a dialect continuum, and neither might be the direct ancestor of Ossetian or Jassic. Alanic would then be made into an etymology-only code, xln-ala. Alternatively, a new language code could be created, ira-sma-pro, and xln removed all together, but I think the former the better option. @-sche, Tropylium --Victar (talk) 20:30, 3 April 2018 (UTC)Reply
There are a couple problems here that I see just from a brief reading of the discussion. You want us to change to a name that is very rarely used instead of a more common one, and also switch to considering it a protolanguage rather than an attested one, despite the fact that it is actually attested (yes, we do that for Proto-Norse, but it is not ideal and is in large part because, as is relevant here, we try to use the most commonly used names where possible). —Μετάknowledgediscuss/deeds01:19, 5 April 2018 (UTC)Reply
Thanks for the reply, @Metaknowledge. 99% of the entries I'll be entering will be reconstructions, and most will be derived from Ossetian, not Alanic or Sarmatian borrowing. I think there is a ton of precedence for using alternative names for codes on wikt, but I'll concede that I can't think of any example of using the language code of a dialect to refer to a whole dialect family. I'm not opposed to using ira-sma-pro instead, but I do still think then the xln code should be discontinued, because Alanic, Sarmatian and Proto-Ossetic should all be under the same code. --Victar (talk) 04:13, 5 April 2018 (UTC)Reply
To give an example of what I had in mind for formatting descendant trees:
The proportion that are reconstructions is irrelevant unless it's 100%. When we have mainspace entries, we should avoid assigning them to protolanguages (which are technically hypotheses) wherever possible. We always try to use to the most common, unambiguous name possible, and if you know of any exceptions, we should see if they ought to be fixed. Basically, I think you're conflating the needs of descendant trees and the criteria for determining what ought to be a separate language. Bear in mind that regardless of what codes and names are, you can always structure descendant trees to show distinct dialects or sublects (Crom daba has done this quite fruitfully). —Μετάknowledgediscuss/deeds04:48, 5 April 2018 (UTC)Reply
@Metaknowledge, I think you're missing the point of my need. I want to create reconstructions of Proto-Ossetic and Sarmatian. Sarmatian and Alanic are well established as two separate dialects. --Victar (talk) 04:59, 5 April 2018 (UTC)Reply
And if they're dialects, they shouldn't have separate codes. Remember, you can still give them separate lines and reconstructions, when and where those are supported by scholarly sources, regardless of the situation with codes. —Μετάknowledgediscuss/deeds06:18, 5 April 2018 (UTC)Reply
Sorry, I tried to indicate xln-ala and xln-sar where etymology-only codes by the dash in them, but I guess that wasn't clear (though I did make that point in my opening statement). We also have oos for Old/Proto-Ossentic, which we could used instead for parent of Alanic/Sarmatian/Ossentic. We currently list it below xln, and I've always considered it a stage between, yet MultiTree seems to use it as their catch-all. Again though, I'm not opposed to using a new ira-sma-pro code. --Victar (talk) 17:34, 5 April 2018 (UTC)Reply
We create etymology-only codes for etymology sections. If there's a language that derives terms from both Alanic and Sarmatian with a meaningful difference between the two, then those codes should exist, but we shouldn't create them just for descendant lists, which can be freely formatted. —Μετάknowledgediscuss/deeds18:39, 5 April 2018 (UTC)Reply
Here's a different idea. Alanic and Sarmatian are both barely attested; from my brief reading of the literature, it seems to be unclear whether or not they represent dialects or fully separate speech communities, and whether they represent the ancestor of Ossetian or a close relative (and given the timescales over which they are attested, they cannot be the same thing as a protolanguage, which is a hypothesis of the most recent common ancestor). The resultant action would be to have separate codes for Alanic, Sarmatian, and Proto-Ossetic, with the former two only in mainspace (in original script, e.g. Greek) and the latter only in Reconstruction space (in normalised form). As always, you can format descendant lists however you like. Does that seem like it would make sense? —Μετάknowledgediscuss/deeds18:39, 5 April 2018 (UTC)Reply
@Metaknowledge: No, I don't agree with that solution. Whether they exhibit the same exact timeframe is irrelevant and labeling Sarmatian and Ossetic as Alanic is inaccurate. What my sources are reconstructing is a common ancestor of all three of these dialectal branches. See https://ibb.co/jFEnSH. --Victar (talk) 19:01, 5 April 2018 (UTC)Reply
Your response is confusing; I did not suggest labelling Sarmatian and Ossetic as Alanic (in fact, I suggested separating all three), and I was under the impression that Proto-Ossetic is the unattested ancestor of all these languages. —Μετάknowledgediscuss/deeds19:30, 5 April 2018 (UTC)Reply
Am I understanding this correctly that it is similar to the problem that he have had/are having with Sanskrit and the Prakrits, namely that there is a dialectal continuum between Sarmatian, Alanic, and the unattested ancestor of Ossetian? —*i̯óh₁n̥C[5]19:24, 5 April 2018 (UTC)Reply
@JohnC5: Not exactly. Sarmatian, Alanic and Ossentic are all largely unattested dialects form a single language of the Middle Iranian period. So what I'm suggesting is that we unify them under a single code and name, and have the dialects differentiated only by etymology-only codes. I'm recommending the name Sarmato-Alanic, which is what I mostly see in literature when referring to them as a whole, but if I had to choose to unify them under one name out of the three, it would be Ossetic, being the only one with modern descendents. What code we use, be it a repurposed one, or a new one, doesn't matter to me. --Victar (talk) 20:26, 5 April 2018 (UTC)Reply
@Metaknowledge: Would you support the idea of unifying Sarmatian and Alanic under Old Ossetic oos (as per MultiTree) with both reconstructed and mainspain entries, and making xln an etymology only code? That should resolve your Proto-Norse argumentment. --Victar (talk) 22:37, 5 April 2018 (UTC)Reply
I see that the name "Old Ossetic" is broadly attested (when spelt correctly), and many sources seem to equate it with Alanic, but that doesn't mean Sarmatian should necessarily be merged as well. If they are indeed dialects as you claimed, then that would be perfectly fine. WP cites EB for the following: "The languages of the Scytho-Sarmatian inscription may represent dialects of a language family of which Modern Ossetian is a continuation, but does not simply represent the same language at an earlier time." If that is true, then Sarmatian should be kept separate. —Μετάknowledgediscuss/deeds23:29, 5 April 2018 (UTC)Reply
@Metaknowledge, there is no code for Sarmatian. If we put Alanic under Old Ossetic as part of a dialectal continuum, Sarmatian, as a dialect thought to be very similar to Alanic, should unequivocally be included. Otherwise it defeats the point. --Victar (talk) 23:49, 5 April 2018 (UTC)Reply
I would be in favor of Old Ossetic for the continuum of Alanic and Ossetic and, if it can be demonstrated as true, Sarmatian. In what way can we adjudicate this Sarmatian situation? As mentioned before, I'm getting a little bit frustrated at continuously running across this "continuum problem" (attested languages descending from unattested near neighbors). There's been a fair amount of research saying that the phylogeny of language change tends to be binary in nature, but that depends on how you look at language continua versus dialectally diverse super-languages. I'd be interested to think about the principled use of language continua in our language data (like "oss-cnt" or the like), not just the "substrata" we use in the etymology-only language data. The question is in the utility of such a demarcation, but the inherent assumption of our current n-ary (or perhaps my theoretically binary) branching system tends to omit this subtlety of language change because frequently these continua are not protolanguages and exist clearly in the data... I dunno. —*i̯óh₁n̥C[5]00:07, 6 April 2018 (UTC)Reply
Ignoring John's tangent... I know there is no code for Sarmatian. That's immaterial; we can make one if we deem it necessary. You claim that Sarmatian is very similar to Alanic, to the point of being a continuum; I know little about this, but found a scholarly source that claims otherwise. Can you respond to that with actual evidence? —Μετάknowledgediscuss/deeds00:16, 6 April 2018 (UTC)Reply
"Sarmatian and Alanic represent a dialect continuum" and "it is difficult to draw the line between Sarmatian and Alanic".[S 1]
"Ossetic [...] is the last remnant of the essentially unknown Middle Iranian dialect area that included Sarmatian, and is said to descend from Alanic."[S 2]
"[Ossetic] is the sole surviving descendant of the Northeast Iranian dialects of the ancient Scythians and Sarmatians and medieval Alans".[S 3]
"Deine klare linguistische Scheidung zwischen Sarmatisch und Alanisch aufgrund der Materiallage nicht möglich ist"[S 4]
Even if Alanic and Sarmatian were divergent enough to call separate languages, that distinction isn't apparent in the little material we have, so to reconstruct them separately at this time would be folly. --Victar (talk) 01:20, 6 April 2018 (UTC)Reply
Thanks, that seems like good evidence for merger. I am now satisfied with having "Old Ossetic" as an L2 header with categorising context labels for the dialects. I would like to wait a couple of days just in case anyone raises an objection, so please ping me with a reminder. Also, please clarify if there are any etymology sections that need to distinguish between the dialects; if not, we can dispense with etymology-only codes and simply remove xln. —Μετάknowledgediscuss/deeds04:44, 6 April 2018 (UTC)Reply
┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ @Metaknowledge, if you have a moment, I would appreciate you making these changes. Thanks. --Victar (talk) 03:10, 16 April 2018 (UTC)Reply
^ Novák, Ľubomír (2013) Problem of Archaism and Innovation in the Eastern Iranian Languages (PhD dissertation)[1], Prague: Univerzita Karlova v Praze, filozofická fakulta
^ Fortson, Benjamin W. (2004) Indo-European Language and Culture: An Introduction, first edition, Oxford: Blackwell
^ Lua error in Module:quote at line 2600: |title= is an alias of |chapter=; cannot specify a value for both
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Some more missing South American languages, 2
Here are a few more South American languages for which we could add codes:
Coeruna / Koeruna (sai-coe), said to be attested.
Conambo language (sai-cnb) is sometimes considered a dialect of the Záparo language, but Loukotka has samples, which show differences: "head" is ku-anak in Zaparo, ku-anaka in Conambo, "eye" is nu-namits (Z) / ku-iyamixa (C), "fire" unamisok (Z) / umani (C), "woman" itumu (Z) / maxi (C). OTOH, telling it apart from what a number of references refer to as the Zaparo spoken in the Conambo river could be non-trivial, so perhaps treating it as a dialect would be easier...
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming siz
We currently call this "Siwa". That's the name of the oasis, but the proper name for the language, which sees far more use in the literature and in general, is "Siwi". —Μετάknowledgediscuss/deeds01:26, 14 March 2018 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming cak
We currently call this "Cakchiquel". Language-learning materials and linguistic literature generally prefer the de-hispanicised spelling "Kaqchikel", as can be seen at Google Books. —Μετάknowledgediscuss/deeds04:41, 17 March 2018 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Converting Proto-Tibeto-Burman to an etymology-only language
I propose that we treat Proto-Tibeto-Burman as an etymology-only variant of Proto-Sino-Tibetan as it appears that the consensus among linguists in the field is that the set of non-Sinitic Sino-Tibetan languages is not monophyletic. (For a parallel, we treat Proto-Baltic as an etymology-only variant of Proto-Balto-Slavic for the same reason.) By the same token, I propose that we change all languages and families currently listed as family = "tbq" to family = "sit" instead. Wyang has said on his talk page that he isn't opposed to the idea, and I don't know who else here is working on Sino-Tibetan issues. What do other people think? —Mahāgaja · talk14:32, 19 July 2019 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Wanji languages
Currently the canonical name of ira-wnj is Wanji which is also used by wbi. One of them should be changed. We also have wny which is Wanyi. DTLHS (talk) 18:27, 22 July 2018 (UTC)Reply
(Also similar: wdd Wanzi / Wandji.) I've renamed ira-wnj to "Vanji", which is the spelling Wikipedia uses anyway. We don't have much content in either language, just a few entries in descendant trees for the Iranian one and a few translations for the Bantu one, but yes, it causes problems (starting with conflated categories) if they have the same name. - -sche(discuss)15:08, 24 July 2018 (UTC)Reply
I've moved all the categories to the new spelling, but it would be great if someone with a bot could change all L2 headers from ==Inupiak== to ==Inupiaq==. —Mahāgaja · talk10:12, 24 October 2019 (UTC)Reply
@Metaknowledge, Erutuon: I don't know if there are any L2 heading left using the old spelling, but there are definitely still translation sections calling it "Inupiak". I don't have a bot (or the remotest idea how to write one); @DTLHS, Benwing2, Ruakh, is this something one of y'all's bots would be interested in doing? —Mahāgaja · talk08:02, 24 December 2019 (UTC)Reply
Support. The phrasing in Liljegren (2016) that you linked to leads me to think that perhaps bsh is the code we should keep, and xvi the one we should retire, although it's wholly arbitrary. Thanks for paying attention to Nuristani, in any case. —Μετάknowledgediscuss/deeds23:07, 3 August 2018 (UTC)Reply
Kata-vari is the larger of the two by at least five-fold, which on one hand would make it the logical encompassing one, but bsh is actually named for its eastern subdialect Bashgali (taken from Dardic), making it a somewhat inaccurate code to begin with, but I don't really care either way. Here is Strand's tree. --Victar (talk) 23:39, 3 August 2018 (UTC)Reply
And here is from Strand's 1974 paper, back when he used to refer to Kati as being the parent language of all three dialects: "Kati (Bašgalī) has three major dialects: Katə́viri, Kamvíri, and Mumvíri. Katə́viri is spoken by members of the Katə́ tribe. It is divided into two major subdialects: Western Katə́viri and Eastern Katə́viri. Western Katə́viri is further subdivided into the dialects of Ramgə́l, Kulám, Ktívi (Kantivo), and Pə́řuk (Papruk)". I'm also fine keeping it named Kati, despite it perhaps being outdated, if that's preferable to others. --Victar (talk) 00:05, 4 August 2018 (UTC)Reply
Support the merger. As for the name: on one hand, "Kati" seems to be about twice as common even when I search for the names together with "Nuristani" to filter out the New Guinea-area language, but on the other hand the absolute number of words that mention either name is small, so if "Kati" is dated and "Kamkata-viri" is preferred these days, and "Kamkata-viri" would also make clearer to readers, etc that we're considering all the dialects under that one header (and that we're not talking about kti, Muyu / Kati), then go ahead and use "Kamkata-viri", (with the other names as alt names). - -sche(discuss)00:58, 7 August 2018 (UTC)Reply
Renamed. Searching the site, the only pages I see using "Ruund" are categories (in family trees) and Reconstruction pages (using {{desc}}), which should update automatically; I think there is, therefore, nothing that needs changing(?) besides the three modules I just changed. - -sche(discuss)08:35, 11 January 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Dungan is technically Mandarin, or a dialect of Mandarin
Hi, I'm not sure whether this is the right venue for this discussion, but I would like to bring up Dungan, which is spoken in Central Asia. According to Wikipedia, Ethnologue, and Glottolog, this is a Chinese language, specifically Mandarin. There are only 20 entries in Wiktionary that are for Dungan, and all of them are Mandarin words, with some from the Gansu and Shaanxi dialects. The difference, however, is that, Dungan is written in Cyrillic. In Wiktionary, all Chinese dialects are merged into one single Chinese entry, and pronunciations are listed. Shouldn't we do that, or at least partially, for Dungan? Please feel free to comment. Thanks. --Mar vin kaiser (talk) 14:54, 16 July 2016 (UTC)Reply
Re venue: this is indeed a venue where discussions of merging/splitting language codes and categories and entries take place. I tend to put my "biggest" proposals (ones that needed votes in the past, or that concern major or controversial languages that I suspect will need votes) in the WT:BP, but here is OK.
Wiktionary:About Chinese#The_Chinese_lects and Wiktionary:Votes/pl-2014-04/Unified_Chinese intentionally left lects that don't use Hanzi separate from Unified Chinese, but I don't know if that was because Chinese editors felt they should never be merged, or just felt that merging them would be difficult and best attempted after everything else had been merged. It would obviously be possible to merge Dungan and other such lects if Chinese editors wanted to; we have plenty of other languages which use multiple scripts (e.g. Afrikaans). However, the various Chinese lects which are distinctive to the point of potentially being not-mutually-intelligible when spoken were able to be unified here because they share a written form in which they are theoretically mutually intelligible. If Dungan is potentially not intelligible with lects from other areas (lects that differ from Mandarin enough that speakers don't understand it without study) in either speech or writing, then what would be the basis for unifying them? - -sche(discuss)15:42, 16 July 2016 (UTC)Reply
Actually, Shaanxi and Gansu Mandarin is mutually intelligible with Dungan. Furthermore, a large majority of Dungan vocabulary is from Chinese, which therefore, has Chinese equivalent entries written in Chinese characters. There are Russian and Turkic vocabulary. My suggestion is to leave Dungan loanwords from Russian and Turkic as written in Cyrillic, and merge the Dungan Chinese words with Chinese entries, and perhaps leaving the Cyrillic entry of those words like how Chinese pinyin and Japanese romaji are left. --Mar vin kaiser (talk) 15:53, 16 July 2016 (UTC)Reply
Also, I speak Mandarin, and I tried listening to Dungan videos in Youtube. They're actually understandable for the most part. As in I can write down what they're saying in Chinese, except for some words though (presumably Russian and Turkic loanwords). --Mar vin kaiser (talk) 16:04, 16 July 2016 (UTC)Reply
The main caveats are that Cyrillic is apparently the standard script for Dungan, unlike with Pinyin or Romaji, and there may be some vocabulary that only exists in Cyrillic. I suppose the writing systems for Hokkien might be analogous, though. Chuck Entz (talk) 01:07, 17 July 2016 (UTC)Reply
This doesn't feel right to me. I think it would be like folding Maltese into Arabic, or merging Hindi and Urdu. I foresee a lot of complaints from anons if entries like дянхуа and شِيَوْ عَر دٍ have a ==Chinese== heading, and I would find it disconcerting myself, too. And what would the definition then say? {{lb|zh|Dungan}}{{form of|Cyrillic script|電話|lang=zh}}? I think readers would find that more confusing than helpful. And then what about the Russian and Turkic loanwords that don't exist in China? They would have to have full definitions without a link to a Hanzi entry, and that would probably baffle readers even more, despite the {{lb|zh|Dungan}} tag. —Aɴɢʀ (talk) 14:14, 17 July 2016 (UTC)Reply
Hindi and Urdu should be merged — they are separate for political reasons. Remember, we allow for Afrikaans entries in Arabic script, Old French entries in Hebrew script, and other odd happenstances of historical script usage. We can continue to use the Dungan header for words only existing in Cyrillic form, just like I believe we do for Min Nan. —Μετάknowledgediscuss/deeds16:10, 17 July 2016 (UTC)Reply
Undecided for now. There are pros and cons. Cyrillic and Arabic spellings could potentially be added to each Mandarin standard pinyin syllable (non-standard could also be considered if confirmed). Multisyllabic only for confirmed ones.
Mandarin pinyin (with tone marks and monosyllabic tone numbered syllables), Min Nan POJ, zhuyin characters have not been "unified" under the Chinese umbrella for various reasons. Some are described above.--Anatoli T.(обсудить/вклад)07:39, 18 July 2016 (UTC)Reply
I think it would be more convenient for editors if Dungan were part of unified Chinese, since it would be easier to edit. It would just feel like too repetitive if I made a new Dungan entry that technically already has an equivalent Chinese entry. --Mar vin kaiser (talk) 08:32, 18 July 2016 (UTC)Reply
@Wyang: if you still support folding Dungan into Chinese, how best could that be accomplished? Make the attested Cyrillic (and Arabic?) forms soft redirects...? (Would the L2 header of e.g. фонзы be "Chinese" or remain "Dungan"?) What is to be done if, as several users worry above, there are loanwords that don't have Han-character representations? - -sche(discuss)04:48, 21 January 2018 (UTC)Reply
@-sche I'm now also concerned about the complexity of this... a possible solution is to have {{zh-pron}} link to the Dungan word, but keep the Dungan L2 heading, and treat Dungan as a full language with the Cyrillic entry showing full usage examples, etc. and linking back to the Hanzi form, perhaps on the headword line. Wyang (talk) 07:03, 21 January 2018 (UTC)Reply
(I just noticed, phōng-kó uses the Chinese header. I think it should use the Min Nan header, similar to how pinyin entries use the Mandarin header. But that's a separate discussion, I think. —suzukaze (t・c) 07:46, 21 January 2018 (UTC))Reply
We had a discussion on this but I don't remember where. I'm OK to use the Chinese L2 header on all romanised entries if they link to (and consequently, have a written form in) hanzi. --08:03, 21 January 2018 (UTC)
As Cyrillic script does not imply tone and merges some consonant, the Cyrillic form is generated from pinyin-like romanisation instead. We need Wiktionary:About Dungan to document it. Also, current pinyin-like romanisation include some redundant vowels.
IPA value are from Wikipedia and needs checking.
I don't know how neutral tone work, nor whether there're tone sandhi or erhua in Dungan.
Some values can not be generated by current pinyin-like romanisation. e.g. чў=q+u (not ü) which is not a valid Pinyin syllable, and ңыйлу (="ng+er+y" lou).
We need a module to transcript Cyrillic Dungan entries.
I oppose the use of tones in the transliteration for Dungan if it's reintroduced. The Cyrillic spelling have no tones and there should be no tone numbers or marks in the romanisation. Further, the transliteration should show what's actually written, not matching the Mandarin pinyin, e.g. йүян(yüi͡an) is actually "yüyan" (without a space), not "yu3 yan2" but perhaps, there's no need to provide any additional transliteration. --Anatoli T.(обсудить/вклад)10:56, 24 January 2018 (UTC)Reply
We have tone and length annotations for every other language whose orthography does not reflect relevant phonemic differences. Why should this be different? Korn [kʰũːɘ̃n] (talk) 12:08, 24 January 2018 (UTC)Reply
@Atitarev: I don't think the Cyrillic in Dungan should be referred to as a transliteration, but as a script of its own. Therefore, I agree with you that tones shouldn't be put in the Cyrillic writing precisely because it is a script, and it is written without the tone, but the IPA should definitely provide the tone, since the script traditionally doesn't write it down. --Mar vin kaiser (talk) 06:30, 25 January 2018 (UTC)Reply
@Mar vin kaiser: I don't think Atitarev is referring to the Cyrillic as transliteration. He's talking about the romanization of the Cyrillic. I found out from Omniglot that there are two conventions used for indicating tones: -, ъ, ь or I, II, III. We should probably adopt one of these two. — justin(r)leung{ (t...) | c=› }07:05, 25 January 2018 (UTC)Reply
@Justinrleung: I see. Then it's settled then, I guess. For cognates with other Chinese languages, a unified Chinese entry will be used with the Cyrillic entry still existing the same way POJ entries in Min Nan still exist. The Dungan portion of the pronunciation would then contain the Cyrillic entry or the Cyrillic script with tone marks? What do you think? Maybe the Cyrillic script with tone marks, but it would link to the Cyrillic entry without tone marks? --Mar vin kaiser (talk) 07:20, 25 January 2018 (UTC)Reply
@Metaknowledge: There are too many unanswered questions, such as tone sandhi, erhua and that (now missing but I have a copy) dunganDictionary.html dictionary talks about more tones, e.g. Ia and Ib, not implemented. User:Justinrleung might have a Chinese source, which has more details about the pronunciation. The Russian dictionary could be used for vocabulary but its pronunciation description is simplistic. --Anatoli T.(обсудить/вклад)04:21, 21 January 2020 (UTC)Reply
@Justinrleung If anyone needs it, I can email the original dictionary (the HTML file and its stylesheet). The internal references are broken (when you try to click on the up arrow to see also). I've got the Dungan-Russian dictionary (pdf) as well - it's less user-friendly and you can't copy most of Dungan terms. --Anatoli T.(обсудить/вклад)05:19, 21 January 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming [lel]
We currently refer to the other languages most commonly referred to as "Lele" with parenthetical geographic disambiguators (Lele (Guinea), Lele (Papua New Guinea), Lele (Chad)). I suggest we do the same for [lel] and change "Bashilele" to "Lele (Congo)". There is an argument that "Lele (Congo)" should be "Lele (Democratic Republic of the Congo)", but that seems too long and none of these languages are spoken in the Republic of the Congo, so I think it's safe. —Μετάknowledgediscuss/deeds05:30, 10 March 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming the Sangu languages
We currently call [snq] "Chango", which is quite infrequently used. We could go for "Isangu" (with the language prefix), because there is another Bantu language called Sangu — or we could bite the bullet and go for geographical disambiguation with "Sangu (Gabon)" for [snq] and "Sangu (Tanzania)" for [sbp], which is what Wikipedia does, and is probably least misleading. —Μετάknowledgediscuss/deeds04:29, 31 March 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Merge [xma] into [ziw]
This is a case where a single language spoken in two countries has been given two codes. Mushungulu [xma] is the dialect of Zigula spoken in Somalia. Maarten Mous says: "I interviewed some of the refugees [from Somalia into the Zigula homeland in Tanzania]. They claim there is no difference between the way they speak Zigula and the Zigula of those who never left the area; the [Tanzanian] Zigula speakers present agreed to this. Noting down some expressions, I had the same impression when comparing this to my Zigula field notes from 1994." —Μετάknowledgediscuss/deeds19:45, 31 March 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Split Bafaw and Balong
Currently, the code [bwt] is called "Bafaw-Balong" in an attempt to cover the two dialects. However, these "dialects" are quite divergent, though fairly closely related. Maho (2009) separates them, as does the Linguistic Atlas of Cameroon. Emmanuel Chia suggests that heavy borrowing has confused linguists into considering them the same language in The Bafaw Language (Bantu A10) (which treats Bafaw and not Balong, considering it completely separate). I suggest that we rename [bwt] to "Bafaw" and make a new code for Balong, maybe bnt-bal. —Μετάknowledgediscuss/deeds20:12, 31 March 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Adding Lwel and Mpiin
The Boma-Dzing languages are a bit of a mess, as Dzing was given a code but none of its sister languages were initially. Lwel is considered a distinct language by Maho (2009) and by {{R:nzd:Crane}}, and seems to have a grammar (Éléments de grammaire morphologique de la langue lwel) that I can't find anywhere, but citations to it seem like a distinct language. I suggest a new code, bnt-lwl. —Μετάknowledgediscuss/deeds 21:02, 31 March 2020 (UTC)
As for Mpiin, it's also considered distinct by Maho (2009) (and, I should add, these opinions are upheld by Van de Velde et al. (ch.2, by Hammarström)), and a selection of words in Koni Muluwa (2010) seem similar to, but consistently distinct, from other Boma-Dzing languages. I suggest a new code, bnt-mpi. —Μετάknowledgediscuss/deeds05:44, 1 April 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Retire [mct]
This code for "Mengisa" is listed as a duplicate code by WP. Van de Velde et al. (ch.2, by Hammarström) say: "The Eton that most Mengisa now speak is not linguistically remarkable and therefore we count it as an Eton variety". (The Mengisa originally spoke [leo].) —Μετάknowledgediscuss/deeds00:46, 1 April 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Merge [nzu] and [ebo] as Central Teke
Van de Velde et al. (ch.2, by Hammarström) say: "are commonly enumerated separately but the recent comparison by Raharimanantsoa (2012) [which I cannot find online] shows that such a distinction is untenable, wherefore we count them as the same". I am agnostic as to which code to choose, but the language should be called "Central Teke". —Μετάknowledgediscuss/deeds00:55, 1 April 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming [auj]
Currently "Augila"; many spellings exist, but all modern linguistic work on the language in English (and most of the material on the town itself) refer to it as "Awjila". —Μετάknowledgediscuss/deeds23:10, 19 March 2020 (UTC)Reply
@Benwing2: There are a bunch of language renames on this page that I'd like to do (this being one of them). If I just ignore the categories, will your regular bot runs create the new ones and delete the old, or should I move the categories instead? —Μετάknowledgediscuss/deeds04:17, 11 May 2020 (UTC)Reply
@Metaknowledge I do a bot run every 3 days to create new categories, but it doesn't delete old ones. I have periodically deleted old categories listed in CAT:Empty categories, but I don't do it on a regular basis. I *think* all the old categories will eventually end up on that page. It's probably better to move the categories and manually fix the lemmas belonging to the renamed languages to have the new language name in their headers. For languages with very few lemmas it's probably not a big deal to do it manually; otherwise, let me know and I'll do it by bot. Benwing2 (talk) 04:29, 11 May 2020 (UTC)Reply
Yeah, none of them are a pain in the ass individually, I was just hopeful due to how many languages there are to rename. (Awjila only has one entry, but I had to move 10 categories for it!) Thanks for the offer, I'll let you know if any have a significant number of entries. —Μετάknowledgediscuss/deeds04:36, 11 May 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming the Kota languages
The name "Kota" is commonly used for two very small languages, so this seems like another good opportunity to use parenthetical disambiguators rather than force one to use a name with a language prefix, rarely encountered in the literature. I suggest that we rename [kfe] from "Kota" to "Kota (India)", and [koq] from "Ikota" to "Kota (Gabon)". —Μετάknowledgediscuss/deeds05:35, 10 March 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Merge [lal] into [nxd] and rename [nxd]
Van de Velde et al. (ch.2, by Hammarström) say that "studies in the field emphasise that Lalia is simply a variety in the Bangando area with no special status vis-a-vis other Bongando varieties". While we're at it, we currently call [nxd] "Longandu", which seems to be one of the rarer names. In the tradition of lopping off language prefixes and matching the usual final vowel, I suggest we rename it to "Ngando". —Μετάknowledgediscuss/deeds01:21, 1 April 2020 (UTC)Reply
Merged. As for the renaming, the issue is more complex than I had realised: there is also [ngd] (Ngando language (Central African Republic)), which we perplexingly call "Bagandou". We could change the parentheticals later, but for now I'm going with "Ngando (Congo)" for [nxd] and "Ngando (Central African Republic)" for [ngd]. —Μετάknowledgediscuss/deeds21:38, 11 May 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming [ngc]
We currently call this "Longandu". I assume the reason for avoiding "Ngandu" was originally potential conflict with [bgf], which we call "Bangandu". Given that "Bangandu" seems like a perfectly fine name, I suggest we rename [ngc] to "Ngandu". —Μετάknowledgediscuss/deeds01:27, 1 April 2020 (UTC)Reply
Well, I seem to have gotten these confused while attempting to clean them up, which is as good an argument as any for better names. We actually call [ngc] "Lingombe", from which we should remove the language prefix. Unfortunately, that brings it into potential conflict with the poorly named "Bangando-Ngombe" [nmj], variously considered a dialect of Bangandu [bgf]. I am instead renaming it to "Ngombe (Congo)" for maximal clarity, and will create a new thread to discuss [nmj]. —Μετάknowledgediscuss/deeds23:17, 11 May 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Merge [gti] into [nyc]
Van de Velde et al. (ch.2, by Hammarström) say: "the only source on this language (Hackett and van Bulck 1956:74) has it as a dialect of Nyanga-li [nyc]." If that's the sole source, I have no idea why it ever got a code. —Μετάknowledgediscuss/deeds01:33, 1 April 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Merge [ppp] and [lnz] into [yaf]
Van de Velde et al. (ch.2, by Hammarström) say: "Pelende and Lonzo denote political rather than ethnolinguistic sub-entities of Yaka [yaf] ... as explicit linguistic data, whenever available, confirms". —Μετάknowledgediscuss/deeds01:38, 1 April 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Merge [njd] into [kde]
Van de Velde et al. (ch.2, by Hammarström) say: "the field research of Kraal (2005 1–7) finds this distinction [between Ndonde and Makonde] untenable from a linguistic and ethnographic point of view". "Makonde" [kde] is the usual term. —Μετάknowledgediscuss/deeds02:02, 1 April 2020 (UTC)Reply
Latest comment: 4 years ago2 comments1 person in discussion
The following discussion has been moved from Wiktionary:Requests for moves, mergers and splits ([[Special:PermanentLink/59536661#Merge_Shona_dialects_[twl],_[mxc],_[twx]_into_Shona_[sn]|permalink]]).
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Merge Shona dialects [twl], [mxc], [twx] into Shona [sn]
Ehret and Kinsman (1981) "Shona Dialect Classification and Its Implications for Iron Age History in Southern Africa" have helpful trees of the Shona dialects, which confirm that [twl], [mxc], [twx] are all mutually intelligible to a high degree with the other dialects of Shona [sn]. (In fact, [twl] and [twx] are subdialects of the larger dialects that most Shona dictionaries treat, and [mxc] isn't even a valid clade, but merely a geographic convenience.) —Μετάknowledgediscuss/deeds02:17, 1 April 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Lala languages
The Bantu Lala language (South Africa) is certainly not the same as Zulu, as should be clear just from the sample in the WP article. Unfortunately, our current "Lala language" will have to be moved to make way. In the emerging tradition of geographic parentheticals, I suggest that [nrz] be renamed to "Lala (New Guinea)" and a new code, maybe [bnt-lal], be created for "Lala (South Africa)". (As a side note, there are also Lala-Roba and Lala-Bisa, which thankfully have hyphenated disambiguators.) —Μετάknowledgediscuss/deeds02:33, 1 April 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Merging [nyr] into [nih]
There's a lengthy SIL report on the Nyiha languages, and skimming it suggests that the Malawian ([nyr]) and Tanzanian ([nih]) dialects are mutually intelligible. Chances are that the codes haven't been used to separate the two countries' dialects in a consistent way anyway, because they've just been called "Shinyiha" (with the language prefix) and "Nyiha" (without) — the latter is what we should use. —Μετάknowledgediscuss/deeds05:21, 1 April 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming [kmw]
We currently call [kmw] "Kikumu"; the language is most commonly referred to as "Komo", but we already have [xom] "Komo", which is not easily disambiguated (it's spoken in three countries and belongs to an obscure group). We can still do better than "Kikumu", though — I suggest we lop off the language prefix and call it "Kumu", which is more common than the name we currently use. —Μετάknowledgediscuss/deeds06:10, 1 April 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming [lwg]
We currently call this "Oluwanga"; we should drop the language prefix, as we normally do and as matches the scholarly literature, and call it "Wanga". (Note that we need to reconsider the Luhya codes altogether, but that can be dealt with later.) —Μετάknowledgediscuss/deeds20:23, 1 April 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming [blv]
We currently call [blv] "Bolo", which refers to a peripheral dialect of this language. This SIL report says that a conference decided on "Kibala", and Angola's Institute of National Languages followed suit (the alternative suggestion of "Kibala-Ngoya" having been rejected due to its historically derogatory usage). I suggest we do the same and use "Kibala". —Μετάknowledgediscuss/deeds20:47, 1 April 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming [syx]
We currently call [syx] "Shamay", which is a rather rare name for it. WP uses "Samay", but I suggest we follow Van de Velde et al. (ch.2, by Hammarström) and use "Osamayi". —Μετάknowledgediscuss/deeds20:57, 1 April 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Merge [xdo] into [mho]
Van de Velde et al. (ch.2, by Hammarström) and indeed every linguistic source I can find has this as synonymous with or a dialect of Mashi [mho]. I can only see a bit of the relevant part in Laranjo Medeiros (1981) VaKwandu on BGC, where he seems rather confused about the language in general, but there may be useful information there if someone can see it (or if anyone still has academic libraries operating in their part of the world). —Μετάknowledgediscuss/deeds21:16, 1 April 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Merge [yax] into [mck]
The Yauma are ethnographically distinct, but speak a Mbunda dialect according to Fleisch (2000). I see a journal article in 1970 that says Yauma is synonymous with "newer Mbunda" (in contrast with "Old Mbunda"), and I'm not really sure what's going on there, but it also supports a merger. I cannot find Yauma mentioned in the usual sources otherwise. —Μετάknowledgediscuss/deeds21:32, 1 April 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming fay
"Southwestern Fars" is a really awful name. Fars is not a macrolanguage as the name would suggest, but a province of Iran, and there are other lects spoken in southwestern Fars. We would do much better to use the unambiguous name "Kuhmareyi", as used by Wikipedia. —Μετάknowledgediscuss/deeds19:44, 9 December 2017 (UTC)Reply
hi, Metaknowledge. At the beginning when I wanted to create the category for the language I searched online to see if I can find Kuhmareyi. I could not find online sources mentioning Kuhmareyi other than Wikipedia. Online sources mostly mention Southwestern Fars. I know Fars is a province. ethnologue and some other sites use this name. I am trying to find the name Kuhmareyi in the book that I have, A treasury of the Dialectology of fars. I should also mention that I am very cautious in adding the new words and double check the words and have not recorded some words that can not be found in other sources. However the book is written by a professional linguist and is funded or guided by Iranian Academy of Persian Language and literature. I think I saw Kuhmareyi in one online source. I have not found the name "Kuhmareyi" So far but the name seems correct.--Eeranee (talk) 21:09, 9 December 2017 (UTC)Reply
I can find online sources in Persian using Kuhmareyi which might not be reliable but I could not find a source written in English using Kuhmareyi. If anyone thinks kuhmareyi is more correct we can use it--Eeranee (talk) 22:06, 9 December 2017 (UTC)Reply
Searching Google Books for both names, I find nothing that would help us much to decide one way or the other: nothing using "Kuhmareyi", and only a couple of general reference works on world languages that mention "Southwestern Fars" in giant lists of languages, probably just copying Ethnologue. (Wikipedia curiously says "the southwestern [Fars] dialects can be divided into three families of dialects according to geographical distribution and local names: Southwestern (Lori), South-central (Kuhmareyi) and Southeastern (Larestani)", as if calling it "southwestern" might be slightly confusing/misnomial.) @ZxxZxxZ, Vahagn Petrosyan, do either of you have a preference or knowledge of which name is more common or appropriate? - -sche(discuss)19:19, 17 December 2017 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming [chw]
Currently called "Chuwabu". There are quite a lot of spellings, influenced by English, Swahili, Portuguese, and Africanist notation, and none of them are unquestionably more common, but this shows that the case for "Chuabo" is the best, which is supported by paging through the Google Books results as well. —Μετάknowledgediscuss/deeds17:11, 5 May 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming [nmj]
We currently call [nmj] "Bangando-Ngombe". It is variously considered a dialect of Bangandu [bgf]: Wikipedia cites Glottolog to support a claim that the code is "spurious", and the Atlas linguistique de l'Afrique centrale calls it "une variété très voisine" to Bangandu. Without a clearer statement about mutual intelligibility or specific fieldwork, I am hesitant to merge it. However, the current name, which hyphenates a (differently spelt!) version of the macrolanguage with the dialect's name, is clearly undesirable. I suggest we rename it to "Ngombe (Central African Republic)" (the national parenthetical serving to disambiguate from [ngc] "Ngombe (Congo)"). @-sche —Μετάknowledgediscuss/deeds23:24, 11 May 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming the Manda languages
We currently call [mgs] "Nyasa", and it came to my attention due to the entry mtua, which User:Wikitiki89 mistakenly created as a "Nyasa" entry, based on an obsolete dictionary of Chichewa that calls it "Kiniassa". Nowadays, Nyasa is usually a name for a group of closely related languages including Chichewa (Nyasa languages) but WP claims it is also used for mjh. Following WP, we could call it "Manda (Tanzania)", rendered problematic by the two other Manda language possibilities, which we call "Manda" and "Australian Manda" (!). I propose that we go all in for national disambiguation and make those parenthetical as well, as "Manda (India)" and "Manda (Australia)". @-sche —Μετάknowledgediscuss/deeds06:18, 28 June 2018 (UTC)Reply
Just for the record, I did a lot of research to determine what the correct language code was for the language described in that dictionary, even referring to this map (source) and strongly considering "mjh". Thanks for finally sorting it out! Note also, there is a "Nyasa" translation at water, which does not align with Chichewa "madzi" (which the dictionary I used gave as "madsi"). --WikiTiki8915:07, 28 June 2018 (UTC)Reply
@Wikitiki89: Next time, try Google instead of poring over maps. :) I was already aware of this dictionary (and its miserable orthography), but searching "Rebman Kiniassa" gets you the Wikipedia article for Johannes Rebmann, which in turn tells you this is Chichewa. As for the "Nyasa" translation, masi... I really don't know what language that should be. The word for "water" in mgs is máchi. —Μετάknowledgediscuss/deeds17:16, 28 June 2018 (UTC)Reply
WP says Australian Manda (zma) is also called "Menhthe", but the only published resources on it I could find used Manda. To distinguish it and [mgs] and [mha], the proposal of [mgs]="Manda (Tanzania)" and [mha]="Manda (India)" and [zma]="Manda (Australia)" sounds good. (As an aside, there is also a "Ma Manda" language.) - -sche(discuss)03:11, 21 January 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming mjh > Merging mjh
Tangentially relevant to the discussion of languages named "Nyasa" above, mjh is not only one of those, but is also called "Mwera" (a name currently occupied by mwe, which can't even by distinguished by a parenthetical, because both languages are from Tanzania!). Maho's Guthrie List, the standard list used by Bantuists, calls it "Mbamba Bay Mwera", but the only hits for that string on Google Books are of that very list. WP chooses to simplify this as just "Mbamba Bay", which is both the most unambiguous option, and also an option that seems to be used only there. I'm really unsure what to call it — anything but our current name of "Nyanza", which refers to the lake and offers only more confusion. —Μετάknowledgediscuss/deeds06:27, 28 June 2018 (UTC)Reply
@Metaknowledge Can they be [called by whatever name is most common, even though homographic to another language, and then] distinguished by linguistic family, like WT:RFM#Canonical_name_of_"fan"? Say, "Mwera (Nyasa)" vs "Mwera (Ruvuma)" or whatever? Btw, we should standardize on putting the family in parentheses (where we also put geographic disambiguators), which will require renaming some languages that were named with the family first, like Papuan Mor (but not Sepik Iwam, which is apparently normally called that, to distinguish it from the other Iwam that is also Sepik). - -sche(discuss)07:05, 17 January 2020 (UTC)Reply
I don't know anything about these languages, but if indeed both mjh and mwe are usually called Mwera and both are spoken in Tanzania, then I'd support "Mwera (Nyasa)" and "Mwera (Ruvuma)". —Mahāgaja · talk07:29, 17 January 2020 (UTC)Reply
@-sche, Mahagaja: I bring good tidings! Van de Velde et al. (ch.2, by Hammarström) report that "the examination by Ebner (1955: 41–43) shows the language to be a variant of Chewa/Nyanja on the opposite side of the lake". We no longer have to worry about the name, because we can merge [mjh] into [ny]. —Μετάknowledgediscuss/deeds01:55, 1 April 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Merging Gaulish and Lepontic
I'd like to address merging Lepontic with Gaulish because it is already included in the English article provided and the terms are frequently the same except for some differences in endings development without any necessary declension change. Comparing this to Latin, as it is merged with Old Latin, this may also be extended to Gaulish and Lepontic. Furthermore, Cisalpine Gaulish is already encompassed by Gaulish even though it was written in the Old Italic alphabet just like Lepontic. HeliosX (talk) 19:55, 15 January 2020 (UTC)Reply
I generally push for a conservative treatment for extinct languages with small, finite corpora. We can afford to over-split a little, if that even is the case here, and be no worse off for doing so. —Μετάknowledgediscuss/deeds03:13, 17 January 2020 (UTC)Reply
Meh. I don't object, but I'm not convinced it's necessary. We're talking about a time difference less than that between modern and Middle English, many of the differences seem to be in whether various morphemes remain attested into the present in various positions (or, conversely, whether modern ones happened to be found in the sparse early corpus), spelling variations are clearly attributable to texts being written by potentially non-fluent outsiders vs modern linguists and natives, and modern Omagua is small and dying. (Fox is another language which underwent changes in a relatively short time period that might be considered substantial, but it seems to still be treated as one language.) It seems to me like we could adequately, and perhaps even more sensibly, handle the differences by labelling words and spellings obsolete where necessary, lemmatizing the modern forms (and conversely labelling words/forms as modern developments in cases where we can tell that they are as opposed to that they just weren't attested in the old texts). - -sche(discuss)21:12, 18 February 2020 (UTC)Reply
Some languages do change a lot faster than others due to social pressures (hence why glottochronology doesn't really work). I found the differences in morphology and lexicon significant, especially in the hopes that someday, someone will build the infrastructure for (modern) Omagua here. But if you're not convinced, I'm fine with shelving this; I certainly won't be working on Tupian stuff myself. —Μετάknowledgediscuss/deeds03:11, 19 February 2020 (UTC)Reply
Charles Grimes says in Spices from the East: Papers in Languages of Eastern Indonesia: "This speech variety has been extinct since 1974, when the last speaker died. No clues other than the name of a stream east of Kayeli called Moksela, give any indication as to where it was spoken or what it was like. If it was spoken from the stream by that name eastward, then chances are likely that it was also a variety of the Kayeli language. People in the Kayeli area remember nothing more than the name of the language, who in the community spoke it […]" (I cannot view beyond this in the Google Books preview.)
"...who in the community spoke it before they died, and that it was somehow different enough to have its own identity." is the rest of the sentence, I managed to coax Google into telling me. The name seemed familiar, as if it had been in one of the wordlists I've been looking at recently, but I just went back over them and searched through various other sources and indeed the only mentions of it I find all just say it's extinct and not recorded; how sad. Removed. - -sche(discuss)08:19, 10 May 2017 (UTC)Reply
In this vein, Makolkol [zmh] is claimed to be extinct (per Wurm 2003, after having 7 speakers in 1988) and apparently unattested (per Stebbins 2010). Harald Hammarström and Sebastian Nordhoff accept this conclusion in Melanesian Languages on the Edge of Asia, but it may be a cautionary tale instead, because an article in LoopPNG from 2016 says five Makolkol still live, and even provides words(!), saying it is related to Simbali: "mam, meaning father, and nan, meaning mother". A 2005 article in Anthropological Linguistics (volume 47, page 77) agrees on the relation to Simbali: "[…] Makolkol (extinct), […] is locally understood to have been a 'mixed language' combining Simbali and Nakanai (an Austronesian language on the northern side of New Britain)." I suppose the code should be left alone for now, pending further data. (There were widely varying estimates of how many speakers it had earlier in the 20th century, and fanciful tales of who they were, "headhunters" or "giants" who "lived in trees" and who no white person had survived meeting at first.) - -sche(discuss)08:43, 10 May 2017 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Even more languages without ISO codes, part 5
This next batch is of languages from lists other than Ethnologue and LinguistList. As before, I've tried to vet them all beforehand, but I will have doubtlessly made some mistakes. NB if you want to find more: I've avoided dealing with most of the Loloish languages, because all the literature seems to be in Chinese. —Μετάknowledgediscuss/deeds04:54, 6 July 2016 (UTC)Reply
Wikipedia suggests this might be only attested vi placenames. I have seen or created codes in some similar cases, and am willing to create one here, if there are not objections... - -sche(discuss)00:17, 2 August 2020 (UTC)Reply
Wikipedia and various of its cited references and authorities say this and Tulua are the same language, but we don't have a code for Tulua ether, so something should still be added. Wikipedia suggests Tulua is a more-used name,s perhaps aus-tul? (Other names I spot in literature: Toolooa, Dulua, Dapil, Narung, Dandan. AIATSIS also lists Gureng Gureng, Goeng, Koreng Goreng, and Taribulung, which however seem to result from confusing it with Taribelang language.) Wikipedia also, confusingly, says it is extinct but that the Australian Priority Languages Support Project has it as a priority as an endangered language "where there are living speakers". A short search turned up A Dictionary of Non-Scientific Names of Freshwater Crayfishes (1994), which has "wunmeen PARASTACIDAE, "crayfish," prbably = Cherax gladstonensis. [Australia: Queensland; Boyne River, Toolooa or Dandan Aboriginal tribe (=Tulua Aboriginal tribe of Tindale, 1974: 186 [...])] Curr, 1887(III):124, Sta. 161." The cited work, the very old report by Curr (who doesn't provide the tribe name per se, and says the list was given to him in an illegible hand), indeed lists that as a word, and "water" as doomoo and "fire" as wi. - -sche(discuss)18:53, 1 August 2020 (UTC)Reply
Done, although AIATSIS says that this is Ethnologue's Uradhi language... but that is a cluster of lects, of which several already have separate codes... although the ISO and Glottolog seem to split them differently. Messy. - -sche(discuss)19:20, 1 August 2020 (UTC)Reply
@-sche, back when I suggested these Australian languages, I included the codes for the Tasmanian languages that Bowern (2012) teased out of various wordlists. At the time, I was ignorant of the fact that there is an ISO code, xtz, for a language called "Tasmanian", and we have a few words in it. There was no single Tasmanian language, so I think this code should be retired and the words sorted into their respective languages by Bowern's scheme. —Μετάknowledgediscuss/deeds05:28, 3 May 2017 (UTC)Reply
It lacks a Wikipedia article but is documented by Jacobs, Texte et lexique indanga (2002). fr.Wikt already has a word from it. OTOH, fr.WP considers it a regional variant of Tetela. And fr.Wikt does have a tendency to treat dialects as language, also splitting e.g. Alsatian German from Alemannic German, Hoanya from Papora, etc. - -sche(discuss)05:21, 29 August 2016 (UTC)Reply
Well, it's definitely part of the dialect continuum known in Guthrie as C.70, which has 8 ISO codes that cover it rather poorly (this is a typical situation with Bantu languages, which really need their own overhaul at some point). I see that its word for "water" is bash in that reference, rather different than Tetela proper ashi. We have to draw lines somewhere, and I can't figure out where Indanga would be merged, so I suppose a new code is in order. —Μετάknowledgediscuss/deeds05:30, 4 October 2016 (UTC)Reply
Many sources seem to accept the existence of three Naish "languages" and we have codes for the other two (though Na (called here "Narua") lacks a WP article, as the editors there seem unconvinced by its separateness). Done (as tbq-laz) —Μετάknowledgediscuss/deeds01:58, 5 April 2018 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Merge [dov] into [toi]
Van de Velde et al. (ch.2, by Hammarström) say: "Dombe is a derogatory nickname for Tonga found in Hwange district of Zimbabwe". This raises an issue with [toi], which we currently call "Tonga (Zambia)" (to distinguish it from "Tonga (Malawi)" and "Tonga (Mozambique)") — it's spoken in Zimbabwe too. The WP article is called "Tonga (Zambia and Zimbabwe)", which is awfully awkward name for a language spoken by 1.5 million people. I think we may have to cut our losses and stick with "Tonga (Zambia)" even after the merger. —Μετάknowledgediscuss/deeds01:51, 1 April 2020 (UTC)Reply
At least I would assume it's spoken primarily in Zambia, as in maybe at least 67% of speakers are in Zambia. But why not use subfamilies instead of countries as disambiguators? [toi]/[dov] could be "Tonga (Botatwe)", [toh] could be "Tonga (Southern Bantu)", and [tog] could be "Tonga (Nyasa)". —Mahāgaja · talk06:41, 15 June 2020 (UTC)Reply
@Mahagaja: "Primarily" is certainly correct here. But the idea of using subfamilies is an intriguing suggestion that I hadn't considered. My only fear is that those subfamilies are not very well known by any terms, as such classification is relatively recent, and to someone unfamiliar with the classification system, the word "Botatwe" would be meaningless and the concept of "Southern Bantu" would be ambiguous. —Μετάknowledgediscuss/deeds06:51, 15 June 2020 (UTC)Reply
I would personally prefer the more accurate label Tonga (Zambia and Zimbabwe), even if it's a bit clumsy. But I'm not opposed to keeping it as Tonga (Zambia) if you prefer. What I'm confused about is why you want to merge [dov] and [toi]. The chapter you're referencing splits [dov] as a separate language Toka-Leya, which suggests that language isn't mutually intelligible with Tonga. Smashhoof (Talk · Contributions) 17:25, 15 June 2020 (UTC)Reply
@Smashhoof: Thank you for checking! I read that footnote as meaning that the "Toka-Leya dialects" were to be thought of as dialects of Tonga, but I see that Hammarström does give the sum of Toka, Leya, and "Dombe" its own line as a language. I went to check Hachipola (1991), who did a study on eight Tonga-group lects, including Toka, and says: "[…] Toka informants all insisted that they were all legitimately Tonga. Valley speakers in particular consider their speech as the 'purer' form of Tonga of which Plateau is the corrupted version [MK: note that Plateau is standard Zambian Tonga]. The speakers of Plateau, Valley and Toka can also understand speakers of Ila, but with some difficulty. However, conversation can be carried out between, say, a Plateau speaker and an Ila speaker each using their own speech form. This is also true of Lenje on the one hand and Plateau on the other […]". I glanced through the lexical items in the second half of the dissertation and found that these three lects were quite similar, and Toka might be closer to Valley Tonga than to Plateau Tonga. I'm really not sure where to draw the line, especially since we (by default) consider Valley and Plateau to both be [toi]. What do you think? —Μετάknowledgediscuss/deeds19:17, 15 June 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Adding Ashraaf
Ashraaf is the sister language to Somali, and currently included in its code despite being mutually unintelligible. Wikipedia has an entry at Ashraf dialect, despite the fact that they quote two sources that make the case for it being its own language, and their references use the usual spelling with a double a. It is somewhat poorly documented (there is no dedicated grammar), but multiple scholarly sources include vocabulary. I suggest the code cus-ash. —Μετάknowledgediscuss/deeds01:33, 11 August 2020 (UTC)Reply
I am inclined to agree/support; Green and Jones' piece outright says, elsewhere from the snippet WP quotes, that "despite all classifications of Ashraaf treating it as a dialect of Somali, our Marka speakers have intimated to us that both Marka/Somali and Marka/Maay intelligibility presents a challenge" and "despite the fact that some classifications treat Ashraaf as a dialect of Somali, Marka and Somali appear not to have a high degree of mutual intelligibility"; they say plurals are almost all different from Somali, and singulars also show some differences. (They mention the otherNames = "Af-Ashraaf", and the varieties {"Marka, Lower Shabelle"} and "Shingani", saying — as you probably saw, but as I will mention here for easier findability later — that for Shingani there is one 'theoretical article' on 'theme construction', Ajello 1984, one short grammatical sketch, Moreno 1953, and one book of pedagogical material, Abo 2007, and for Marka there is an unpublished grammatical sketch, Lamberti 1980, and an article on verb inflection, Ajello 1988.) And I see that Maay already has a separate code, reasonably (the IEL says standard Somali is "difficult or unintelligible to Maay and Digil speakers" unless they've learned it). The only thing that gives me pause is that, as they admit, many prior (albeit non-Ashraaf-specific) classifications have viewed Somali as a single language, as WP claims speakers also view it, despite the lack of mutual intelligibility. - -sche(discuss)20:46, 11 August 2020 (UTC)Reply
That's a matter of linguistic nationalism, coupled with an unfortunate habit of scholars like Lamberti to count everything spoken by people who identify as belonging to Somali clans as "Somali dialects". Maay, Jiddu, Garre, and many others were in this boat, but have all gotten ISO codes, except for Ashraaf. —Μετάknowledgediscuss/deeds21:07, 11 August 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Meänkieli (fit) and Kven (fkv) into Finnish (fi)
Finnish dialects
I think that linguists consider these to be dialects of Finnish, so that would make these pluricentric standards of a single language. I don't know if keeping them separate would hold any value? —CodeCat14:05, 15 March 2014 (UTC)Reply
The impression I get from the example at w:Meänkieli is that the differences are very minor, no more than there might be between Croatian and Serbian. I notice systematic loss of -d- and Finnish -ts- corresponds to -tt- in Meänkieli. They definitely look mutually intelligible. Kven looks a little more different, but it might also just be the spelling; I don't know how hard it would be to the average Finnish speaker. —CodeCat23:26, 23 March 2014 (UTC)Reply
I know maybe a dozen words of Finnish, so I can't judge for myself, but the impression I get from the Wikipedia articles is that there's an equal or greater range of variation between dialects in Finland as there is with these dialects- if these dialects were on the other side of the Finnish border, they would probably be considered just part of the normal dialectal variation (I'm sure there are some differences due to their isolation from the influence of standard Finnish, as well). They have special status because they're in Sweden and Norway surrounded by Swedish and Norwegian. Chuck Entz (talk) 23:53, 23 March 2014 (UTC)Reply
Finnish wasn't even a single language to begin with originally. There's several dialect groups that form a continuum, but it's not easy to draw clear lines. Savonian (eastern) dialects for example might well be closer to Karelian (considered a separate language) than they are to western Finnish. —CodeCat00:10, 24 March 2014 (UTC)Reply
My impression is the same as Chuck's, that these could be merged. By my (quick) count, we have 11 Meänkieli entries and 14 English entries with Meänkieli translations, and 19 Kven entries and 8 entries with Kven translations. - -sche(discuss)02:45, 24 March 2014 (UTC)Reply
Blue indicates areas where Finnish is spoken by the majority, and green indictes minority. Meänkieli and Kven are considered Finnish on this map
I somehow missed this discussion when it was active, but better later than never. I have the following comments:
The map is outdated. There's practically no Finnish-speaking population left in the areas which were annexed by the Soviet Union during and after the WWII. The map on the right is more up-to-date.
There's some Ingrian population left in the St. Petersburg area, but their number and share of population (less than 0,5‰ in Leningrad oblast) is drastically reduced due to 1) inflow of Russians to St. Petersburg, 2) Stalin's terror in the 1930's and 3) emigration to Finland between 1990 and 2011.
I'm not sure of Kven-speakers, but the speakers of Meänkieli tend to be quite strong in their opinion that they are not Finnish-speakers. It is probably true that if the border were in another place, Meänkieli would be considered a Finnish dialect. But then again, it would hardly be the same language as it is today - it would have preserved less archaic features and there would be much less Swedish influence in it. If ISO regards it a language, how could we be wiser?
Meänkieli is an official minority language in Sweden, and is regarded as distinct from Finnish which also has a (separate) minority language status there.
"Finnish wasn't even a single language to begin with originally." -- Show me one that was!
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Renaming svm
From ‘Molise Croatian’ to ‘Molise Slavic’. It seems most scholars in the field, excluding some in Croatia itself, have been switching to the latter name; see, for instance, the papers of Krzysztof Borowski or the more recent works of Walter Breu. The endonym is naš jezik(literally “our language”) or na-našo(literally “in our (manner)”), without any specific ethnonational designation; the other names are apparently recent impositions. For this see for instance Sujoldžić 2004:
Along with the institutional support provided by the Italian government and Croatian institutions based on bilateral agreements between the two states, the Slavic communities also received a new label for their language and a new ethnic identity — Croatian — and there have been increasing tendencies to standardize the spoken idiom on the basis of Standard Croatian. It should be stressed, however, that although they regarded their different language as a source of prestige and self-appreciation, these communities have always considered themselves to be Italians who in addition have Slavic origins and at best accept to be called Italo-Slavi, while the term »Molise Croatian« emerged recently as a general term in scientific and popular literature to describe the Croatian-speaking population living in the Molise.
Slavomolisano: […] In scientific work, this name is predominantly used, either in its Italian form or in translations. As the language is "genetically" affiliated to the Serbo-Croatian macrolanguage with its dialectal continuum and the problems of its segmentation, a denomination, referring to one of the individual Standard languages of this group, e.g. Croatian or Bosnian, should be avoided, the more so, as its individual character is mainly
due to the language contact with Italian and its dialects, especially that of Lower Molise.
Ethnologue, Glottolog, and SIL (as well as Wikipedia) all followed the ISO’s lead and list the language under ‘Slavomolisano’, the Italian form of ‘Molise Slavic’ (which would also be fine). AFAIK I’m the only contributor to Wiktionary in this language to date. — Vorziblix (talk · contribs) 19:57, 15 April 2018 (UTC)Reply
If most scholars as well as Ethnologue, Glottolog, SIL, and Wikipedia all call it Slavomolisano, shouldn't we do the same, rather than call it Molise Slavic? —Mahāgaja (formerly Angr) · talk11:46, 19 April 2018 (UTC)Reply
Sure, that would also be fine. The two are used pretty much interchangeably in scholarly publications (so the ISO note says ‘either in its Italian form or in translations’). It’s a bit of an odd situation, given the most prominent scholar in the field (Breu) submitted the language under ‘Slavomolisano’ to the ISO (hence the adoption by all the other organizations) but uses ‘Molise Slavic’ in his own recent English publications. But if you think it’s preferable to directly follow Ethnologue and the others I wouldn’t object. — Vorziblix (talk · contribs) 12:57, 21 April 2018 (UTC)Reply
On Google Books, "Nama language" and "Khoekhoe language" are about equally common. This is misleading, however, because many of the hits refer one of multiple Khoekhoe languages, as the term is used somewhat more broadly by some authors. Further evidence of this is supplied by "speak Nama" being much more common than "speak Khoekhoe" on Google Books. If we restrict ourselves to linguistic literature, "Nama language" is significantly more common than "Khoekhoe language".
In short, we do sometimes use a less common name for a language in order to disambiguate or avoid a name widely considered offensive. The name "Nama" is less inclusive, but also slightly less ambiguous, and seems to be the most common name in usage, pace Wikipedia. —Μετάknowledgediscuss/deeds18:52, 18 March 2019 (UTC)Reply
The native name is Khoekhoegowab (literally "Khoekhoe language"). "Nama language" may have been more common in the past, but the standardized language today is called "Khoekhoe" or "Khoekhoegowab". The dictionary I have says that "Khoekhoegowab is the Language of mainly the Damara, Haiǁom and Nama." In The Khoesan Languages (2013, Routledge Language Family series), they exclusively refer to the language as Khoekhoe; however, they distinguish between Namibian Khoekhoe (Nama/Damara) and Haiǁom/ǂĀkhoe. "Khoekhoe(gowab)" does seem to be a more accurate and preferred term to me. Smashhoof (talk) 21:43, 18 March 2019 (UTC)Reply
If you search a digital copy of The Khoesan Languages, you'll see that different authors use different terminology in the book. The sections concerning the language in question call it Khoekhoegowab [well, usually just Kh. for short], which is far less common than either Khoekhoe or Nama and little used in non-scholarly English. —Μετάknowledgediscuss/deeds00:58, 19 March 2019 (UTC)Reply
Looking in my copy of The Khoesan Languages, I see a few usages of "Khoekhoegowab", but "Khoekhoe" seems to be used more. "Khoekhoe (N.Kh.) is strictly a suffixing language...", "Khoekhoe categorizes nouns according to ...". The whole morphology section on the language seems to use "Khoekhoe". The syntax section uses "N.Kh." for "Namibian Khoekhoe(gowab)". Regardless, since the standard language is called Khoekhoegowab, I think it would be best to call it Khoekhoe, as that seems to be equivalent and more common than the full Khoekhoegowab in English. Also, Wikipedia uses the term Khoekhoe (see Khoekhoe language), so it would also make sense to keep the same terminology between wikis. Though, that article does use the term "Nama" more often than "Khoekhoe", which is a bit odd given the title is "Khoekhoe language". Smashhoof (talk) 02:49, 19 March 2019 (UTC)Reply
I asked someone who knows more about this and they said that Khoekhoegowab is the standard language, taught in schools and used in media, but Nama/Damara is the colloquial spoken language. Locals refer to it as Damaranama, Namadamara, Namataal, or Namagowab. Smashhoof (talk) 23:06, 18 March 2019 (UTC)Reply
I would like to propose splitting HWC (Hawaiian Pidgin) off of EN (English) due to the following reasons:
1. The discussion in Template talk:hwc is outdated. It took place back in 2012/2013. This is before the US Census recognized Pidgin as a separate language in 2015 (2-3 years later). As such, the general linguistic view of HWC's status may have changed as a result of federal recognition.
2. Many of the arguments these people used were flawed (at least in my opinion).
- They implied Hawaiian Pidgin shouldn't be classified as a creole due to its high levels of mutual intelligibility with Standard English, even going as far as to compare it with "gangsta slang" from NYC or LA. The problem with this argument is that intelligibility doesn't define a creole. Creoles are naturalized pidgins, and pidgins are: "an amalgamation of two disparate languages, used by two populations having no common language as a lingua franca to communicate with each other, lacking formalized grammar and having a small, utilitarian vocabulary and no native speakers". That perfectly defines the origins of Hawaiian Pidgin. It formed as a result of many immigrant groups learning English. In fact, Tok Pisin has a very similar origin.
- The central claim that Hawaiian Pidgin can be understood by English speakers is only a half-truth to begin with. Many English speakers say they understand "Pidgin", but that's because the majority of Pidgin speakers mix a lot of their speech with Standard English, even more so in formal contexts. Full-on Pidgin can be a tough nut to crack. Furthermore, it's interesting that they use the argument of mutual intelligibility on Wiktionary, the same website that lists Bulgarian and Macedonian as separate languages. Those two languages are more similar to each other than Standard English is to Hawaiian Pidgin, yet for some reason, the BUL-MAC languages are separated meanwhile HWC is listed as dialect of EN.
3. Even in Wiktionary, it gets referred to as a "creole language". In the page Hawaiian Pidgin, it literally says "A creole language based in part on English used by most "local" residents of Hawaii. It's kind of odd that Wiktionary defines it as a creole language, but doesn't actually list it as one.
I started that discussion in 2012, and I don't really agree with how I approached it then. Hopefully I can do a better job now. Your first and third points aren't really relevant here: the political status of hwc doesn't have any practical bearing on its linguistic status, and our entry has no authority over our language treatment; it could well be changed. But your second point gets to the crux of the matter, and I am no longer certain how to proceed. Ultimately, we need to identify texts or recordings that we can agree are "pure" hwc, as opposed to being mixed with (colloquial) English. You mention Bulgarian and Macedonian; one reason it's easy to separate them is that there are lots of texts in each, clearly belonging to one or the other. Mutual intelligibility may be one of our best guides if we can't even decide what to put in which category. —Μετάknowledgediscuss/deeds02:45, 11 September 2020 (UTC)Reply
My point wasn’t that the political status changes the linguistic status. I tried to say that federal recognition was a result of linguists changing their attitudes towards HWC, since the government wouldn’t randomly recognize it out of nowhere if it weren’t for linguistic perception. That’s why I thought the discussion needed to be revamped because there’s likely been a change in linguistic consensus from 2012 to 2020, if we’re seeing different developments like this.
And fair point about the third argument. But regarding the second one, I do think there are texts that are in pure hwc, such as the Hawaiian Pidgin Bible (Da Jesus Book). The book “Pidgin Grammar: An Introduction to the Creole English of Hawaii” also showcases pure Pidgin (although the explanations are in English). — Coastaline (talk) 04:27, 11 September 2020 (UTC)Reply
Just to clarify: even if there's consensus to reinstate Hawaiian Pidgin as a separate language, the template {{hwc}} will not be recreated, because that's not how we work with languages anymore. Rather, an entry for hwc will be added to Module:languages/data3/h. I don't know anything about HP beyond what Wikipedia tells me, but if the entire Bible is written in the language illustrated by File:Hawaii Pidgin crop.jpg, then I do think it's appropriate to call that a separate language rather than a variety of English. It may not be as different from standard English as Bislama is (as Meta said 8 years ago), but I do get the impression that it's about as different from it as Jamaican Creole is, one big difference being that Jamaican spells things phonetically while Hawaiian sticks closer to standard English orthography (Jamaican Creole fuud vs. HP food), which makes HP easier to read for English speakers. —Mahāgaja · talk15:24, 11 September 2020 (UTC)Reply
Poking around the copy Google Books has snippets of, it does seem to all be like that, including e.g. "But wen da police guys wen come to da jail, dey neva find Jesus guys ova dea. So da police guys wen go back an tell da leadas: 'we wen find da jail door still yet stay lock, an da security guard standing by da doors, but wen we wen open um, nobody stay inside.' Wen da leada guys hear dat, da captain fo da temple guard an da main priest guys wen come mix up, an dey wen tink plenny wat goin happen." (I also found a book called "Pidgin to Da Max" which mentions a number of loanwords from Samoan, etc, and has its own usexes like "aalas dollahs means no mo' kala", and suggests the pronunciation of some words differs from English even when the spelling does not.) On one hand, it seems mostly intelligible, and there are substantially less intelligible dialects of English that we clearly treat as such; on the other hand, I see the argument above that it has a different origin. I have no strong opinion, but will say that on a practical level we do a poor job of covering most things we've merged into English, often having only a little vocabulary, little information on e.g. how verbs conjugate, and few usexes, since putting "wen da police guys wen come to da jail, dey neva find peopo" as a usex under police#English (as long as Hawaiian Pidgin is considered English) would probably be frowned on as inappropriately niche. - -sche(discuss)09:42, 12 September 2020 (UTC)Reply
Thank you for sharing your thoughts. I’d just like to mention that earlier, Metaknowledge said Bulgarian and Macedonian are separate languages since both lects can be easily identified in writing. But I do in fact believe Pidgin can be identified as well. Sure, not all Pidgin books spell things the same way but it generally keeps basic rules (such as “where” being “wea” and “the kind” being “da kine”). Pidgin writing as a result should be easily identifiable from American English. It’s kind of like Neapolitan, since it’s easy to identify it from Standard Italian but not everyone spells its words the same way, even among those who speak the same dialect of Neapolitan. Much like Pidgin, it’s not standardized but it still can be differentiated. — Coastaline (talk) 01:06, 13 September 2020 (UTC)Reply
@Mahagaja, @-sche In September, I requested that Wiktionary should add hwc as a language. However, the discussion died down and the status quo continued. Once again, I'd like to propose recognizing hwc as a separate code from en. The main counter-argument to my first post was that there's no agreed-upon way of separating pure Pidgin from Standard English. I actually disagree with this claim, because material in pure Pidgin does exist. The entire new testament was translated into Hawaiian Pidgin, and they're actually close to finishing the old testament. If you go to a Hawaii bookstore or search in Google Books, you can find more literature in pure Pidgin. The spelling tends to be uniform for most common words, with "where" being "wea" or "the" being "da". "Hamajang" and "da kine" are also uniform, and so are most Pidgin-exclusive terms. Spelling variations may occur but I don't think it stops it from being listed as a language. Wiktionary lists Neapolitan as one, yet spelling in that language is not standardized and many variations of the same word/pronounciation occur.
So what are your thoughts? Also, in case there's any confusion, Coastaline was my previous account. I switched to this current one since I recently recovered it. ― Haimounten (talk) 21:15, 16 October 2020 (UTC)Reply
@Haimounten: These discussions can sometimes take a while. I'd encourage you to merge this into the original conversation and ping people who contributed, rather than creating a new section, which may confuse people new to this conversation. I think I now lean toward reinstating the code, but I still want to hear others' positions. —Μετάknowledgediscuss/deeds21:44, 16 October 2020 (UTC)Reply
It's a grey area: much of the vocabulary is either identical to, or readily intelligible to a speaker of, standard English, and because Wiktionary is a dictionary (with entries for individual words) rather than a collection of long texts written in the lect (the way a hwc.Wikisource or even hwc.Wikipedia would be), differences in things other than vocabulary (such as grammar) are relatively less prominent / significant. However, I'm inclined to treat it as its own language (a) because of the argument that its origin suggests it's a distinct lect rather than a variant of English, and because treating it as its own thing would be consistent with how we treat Jamaican Creole or West African pidgin(s) as their own lects, and (b) because we're not meaningfully covering it as English, and I can't see us starting to: I can't see anyone putting up with wen da police guys wen come to da jail, dey neva find peopo as a usex on police, and we don't do a good job of indicating when a "general English" word is also used in a dialect, so while peopo might have its own entry with {{lb|en|Hawaiian Pidgin}}, it's unclear how anyone would know that e.g. police was used in Pidgin, since we wouldn't label it {{lb|en|US|ncluding|Hawaiian Pidgin|and|Midwest|and|NYC|and|New England|and|Southern US|and|AAVE|Canada|UK|Ireland|Australia|NZ|India}}. And we have someone who wants to add content in it and it does have its own ISO code, which we're no longer sure of our decision to retire. - -sche(discuss)05:26, 3 November 2020 (UTC)Reply
I'm still in favor of including hwc as a separate language, if for no other reason than that "wen da police guys wen come to da jail, dey neva find peopo" is not a sentence of English, not even a nonstandard variety. —Mahāgaja · talk12:48, 4 November 2020 (UTC)Reply
OK, I restored it, under code hwc and name "Hawaiian Creole" (since Ngrams suggests that's more common, and WP suggests it's also more correct, than Hawaiian Pidgin). I will say, though, there are things less "English" than "wen da police..." which we treat as en, like various dialects of England. (To pluck two examples out of the texts in the EDD, "A dooat like t'thowts a bin ower gyversum an hankeran eftre it", "Lükee zee tü 'er, 'er'th agot a rat! My eymers, 'ow 'er shak'th 'n!" These are far from the most extreme examples, just two examples I found in a quick search.) - -sche(discuss)17:57, 4 November 2020 (UTC)Reply
I would not necessarily be opposed to creating language codes for some traditional dialects of England as they are certainly as different from standard English as Scots and Yola are. —Mahāgaja · talk19:22, 4 November 2020 (UTC)Reply
The names of most languages which belong to the Northern branch of the Jê family as they are now are not the ones actually in use (either by linguists or by their speakers), and for a couple of languages even Glottolog (let alone Ethnologue) gets it right. I suggest:
renaming Suyá (suy) as Kĩsêdjê (of which Suyá should be an alias);
renaming Kayapó (txu) as Mẽbêngôkre (of which Kayapó and Xikrin are two dialects) — if this is deemed too diacritic-heavy, Mẽbengokre is also acceptable, but Mẽbêngôkre is the standard in the field nowadays;
creating a code for Proto-Northern Jê, which would be a daughter of Proto-Jê and an ancestor of all the aforementioned languages as well as Krahô and Canela.
I would also be happy to know if I can make any of the above changes myself without being an admin (I am new to Wiktionary so not everything is clear to me yet). Degoiabeira (talk) 01:19, 23 August 2020 (UTC)Reply
In answer to the last question, on a technical level you can't make these changes without being an admin or Template Editor, since they require changing Module:languages. But at least three admins have eyes on this now (User:Mahagaja, User:Metaknowledge, and me) so any changes that need to be made will get made after discussion here. :)
As for the suggested renames: AFAICT Suyá is much more common than Kĩsêdjê (or Kisedje) overall, though most occurrences refer to the tribe, and many of the occurrences which refer to the language are from 25+ years ago, as you brought up on Mahagaja's talk page. Even then, judging by google books:Suyá OR Kĩsêdjê OR Kisedje language, the two names seem roughly equally(?) common in books from the last 25 years. On one hand, Kĩsêdjê being the autonym speaks in favor of it; on the other hand, Suyá being historically more common speaks in favor of it. (From the perspective of ease of input, both names contain diacritics, so neither is per se easy to type.) The situation with Kayapó / Mẽbêngôkre is much the same.
For Parkatêjê, it seems like that name might indeed be most common, aided by the fact that there are so many alternatives ("Gavião Perkatêjê", bare "Gavião", "Gavião do Pará", etc) that no one name is that common AFAICT.
For Apinayé, judging by both books and scholarly journal articles, Apinayé seems to be the most common name even in recent sources, and it is also most common among the sources Glottolog lists which are specifically about the language and which use one spelling or the other in their names, so that is fine as-is and should not be renamed.
I will look into Pykobjê and Tapayuna.
Creating a code for Proto-Northern Jê seems like a good idea (there are reconstructions of it), especially if you will be adding entries in it or mentioning it in etymologies of other words. :)
I assume the other admins I mentioned above are following this, having participated in it when it was on the user talk page of one of them, but I'm going to re-ping User:Ungoliant MMDCCLXIV because I recall that pings must(?) be added in the same line as signatures in order to work. - -sche(discuss)07:48, 29 August 2020 (UTC)Reply
Not much more than what you’ve researched yourself. Ultimately each language will have to be researched individually to find its ideal canonical name. I do recall reading that our canonical name of a native Brazilian language was considered an ethnic slur by the people who speak it. I think it was one with a bird’s name, maybe Gavião or Urubu Kaapor.
I’ll add that I believe the main criterion in choosing a canonical name should be the name used in English-language media only. If multiple English-language names exhibit roughly the same amount of use, only then should its status as an autonym or not be considered as a “tiebreaker” (as should practical concerns such as the non-use of diacritics). — Ungoliant(falai)22:21, 29 August 2020 (UTC)Reply
Thanks for looking into it. The situation with Kayapó vs. Mẽbêngôkre is arguably not the same, however: the label Kayapó has been ambiguously used for designating both (i) a specific Mẽbêngôkre-speaking people (that is, to the exclusion of the Xikrin) as well as the dialect they speak and (ii) as a synonym of Mẽbêngôkre (that is, encompassing both the Kayapó and the Xikrin). In this sense, Mẽbêngôkre is preferrable because it helps avoid this ambiguity. Kayapó and Xikrin could be listed as varieties of Mẽbêngôkre (perhaps as different between each other as are Quebec French and Hexagonal French). Regarding the remaining comments, I realize I might be an interested party (I've been doing comparative and some descriptive work on Northern Jê for ~6 years and I've heard too many linguists' and native speakers' comments regarding their preferred ways to spell the names of these languages), so I guess I'll just wait for a consensus among unbiased Wiktionarians to form. I can also confirm that the Proto-Northern Jê reconstructions you might have seen out there are most probably mine and yes, I would be willing to add the respective entries (guess it's not a problem as long as the reconstructions are published in peer-reviewed outlets). Degoiabeira (talk) 18:43, 4 September 2020 (UTC)Reply
@Degoiabeira I got a bit distracted, but returning to this, I added a code for the family Northern Jê, sai-nje, and a code for the reconstructed language Proto-Northern Jê, sai-nje-pro. You may already be familiar with how to use them, I don't know, but if not: reconstructions in it would go the Reconstruction namespace (Reconstruction:Proto-Algonquian/askyi is an example entry in that language), and are mentioned in etymologies in the way you see in e.g. hàki, while the family could can be used if you don't know the precise language (as in sagamité). - -sche(discuss)10:12, 12 September 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Chuvantsy and Chuvan
Hi, I propose renaming Chuvantsy (xcv) to Chuvan. The pro of this renaming is the consistency between the WM projects (both the English wikipedia and Wikidata favour this name). Furthermore it's a more simple name, that is easily deductible from the native speakers' ethnonym. I don't personally see any downside to this, since all the primary (descriptive) sources on this language are written in Russian, and the term Chuvan seems to be just as, if not more, often in use as Chuvantsy. Thadh (talk) 23:43, 22 March 2021 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Adding Fogaha
It's time to start chipping away at the mess that is our Berber codes. Wikipedia considers the Sokna language to be spoken in three oases: Sokna, Fogaha, and Tmessa. According to van Putten, the claim of any Berber variety spoken in Tmessa is wholly unsubstantiated. That leaves Sokna and Fogaha, which were each documented by different Italian scholars using their problematic style of notation, but do seem to be significantly different nonetheless. I suggest that we add a code ber-fog for Fogaha (variously spelled "Fuqaha", "Foqaha"; the definite article is often tacked on, including by its only documenter Paradisi, who wrote "El-Fogaha", but this seems contrary to one's expectations when it comes to alphabetisation). —Μετάknowledgediscuss/deeds07:39, 10 August 2020 (UTC)Reply
Support. (I can even find such spellings as "Fojaha", "Al-Fojaha", "Fodjaha",...) Off of the immediate topic, but re Tmessa: I can find quite a few overviews which mention Tmessa, but AFAICT none provide detail about it. The Oxford Handbook of African Languages (Rainer Vossen, Gerrit J. Dimmendaal, 2020), page 285, says it is merely undocumented, not nonexistent, as part of a lengthy lament about Berber ("an unconvincing classification has been provided by Aikhenvald and Militarev (1991), [...] at many points this classification seems to be arbitrary. Moreover, at points it classifies dialects which are fully undocumented (e.g. Tmessa in Libya) or which are not Berber at all (e.g. Tadaksahak, which is Northern Songhay, and the Kufra oases, which are Teda-speaking). Unfortunately, some of the main lines of this classification have been taken over by the Ethnologue"). I reckon these changes are fine, though, since if any documentation subsequently emerges attesting Tmessa, it'll either show that it should be (re)included in an existing code or given its own. - -sche(discuss)21:28, 11 August 2020 (UTC)Reply
Here's a blogpost that explains the deal with Tmessa. As for the spellings with (d)j, they merely betray people who have made a faulty transcriptional assumption because they can't actually read the Arabic script... —Μετάknowledgediscuss/deeds22:17, 11 August 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Overhaul of Berber language codes
The current state of our Berber (Amazigh) codes is the ISO codes without any modifications made to their scheme, depsite its many problems. The Berber languages have comparable diversity to Romance, but specialists are distinctly uncomfortable with the idea of dividing up what is mostly a dialect continuum into over 25 languages, as Ethnologue does, despite the fact that nobody disputes this practice for Romance (even though on grounds of mutual intelligibility, it's hard to say why, say, Galician and Portuguese should be considered separate). For the most part, I think we should follow the Romance practice and be (mild) splitters; in cases like Tashelhit and Central Atlas Tamazight, you wouldn't gain much by merging (as most entries would still be on separate pages), and the remarkable internal dialectal diversity in these "languages" means that good coverage will still require a great deal of dialect-tagging. I have worked through the Berber classification used by Kossmann (2013) below, using his numbered "blocks", with my notes on changes to our scheme that I recommend.
Two languages, Zenaga [zen] and Tetserret [tez], both separate from the continuum and each other. No changes needed.
The Tuareg group, which has a macrolanguage code [tmh], but also codes for Tahaggart [thv], Tawallammat [ttq], Tayart [thz], Tamasheq [thq]. Tamasheq is meant to encompass both Adagh and Taneslemt, and the sole sedentary Tuareg lect, Ghat (spoken in Libya) is missing a code. I struggled to find data on mutual intelligibility; Adam (2017) indicated that speakers from Ghat had a lot of trouble understanding non-Libyan dialects, but that this might be a recent phenomenon. There is a written Tuareg standard using Tifinagh, although it does not indicate vowels and does not seem widely used any more. Clearly, the current state is redundant, and I think collapsing into a single "Tuareg" would be workable, but perhaps not ideal.
South-Central Moroccan group, Tashelhit [shi] and Central Atlas Tamazight [tzm], both written in Tifinagh. There is also a poorly defined Judeo-Berber [jbe], which I don't want to get into (but reminds me of the issues with Judeo-Arabic). These are all fine, but ISO also gave a politically-motivated code for "Standard Moroccan Tamazight" [zgh], which is a redundant literary koiné that I have already posted a request to delete above.
Northwestern Moroccan group, Ghomara [gho] and Senhaja de Srair [sjs]. No changes needed.
The Zenatic group, and here is where things get messy.
The existing codes in the West are Tarifit [rif], Chenoua [cnu], Tachawit [shy]. Missing are Iznasen (sometimes considered "Eastern Riffian", so presumably merged under [rif]) and Eastern Central Atlas Tamazight (a kind of contact language between [tzm] and [rif], presumably merged under [tzm]).
In the Moroccan and Algerian oases, Tagargrent [oua], Tugurt/Temacine [tjo], Taznatit [grr], Tumzabt [mzb], and Tidikelt [tia] (which is almost undocumented!) all have codes, thus excluding Sud-oranais (principally the Figuig dialect, which has a grammar). Mutual intelligibility in this cluster of dialects is known to be high, and this seems ridiculously oversplit; I would recommend one code, although I'm not sure what to call it. Ethnologue calls it "Mzab–Wargla", which are only two of the dialects and is obviously not great, while Kossmann calls it "Northern Saharan oasis" Berber, which is highly clunky, and my "North Saharan Berber" is a protologism.
In the East, there is a single code for Sened [sds], an extinct dialect in Tunisia, and no codes for the extant Tunisian varieties (principally Djerbi), nor for the closely related Zuwara dialect over the border in Libya. I recommend that [sds] be repurposed as "Tunisian Berber", and a new code be added for Zuwara (also known as "Zuaran"), which has a grammar by Mitchell; this could be considered oversplitting, but it would be odder for "Tunisian Berber" to have entries that are almost all from Libya.
The remaining languages of the East belong to various blocks, and are mostly covered appropriately: Kabyle [kab], Nafusi [jbn], Siwi [siz], Sokna [swn], Ghadames [gha], and Awjila [auj]. The only one missing is (the presumably extinct, and thus finite) Fogaha, which I recommended for a new code in its own section above.
Feedback would be appreciated, especially on the naming of North Saharan Berber. Note that I intend to give etymology-only codes for all the dialects that end up getting merged. —Μετάknowledgediscuss/deeds22:04, 10 August 2020 (UTC)Reply
Re North Saharan Berber: I think naming it "Northern Saharan Berber" is OK; we're not inventing the grouping, it's a descriptive name, and it's similar to and simply trims an excess word off of an existing name used in literature. - -sche(discuss)21:06, 11 August 2020 (UTC)Reply
Reviewing Tuareg further, I see that Heath (2005) refuses to take a stand: "it is very difficult to decide whether we are dealing with a single "Tuareg" language (with many dialects), or two or more languages (each with some internal dialectal variation)." That said, he endorses the four-way distinction (where "Tahaggart" should be renamed to "Tamahaq" to include the Kel Ajjer), supports the bundling of Tamasheq, and makes no mention of Ghat, presumably lumping it into Tamahaq as well. The 4 lects with codes all have good dictionaries (ttq and thz share a dictionary, but all forms are marked for which lect they belong to), and 3 of the 4 have grammars, so they won't be hard to sort out or attest. As a result, I now definitively lean toward retiring the macrolanguage and keeping the split codes. —Μετάknowledgediscuss/deeds06:26, 26 August 2020 (UTC)Reply
I still haven't dealt with this, and I found reason to merge Tuareg instead of splitting it. Sudlow (2008) says: "Tamasheq has split into three major dialects which the Tamasheq themselves refer to as ‘sha’, ‘za’ and ‘ha’, or Tamasheq, Tamajeq, and Tamahaq." His grammar itself supports the utility of that view, as it treats two dialects spoken together (often in a mixed form) in Burkina Faso, which would have to go under two different language codes in the splitter scenario. We would have to dialect-tag our entries carefully, of course. —Μετάknowledgediscuss/deeds23:33, 8 March 2021 (UTC)Reply
Kossmann, who is a leading Berberist, offers this judgement on Tuareg: "In spite of important differences, it would be exaggerated to consider these variants distinct languages." This makes me feel very secure in merging them, so I will finally get all of this in working order soon. —Μετάknowledgediscuss/deeds19:41, 24 August 2021 (UTC)Reply
I have made all the changes recommended in this section. I chose [mzb] as the code for Northern Saharan Berber, and preserved the Tuareg dialect codes as etymology-only codes (adding an etymology-only code [ber-ght] for Ghat). I added Zuwara as [ber-zuw]. I also merged Judeo-Berber into Tashelhit, following our practice with Judeo-Arabic (see Category:Judeo-Berber). —Μετάknowledgediscuss/deeds05:37, 25 August 2021 (UTC)Reply
So long as the child entries are on the Sanskrit entry, I don't mind if they're deleted, not because they're unsourced, but because Proto-Dardic is controversial. --{{victar|talk}}06:33, 6 September 2021 (UTC)Reply
To clarify, I did not nominate them for deletion simply because of them being unsourced, but because of the language: seeing that some Proto-Indo-Aryan entries have also been deleted lately. ·~dictátor·mundꟾ06:56, 7 September 2021 (UTC)Reply
@Kutchkutch: As far as the descendants of categorized Proto-Dardic entries are concerned, there are no more any instances of a Sanskrit entry wanting Dardic descendants. We have thus far spotted 4 Proto-Dardic entries ([1], [2], [3], [4]), and all of them are prepared for deletion. ·~dictátor·mundꟾ13:57, 6 September 2021 (UTC)Reply
@Metaknowledge: The descendants had already been moved one month ago, but I am not sure if we should get rid of the lang code, as it’s used in the Descendants section. @Bhagadatta, Kutchkutch, Victar: While the Proto-Dardic entries are to be deleted, should we keep the reconstructed forms in the Descendants section? @Svartava2 has removed the terms without prior discussion. ·~dictátor·mundꟾ06:48, 9 October 2021 (UTC)Reply
@Inqilābī, Svartava2: Even if the protolanguage code is removed and the reconstructed forms in the Descendants section are not needed, the Dardic label would still be necessary to separate those languages from other Sanskrit descendants. Kutchkutch (talk) 13:23, 9 October 2021 (UTC)Reply
@Metaknowledge: Proto Dardic did most likely exist but there aren't enough published sources to cite while creating reconstructions. The study in the area is quite murky, with some authors even confusing them with Nuristani languages (Nuristani languages are geographically close to the Dardic languages but form a separate subfamily within Indo-Iranian unlike Dardic which is part of the Indo-Aryan subfamily). So yeah, Wiktionary would presently lose nothing by removing the inc-dar-pro code. It's like having Proto-Bengali-Assamese or Proto-Tamil-Kannada. If some good reliable sources that deal in Proto-Dardic become available later on, we can always add the code back. -- 𝓑𝓱𝓪𝓰𝓪𝓭𝓪𝓽𝓽𝓪(𝓽𝓪𝓵𝓴)13:02, 23 October 2021 (UTC)Reply
@Kutchkutch for now it's ok to change all these to Dardic:, so that for is little change (just Proto- and recons term/[Term ?] removed) and because we wouldn't have to change the ordering like if there is "{{desc|inc-dar-pro}}" before Gujarati we'd have to move the, say, Kashmiri descendant after Gujarati one in alphabetical order; hence its easier to be done by bot (most of these replacements were done by NadandoBot). From now it's up to the editor what to do; like with Bengali-Assamese: — diff and * {{desc|as}} * {{desc|bn}} * {{desc|syl}} are both valid and neither is wrong. However I do think that if there is only one desc like it would be better to show that one separate without label Dardic:. Svartava2 (talk) 15:09, 25 October 2021 (UTC)Reply
Unattestable or SOP language name fudged by the for-profit language database Ethnologue; the linked Wikipedia pages will always stay stubs. The Verkehrsanschauung sorts this under Algerian Arabic. There is a continuum with Moroccan Arabic. Pinging @Fenakhay following our considerations at Talk:زرودية; nobody has ever thought such a category. One can attest the term “Saharan Arabic” but this is of course not meant to denote one language. As distinguished from Hassānīya Arabic one could need codes eastwards for some dialects spoken in the Sahara and Sahel, like for Malian and Nigerien Arabic – Chadian Arabicwe have as shu. Fay Freak (talk) 13:31, 5 April 2020 (UTC)Reply
I think the idea here is that Algerian Arabic as spoken in Algiers belongs on a continuum that speakers in the southern and western desert of Algeria (inclusing many who are not ethnically Arab) do not belong on. I don't see the use of this code, though, because that speech definitely counts as Hassaniya. (As for the other countries on the fringe, we seem to count Nigeria and Cameroon under Chadian Arabic, which is less than ideal, but then again, I think that our whole system needs to be more like Chinese to be functional at all.) —Μετάknowledgediscuss/deeds18:18, 5 April 2020 (UTC)Reply
Latest comment: 3 years ago2 comments2 people in discussion
The following discussion has been moved from Wiktionary:Requests for moves, mergers and splits ([[Special:PermanentLink/62589227#Merge_[aec]_“Saidi_Arabic”_into_[arz]_“Egyptian_Arabic”|permalink]]).
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Hardly attestable language name. Occurrences for it turn out as “Saʿiidi Arabic” for example. “Upper Egyptian Arabic” (I don’t see use of “Upper Egypt Arabic”, but this is a detail) is sometimes posited, variously ill-defined because of superstratum influence from the Nile Delta, but this is usually not accepted as distinct language and can well be an sum-of-parts term; Arabic Wikipedia clearly says Ṣaʿīdīy Arabic is “within the Egyptian Arabic dialects”. There are but some isoglosses dividing Arabic-speaking Egypt latitudinally, but so one can shed urban and bedouin speech and Muslim and Christian speech. English Wikipedia says “the realisation of /q/ as [ɡ] is retained (normally realised in Egyptian Arabic as [ʔ]” but this is the speech of the desert vs. the speech in the largest cities and [ɡ] for ق(q) can be found in northern Egyptian bedouins. Most natives of Upper Egypt would use the code for Egyptian Arabic, and I would too sometimes with label “Upper Egypt” or more specific labels. I have created the single entry in “Saidi Arabic” and only because there was this code. Fay Freak (talk) 13:54, 5 April 2020 (UTC)Reply
Latest comment: 3 years ago6 comments2 people in discussion
The following discussion has been moved from Wiktionary:Requests for moves, mergers and splits ([[Special:PermanentLink/62589227#Merge_[ayn]_“Sanaani_Arabic”,_[acq]_“Ta'izzi-Adeni_Arabic”,_[ayh]_“Hadrami_Arabic”_into_a_new_“Yemeni_Arabic”|permalink]]).
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Merge [ayn] “Sanaani Arabic”, [acq] “Ta'izzi-Adeni Arabic”, [ayh] “Hadrami Arabic” into a new “Yemeni Arabic”
The categories are ill-defined; the second one is unattestable. Yemen is a dialectologically complicated area – the material is even partially collected outside of Yemen, as for instance the Dictionary of Post-Classical Yemeni Arabic 1990 is surveyed in Israel – with many isoglosses and the number of distinct languages is multiplied at will if one is oblivious of SOP designations of lects so for this reason, apart from nobody being able to safely use these categories, it is not wise to distinguish at the L2 level already. If you look into the Dialect Atlas of North Yemen and Adjacent Areas published by Brill four years ago and probably also Saint Elbakyan you will forget those categories and won’t be able to map content onto the codes. Apart from that the distinction is inconsistent with the fact that we have a code for “Judeo-Yemeni Arabic” jye – as if Jewish speech in Yemen were one language while Muslim speech were three! I just mention here that I find “Judeo-Arabic” and its sub-languages suspicious, it could all be only Classical Arabic or Arabic dialects with some peculiarities written in Hebrew script, remaining from a time when Wiktionary did not use {{spelling of}} for this purpose. The Jews might deal with it themselves. Fay Freak (talk) 14:23, 5 April 2020 (UTC)Reply
I pinged you in a separate discussion above about Judeo-Arabic. As for Yemen... it is true that the codes are oversplit, but it is also true that the WAD attests to the fact that when Hejazi and Omani speakers disagree about a word, chances are that North Yemen agrees with the Hejaz or Egypt, and that South Yemen agrees with Oman. The existence of Piamenta's dictionary gives me hope for treating Yemen as a unit, but I don't know how he does it (do you have a PDF?). —Μετάknowledgediscuss/deeds18:33, 5 April 2020 (UTC)Reply
I find this “South Yemen” and “North Yemen” part difficult and think that one needs an elevation profile of Southwest Arabia to dissect the Arabic dialects of and close to Yemen. Fay Freak (talk) 18:41, 5 April 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Yavapai vs. Havasupai-Walapai-Yavapai
We are currently treating Yavapai nai-yav as a separate language from Havasupai-Walapai-Yavapai yuf, but some of our yuf entries (e.g. 'ha) are tagged as Yavapai. Is there any difference between nai-yav and the Yavapai dialect of yuf? If not, can we merge nai-yav into yuf? And if we do want them separate, can we rename yuf "Havasupai-Hualapai" (as per Wikpiedia) or "Havasupai-Walapai" rather than keeping a different language's name in there? We only have one entry for nai-yav, namely ke#Yavapai. @-sche seems to be the only person who's dealt with these languages in the past. —Mahāgaja · talk11:07, 30 August 2021 (UTC)Reply
Interesting; I dug into the edit histories, and I added Yavapai in February 2016, at which time I also added a word. I added Walapai-and Yavapai-specific content under the trinitarian name Havasupai-Walapai-Yavapai a couple months later; I must've been looking to add the Walapai content, saw we had a code for Havasupai-Walapai-Yavapai, since Stephen G Brown had previously added Havasupai-Walapai-Yavapai content like T:yuf-personal pronouns and entries forthose pronouns without dialectal labelling, and forgot about the Yavapai-specific code. Wikipedia has Havasupai-Walapai and Yavapai as distinct but similar languages (a 1978 United States Indian Claims Commission decision says "the Walapai and Havasupai [...] spoke a language and had a culture very similar to the Yavapai"); OTOH, Thomas Sebeok's Native Languages of the Americas, volume 1 (2013), page 467, cites them as an example of ethnic groups that fought "even though [...they] spoke the same Yuman language". Christopher Moseley's Encyclopedia of the World's Endangered Languages (2008) calls the language "Upland Yuman" ("Upland Yuman is a Yuman language"), spoken by the "Hualapai, the Havasupai, and the Yavapai, the last traditionally divided into four regional subtribes. Each community speaks a distinct variety, with the Yavapai varieties forming a well defined dialect, although all varieties are mutually intelligible with little difficulty." I suppose they should be merged. - -sche(discuss)18:53, 30 August 2021 (UTC)Reply
Knaanic hasn't been definitively proven to exist as a distinct language yet, and its few attestations (and entries here) seem indistinguishable from standard Slavic languages in Hebrew transcription. Here's a pretty good article by Dovid Katz on the topic. airy—zero (talk) 03:22, 4 November 2021 (UTC)Reply
Knaanic is a useful holding category for the scattered attestations of West Slavic words in Hebrew script, which have to be included in the dictionary some way or another. This discussion doesn't belong at RFDO, because there is no chance we will delete all those entries, which pass CFI and definitely belong here. The question is whether we want to merge Knaanic into another language code, presumably Old Czech, although in all honesty, I rather doubt that all the Knaanic attestations can really be called Old Czech in good faith. The difficulty of deciding where they ought to go, and whether people would even think to look for them there, is why Knaanic exists as a separate language here in the first place. —Μετάknowledgediscuss/deeds06:11, 4 November 2021 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Adding Early Modern Korean (ko-ear) as an L2
(Notifying TAKASUGI Shinji, Atitarev, HappyMidnight, Tibidibi, Quadmix77, Kaepoong, Mahagaja, Metaknowledge): @Fish bowl, @Solarkoid, @Lunabunn, @Echo Heo (feel free to tag anyone else). Currently, Early Modern Korean (EMK) is listed as an etymology-only language, limiting its usage as a header and leading all EMK terms to be listed under the "Korean" header with an "Early Modern" label. However, through several discussions on the matter (thank you @ Tibidibi), it's been made more and more clear, that it needs to be separated out from Modern Korean in order for its coverage to be done well. Here are a few reasons (feel free to make any corrections if needed):
Early Modern Korean used a rather different orthography compared to Korean with letters like ㅼ(st) and ㆉ(yoi) becoming obsolete after that time period, leading Modern Korean speakers without training in the language to struggle to comprehend it.
There are terms that are solely used in EMK such as 툐상(tyosang) that deserve to have their own space and coverage rather than thrown under the "Korean" header and left alone.
Currently our etymologies are unclear in terms of showing the development of words from Middle Korean (MK) to EMK to Modern Korean (MoK) such as in ᄃᆞ리〮다〮(tòlítá), or often skipping EMK altogether due to the lack of clarity. Additionally, with words such as 구무(kwumwu), we'd be able to show more clearly the relation between EMK terms and their dialectal descendants.
Separating EMK out will allow us to more easily create pronunciation & transcription modules for the language (such as what was done with Jeju in Module:jje-pron & Module:jje-translit), rather than the somewhat inconsistent transcriptions that have been used in the past or attempting to include all the EMK use-cases within the current system.
Overall, having Early Modern Korean as its own L2 would greatly improve its coverage if done well, make life easier for the Koreanic editor community, make our content clearer and more precise for readers, and finally put Wiktionary in line with Koreanic Linguistics as a whole. AG202 (talk) 09:06, 23 January 2022 (UTC)Reply
@AG202Strong support on all points. As a way of division, I suggest that:
The split between Middle Korean and EMK is generally a hard line at 1600, but the few texts that use MK-only characters may be treated as MK.
The split between EMK and modern Korean is more fluid.
Texts made by foreign learners of the language in the late nineteenth century and onwards are considered modern Korean, despite being written in an EMK orthography, because they are often the first attestations of archetypally modern Korean words not attested in EMK, such as 때문(ttaemun) and 언니(eonni).
개화기 국어 (開化期國語, gaehwagi gugeo), the language of “modernizing” texts from the period from c. 1890 to 1910, is considered modern Korean despite having EMK orthography, because they are the first attestations of many of the Japanese orthographic borrowings (wasei kango) totally absent in EMK but a defining feature of modern Korean vocabulary.
“Traditional” texts from the 1890–1910 period, such as personal letters of provincial people less affected by Westernizing trends or traditional novels untouched by Western and Japanese influence, are still considered EMK. The quotation at 섬섬옥수(seomseomoksu), despite the late date, is a typically EMK material and would hence be quoted at Early Modern Korean셤셤옥슈(syemsyemwoksyu).
Being that it's been three weeks, the only active EMK editor is in strong support, and there hasn't been any opposition (along with tagging all of the Korean editors in the working group), I'd like to close this as RFM-Split. @Erutuon, @Surjection, if you have the time, please. Edit: Apologies for forgetting to sign, and thank you @ J3133, for pinging them for me. AG202 (talk) 01:07, 13 February 2022 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Formosan Kulon-Pazeh (uun) split to Kulon (uon) and Pazeh (pzh)
According to this ISO change request (Kulon-Pazeh Change Request Documentation), Kulon (uon) and Pazeh (pzh) are now separate languages. How do we update Wiktionary to add data specifically for pzh from now on (given that data is nonexistent for the long-extinct Kulon language). I'm in the process of adding lemmas and example sentences for several of the western Formosan languages. Kangtw (talk) 19:57, 3 April 2022 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Moving Finno-Ugric families to Uralic families
We don't recognize Finno-Ugric as a valid family; just Uralic. Hence urj is a valid code, but fiu isn't. Nevertheless, we're using fiu- as the prefix for four branches: fiu-fin for Finnic, fiu-mdv for Mordvinic, fiu-prm for Permic, and fiu-ugr for Ugric. I propose we use urj- for these instead, thus moving as follows:
fiu-fin → urj-fin
fiu-mdv → urj-mdv
fiu-prm → urj-prm
fiu-ugr → urj-ugr
At the same time, we should move the codes for the corresponding protolanguages:
fiu-fin-pro → urj-fin-pro
fiu-mdv-pro → urj-mdv-pro
fiu-prm-pro → urj-prm-pro
fiu-ugr-pro → urj-ugr-pro
as well as the code for the etymology-only lect Proto-Finno-Permic:
That seems like a lot of disruption for a small theoretical benefit: we've always used codes like aus, cau, nai and sai that we don't recognize as families for making exception codes, so it's not a huge violation of our naming logic. In this case, though, it looks to me like we don't recognize fiu more because it's too much like urj, not because it's invalid, per se (though I don't know a lot on the subject). We do have gmw-fri rather than gem-fri, for instance. Of course, I'd rather follow those who actually work in this area- especially @Tropylium. Chuck Entz (talk) 20:59, 23 December 2016 (UTC)Reply
I have no opinion on moving around family codes either way (it doesn't seem they actually come up much, whatever they are), but if we start moving around the proto-language codes, I would like to suggest simple two-part codes. Proto-Samic and Proto-Samoyedic are already smi-pro and syd-pro, so is there any reason we couldn't make do with e.g. fin-pro, fpr-pro, ugr-pro etc.?
Also, as long as we're on this topic, at some point we are going to need the following:
Proto-Mansi: (ugr-/urj-?)mns-pro
Proto-Khanty: (ugr-/urj-?)kca-pro
Proto-Selkup: (ugr-/urj-?)sel-pro
No rush though, since so far we do not even have separate codes for their subdivisions. The only distinction that comes up in practice is distinguishing Northern Khanty from Eastern Khanty (Mansi and Selkup only have one main variety that is not extinct or nearly extinct). --Tropylium (talk) 10:53, 24 December 2016 (UTC)Reply
There is, actually, a reason: our exception codes are designed to avoid conflict with the ISO 639 codes, so they start with an existing ISO 639 code or a code in the qaa-qtz range set aside by ISO 639 for private use. fin is one of the codes for the Finnish language. fpr and ugr are apparently unassigned- for now. As for the three proto-language codes, those don't need a family prefix because they already start with an ISO 639 code. Chuck Entz (talk) 13:06, 24 December 2016 (UTC)Reply
The only codes in the qaa-qtz range we actually use are qfa as a prefix for otherwise unclassified families and qot for Sahaptin (a macrolanguage that wasn't given an ISO code of its own), right? —Aɴɢʀ (talk) 07:55, 26 December 2016 (UTC)Reply
Right. And I didn't notice that we used qot although it is not an ISO code; it seems we followed Linguist List in using it. For consistency, I suggest changing it to fit our usual scheme, so nai-spt or similar (nai-shp is already in use as the family code). - -sche(discuss)05:10, 27 December 2016 (UTC)Reply
Support, for consistency. fiu is different from nai, because fiu is [a supposedly genetic grouping which is] agreed to be encompassed by a higher-level genetic family which also has an ISO code (urj), and that code can be used if we drop fiu. nai and sai are placeholders rather than genetic groupings, and they're useful ones, because If we dropped them we'd had to recode everything as qfa- (and might conceivably run out of recognizable/mnemonic codes at that point). - -sche(discuss)05:10, 27 December 2016 (UTC)Reply
I've recoded Sahaptin and all the Finno-Ugric lects except fiu-fin-pro which requires moving a lot of categories, which I will get to later. - -sche(discuss)17:20, 8 March 2017 (UTC)Reply
I can volunteer to carry out the move from fiu-fin-pro to urj-fin-pro if there's still consensus that it should be done (which there is, as far as I can tell). — SURJECTION/ T / C / L /09:19, 12 January 2023 (UTC)Reply
Now there are lots of module errors in Cat:E as a result of these language code changes. It might be easiest to fix them by bot. @DTLHS, what do you think? — Eru·tuon22:45, 9 March 2017 (UTC)Reply
@DTLHS: Three incorrect language codes remain, I think: fiu-ugr, fiu-fpr-pro, urj-fin-pro. I couldn't figure out what fiu-fpr-pro should be; it seems to refer to Proto-Finno-Permic, but I searched various language data modules and didn't find a match. Is there someone who can look through and fix the remaining module errors that relate to incorrect language codes? @Tropylium, @Angr, @-sche? — Eru·tuon04:52, 11 March 2017 (UTC)Reply
The categories were easy to deal with: you just change the {{derivcatboiler}} to {{auto cat}} and the template plugs in the correct language code, if it exists. That also makes it a quick way to check whether there is a correct language code. by the time I finished that, there were only a dozen or so entries left in CAT:E due to everyone else's efforts, so I finished off the remainder by hand. It would have been easier if there hadn't been hundreds of other module errors cluttering up CAT:E- yet another reason for you to be more careful. Chuck Entz (talk) 21:10, 11 March 2017 (UTC)Reply
I apologise for not catching and fixing those uses at the time I renamed the codes. I searched all pages on the site for each of the old codes, and some pages turned up [including pages where the codes were used inside some templates, and I fixed those pages], so I forgot to also do an "insource:" search to catch other uses inside templates like {{m}}. We so rarely change language codes compared to changing language names. - -sche(discuss)23:36, 11 March 2017 (UTC)Reply
prm-koo for Old Komi (Writing systems: Perm and Cyrs; Family: urj-prm)
prm-kya for Komi-Yazva (Writing system: Cyrl; Ancestor: prm-koo; Family: urj-prm)
Furthermore, the newly made prm-koo should be given as the ancestor of Komi-Permyak (koi) and Komi-Zyrian (kpv). The writing system Perm should be removed from both these languages.
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Mari phylogeny
It seems Eastern Mari (chm) is treated as an ancestor of Western Mari (mrj). Historically, that makes little sense though, because both are different written standards of the same continuum, so this is about akin to setting Nynorsk as a descendant of Bokmål, Komi-Permyak as a descendant of Komi-Zyrian or setting Livvi as a descendant of Karelian. I think we should set both as direct descendants of Proto-Uralic (urj-pro), or, alternatively, create a code for Proto-Mari. I know too little about the history of Mari languages to say anything sensible on that topic. Anyway, pinging some editors that might be interested, @Tropylium, Atitarev, -sche. Thadh (talk) 18:03, 20 February 2022 (UTC)Reply
Yeah, that was the minimum requirement. I thought maybe if there were any solid Proto-Mari reconstructions, we could add a code for that, but seeing as nobody answered it's better to leave that for the future. Thadh (talk) 07:18, 1 June 2022 (UTC)Reply
@-sche, Tropylium, Mahagaja Would anyone oppose moving chm to mean "Mari" (as a family), creating chm-pro for Proto-Mari and adding the iso-intended code mhr for Eastern Mari? If I understand correctly, @Surjection has volunteered to handle the bot jobs if needed. Thadh (talk) 18:52, 7 November 2022 (UTC)Reply
Literature on the form of common Mongolic spoken between the 13th and 16th centuries increasingly uses the term "Middle Mongol" rather than "Middle Mongolian". This is primarily because it was the ancestor to several extant Mongolic languages, not all of which are spoken by people who we would usually call Mongolians (though they are sometimes considered part of the Mongol people - especially historically). For example, Buryat (spoken in Buryatia, just north of Mongolia) and Kalmyk (spoken in Kalmykia, around 4,000km to the west of Mongolia). There are several others.
I also feel that this would (slightly) reduce the faulty implication that Mongolian is the "primary" descendant of Middle Mongol, too. To give a comparison, using the name "Middle Mongolian" is roughly equivalent to using the name "Old Russian" for Old East Slavic: obviously less than ideal.
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Rename Ulukwumi (ulb) to Olukumi, add Yoruba as an ancestor for Lucumí (luq), & the case of Itsekiri (its)
As we're moving towards the creation of a more robust coverage of Yoruboid languages, see: agutan for a rough start, we'll need some changes made earlier rather than later.
Lucumi (luq) should be changed to have its name have the accent mark on the "i" becoming "Lucumí", and then being that it is a liturgical language derived from Yoruba, it should have Yoruba as its ancestor.
Additionally, there's the case of Isekiri (its), which looks as if it's spelled "Itsekiri" the most in English, including by the main driver of teaching the language nowadays "Learn Itsekiri". The spelling "Isekiri" seems to come from the spelling in the language which involves a ṣ, but being as that does not exist in English, it's often been adapted as a "ts". At the moment, I am leaning towards the "Itsekiri" spelling, though I would like more input if possible.
Those are the changes that we need for now, and hopefully we can continue increasing the coverage of these languages. AG202 (talk) 20:30, 6 July 2022 (UTC)Reply
Sorry ooo, I don't usually pay attention to the notifications from Wiktionary and Wikipedia, I always think that it'll be difficult to use. About Ìṣẹkírì, in my opinion I think that you write it as Ìtsẹkírì because it isn't normal for them to write it with "ṣ" like us Yorùbá do. Egbingíga (talk) 02:44, 16 September 2022 (UTC)Reply
I went ahead and changed Olukumi, per the persuasive evidence above. Not sure whether Lucumí really needs the accent; ngrams might suggest Lucumi was recently overtaken by Lucumí but it's not as overwhelming as with Ulukwumi not even getting enough hits to plot. - -sche(discuss)23:25, 23 July 2022 (UTC)Reply
@-sche Thank you for the changes! Hmmm for Lucumí, I'm still leaning towards including the accent, but I see your point. I do wish that there were more folks active here to comment on this though. AG202 (talk) 19:52, 13 August 2022 (UTC)Reply
Thanks! Lucumi → Lucumí is also now done. Closing this discussion since I think everything was covered, and will archive in a week. AG202 (talk) 03:11, 1 June 2023 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Canonical name of "fan"
Our canonical name for fan is "Fang (Guinea)", which is unfortunate since it isn't spoken in Guinea. It's spoken primarily in Gabon and Equatorial Guinea. I'd recommend calling it "Fang (Gabon)" since there seem to be more speakers in Gabon than in Eq.G. and since the name of Gabon is shorter. —Mahāgaja · talk 11:53, 17 October 2019 (UTC) I'd recommend calling it "Fang (Equatorial Guinea)" since according to Ethnologue there are more speakers in that country than in Gabon. —Mahāgaja · talk12:09, 22 October 2019 (UTC)Reply
The fact that there are even some speakers in Cameroon (according to Wikipedia) but it's not the same as "Fang (Cameroon)" is icing on the confusion-cake... and they're both Bantoid languages, so disambiguating by family doesn't help, and both spoken in Central Africa, so we can't disambiguate by mere region as we sometimes do. - -sche(discuss)16:00, 24 October 2019 (UTC)Reply
Ah, true; I saw that Wikipedia classified Fang language (Cameroon) as "Western Beboid" but caveated that it was not necessarily a valid family, but if Beboid overall is valid, then that works and is clearer (IMO) than picking just one of the countries to list. I would support renaming them in that way. The only other languages that come to mind which are disambiguated by family are "Austronesian Mor" and "Papuan Mor" both spoken in West Papua), which are mentioned on WT:LANG; probably we should change those to "Mor (Austronesian)" and "Mor (Papuan)" for consistency. - -sche(discuss)18:00, 25 October 2019 (UTC)Reply
I was going to rename Austronesian and Papuan Mor to use the 'parenthetical, postpositive' naming format for consistency with the Fangs and with how languages with country-name disambiguators are named, but I notice we also have e.g. Austronesian and Papuan Gimi, Austronesian and Sepik Mari (and some other Maris), and several other such languages, and I don't have time to rename all of those, so consistency will have to wait. (There is also "Sepik Iwam", but it appears to actually get called that, to distinguish it from the other Sepik language called Iwam which is spoken in the same place and belongs to the same Iwam subfamily of Upper Sepik. Confusing!) - -sche(discuss)08:41, 28 November 2019 (UTC)Reply
The reason I haven't archived this is that several other languages, mentioned above, still need to be renamed to fit the format of the other languages. I or someone else just need(s) to find the time... - -sche(discuss)01:48, 6 August 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Proto-Witotoan and Proto-Huitoto-Ocaina
Currently, the way they are treated makes no sense: Proto-Huitoto-Ocaina is handled as the parent of Proto-Witotoan and nothing else. In our word list, Proto-Witotoan is used to denote Proto-Bora-Witoto, which is a macroreconstruction that is very speculative. I propose we merge these into one language, Proto-Witotoan, which seems to be the more common term.
If I recall and understand correctly (?) the current setup is based on your request here, where I noted the issue of the broader/older grouping having been set in some entries as the child of a smaller/younger grouping. No objection to changing it. I see that a handful of Murui Huitoto entries use Proto-Witotoan in their etymologies, but these should be unaffected by removing Boran from its scope. - -sche(discuss)16:53, 9 May 2023 (UTC)Reply
Oof, I didn't even recall that discussion - I guess I was a bit too hasty, sorry. If there are no further objections, I'll just manually remove and/or change the reconstructions from the Murui Huitoto etymologies and fix this. Thadh (talk) 17:05, 9 May 2023 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
West African Pidgin English varieties
Ethnologue has assigned codes to some but not all of the varieties of West African Pidgin English, and we in turn have incorporated some (e.g. pcm) but not all (e.g. not gpe) of those codes. As WP notes, the "contemporary English-based pidgin and creole languages are so similar that they are sometimes grouped together under the name 'West African Pidgin English'" (a name which also denotes their predecessor which developed in the 1700s). WP's examples are illustrative, particularly in that its Ghanaian and Nigerian Pidgin English examples are identical. I propose to merge at least the following three varieties into wes, renaming it "West African Pidgin English":
Ghanaian Pidgin English (gpe)
Nigerian Pidgin English (pcm)
Cameroonian Pidgin English (wes)
We could also discuss whether or not to merge Sierra Leone Krio (kri, which WP notes its often mistaken for English slang due to its similarity to English, but which has a somewhat distinct alphabet), Pichinglis / Fernando Po Creole (fpe), and Liberian Kreyol / Liberian Pidgin English (lir). - -sche(discuss)21:11, 11 August 2015 (UTC)Reply
The question is a very complex one. Firstly (but of least importance), scholars are divided on which lects have creolised and which have not, but it is generally agreed upon that at least some of the language you mentioned are not pidgins, which would make the name "West African Pidgin English" somewhat of a misnomer (the more neutral name "Wes-Kos" have been suggested as an alternative, but even linguists haven't fully adopted it). Secondly, all these lects are remarkably similar on a lexical level, but that's unsurprising; after all, they resulted from separate but very similar language contact events, and then probably modified each other (one scholar posits that Krio and Cameroonian Pidgin English relexified each other to some degree after pidginisation). The similarities are also obscured by the fact that there is nothing close to an agreed orthography for most of these, and pronunciation does differ a bit across West Africa. Linguistically, I'd probably merge them all, but practically that may not be the best decision. I know we have entries in pcm, but probably next to nothing for the rest, and if somebody wants to add them, given how each lect is very neatly assigned to a certain West African country, at least it won't be confusing for them to do so. Conclusion: the literature is schizophrenic, the lects mutually intelligible, and the existing situation remarkably unproblematic. Therefore I abstain. —Μετάknowledgediscuss/deeds21:19, 16 August 2015 (UTC)Reply
I see no evidence that this exists as a separate language, and move that it be merged with tr. The literature which references it seems to describe the dialect of Turkish which may be spoken by Gagauz people in the Balkan Peninsula. —Μετάknowledgediscuss/deeds20:17, 3 July 2016 (UTC)Reply
Wikipedia, citing Ethnologue, insists that Balkan Gagauz Turkish, Gagauz, and Turkish are all separate, and a few sources do seem to take that view, e.g. Cem Keskin, Subject agreement-dependency of accusative case in Turkish, or, Jump-starting grammatical machinery (2009) speaks of "Balkan Gagauz Turkish, Gagauz, Turkish, Iraqi Turkmen, North and South Azerbaijani, Salchuq, Aynallu, Qashqay, Khorasan Turkic, Turkmen, Oghuz Uzbek, Afshar, and possibly Crimean Tatar". Other references speak of Balkan Gagauz Turkish as a variety of Gagauz, e.g. James Minahan's Encyclopedia of the Stateless Nations says "The Gagauz speak a Turkic language [...] also called Balkan Gagauz or Balkan Turkic, [which] is spoken in two major dialects, Central and Southern, with the former the basis of the literary language. Other dialects [include] Maritime Gagauz" (which comports with w:Gagauz's list of its dialects). Matthias Brenzinger's Language Diversity Endangered also treats Balkan Gagauz "or slightly misleading, Balkan Turkic" in his entry on Gagauz, but says it that the Balkan "varieties might deserve the status of outlying languages but very little information is available about them." (A few generalist references seem to subsume all gag into tr.) I would leave them all separate, pending more conclusive evidence that they should be merged. - -sche(discuss)23:58, 3 July 2016 (UTC)Reply
I think there's some confusion about what exactly we're talking about, and whether it's Gagauz or Turkish. Just because they use the term "Balkan Gagauz Turkish" doesn't mean that they're referring to the language with ISO 639-3 code bgx. When I look at who's citing the references listed for bgx at Glottolog, Manević (the reference for its classification) is cited in papers clearly talking about the dialects of tr. These are the only actual words attributed to this lect that I can find. —Μετάknowledgediscuss/deeds00:33, 4 July 2016 (UTC)Reply
@Tropylium, on the subject of Turkic languages spoken in Europe, do you know anything about this one, and about its differences or similarity to Gagauz and standard Turkish? - -sche(discuss)01:08, 11 May 2017 (UTC)Reply
I'm not previously familiar with this dispute, but here are a few handbooks on the topic:
Menges in The Turkic Languages and Peoples has the following slightly complicated quote (p. 11): "The Turkic languages spoken farthest west are the Balkanic dialects of Osman and Gagauz in Bosnia, Bulgaria and Macedonia. These seem to form two groups, one of possibly pre-Osman origin, and a later Osman one. To the former belong the Gaǯaly in Deli-Orman (Eastern Bulgaria), who, according to V. A. Moškov, are descended from the Päčänäg, Uz, and Torci (?), the Surguč, numbering about 7000 people in the district (vilājät) of Edirnä, who call themselves Gagauz. In Moškov's opinion, they, too, go back to the Päčänägs (?) and the Macedonian Gagauz; they number ca. 4000 people in southeastern Macedonia." — It seems clear that some group(s) corresponding to "Balkan Gagauz" is being identified here, but I am not even sure how to parse the sentence structure; e.g. are "Uz" and "Torci" some of the pre-Osman Turkic groups, or some of the alleged ancestors of the Gaǯaly? ("Osman" is, of course, Turkish.)
Hendrik Boeschoten in a classificatory chapter in Routledge's The Turkic Languages mentions that "a few speakers [of Gagauz] in northern Bulgaria, Romania and Greece, adhere to the Orthodox faith, and have their own history." This again seems to refer to "Balkan Gagauz", but with no indication of being its own language.
So far I would gather from this that "Balkan Gagauz" is at most a sister language of "non-Balkan Gagauz", and perhaps indeed just a different dialect group (perhaps one whose features are not reflected in written standard Gagauz). But the Manević 1954 paper would be more informative on this topic, if anyone wants to hunt it down. --Tropylium (talk) 11:55, 11 May 2017 (UTC)Reply
I think Balkan Gagauz should be merged with gag, especially since it contains no entries. The few terms that would be specific for Gagauz spoken outside of the traditional Gagauz area in Moldova/Romania/Bulgaria can be dealt with within gag entries. The only thing is that some etymologies of other Turkic languages sometimes refer to Balkan Gagauz instead of Gagauz, because editors didn't know the difference between two. Otherwise I don't see any problems with merging them two.
@Metaknowledge This is a hard question, I can offer only guesswork.
I can't find any good maps for the distribution of Gagauz and (Muslim) Turks proper in the Balkans, most don't show Balkan Gagauz at all although we know they exist at least in Bulgaria and Macedonia.
It seems that they are not easily separated geographically from Muslim Turks although they presumably live in different localities. I'm guessing this means that their languages ("Balkan Gagauz Turkish" and "Rumelian Turkish") could be the same, although maybe only the latter call their language "Turkish", so I guess that they (would?) use Standard Turkish in education and administration.
This would be a good argument to merge Balkan Gagauz into Turkish, except that this paper shows that Balkan Turkic (if this really is a single language) is quite distinct from Anatolian Turkish and perhaps worth considering a different language. Baskakov also considers Balkan Turkish and (Moldovan) Gagauz to form a clade within Oghuz and Anatolian Turkish and Azerbaijani to form another. Crom daba (talk) 21:35, 30 September 2018 (UTC)Reply
Merge / delete it. The distribution of the name, the way it is “mentioned”, points towards it being a ghost language. The name is not attestable as used by anyone having particular information about it; nobody can add anything under it either in such a situation where it is a content-filled concept for nobody. Its alleged synonyms “Balkan Turkish” and “Rumelian Turkish” show it is just an SOP term for Turkish as spoken on the Balkans respectively Rumelia, i.e. remnant speakers of the Ottoman rule. German Balkantürkisch, distinguished from Türkeitürkisch as a regiolect. Fay Freak (talk) 13:38, 2 December 2020 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Marrithiyel
Maridan [zmd], Maridjabin [zmj], Marimanindji [zmm], Maringarr [zmt], Marithiel [mfr], Mariyedi [zmy], Marti Ke [zmg]: should these be merged? References speak of a singular Marrithiyel language. - -sche(discuss)21:30, 20 July 2016 (UTC)Reply
We did delete Hawaiian Pidgin English in the past though (see Template talk:hwc). I don't see how this case is any different. -- Pedrianaplant (talk)
I know we did, but I didn't participate in that discussion (only 3 people did), and I disagree with it too, probably even more strongly than I disagree with merging cpi. —Aɴɢʀ (talk) 17:02, 8 February 2017 (UTC)Reply
Basically, this is a terminological problem. There may have been a true pidgin in each of these cases, but it has not been recorded. What is called a pidgin in many descriptive works is instead a dialect of English that is very easy to understand, nothing like the real English-based pidgins and creoles that I have studied. If you look at the actual quotations used to support lemmas in Chinese Pidgin English, you find that it is Chinese English. Support merge, but leave [cpi] as an etymology-only code. —Μετάknowledgediscuss/deeds23:16, 8 February 2017 (UTC)Reply
At least some texts seem very distinct, to the point of unintelligibility; consider "Joss pidgin man chop chop begin" (Whedon's translator begins chopping things? or "god's businessman begins right away"?). On the other hand, other sentences given by Wikipedia are quite intelligible...and possibly not attestable under the stricter CFI to which English is subject. I'm not sure what to do. (Our short previous discussion also didn't reach a firm resolution.) - -sche(discuss)17:46, 8 March 2017 (UTC)Reply
Pray somebody add |scripts = {"Narb"} to Module:languages/data3/x after line 1026 for xna. (Otherwise mentions of words in it are shown in slanted letters.)
Good catch. I've updated Common.css and Mobile.cc and set it to display rtl. Sadly, it seems there are no fonts that display it. If you or I could find a good image of what the letters are supposed to look like, I might have time to make a basic font iff the letters don't have to be joined the way they do in Arabic. - -sche(discuss)22:08, 18 November 2018 (UTC)Reply
I as an Archfag recently had a great update three weeks ago that adds displaying support for Old North Arabian, amongst other things like which improved Arabic and Syriac script rendering everywhere. gucharmap calls the name of the font by “Noto Sans Old North Arabian”, which I find in the filelist of the noto-fonts package. @-scheFay Freak (talk) 22:29, 18 November 2018 (UTC)Reply
I think everything under Category:Old North Arabian script languages should be “Ancient North Arabian” (xna), it is to wonder that Dadanitic (sem-dad), Hismaic (sem-his), Safaitic (sem-saf), Taymanitic (sem-tay), Dumaitic (sem-dum), Hasaitic (sem-has), Thamudic (sem-tha) are separate languages on Wiktionary (some also with no script assigned). (Prolly someone went through some lects and added all he found.) Those lects are at a level of attestion or study where it does not even matter whether they are dialects or languages, and “Thamudic” is even a collective term for any of the Ancient North Arabian lects not further classified. Many inscriptions cannot be classified unto more specific lects anyway (you know, people also were nomads and wrote graffiti here and there) and they can only be entered as “Ancient North Arabian”. With words being found randomly and in concise consonantal writing I don’t see why one would pursue separation other than by stating the find spot.
Also, “Qatabanian” (xqt), “Sabaean” (xsa), “Minaean” (inm), “Harami” (xha, redirects to “Minaean” on Wikipedia), Hadrami (xhd) – likewise otiose distinctions, regarding form and amount of attestion of Epigraphic South Arabian, as the name says only epigraphically attested, without any vowels –, have been unpopular in use already, entries and etymologies use the header “Old South Arabian” (sem-srb). I suggests to cross out those. Etymology-only is possible so one can use those in {{cog}} when in an individual case a word is known to be attested as of one of the dialects. North Arabian epigraphy categorization is more complex and it is better anyway to mention in each etymology where a lexeme has been encountered.
Himyaritic (sem-him), as an attested language, is rather mythical because the Ḥimyarites wrote Sabaean. Wikipedia mentions “three Himyaritic texts”, at the same time in the Encyclopedia of Arabic Language and Lingustics s.v. we read about two: “It is not even possible to establish whether they were written in the same language. The first text dates from around 100 C.E. and the second from around 300 C.E.” And about the secondary material from Early Medieval Arabs: “It is easy to see that quotations from Himyaritic offer very different readings according to the manuscripts.” Or according to others, mentioned in the EALL, Ḥimyarite is the same as Arabic, only with peculiar features (which might as well derive from Arabicized transmission, or later language fusion or whatever, much that could fool us). It could be grouped with those spurious languages if this category held languages from Antiquity.
Gurage is according to Wolf Leslau, it’s most eminent scholar, one language with twelve dialects; others share this view. The material for this language, particularly by Leslau across his works, only lists words as “Gurage”, without qualifying if they are “Inor”, “Mesqan” or some other Gurage, so on Wiktionary one cannot simply give “Gurage” words (which has recently been done in Semitic comparisons by abusing the code of the largest dialect Sebat Bet Gurage, in spite of the source saying “Gurage”). The following dialects I find on en.Wiktionary as languages: Kistane/Soddo (gru), Mesqan (mvz), Sebat Bet Gurage (sgw), Silt'e (stv), Inor (ior), Muher (sem-mhr), Mesmes (mys), Chaha (sem-cha), Wolane (wle), Zay (zwa); some of these are considered subdialects of Sebat Bet Gurage. There are more I don’t find on Wiktionary. It’s perhaps like with the Aramaic dialects yore or the Low German dialects today. People publish Westphalian dictionaries but it’s still Low German and so treated by Wiktionary. I suspect that instead of holding controversial subdivisions deriving from Ethnologue we should, holding to the sources, keep the Wiktionary-language level higher. The source for a certain word can be further qualified by labels as with Coptic. I mean that with language, unlike with biological taxonomy, one cannot simply assume that distinctiveness of a taxon is ascertained by experiments and then authoritatively published in some reference. As the individual forms are described in this dictionary, one must weigh if the data allows distinction at all. Currently it looks to me that hence Gurage must be lumped; I don’t know if, with new data or emerging different literary standards, separating the lects with separate codes will later be convenient (the increase in language material will be disappointing and unlikely someone will come and add Gurage in thousands of entries anyway, let’s be realistic), but I doubt that it would be comfortable. See also Why is Old Novgorodian a separate language in Wiktionary? This is the question: Is the difference in data enough to justify separation? The actual language-dialect distinction does not matter, it must be seen functionally, for dictionary purposes, for dictionary purposes. And if linguists publish material as “Gurage” the distinction is probably not good for Wiktionary headers. Isn’t it out of scope of Wiktionary to distinguish lect clusters when they are generally unwritten and chiefly written by and variously lumped and splitted by linguists? That’s a difficult question. Also I fear that such distinctions might be precisely the cause why nobody comes and pours out his rich Gurage knowledge. An adept would not be sure to distinguish, pendulating between two extremes, not witting if he should split as much as he can by all kinds of criteria or if to standardize and to abstract. To help though first all mentioned codes need the Ge'ez and Latin script both assigned, and the macrolanguage created. Maybe there will be late order from early ambiguity. Though I would perhaps do the order by lumping and labelling by location, were I that certain aficionado.
This discussion really belongs at rfm, because that's where we normally discuss changes to whether or how we recognize a language. The Grease pit is for discussing how to implement something along those lines- not whether it should be implemented. The other option would be at the Beer parlour, but this seems like something that would benefit from the more specialized focus of rfm. Chuck Entz (talk) 03:39, 14 November 2018 (UTC)Reply
Some prior discussion of Thamudic et al is on Category talk:Hismaic language; IIRC they were separated because literature does mention them as distinct entities, but if they were very similar or often treated as one language, and especially if there's difficulty in assigning specific texts to specific ones due to similarity, that would be an argument for reversing that decision and going back to the conservative approach of treating them all as one language with 'dialect'/'region' labels where appropriate. (As to the venue, yes, these discussions tend to happen on RFM for quirky historical reasons — originally the discussions entailed actually merging or splitting language templates — although some have proposed the Beer Parlour as a more logical venue. There are minor benefits and drawbacks to either venue; this venue does have the advantage that discussions stay on the page until resolved.) - -sche(discuss)17:20, 14 November 2018 (UTC)Reply
Who is likely to have access to resources on Africa's Semitic languages that could help judge what to do with Gurage? User:Metaknowledge, User:Wikitiki89? Wikipedia insists "The Gurage languages do not constitute a coherent linguistic grouping", which seems incompatible with merging them. William A. Shack, in his book on The Gurage, writes that "each Gurage dialect is usually understood only by its own speakers, and there is a rough correlation between the contiguity of dialect groups and the extent to which their dialects are mutually intelligible." (Steven Danver, in his (general-focus) encyclopedia, says "the languages of the different groups of Ethiopian Gurage are seldom mutually intelligible.") Marvin Lionel Bender, in his 1976 Language in Ethiopia, says "Although seventeen varieties of Gurage dialects are listed, mutual intelligibility reduces this to four languages and three dialect clusters as follows (Hetzron classification): Gogot, Misqan, Muxir, Soddo East Gurage (Inneqor, Silti, Urbareg, Weleni, Zway) Central West Gurage (Chaha, Gumer, Gura Izha) Peripheral West Gurage (Ener, Geto, Indegegn, Innemor)" However, his very next sentence is: "Gogot, Muxir, Soddo comprise a geographical (non-genetic) grouping of non-mutually-intelligible languages known as 'North Gurage'", all of which seeems to suggest that merging all of the Gurages would not be sound. - -sche(discuss)17:28, 14 November 2018 (UTC)Reply
The cited grouping of course adds to the confusion. Three languages, but four dialects clusters, not mentioning their intersections? Well, we will not find out how one should see them without deep-diving. But the question is which direction Wiktionary should go: likely the current division is not correct. Should Wiktionary just add all possible splits so they can be cleaned up later when someone would commit himself to add the whole Gurage and judge about which distinctions are most convenient or should we have one macro-code because distinction is hopeless? The reason why I have even mentioned Gurage is that for example Leslau’s Etymological Dictionary of Geʿez which I like to use just gives words as “Gurage”, which sounds like there is a common vocabulary. Fay Freak (talk) 14:43, 15 November 2018 (UTC)Reply
Perhaps you can deduce from Leslau's literature list which Gurage language he gets his data from? He seems to have written an etymological dictionary of Gurage as well, presumably its foreword could clear things up.
As a volunteer project (run on fancy), we really have no other choice than to wait for someone to investigate the matter deeply and order the languages in a manner that facilitates their lexicographical work.
@Fay Freak, -sche: A bit late, but here are my responses to the three outstanding problems (your #2–4):
It is fairly evident that Ancient North Arabian is not a single language, and I advocate that sem-xna be abolished rather than the specific language codes; read Al-Jallad (2018), "What is Ancient North Arabian?". He sees Safaitic (which he has written a grammar of) and Hismaic as being of the same continuum as Old Arabic, but they are obviously too distinct from Classical Arabic for lexicographical purposes. He supports the distinctness of the others as languages, and of the various "Thamudic" lects. Based on Al-Jallad, I would prefer we split Thamudic B, C, D, etc as necessary; each language will have a very small corpus, but it seems like the most honest way to do it, and if more inscriptions are found, the lettered Thamudic wastebaskets will probably get their own names as the others did.
Old South Arabian is also not a single language, though Sabaean was the standard that the other lects imitated, and I advocated that sem-srb be abolished as well. Multhoff (2019) in The Semitic Languages makes the case for four distinct languages: Sabaean, Minaean, Qatabanian, and Hadrami. She makes no mention, however, of Harami. Macdonald (2000), "Reflections on the linguistic map of pre-Islamic Arabia" explains that "Harami" is a name given to a few Sabaean texts that seem to have been contaminated by other Semitic languages, which is not at all an unusual feature and not unique to that site, so I suggest we remove that code.
As for Himyaritic, I now think I was wrong to include it. There are three texts often attributed to it, but see Stein (2008), "The ‘Himyaritic’ Language in Pre-Islamic Yemen", which makes a strong argument to consider these as simply very late examples of Sabaean, which is indisputably the language of the other texts of the region in that script.
Finally, for Gurage, the chief problem is that some scholars follow Hetzron in saying that Gurage is polyphyletic, in which case lumping would be committing a grave error (and the same charge has been levelled for Aramaic, with perhaps more evidence). Meyer (2011) in the International Handbook does seem to support the unity of Gurage, and treats the lects together, which gives me hope for lumping, but he is unwilling to commit to whether they should be considered dialects or languages. I think your Gurage-adding genius is mythical, so we have to choose which is least bad: many languages with scanty coverage, because their forms may be similar to forms entered under a different L2 header; or one Gurage language with decent coverage, but many forms that are not marked for what dialect they belong to and therefore a poor resource. I hesitantly support merger, given those choices. —Μετάknowledgediscuss/deeds03:13, 10 August 2020 (UTC)Reply
An addendum: "Hadrami" is a terrible name for xhd, and invites confusion with Hadrami Arabic. Wikipedia uses "Hadramautic", but N-grams and a quick literature review suggests that "Hadramitic" is more common. @Fay Freak, -sche again (yes, I know I'm pestering, but I don't want to move forward on all this alone, both because I am fallible and because some of these, particularly splitting OSA, would require a bit of work, although in that case there is an online corpus that will help immensely). —Μετάknowledgediscuss/deeds02:40, 17 August 2020 (UTC)Reply
Re North Arabian: Many works I browsed through speak of Old North Arabian as a unit with dialects, but also carefully specify what lects (including Thamudic B vs C, etc) words are attested in. Some imply, in their presentations, that a large number of words are identical between dialects, at least in the sample of vocabulary that they're treating (e.g., the pronouns treated in Roger D. Woodard, The Ancient Languages of Syria-Palestine and Arabia (2008), pages 197-198), though this seems to be because the authors are presenting 'normal', normalized and romanized forms, given Al-Jallad's evidence that words (even the supposedly distinctive definite article) varied not just among dialects but even within the writings of individual speakers. The native script also loses many possible differences in pronunciation, but then, we are a written, writing-based dictionary. I find slightly more works speaking of "Ancient North Arabian dialects" than "Ancient North Arabian languages", and the fact that some authors have argued the varieties are the same language not only as each other but even as Arabic itself does suggest a high degree of similarity (or that the scholars in question are lumpers). As we're dealing with small, extinct and apparently clearly delineated corpora, it seems like the conservative approach of treating each under its own L2 could be better, and we could retire xna ... unless we need it as a wastebasket for unsorted things, which Al-Jallad (and Fay Freak, above) suggests we would. (Bah, It's messy business, deciding what's a language and what's a dialect...) I will try to dig into the rest later. - -sche(discuss)04:10, 19 August 2020 (UTC)Reply
Well myself I have added Sabaean, Minaean, Qatabanian entries meanwhile, understanding and quoting a few inscriptions, although apart from some occasional features I noticed little how such an inscription can be classified as either, other than by provenience or rulers or gods mentioned—but that must be due to my blasé comparative approach that also makes me read Romance without recognizing the individual language. So somehow the volition to a merge is gone, though the lumping codes “Old South Arabian” and “Old North Arabian” must be kept for inscriptions no one has classified. Both are useful.
For Himyaritic, however, nothing is left. As here said already, the three alleged Himyaritic inscriptions don’t even need to be in the same language, and they aren’t even from anything to be called Ḥimyar (there are “Lesser Himyarites” and “Greater Himyarites” and the ethnic identity is fragmentary, too, by the way). In the “Critical Reevaluation” of the Ḥimyaritic language – cited by Wikipedia on Himyaritic language one does not know what for: their “undeciphered-k language” header recently introduced is surely a made-up term, oddly suggesting that these inscriptions are yet another language when those “k-language” inscriptions are exactly those otherwise claimed for Himyaritic, so we see Wikipedia editors had no clue and phantasize together languages due to their disdain for primary sources – helpfully includes a map, also coming to the conclusion “we have no reason to assume the existence of some “non-Ṣayhadic” language in pre-Islamic Yemen that was spoken besides the (Late) Sabaic idiom known from the inscriptions.” That from the fact that “Himyaritic” words typically given from Arabic sources are all also found in Sabaic, and the grammar found in the three inscriptions, including the prefixed instead of postfixed article which is only found in two of them, is too either found in Sabaean or can well be ascribed to their being poetry, which is also the reason for their being poorly understood. Many Arabic poems are also hard to understand and mostly helped by the copious material for the language which is not the case for languages with so limited a corpus, like Old South Arabian. Even in the Digests, Latin prose, not all passages are of discoverable meaning.
What would hinder man though to add understood words with quotes from the ominous inscriptions as Sabaean? Or anything from Arabic sources transmitted as Himyaritic instead of Arabic as Sabaean? For there is no evidence for it being a particular language. You see, from the corpus-based standpoint Wiktionary takes Himyaritic must go. Nothing can get the header “Himyaritic”, it can only be mentioned at Sabaean or Old South Arabian entries that Himyaritic nature is suggested by those who have come to believe in this extraordinary claim for which extraordinary evidence is not provided. Fay Freak (talk) 04:18, 6 August 2021 (UTC)Reply
I went on and moved our only “Ḥimyaritic” entry after that famous sentence to Yemeni Arabic in which the word طَيِّب(ṭayyib) for “gold” turns out otherwise known, and to be nothing else than Classical Arabic طَيِّب(ṭayyib, “good”) meaning “refined” and therefore gold, while Old South Arabian could not have developed such sense, so it is clear the famous quote one has been so inept to classify is at best only macaronic Sabaean-Yemenite Arabic. It is well put by Marijn van Putten:
The Arab grammarians were interested in describing correct usage of language of Classical Arabic. It is quite clear that Himyaritic (and by extension Yemeni Arabic) did not fall in the category of 'correct usage'. Within this context, it is of course not surprising that anything that is "wrong" and from Yemen might be denoted as Himyaritic. This would then include both varieties of Yemeni Arabic and some surviving vestiges of Ancient South Arabian.Fay Freak (talk) 04:59, 13 September 2021 (UTC)Reply
Now also in a new article by Koutchoukali like communis opinio, though his blogs transpire by him stalking Wiktionary: later Muslim historians would refer to anything related to South Arabia’s pre-Islamic history as “Himyaritic,” all memory of its other states having passed away.Fay Freak (talk) 01:01, 18 September 2021 (UTC)Reply
@-sche, Theknightwho: So are we gonna delete Himyaritic now? I promise there is nothing to add in it, and I have ever been averse to use the code, as on بطيخ; everything one needs to know about it is in the dictionary entry Himyaritic. Everything else in this five-year old topic is resolved:
Old South Arabian and Ancient North Arabian are both necessary evils in addition to individual languages of the Old South Arabian and Ancient North Arabian families because inscriptions are all over the place and sometimes not informative enough for specific identification, so we can’t remove the languages like we have removed Nahuatl, but like for Nahuatls the language treatments use to be sharp. Gurage nobody will reliably untangle in the near future; we may have a family under sem-eth for the “Gurage cluster”. Fay Freak (talk) 03:07, 6 December 2023 (UTC)Reply
No objection to renaming Arawak, but I'm not sure about Kalhiphona, which seems to be quite rare even on a Google web search, and which seems to invite as much possible confusion (in its various spellings) with the various spellings of Garifuna as it avoids with other "Carib"s. - -sche(discuss)06:56, 19 September 2018 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Lenape
The decision to deprecate Lenape as a canonical language needs to be revisited. It seems to me that removing Lenape as a canonical language is like removing French as a canonical language and instead having only Quebecois and Parisian. Classifying Wënami and Munsee as seperate canonical languages is a narrow linguistic differentiation that quickly spills into the political. In a 2013 discussion (archived at Category talk:Unami language) the difference between Wënami amd Munsee was compared to English and Scotts. This is a faulty comparison because the political and cultural situation is in no way analogous (and language versus dialect is always a deeply political not just academic subject). The continuum of dialects once spoken in Southern New York, New Jersey, Delaware, and Pennsylvania were on the brink of disappearing forever. What we are witnessing now is a process of standardization that could revive the Lenape language. I hope the Wiktionary community would not want to contribute to fracturing such efforts. Unnecessary rigidity about what is to be considered a canonical language could jeopardize such standardization and revival. I strongly urge everyone to reconsider having a canonical language category for Lenape, and specifying the dialect of origin for specific terms, usages, and grammatical conventions. Note that all orthographies for Lenape are relatively recent inventions based on various European languages, and that standardization here is more than warranted. Andreas.b.olsson (talk) 05:17, 29 July 2021 (UTC)Reply
This is tricky. Some people want to document each language as such, especially with an eye to their past as mutually unintelligible languages (Marianne Mithun, The Languages of Native North America, 2001, page 331; Siebert even seems to suggest Munsee was more closely related or similar to Mahican than to Unami) with extensive differences in orthography and phonology (e.g. in having l vs r in their reflexes of PA's *r/*l phoneme). (Archaeological as well as linguistic evidence suggests the distinction between the two groups goes back to prehistoric times.) Merging northerly Munsee into the now more dominant southern Unami could complicate documentation of that critically endangered language, which some people are studying and learning from its remaining speakers in Canada. On the other hand, the most prominent revitalization efforts in the United States, based mostly on southern Unami, do seem to speak of Lenape as a single language (and FWIW, Canadian Munsee efforts, albeit with a different spelling, also seem to speak of a Lunaape language), apparently aiming to standardize the two into one language with an eye towards keeping it alive into the future. We have to think carefully about what to do here. - -sche(discuss)17:52, 29 July 2021 (UTC)Reply
I have nothing to contribute beyond the wish for some precision in the etymologies of toponyms, which can easily be accomplished with qualifiers or labels without complicating the creation of good language entries. DCDuring (talk) 18:33, 29 July 2021 (UTC)Reply
You say on Talk:mochipwis that "I'm a learner of modernized Lenape, a language in the process of being revived. I am not an expert in the differences between Wënami versus Munsee. What I do know is that there are no first-language speakers of either dialect left." This goes to the heart of the issue, I think: Wiktionary does not only cover modern languages as they are currently spoken, but also covers languages that existed in the past. Although (as noted in the earlier discussion) I initially created "Lenape" content under a unified code, following the "lumper" approach of the US revivalists, the differences between the lects are (as Chuck noted in that same disscusion) historically extensive, to the point that the modern linguistic literature I've been able to find that says anything about their intelligibility accepts Mithun's statement that they were mutually unintelligible, which militates against combining them. (OTOH, if modern speakers are trying to merge them, that militates in favor of a merger.) Wiktionary does merge e.g. many Sinitic languages even when they're mutually unintelligible when spoken (and conversely we split e.g. Bokmal and Nynorsk even though they're not merely mutually intelligible but the same language), so either approach could be made to work. - -sche(discuss)19:29, 29 July 2021 (UTC)Reply
@-sche: What about recognizing all three codes? We could have Lenape (either using del or creating a new code for it, e.g. del-len, so as to keep del for the family) for the modern language undergoing revitalization, and also have unm and umu for the historical stages. —Mahāgaja · talk19:57, 29 July 2021 (UTC)Reply
Adding a link here to the important Lënape Talking Dictionary maintained by the Tribe of Delaware Indians. It should be recognized that the code del (for Delaware) is potentially problematic and controversial. If I say in New York City “I speak a little Delaware” I’m likely to get quizzical looks. But if I say “I’m learning Lënape” an educated resident of the city is likely to remember social studies classes in elementary school, contextualize, and understand what I mean. Wouldn’t it be better to have a new code for modernized Lënape that contains the letters in Lënape? Andreas.b.olsson (talk) 21:56, 29 July 2021 (UTC)Reply
We use the ISO 693-3 codes for languages, so that's out of our control. They're not going to change the code; they've refused for www, which has some significant real-life problems. In practice, the codes may sometimes look like abbreviations, but in theory, there's no necessary connection; it's just a three letter code.--Prosfilaes (talk) 01:30, 30 July 2021 (UTC)Reply
@Prosfilaes there is no ISO designation for the Lënape variants. So what language does the ISO code del stand for? Old Wënami? Mondern Lënape? Munsee dialects still spoken in Moraviantown? Andreas.b.olsson (talk) 02:50, 30 July 2021 (UTC)Reply
@Andreas.b.olsson: In the ISO itself, del stands for a macrolanguage called Delaware whose individual languages are Munsee umu and Unami unm (see https://iso639-3.sil.org/code/del). We're not obliged to use ISO's names, though, so we are free to call del Lenape or Lënape rather than Delaware. Don't get hung up on the letters used for the code; they don't have to bear any relation to the name of the language (xcl stands for Old Armenian, for example, and the code for Mapudungun is arn). —Mahāgaja · talk07:40, 30 July 2021 (UTC)Reply
@Mahagaja:Fair enough about the abstraction. I did not know that ISO had a more complex sub-categorization. Thank you for providing the link and helping my knowledge about these standardizations grow. My concern was related to the fact that del is derived from Delaware – the European name for the region – and not Lënapehokink, the Lënape name. However, as far as I understand the Lënape are themselves comfortable with the syncretic designation and retain it in their name (Nation of Delaware Indians). I should therefore not assume any offending connotation in the designation del. I’m trying to reach out to the Oklahoma branch of the Lënape. Let’s see what they say about this issue. As a dominant constituency they should have a say about this actively evolving linguistic situation. On an additional note, there is a significant difference in the orthography being actively taught by the mainly Munsee branch in Moraviantown Canada.Andreas.b.olsson (talk) 08:49, 30 July 2021 (UTC)Reply
I’m not sure there is anything wrong in the abstract splitting of the dialects/languages. It’s the naming in Wiktionary for these lects that does not reflect how these lects are referred to in actuality (where there is unfortunately ambiguity). A proper and respectful name is needed for the following variants:
Variant currently taught at Talking Lënape Dictionary and extensively documented by Jim Rementer et al in recent history (late 20th century).
Variant currently taught in Moraviantown where there seems to be wide variations between the last handful of fist-language speakers. I’m not sure any of them are still alive but the community seems to be sticking with their own historical orthographical variant rather than using Rementer’s et al. revised system (the one I’m learning and using).
The modern variant in Moraviantown seems to be referred to alternatively as Lunaape and Munsee Language. The modern variant based on Rementer et al work is consistently referred to as Lënape or alternatively without the schwa (Lenape). Andreas.b.olsson (talk) 13:06, 30 July 2021 (UTC)Reply
For the purpose of historical linguistics Old Lenape can be divided into:
Southern Unami
Northern Unami
Munsee
Note here that Unami seems to be a term derived from historical linguistics. I believe it should not be used to denote any of the revitalized lects currently in use. However, the Stock-Bridge Munsee do refer to their Lunaape variant in English as the “Munsee language”. Andreas.b.olsson (talk) 14:54, 31 July 2021 (UTC)Reply
Here is what I propose in terms of naming, categorization, and ISO encoding.
Old Lenape: del
Munsee: umu
Unami: unm
Lënape: lnp
Lunaape: lne
LNP and LNE are not used yet by ISO 639-3. Andreas.b.olsson (talk) 15:29, 1 August 2021 (UTC)
North and South Unami can be treated as minor variants (i.e. dialects). Munsee is treated as distinct, leaving the question open about whether it was more closely related to Mohican. Andreas.b.olsson (talk) 15:35, 1 August 2021 (UTC)Reply
We can't be making up our own three-letter codes, but we can put them after a family code, separated by a hyphen. So if we wanted to, we could in theory create alg-lnp, alg-lne, etc. —Mahāgaja · talk10:08, 2 August 2021 (UTC)Reply
Why can’t we be making up codes if they are not used? This can be in conjunction with ra request to ISO that these lects be added with those codes. It’s one thing to request a change, another the treat the ISO 639 as extensional. With its increased influence Wiktionary ought have some influence on ISO. Andreas.b.olsson (talk) 16:40, 2 August 2021 (UTC)Reply
We can't make up codes because ISO might assign them to some other languages in the future. I think you might be overestimating Wiktionary's influence; also, the ISO committee would take a whole lot of convincing before they agreed to two more codes. Considering you started out this discussion arguing against treating Unami and Munsee as two different languages, it seems like it would be quite a stretch to convincingly argue that they are, in fact, five different languages. From everything you've explained above, I'm starting to think we should recognize only del as a canonical language and reassign unm and umu to be etymology-only codes for subvarieties of del; if necessary, we can add more etymology-only codes (since these are local to Wiktionary and have nothing to do with ISO) for other chronolects and regiolects. —Mahāgaja · talk18:38, 2 August 2021 (UTC)Reply
@Mahagaja I think your low-balling Wiktionary’s growing influence, and a few people’s ability to influence seemingly immobile institutions. My point has evolved based on @-sche’s input about linguistic historicity and the strong difference between orthography used by those who learn Lunaape versus Lënape. Importantly, there is extremely strong merit in treating the language recorded by Zeisberger et al in the 18th century as a different language. It is grammatically quite distant from Lënape and Lunaape. Note that the del, unm, and umu seem based on the historical study by Moravian missionaries (who referred to the language as “Delawarische Sprache”). There needs to be a clear distinction between these historical lects (which I’m referring to as “Old Lenape”) and the languages spoken today. Andreas.b.olsson (talk) 03:26, 3 August 2021 (UTC)Reply
Seeming immobile to whom? They've seemed relatively responsive to me; they offer a final decision on most requests in an annual wrapup. I certainly question any need to bluster ahead and then expect SIL to follow; that seems likely to be counterproductive.--Prosfilaes (talk) 09:31, 4 August 2021 (UTC)Reply
I would also note here that the German entry in Wikipedia is titled “Delawarisch” and talks about it first as a single language, and then alternatively two languages depending on one’s view. The equivalent English entry is “Delaware languages”. However, in both entries the focus is on the language(s) studied by Moravian missionaries et al, and then linguists studying those initial European studies (which haphazardly invented wildly different orthographies based on the authors own mother tongues). @-sche is right: it’s complicated. Here are the facts:
Old Lenape is grammatically very different from modern variants.
There are two very different modern orthographical variants.
So if we are going to base decisions on precedent – which one ought to - then the way Wiktionary has approached the orthographically distinct Norwegian Bokmål and Nynorsk speaks for having distinct language codes for Lënape and Lunaape regardless of whether they should be treated as the same language or not. Unfortunately, del really in the end refers to Old Lenape since the ISO further subdivides it into Unami and Munsee, which are really references to the language(s) studied by Zeisberger’s et al and those who continue to study their studies. To use it for Lënape and Lunaape would be confusing and improper.Andreas.b.olsson (talk) 04:25, 3 August 2021 (UTC)Reply
This is not Norwegian, and there's no other language that is presented like Norwegian. The closest are Ottoman Turkish and modern Turkish and Urdu and Hindi, and in both of those cases you're looking at different scripts, Arabic and Latin or Devanagari plus vocabulary divergence. One has to look at all the precedent, not just one example.--Prosfilaes (talk) 09:31, 4 August 2021 (UTC)Reply
IMO, as someone who as added a good portion on the Munsee entries, unless someone can demonstratively show the need, I would say the way things are set up right now is perfectly fine. More work should be made to show the linguistic differences before any new codes are created. This all seems like hypothetical conjecture to me. --{{victar|talk}}05:25, 3 August 2021 (UTC)Reply
Having looked closer at Zeisberger’s work, I should say that how distant Old Lenape is from Lënape and Lunaape is uncertain to me. There seems to have been a lot of simplification in how verbs are used, with certain modes falling greatly out of favor (e.g the subordinate mode). The issue is complicated by the very different orthography used by Zeisberger, who used German inspired phonological conventions. If Wiktionary were to want to document these changes though, it ought to leave a space for Old Lenape. The issue reminds me of Swedish, which went through a drastic change in orthography in the late 19th and early 20th century.
I think what I get back to is that Unami is not the proper L2 name for a language. The proper name for this language is Lënape. So even if we keep the coding structure in tact, the name of that lect needs to be changed (Unami => Lënape). Andreas.b.olsson (talk) 09:24, 3 August 2021 (UTC)Reply
Out of curiosity @Victar, how come you chose Munsee as the name of the L2 lect and not Lunaape? I see that the community in Moraviantown uses either designation. Nonetheless, what made you chose one over the other? If you are a member of the Stockbridge-Munsee community, please help me better understand and my sincerest apologies for potentially overstepping. Andreas.b.olsson (talk) 09:36, 3 August 2021 (UTC)Reply
I guess another thing that bothers me is that the code unm – however abstract in the mind of the Wiktionary community – carries the stigma of a split. It’s continued use will cause continued sclerosis. It’s does not seem a term that all parties – Lënape in Oklahoma, Munsee in Canada, and enthusiasts like myself in New York City – can all rally around. It seems to me that there is something quite new going on, an amalgamation and standardization that should be captured in vivo by Wiktionary. Andreas.b.olsson (talk) 10:14, 3 August 2021 (UTC)Reply
(Sorry, I was busy/distracted.) If we could reach out to Lenape folks, including Munsee speakers or learners, for clarification of whether they consider the lects one language (in particular, if Munsee consider their language to be the same as Unami, or if it is only the US-based folks running the Unami-based Lenape Talking Dictionary who name it as if it speaks for all Lenape), that'd provide useful evidence: if speakers want to consider the lects one language, it'd be evidence that we should merge (and distinguish with qualifiers, etc); otherwise, the current setup is fine. It seems some of the objection comes down to just the names, preferring "Lenape" to "Unami"; in another discussion Andreas notes that speakers would sooner say the equivalent of "I speak Lenape" than "I speak Unami", but this is not evidence that they are one language (consider various Dogon languages, which are all Dogons though not all mutually intelligible, including Bangime language whose speakers consider themselves Dogon who "seem unaware that [their language] is not mutually intelligible with any Dogon language"). Simply renaming "Unami" to "Unami Lenape" (and "Munsee" to "Munsee Lenape") is an option, although we usually try to use the most commonly used names for languages, which may be the current names. The shape of the codes is a non-issue since they're mostly not visible to readers, only editors. This discussion does make me realize that they're reader-facing in the names of certain categories, and that this is probably confusing, but that's a general issue not specific to Lenape. (How many people looking at our Khanty content would guess that the list of berries is at "Category:kca:Berries"? Shouldn't we always use language names?) There's no linguistic basis for introducing more than one new code (e.g. a five-way split!), as floated above. - -sche(discuss)02:24, 5 August 2021 (UTC)Reply
I’m the newcomer here @-sche. Though I have used Wiktionary for a long time, I did not prior to this participate in its maintenance. I will defer to you all on the best back-end encoding, and whether proliferating codes for anticipated future linguistic differentiating makes little sense. My objection – and this was not clear to me before either – seems reduced to the fact that “Lënape” was stripped from how the lect(s) are referred to. I think there is sufficient online evidence to make the affirmation that the so called Unami branch is almost exclusively referred to as Lënape. Therefore I think it would be warranted to make this change as soon as possible. Having looked at quite a few of the “Unami” entries it seems to me that they are almost all derived from the work of the Talking Lënape Dictionary project. Hence the issue of the distance between the lect recorded by Zeisberger et al versus the more recent collected speech samples of the Talking Lënape team is mute, and can be resolved later. Note though this may mean having to encode for another lect at some point in the future (especially since the spelling used by the Moravian missionaries was based on German and very different). I agree fully that the Munsee community should have a say in whether to treat their lect as the same as that taught by the Talking Lënape Dictionary project. It should be noted though that it can be factually stated that they have to date used a very different orthography in lessons and materials they have posted online. Andreas.b.olsson (talk)
@-sche, @Mahagaja, @Victar can we split out the decision about the naming for the lect with the code unm and rename it from Unami to Lënape? There is overwhelming evidence that Lënape should be the proper name of the language, and not Unami. The code issue and whether to treat the Munsee branch as a separate language (and what to call it) can be dealt with separately and later.
@-sche, @Mahagaja, @Victar asking again about splitting out the issue of the language's proper name and making an initial narrower decision. I would like to continue building out Wiktionary's Lënape entries, but first I would like to make sure the language currently referred to as "Unami" is referred to by its proper name. Andreas.b.olsson (talk) 13:01, 12 August 2021 (UTC)Reply
AFAICT "Lenape" is a far more common name than "Lënape", so renaming "Lenape" to "Lënape" does not seem appropriate. And because Unami is only one variety of Lenape, renaming Unami to Lenape/Lënape would be confusing. The difficulty seems to be that the main US-based language project produces its Talking Dictionary seemingly based on Unami but named as if it's one overarching "Lenape"/"Lënape" language. (I know this kind of thing has come up before, where a prominent dictionary of what it considers "a language" doesn't specify which of two actually-distinct lects its words are, or conflates them; I will track down other examples if necessary, but it seems tangential.) I don't want to stand in the way of a Native language revitalization effort, so the above-discussed option of leaving Unami and Munsee for the historical (and historically not mutually intelligible) varieties and making Lenape a language code (rather than a family code) for the "unified" language has some appeal, but it would entail duplicating a lot of content which is currently Unami. There is also the option of renaming Unami to "Unami Lenape" to get the "Lenape" name in there. - -sche(discuss)19:19, 30 August 2021 (UTC)Reply
Again, I agree -- renaming Unami to Lenape would simply cause confusion whilst adding no benefit and renaming it to "Unami Lenape" does not serve to disambiguate. --{{victar|talk}}20:13, 30 August 2021 (UTC)Reply
I’m not sure I understand for whom this would be confusing @Victar. I agree @-sche that Lenape may be simpler to type than Lënape. Anyway, the schwa indicator is optional in the dominant modern dialect of Lënape. I have a deep hunch here that Linnaean taxonomic orthodoxy and perpetuation of erroneous past anthropological claims are at fault for the strict linguistic distinction between Unami and Munsee. The more I study Lënape, the fewer distinctions I see in the various dialects except for in the choice of orthography (which noticeably stems from quite recent European attempts to record an unwritten language). Andreas.b.olsson (talk)
I suspect that Wiktionary is perpetuating a European construct erected between the late 18th and early 20th century. Of course, these faulty past errors of European linguists have had real effect, seemingly causing “two languages” to appear to exist. I need to further study the Munsee branch to substantiate my claims.
I continue to believe Unami is a culturally and linguistically inappropriate term for any dialect of Lënape. If the language spoken by the descendants of the Unami tribe of the Lënape were to have any other name than Lenape – which I insist is the appropriate designation – it might be Delaware. This is a name I believe the (predominantly) Unami descendants living in Oklahoma accept and sometimes use in English despite its European origins.
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Arawak and Carib again
Resurrecting an old discussion on the same subject above, I propose we rename Arawak (arw) and Island Carib (crb) to Lokono and Kalinago, respectively. As User:Victar noted in that discussion, Arawak is easily confused with the Arawak/Arawakan family. (Indeed, I have run into many etymologies where someone has mislabelled a word from a different Arawakan language as Arawak). Island Carib, meanwhile, is (1) not a Cariban language, making etymological discussions occasionally confusing; (2) no longer generally called by that name, since the people are now officially called the Kalinago by the Dominican government; (3) easily confused with the unrelated 'true' Carib language, which we call Galibi Carib.
I would also add this latter language (the one we currently call Galibi Carib (car)) to the discussion, and suggest it should be renamed either Kari'na or Kari'nja, names that are both more common in the current literature and closer to the native name (karìna in Courtz’s orthography). (The most common name for this language in the literature is actually simply Carib, but this may itself be confusing.) — Vorziblix (talk · contribs) 15:56, 3 January 2024 (UTC)Reply
Support: Obviously still support renaming both to Lokono and Kalinago, respectively. I can't speak to Galibi Carib as I'm not familiar with the academic literature. --{{victar|talk}}16:22, 5 January 2024 (UTC)Reply
I’ve added most of those missing entries (and improved one or two others), and have also changed Galibi Carib (car) to Kari'na, since I am actively working on that language and would rather make the change while our coverage is still small. I will leave the discussion to run for a few more weeks before changing the others, and we can always also revert/change Kari'na to something else if the disucssion turns that way. — Vorziblix (talk · contribs) 13:42, 16 January 2024 (UTC)Reply
All three of these moves are now complete. The change away from ‘Arawak’ came not a moment too soon; a large amount of references to ‘Arawak’ (arw) in our entries were entirely wrong and needed to be corrected to ‘Arawakan’ (awd), and indeed there may be some remaining references to arw that are still wrong. — Vorziblix (talk · contribs) 06:42, 2 February 2024 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Mansi Languages Split
Since the Mansi language represented in Wiktionary isn't a thing, since it is several languages compiled under one name and code, I would like to propose a split. It would be beneficial to split it along the 4 main dialects, (the ones still extant) Northern, Eastern and (extinct) Western, Southern. The subdialectal markings can be represented with labels, already in place in the current category for Mansi. Ewithu (talk) 18:19, 9 January 2024 (UTC)Reply
"Pomeranian" is essentially a term for the family consisting of the Kashubian and Slovincian languages. In fact, Pomeranian (or its Polish counterpart, język pomorski) has always been used as a synonym of Kashubian (Polish język kaszubski), as witnessed by the 1893 dictionary called Słownik języka pomorskiego czyli kaszubskiego ("Dictionary of the Pomeranian a.k.a. Kashubian language").
There are no written records of an ancestor of both Kashubian and Slovincian, and any attestation of a Pomeranian lect will automatically fall into either of the two categories according to our current handling. However, the two languages do share an ancestor, and this ancestor did influence other languages. As such, it seems only logical to set this language as a proto-language, and rename it to Proto-Pomeranian in reference to its being the unattested ancestor of a language family, rather than being an attested language.
Support I have looked into usage and it appears that this ancestor—not only in English, but also Russian and other current tongues of linguistic science—can only be called “Pomeranian” in the same way as Proto-Slavic can be called “Slavic” language, and it the same way as “Slavic” is any Slavic language, or “Turkic” any Turkic language etc., “Pomerian” is any language of the said group. Fay Freak (talk) 12:35, 11 January 2023 (UTC)Reply
Done. @Vininn126, Thadh, Sławobóg, Silmethule, Gnosandes, Mahagaja, Fenakhay, Fay Freak Currently there is no family corresponding to Proto-Pomeranian and the code for this language is 'zlw-pom', which is exceptional in lacking the '-pro' suffix normally given to proto-languages. I'm thinking we should add a 'Pomeranian' family with code 'zlw-pom' and give Proto-Pomeranian the code 'zlw-pom-pro'. Thoughts? Benwing2 (talk) 07:25, 28 February 2023 (UTC)Reply
@Benwing2 error corrected by adding * (in *žarъ). Pomeranian term I met here — Martynaŭ, V. U., editor (1985), “жар”, in Этымалагічны слоўнік беларускай мовы [Etymological Dictionary of the Belarusian Language] (in Belarusian), volumes 3 (га! – інчэ́), Minsk: Navuka i technika, page 210 -- ZomBear (talk) 03:16, 1 March 2023 (UTC)Reply
The little reality of the field of study does not equal the irreality of the language itself. Subbranch proto forms are often left unreconstructed for being too small a fish when a larger one is available. Fay Freak (talk) 22:28, 4 March 2023 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Merging Proto-Pomeranian
As it stands, Proto-Pomeranian is next to useless and I don't see it becoming any more useful. There is a single lemma that we could treat as Pre-Kashubian. I don't see any reconstructions coming in the future as a lot of linguistics consider Slovincian a dialect of Kashubian (this needs further analysis and it's something I'm working on). The next highest level that linguists agree on is something like Pomeranian-Polabian, they share a lot of sound changes in common, but I think it would make more sense to treat descendants sections with something like:
I support removing Proto-Pomeranian, adding it was a mistake. We have nothing to make it work. It is probably not different from Proto-Polish and stuff like that. Sławobóg (talk) 15:00, 30 January 2024 (UTC)Reply
I'm going to speedy delete this tomorrow - I think everyone who has an opinion has spoken. As to the grouping, I have a new idea that I will bring up elsewhere. Vininn126 (talk) 13:27, 5 February 2024 (UTC)Reply
@Vininn126: It turns out there were at least half a dozen entries linking to Proto-Pomeranian words in either the etymology or the descendants. There were a couple of those where the proto-form was followed by its alleged descendants, which I fixed by simply replacing it with the word "Pomeranian:". The rest need to be dealt with. Chuck Entz (talk) 04:56, 7 February 2024 (UTC)Reply
@Vininn126: Sometimes it takes a few days (widely transcluded modules can take langer- there are still errors popping up from a mistake someone made and immediately fixed last week).
A better method is to use insource:"zlw-pom-pro" in Special:Search with the option for searching all namespaces checked. The single remaining one is in the Reconstruction namespace (There's also the archived rfv discussion for the reconstruction you deleted, but I'm not sure what to do about that). Chuck Entz (talk) 15:42, 7 February 2024 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Splitting Khanty Languages
A long-standing issue with Ob-Ugric Languages as a whole (which I tried pointing out with the request for splitting Mansi as well) is that they are categorized under one language, while they could be considered separate languages, one fact is due to mutual unintelligibility.
But this talk is considering Khanty languages. We could do the way that it is on glottolog with the Khantic being the root of the all other Khanty languages.
You can also find other sources for this on the List of languages of Russia (v2023) (in russian) site with the 143-146 numbers on the list
@Nyuhn I know you've only started to work on the Khanty language not so long ago, but I especially would love to hear your thoughts on this matter :] Ewithu (talk) 13:02, 6 February 2024 (UTC)Reply
Yes almost all western researchers have been on board for decades now with "Khanty" as being at least three languages. Russian scholars are catching up to it too (e.g. a study was recently advertized as "Khanty dialects found to differ more than Slavic languages"). The minimal division would be Northern Khanty, Southern Khanty, Eastern Khanty as still used by e.g. Salminen in Routledge's recent (2023) The Uralic Languages, 2nd Edition handbook. In terms of practicality for someplace like Wiktionary, there's three or four essentially different literary standards for Northern (Kazym, Shuryshkar, Obdorsk; also the by now little used Middle Ob) and two or three for Eastern (Surgut, Vakh-Vasyugan, nascently Salym) which each could be treated as different languages too. The division of Northern is not highly in line with what comparative studies would have to say (many field doculects could not be nicely fit to it) but for anything documented from literary use, this would save a lot of "(Vakh dialect)" "(Obdorsk dialect)" comments. --Tropylium (talk) 12:48, 7 February 2024 (UTC)Reply
If the literary standards represent mutually intelligible dialects I would suggest keeping them unified. This is somewhat similar to the situation with Galician, where there are two competing literary standards ("standard" and "reconstructionist", with the former made to look like Spanish and the latter to look like Portuguese), and we do use labels to handle the differences (and if the same spelling is used for both, there are two verb tables under Conjugation). The main issue I see with creating separate languages is it leads to duplication as you can't simply point the word in one literary standard to another using {{alt form}}. Benwing2 (talk) 21:18, 7 February 2024 (UTC)Reply
I don't know too much about Galician nor Occitian, but if the vocabulary difference would be our biggest problem between the Khanty dialects, we would be ok to keep it unified.
But since between Khanty dialects, even the grammar differs. I don't know how we could handle the different ways nouns decline or verbs conjugate if we keep it unified. Ewithu (talk) 22:10, 7 February 2024 (UTC)Reply
@Ewithu What I mean is, split based on mutual intelligibility, hence maybe a 3-way split or however-many-way split, but not split dialects just because they have different literary standards, if the literary standards refer to mutually intelligible dialects. Grammar also differs among Galician and Occitan standards and it's handled just fine. Benwing2 (talk) 22:49, 7 February 2024 (UTC)Reply
Komi is splitted into three languages, Udmurt is in two, Selkup is also in two, Mari is in three (should be Northwestern one two by the way), Koibal has its own section despite being essentially a Kamassian dialect. I think splitting it into big chunks would really help reduce messiness, especially if there are also going to be dialects written down.
@Kaarkemhveel I am not proposing mixing everything together but splitting based on mutual intelligibility. Glottolog for example has 5 Khantyic languages and 3 Mansic languages. What I'm concerned about is splitting every proposed literary standard into its own microlanguage; this would give us 8+ Khanty languages and I don't know how many Mansi ones. Things like "Koibal has its own section despite being essentially a Kamassian dialect" IMO should not be considered precedents for micro-splitting. Benwing2 (talk) 00:57, 8 February 2024 (UTC)Reply
Right, that sound reasonable. In my head at least it looks like Northern/Eastern/(Salym?)/Southern Khanty and for Mansi either Nothern/Eastern/Western/Southern or Northern/Eastern/Western&Southern (though the last suggestion based rather on their current use status).
Dear EnWiktionary bosses! I must keep you informed that so called 'Khanty', 'Koryak', 'Selkup', 'Nenets' and some other languages which are used to be considered as united languages have never been a matter of fact of the reality but only a product of early-Stalin national policy. Speaking about Khanty subfamily, the main sources have been already provided above. Khanty subfamily consists of at least 4 languages: Vakh-Vasyugan with Vakh literary norm; Surgut with its literary norm; Southern with no actual written use afaik; and Northern with previously used Middle Ob literary norm and currently used Kazym, Shuryshkary and Lower Ob literary norms. All literary standards I mentioned have, for instance, schoolbooks presently being released, some of them I have in paper on my right hand. The languages of Khanty group (as well as Sami to give an understandable parallel) are not mutually intelligible. For instance, Northern Khanty morphology is very close to the Northern Mansi one with 3-4 noun cases while Vakh-Vasyugan has 10-13 cases like Southern Selkup which is believed to be the remaining of the archaic state because of the close contact with Selkups. So people applying for splitting Khanty just range things as the things are, moreover they are ready to work. Grigoriy Korotkih (talk) 09:50, 8 February 2024 (UTC)Reply
Khanty languages could be considered separate languages. Most of the dictionaries describing them through dialects or subdialects (говоры). Those division already presented in the Wiktionary using regional categories.
How it would be beneficial to the Wiktionary to have three Khanty L-2 sections on the "тур" page instead of one?
Also we will need to somehow handle translations and etymology links from other languages, that refer to unspecified Khanty. Nyuhn (talk) 09:23, 9 February 2024 (UTC)Reply
Should we merge East Slavic languages because they have words like нога or зуб then? Don’t different Khanty lects have different inflection paradigms, pronunciation, even use sometimes? Speaking of refering to "unspecified Khanty": wouldn’t it be beneficial to the Wiktionary to specify it, rather than pile everything into "Khanty"? Kaarkemhveel (talk) 12:47, 10 February 2024 (UTC)Reply
So in light of all that has been discussed, we should be good to split both languages, according to the cardinal directions of the North East South, and West. We can use labels to further specify which dialect it came from. Naturally each separate language (represented by the 4 cardinal directions) will have to (if possible to find) have an {{alt form}} section to make switching between dialects (represented by the river they lived beside traditionally) easier.
Support or Oppose? Ewithu (talk) 09:59, 24 February 2024 (UTC)Reply
@Ewithu Northern/Southern/Eastern/Western seems correct for Mansi but for Khanty maybe it should be Northern/Southern/Eastern per Wikipedia; Western Khanty appears to be a larger grouping consisting of Southern and Northern. I agree with the idea of identifying dialects by rivers. (We also ended up doing this for the CAT:Ye'kwana language, inspired by the Khanty and Mansi situation, as using river names was the only unambiguous and accurate way of referring to different dialects that we could come up with.) Benwing2 (talk) 10:20, 24 February 2024 (UTC)Reply
@Ewithu I think what we need to do is as follows (for Khanty, similarly for Mansi):
Assign language codes kca-nor, kca-sou and kca-eas to the split-out languages.
Assign a temporary family code urj-kca to the new Khantyic (Khantic?) family; see below.
Set up tracking for all uses of kca as a language code. (This will make it easier to find the references to this code so they can be changed.)
Split the current {{kca-*}} templates into per-new-language templates. Thankfully there are only 4 templates. I assume the pronouns listed in {{kca-table-ppron}} are Kazym pronouns. I'm not sure if all the Khantyic languages use the same set of cases; if not we need to modify the new declension templates to have the appropriate sets of cases in them.
So I would say split it 4 way, because on one hand, the western dialect is extinct, while the eastern (is also extinct in theory, but not proven yet) is extant, and western has a lack of Cyrillic dictionaries or sources (except that one Gospel of Mathew written in Lower Lozva most likely), while Eastern even has audio pronunciations on Lingvodoc. While I would also say keep it as Central Mansi, because those two languages have very few resources but I don't this is a really valid argument. Ewithu (talk) 06:17, 28 February 2024 (UTC)Reply
@Ewithu My general take is we should go by mutual intelligibility unless there's really a good reason to do it otherwise. Since Glottolog seems to think the two varieties are mutually intelligible (at least that's what I assume by the Central grouping) and there are few resources overall, I think it's easier to have fewer splits; we can always split later if the need arises. But I don't feel super strongly about it; maybe someone else can chime in. Benwing2 (talk) 06:23, 28 February 2024 (UTC)Reply
I can't actually confirm your assertion about Google Books; "Khantic" does seem to occur more often but most of the hits seem to be garbage. Benwing2 (talk) 07:09, 28 February 2024 (UTC)Reply
7 should be done. A few cases were left with etymology templates that used kca as a source, but those should work, as it now means "from one of the Khanty languages". Most references were updated; some others were indiscriminately changed to kca-nor with attention templates, as I was requested to do. — SURJECTION/ T / C / L /22:06, 28 February 2024 (UTC)Reply
Done. The remaining instances in mainspace at least are valid uses of the family code. I'll check the other namespaces (at least the ones that matter) as well. — SURJECTION/ T / C / L /22:22, 28 February 2024 (UTC)Reply
@Ewithu: If the differences in grammar, phonology and morphology of Eastern Mansi and Western Mansi are not too large, then the fact one is little-attested and extinct is all the more reason to keep the two together. Thadh (talk) 09:49, 28 February 2024 (UTC)Reply
Considering how often etymological sources distinguish between Western and Eastern Mansi though, we may want to consider adding them as etymology-only languages. — SURJECTION/ T / C / L /08:13, 29 February 2024 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Some missing Sino-Tibetan languages
Hi, this is a proposal to add three new language codes, relating to a small family of Bai(-adjacent) languages which have been dubbed the Cai-Long languages (from Chinese 蔡龍語支), spoken in western Guizhou.
@Stevey7788 very kindly created an expansive vocabulary list of Greater Bai Macro-Bai languages several years ago, which covers all of these in detail, including a number of sources (all of which are in Chinese, unfortunately).
The only extant member of the family, with approximately 1,000 speakers in western Guizhou, with the numbers having declined from over 10,000 in the 1980s. Just over 950 terms are listed in the vocabulary list, including those of the dialects spoken in Shuicheng and Yangjiazhai.
@Wpi Macro-Bai, and then giving the core Bai languages (i.e. the ones we already have language codes for) their own language family within that simply called "Bai languages". I've checked, and I can't find any evidence the term "Greater Bai" is actually used outside of Wiktionary or Wikipedia, so we should definitely fix that. Theknightwho (talk) 17:17, 16 February 2024 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Splitting Northern Bai properly
Currently, we use the code bfc for Northern Bai, classifying it as one of the Bai languages. However, in 2014, bfc was split into Panyi Bai and Lama Bai ([10][11][12]). Lama Bai was given the new code lay, while Panyi Bai retained the old code bfc. At some point since then we added Lama Bai to our list of languages, but we never renamed bfc to Panyi Bai, which seems to have been an oversight. From the documents, it seems like they still do form a "Northern Bai" grouping, however. Theknightwho (talk) 02:25, 17 February 2024 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Anglicising the names we use for the branches of Min Chinese
On the back of the recent split of Min Nan into a family, I thought it would be a good idea to consider renaming the various branches of Min to the names usually used in English. With the exception of Coastal Min and Inland Min, the names we currently use are transliterations from Mandarin, whereas most sources - including academic sources - prefer to translate them into English. I have no doubt we only do this because the ISO uses the transliterated names, for whatever reason. As such, I propose the following:
mnp: Min Bei → Northern Min
cdo: Min Dong → Eastern Min
nan: Min Nan → Southern Min (note: currently being converted into a family)
czo: Min Zhong → Central Min
cpx: Puxian → Puxian Min (edit: added per the discussion below)
It's difficult to use Google Ngrams with this, since there are a ton of false positives, but the general trend is very strongly towards the English names historically, with the transliterations increasing in popularity in recent publications. However, taking a look at Google Books results, a lot of the transliterations come from poor quality sources, or in many cases texts that simply contain the ISO language names. Google Scholar is somewhat more helpful, though, and shows a clear trend towards the English names even with very recent papers. Theknightwho (talk) 18:02, 4 March 2024 (UTC)Reply
Support; this is also consistent with general practice here at Wiktionary to prefer English terminology rather than native-language terminology. Benwing2 (talk) 00:15, 5 March 2024 (UTC)Reply
Not sure here. Wikipedia uses dashes in the names of many languages (e.g. most of the Yue languages) where Glottolog uses closed compounds. Glottolog uses the form Pu-Xian with a hyphen not en-dash (although Omniglot uses Puxian). In general I prefer less punctuation than more so my gut would say Puxian but properly you'd have to look at scholarly sources to see which one is most common. Benwing2 (talk) 06:19, 10 March 2024 (UTC)Reply
It is because the name 莆仙 Pu(-)Xian is formed from two placenames, 莆田 Putian and 仙遊 Xianyou. This is also the case for the Yue dialects listed on Wikipedia. —Fish bowl (talk) 06:24, 10 March 2024 (UTC)Reply
@Fish bowl "Puxian Min" seems to be a lot more common than "Pu-Xian Min" in English-languages sources, but I agree it should be renamed as well. I'll add it to the list at the top. Theknightwho (talk) 06:34, 10 March 2024 (UTC)Reply
Latest comment: 4 months ago5 comments4 people in discussion
The following discussion has been moved from Wiktionary:Requests for moves, mergers and splits ([[Special:PermanentLink/78966576#Change_"Chinese"_[zh]_from_descendant_of_"Middle_Chinese"_[ltc]_to_ancestor_of_"Old_Chinese"_[och]|permalink]]).
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Change "Chinese" [zh] from descendant of "Middle Chinese" [ltc] to ancestor of "Old Chinese" [och]
@Fish bowl I'm not a Chinese editor, but from a outside perspective, that'd feel more weird to see (how can Chinese be the ancestor of Old Chinese?). It's also make the Chinese lects go like "Chinese -> Old Chinese -> Middle Chinese -> lect", which seems more confusing to me. Honestly, at this rate. I'd just remove Chinese from the family tree entirely with its current treatment. Or, at least make it on the same level as Old Chinese, rather than an ancestor. AG202 (talk) 04:33, 4 March 2022 (UTC)Reply
I don't see a problem with the current setup...? As AG says, making modern Chinese an ancestor of Old Chinese would be weird (and wrong). I think the current setup is fine...? The only awkwardness is that we group all kinds of Chinese under one L2. I'm going to <s>strike</s> this as no action taken so it can be archived soon, since the discussion is very old and this page is very large, but you can reopen it if you object... - -sche(discuss)05:33, 31 March 2024 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Language codes for the proto forms of upper branches of Dravidian (south, south-central, central, north)
Has been discussed many times and was also agreed up to add them like here and here
The proposal was because there are tons of terms which are restricted to certain branches, sometime being only attested in like 4 languages of 1 branch (*cinki for example) in a family of 80+ languages with 4 branches but are reconstructed to the proto stage, {{R:dra:DL}} has many reconstructions for the inner branches.
{{R:dra:Southworth}} mentions many cases where BK reconstructs terms to PD when they are restricted to certain branches like PSD or PCD further saying it should be reconstructed only to the proto branch only. AleksiB 1945 (talk) 13:21, 21 December 2023 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Adding Proto-Bai
While we do currently have a family covering the Bai langauges (sit-bai), we don't have the code sit-bai-pro for Proto-Bai. There are a few publications out there which give Proto-Bai reconstructions, and our Macro-Bai comparative vocabulary list contains 469 from Wang Feng's Comparison of languages in contact: the distillation method and the case of Bai (2006). I'm not suggesting that we blindly create entries for all of these, but we already reference Proto-Bai in four entries anyway: Lama Bai ɕy³³, Southern Bai ɕy³³, Central Bai xuix and Chinese 山(shān), so there's already a need for the code. Theknightwho (talk) 03:13, 17 February 2024 (UTC)Reply
Both of these names are common in the literature, but newer and more professional publications seem to prefer Nùng, which I suspect is down to typesetting no longer being an issue.
It matches our general treatment of other languages in Vietnam, such as Tày, Ná-Meo, Ts'ün-Lao etc.
Other than headings, this spelling is already used in entries in the form of labels (e.g. bân, nãhm, nưhng, slao, slíhm).
@Theknightwho I think you should have waited on this. You only waited a week and no one supported the move. I don't agree with the general principle that we should include all the native-language accents and other Unicode chars in the Wiktionary names of languages. These names should reflect the *English* usage of such names, not the native-language usage. Possibly the name change was justified in this particular case but I don't want any precedent set that would justify e.g. moving the name O'odham to ʼOʼodham (or even worse, a half-ass rename like Kwami did to the Wikipedia article Oʼodham language, which includes the "Unicodified" apostrophe in the middle of the name but omits the apostrophe in the beginning). Benwing2 (talk) 05:02, 25 February 2024 (UTC)Reply
@Benwing2 I'm not trying to set any precedent - I just noted that it's generally spelled with the diacritic in more recent (English) publications, and our entries already spelled it that way outside of the headings. Given no-one opposed it, it didn't seem like an issue. I'll wait longer in future, though. Theknightwho (talk) 05:21, 25 February 2024 (UTC)Reply
RFM discussion: Move Hachijo Japanese to Hachijo language & more (February–March 2024)
Discussion suggested by @Theknightwho. Please share any thoughts you may have.
Recent scholarly consensus is that Hachijo is distinct from Japanese, being one of the earliest branches of if not outright parallel to (Old) Japanese. Hachijo should thus warrant a conversion into a language.
Similarly, Tugaru-ben and Satuma-ben could also be elevated to language level, though the exact classification on Kyuusyuu lects and the affinity of Tugaru-ben with the rest of Japanese is still not widely agreed upon. This is to be discussed. I am personally ambivolent on the issue though this has already been brought up on preliminary discussions on the enwikt Discord.
Support. I am not knowledgeable enough to have an opinion on this, but we will need to make a decision on whether it's a descendant of Old Japanese or not before adding this. (Also, less important, but I'd prefer we used the name Hachijō with the macron.) Theknightwho (talk) 01:50, 19 February 2024 (UTC)Reply
Good choice. I will raise it in a separate thread, but the issue of spelling also affects some of the Ryukyuan languages, where the names we've lifted from the ISO are suboptimal and aren't really used outside of contexts directly related to the ISO standard itself. Theknightwho (talk) 19:40, 20 February 2024 (UTC)Reply
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
Making Sichuanese a full language
We currently handle Sichuanese (cmn-sic) as an etymology-only variant of Mandarin. Four things to consider:
Its treatment in Chinese entries is on the same level as that of Mandarin (proper) and Dungan: i.e. it is listed in the "Mandarin" section, in which "Mandarin" is used in the broad sense. Given that Dungan (which is treated as a full language) is also listed there, there's no reason why Sichuanese should be any different. See 女, which gives Mandarin 女(nǚ), Sichuanese 女(nyü3) and Dungan 女(nü, II). Note that Sichuanese transliteration is already handled automatically, since the transliteration module handles it as an independent lect.
Sichuanese is Southwestern Mandarin, whereas we use Mandarin to refer pretty much exclusively to the modern standard of Beijing Mandarin (except for that heading in the Chinese pronunciation table, already mentioned). They are pretty distinct.
As with Hokkien and Hainanese (as mentioned in #Converting Min Nan into a family), the designation of etymology-only language seems to have been arbitrary, and was likely influenced by the fact it doesn't have its own ISO code.
The line is much blurrier here, but it was raised in a past thread about Shanghainese (re the category Category:Shanghainese Chinese) that categories with names such as Category:Sichuanese Mandarin cause confusion, because of the distinction between Sichuanese [Mandarin] and standard Mandarin as spoken in Sichuan.
Two further considerations:
We might instead want to make Southwestern Mandarin a full language instead, with Sichuanese itself being part of that.
The code would likely need to be changed, as retaining cmn-sic would be misleading if we split it out of cmn.
Either along provincial boundaries and make areal groups (they're all similar enough) or just use Li's atlas. It just feels kind of weird to have "Southwestern Mandarin" as a language since it's very much a linguistic concept — 義順 (talk) 08:13, 25 February 2024 (UTC)Reply
Yes, it would. If you don't see it being practical I can make it practical by adding a few hundred Wuhanese/Kunmingese entries next Sunday. Plus, I really don't see any other functional alternative for resolving this issue — 義順 (talk) 08:25, 25 February 2024 (UTC)Reply
@ND381 I have no opinion - I just want to make sure we don't add Sichuanese and then find we need to move everything to SW Mandarin later on when adding other SW Mandarin lects. If you think they need to be separate languages then that's totally fine. Theknightwho (talk) 08:49, 25 February 2024 (UTC)Reply
Right, thanks for the clarification. I really don't see Southwestern Mandarin needing a language-level code any time in the future, as we have Wuhanese implementation and expansion set as a long term goal (see here) and, as mentioned, Southwestern Mandarin doesn't really act like a single coherent unit outside of linguistics even when compared to ones we have like Xiang or Jin, which have "Hunanese" and "Shanxinese" connotations respectively — 義順 (talk) 08:57, 25 February 2024 (UTC)Reply