Wikidata:Requests for comment/Need comment on English item label and foreign language proper noun

An editor has requested the community to provide input on "Need comment on English item label and foreign language proper noun" via the Requests for comment (RFC) process. This is the discussion page regarding the issue.

If you have an opinion regarding this issue, feel free to comment below. Thank you!

Summary: should we used the translation of the foreign language proper noun as (the primary) English wiki item label

User:Adamant1 keep on changing item Banca Carige (Q3633695) (and may be other item as well , e.g. Credit Agricole Group (Q590952), Cooperative Credit Bank of Rome (Q25060394), etc.) to hoax value "Carige Bank". The bank, which called itself "Banca Cargie" in its native Italian language, also called itself "Banca Cargie" in its English website. The English wikipedia also used "Banca Cargie" as article title. Thus, it just no sense in wikidata the English item label for the item Banca Carige (Q3633695), need to change to self translated (or unpopular fork translation of a proper noun) "Carige Bank".

Please instruct if this happen next time, the issue should report here, Rfc, or admin noticeboard. Matthew hk (talk) 02:27, 14 April 2020 (UTC)[reply]

For source, Reuters also use "Banca Carige" in its English language newswire article. [1] Matthew hk (talk) 03:11, 14 April 2020 (UTC)[reply]
So did The Economist [2] Matthew hk (talk) 05:37, 14 April 2020 (UTC)[reply]
The Guardian used both Banca Carige and Carige Bank [3][4]
WSJ use "Banca Carige " [5]
NYTimes use "Banca Carige " [6] Matthew hk (talk) 13:48, 14 April 2020 (UTC)[reply]
google ngram unexpectantly show the historical full name "Cassa di Risparmio di Genova" is the most popular, Banca Carige the second and not found for Cargie Bank. Feel free to refine ngram search term Matthew hk (talk) 13:51, 14 April 2020 (UTC)[reply]
  • @Jura1: No I don't and I shouldn't have to. Banca is bank in English and it's called the English label. That's a good enough reference for me. There is no guideline about it either. Only a failed proposal that didn't go anywhere. For good reason IMO, because having an English label where your not allowed to use English words would be completely nonsensical and completely worthless. Also, as I told Matthew hk nothing makes the sources he cites authoritative either and I'm not going to get into a pissing contest of "my magazine has more clout then yours does." Further, there are many examples out there of English brand names being translated into other languages (Walmart translated into Chinese being one) and I see no particular, unique reason the English label should be forced to be the dumping ground for every other language because routers or some random paranoid user says so. Lastly, I really don't appreciate Matthew hk's repeated insinuations that my translations are hoaxes. It's unproductive and needlessly pisses me off. Plus, it just makes me less inclined to give a crap about this. Which doesn't help anyone. --Adamant1 (talk) 09:15, 14 April 2020 (UTC)[reply]
I will reference this though, where the Vatican calls the cooperative Bank of Rome, exactly that as an example of how this is mainly just a none issue created by a paranoid user. It would be ridiculous if we had to provide sources every time we wanted to translate a word, a proper noun or not, just to satisfy someone with a personal prefrence against English who thinks people are making things up. Its a translation, not a citation in Wikipedia. It doesnt need to be varified. Otherwise there would be endless edit waring and bickering about what language should be used in what label and who's sources are better etc etc. Seriously. This whole thing is ridiculous. The English label is for English. Get over it. --Adamant1 (talk) 10:24, 14 April 2020 (UTC)[reply]
So, how do you explain Reuters call it "Credit Agricole" [7] not you alleged "common usage" Agricultural Credit Group? We don't carry self-translation in wikidata, but reflection/c&p common usage from reliable source. Please stop label yourself authoritative. Matthew hk (talk) 13:48, 14 April 2020 (UTC)[reply]
Or please provide reliable source for the translation "Dupuy Bank of Parseval" for Banque Dupuy de Parseval (Q2882897) while this seem reliable site in English language https://thebanks.eu/banks/13658, call it Banque Dupuy de Parseval? Would you translated human name? Where did you get the policy to self-translate proper noun. Matthew hk (talk) 13:55, 14 April 2020 (UTC)[reply]
I don't know or care why Reuters calls the bank what they call it, because it's completely irrelevant. Just like it's irrelevant why the Latvian label for potato says kartupelis etc etc. Again, labels aren't like facts in Wikipedia where you need to cite a source to "prove" they are correct. Words are just words. I'm not saying I'm an authority. I'm saying there isn't one. Nor does there need to be. There isn't an authority to "prove" the word kartupelis is potato in Latvian. It would be utterly ridiculous to suggest there is or that we need to consult one, because it's common knowledge. Just like it's common knowledge that the word for bank is bank in English. Not banco or whatever. It would be garbage to suggest kartupelis be changed to potato because potato is used more and that's what some magazine calls it. Which is your only argument here.
Seriously, people create and edit labels all the time, based on how things translate into their language without consulting a professional linguist. They aren't making up words by doing "self-translations." Nor are they "language hoaxers" by calling a potato a kartupelis. It would be utterly worthless nonsense to go on a crusade to route out "language hoaxers", like your doing here, to find out what authority they consulted before using the word kartupelis. Your whole argument is basically based on nothing except baseless, vague, paranoid accusations and a failed proposal. Since you clearly have no ground to stand on, I suggest you let this go and leave English labels in English. Otherwise, I'll report you for edit warring and making slanderous accusations. --Adamant1 (talk) 18:29, 14 April 2020 (UTC)[reply]
I will wait for the boomerang, please go ahead to report to admin and please mention the evidence in this rfc and talk page. Matthew hk (talk) 08:27, 15 April 2020 (UTC)[reply]
If you keep reverting me, continue modifying English labels I've edited in the way that you have, or make any more false claims of "language hoaxing" about me I will. --Adamant1 (talk) 12:20, 15 April 2020 (UTC)[reply]
  • Do you have samples of such translations of company names elsewhere within Latin script languages?
    Obviously there are organizations with official names in several languages. Sample: "World Health Organization" is called "Organisation mondiale de la santé", but that's its official name. Such official names can change.
    Even if it wasn't a proper noun and all uses in English are the one Matthew mentions, this is what would be used. If you want, you could use literal translation (P2441) for a literal translation (to qualify the official name). Maybe "world organization of health" to qualify "Organisation mondiale de la santé"? Admittedly, such literal translations can remind us of Q54. --- Jura 12:43, 15 April 2020 (UTC)[reply]
I only knew UniCredit is known as UniCredit in English, Germany, and Italy and noone call it for literal (self made-) translation as "United Credit ". Seriously, if he just refuse to provide reliable source can i just report as vandalism? Matthew hk (talk) 14:42, 15 April 2020 (UTC)[reply]
@Adamant1: Regarding your first reply to Jura1, keep in mind that the label is the most common name an item is known by. That's why sources are needed. Changing the item's label without first discussing it on enwiki is a terrible idea, unless you provide sources that demostrate you're right. Like said above, some organizations have official names in many languages, but you haven't proved this is the case. Without evidence you're just acting recklessly, and it would be better if you abstain from such actions. Esteban16 (talk) 00:35, 16 April 2020 (UTC)[reply]

@Matthew hk: Where would I even enter a source when changing the label? It doesn't even have an option to leave a changeset comment when writing or changing labels. Which proves how low on Wikidata's priority list that we provide sources before changing label is and how much your blowing this out proportion, because there's literally no way to provide one. If Wikidata cared, there would be. Or is just that we only have to provide a source when people like you come along and make a massive issue out of it? You keep going off about how I should provide a damn source, but you haven't provided one for a policy that says I have to or even how I can. That's on you. Doing something that has literally guideline or rule about isn't vandalism. Feel free to report me though. I could really give a crap at this point, because your just ignoring what I'm saying, repeating yourself, and continue to attack me when I asked you not to. So, I have zero urge to deal with you anymore, because it's pretty clear you don't actually care about resolving it.

@Esteban16: Jura1 wrongly thought that the issue was about statements. As is evidence by the "sources" link the user provided. Which this has nothing to do with. There is no way to add sources to labels. So, no I didn't provide sources originally and I'm not going to. Especially with the way Matthew hk is acting about it. Even if he was being reasonable there is no guideline or even a way to add sources to labels anyway. See my comment about it above. Not that I don't have any sources though. I provided one above. I'm not going to waste my time providing sources for every edit I've done though just because someone with an extremely bad attitude like Matthew hk has a personal issue. Plus, I made the changes to label a pretty long ago and I don't keep records of everything just in-case. It's not on me if Wikidata doesn't provide a source field for labels.

Just to repeat myself so we are clear, I have to think if Wikidata required or cared about providing sources for labels they would have a built in way to do it. That isn't just people calling each other out or users doing "translation hoaxer" witch hunts, like Matthew hk is doing. There would at least be a guideline about it if it's something we should be doing.

As far as you saying it is reckless to change labels or that it should be discussed first on enwiki, that is totally nonsense. There are many perfectly legitimate reasons people change labels and it happens all the time, without issue or anyone caring. Labels are also constantly being created without discussion first or the creator entering sourcing into some nonexistent phantom "provide source here" entry box. Let alone is there anyone like Matthew hk looking over their shoulder badgering them to pony up an "authority" to prove their label is "proper." Most of the time it's done based on nothing except that the person thinks the label is correct. Which is perfectly fine IMO. The labels aren't set in stone and can be edited for a reason. There is no "one and done" rule in Wikidata. Your assertion otherwise and that we need to appeal to a higher authority for permission first before editing them is simply wrong. Just to be clear, I'm 100% fine sourcing things on Wikidata that need sourcing where it provides a way to. That isn't the case with labels though. There is no way to add a source to a label and there's nothing in the guidelines saying that you need to. --Adamant1 (talk) 01:57, 16 April 2020 (UTC)[reply]

What is 100% acting recklessly though is Matthew hk reverting users based on a personal opinion that they don't have a source or because he thinks they are just "language hoaxers." More so because he didn't discuss it first. For all he knew they could have been correct translations, but apparently he thinks it's fine revert someone based solely on paranoia and what Reuters calls something. Both of which are totally reckless reasons to revert someone. --Adamant1 (talk) 03:10, 16 April 2020 (UTC)[reply]
OK. I was aware of the options. I don't think either one would work in this case. That said though a reminder of what possibilities are out there is always good. --::Adamant1 (talk) 05:20, 16 April 2020 (UTC)[reply]

In response to User:ArthurPSmith suggestion in admin noticeboard , the rfc question should more refined as follow:

  1. There are many case lack of English language reliable source for foreign language proper noun, so how these label should be treated . literal translation by the editor is allowed or not?
  2. There are in some case, the foreign language proper noun also became a common usage in English media. For example, the football club is just known as SSC Napoli but never SSC Naples, so did Banco Popolare (Italy) Banco Popular (Spain). But in some case, just like the bank Banca Cargie, despite the usage of "Banca Cargie" is more popular in English according to ngram (even more popular for "Cassa di Risparmio di Genova" not sure why), the usage "Cargie Bank" can also be observed. Even more counter example like "F.C. Internazinoale Milano", is more commonly called "Inter Milan" and some times "Internazinoale" (see all the RM discussion in en-wiki). It don't need discussion if the noun is appeared in English media, then all of them should eligible to label parameter in wikidata. However, how the primary English label be determined? Any metric way to determine the most common usage? Ngram alone or the article title in en-wiki? If have one?


Hope it is more clear to discussion despite offtrack by vandal. Matthew hk (talk) 15:29, 16 April 2020 (UTC)[reply]
  1. There's plenty of examples where a literal translation is completely appropriate. Like with the World Health Organization above. Which is probably why there is no hard and fast rule about it.
  2. Because of the inherent ambiguity (per your examples), lack of varifiability, and uncertainty involved (plus the good faith rule) the best way to handle it IMO is to just go with English in the English label like it was intended and assume good faith (the person had a valid reason to translate it etc etc). I see nothing anywhere to indicate we should do otherwise.
  3. There is no way to determine "the most common usage", nor is that the rule. If so, there wouldn't be any labels for most things, even proper nouns. Organisation mondiale de la santé isn't the common usage, but it's fine to use in Wikidata. Same goes many other proper nouns. Just because it might sound "clunky" or "wrong" doesn't mean it is.
  4. I have as much right to participate in the discussion and voice my opinion as anyone else does. Especially since I was pinged and my actions where being miss characterized. Also, I brought up some valid questions IMO that have yet to be answered. If they were, this probably would have been resolved already, at least on my end, and you'd be free to scream into the void about the big bad vandal without me being involved in it.
  5. It's pretty clear from your attitude everywhere that the only answer your willing to accept is the one that fits your agenda and personal opinion of how things should be done. One user already answered you that it was OK to literally translate proper nouns in some cases and no one else has explicitly said otherwise. Your probably never going to get a 100% clear cut answer that can be applied in all cases because there isn't one. A few people's opinions in an RfC doesn't dictate policy anyway. --Adamant1 (talk) 20:26, 16 April 2020 (UTC)[reply]
You need to understand someone else idea. WHO had official name in other language as it is a special agency of UN. Translation/Mentioning by local media such as Chinese/English/French/German/Korean are readily available, so that easily to find reliable source to prove the common name in their local language. "Banca Carige" are readily used by English media and it is you to insist something else as the primary label of the entry. Lastly, in some case, wikipedia somewhat work on counting the consensus of majority people. And Please find people to support your fringe idea. I think you can somewhat canvass people to rfc BTW. Matthew hk (talk) 20:32, 16 April 2020 (UTC)[reply]
Which idea is it that I'm not understanding? On the WHO, that's exactly my point. If someone sees the WHO mentioned in their local language on their TV and enters that spelling into Wikidata, it's not invalid because we didn't see the TV show and they can't replay it for us. A lot of more minor languages just don't have the easily accessible sourcing ones in the "west" do. That doesn't mean they shouldn't have their languages represented in Wikidata. I wouldn't have sources for something I did six months ago either. That doesn't mean the English spelling of the label should be deleted or changed. Seriously. What "idea" of mine exactly is "fringe"? Everything I stated is a fact. --Adamant1 (talk) 23:38, 16 April 2020 (UTC)[reply]
  • Not sure if it was clear: WHO is an official name that is defined in several languages. Accordingly it's "World Health Organization" and not "World Organziation of Health". One needs to keep in mind that Wikidata is a secondary database. Information should be available elsewhere before being added here. --- Jura 20:35, 16 April 2020 (UTC)[reply]
I 100% agree with that. The issue is that with labels there is no way to show that it is available elsewhere when you enter it. Which isn't on me. A policy of "provide the source when someone questions you", which is how Matthew hk is treating this, wouldn't work for many reasons. I'd also say that many valid translations are added through ways that can't be sourced like using a translation service or having local knowledge. I'm sure that's how most of the Chinese translations are done. They aren't invalid and there's no way to determine who is doing it or using a source we can verify. Except for witch hunts. Which again doesn't work as a policy or general way of dealing with things. It's also not a standard that any other language except English has to live up to. No one is out there edit warring the Latvian language labels to force them into using the Chinese ones. There's no reason English labels should be singled out as the de-facto dumping ground for all other languages. Otherwise, why even have the labels in the first place? I'd seriously like to know, if there purpose isn't to translate labels into those languages, because there is one single correct language for a label, why are they even there?--Adamant1 (talk) 22:18, 16 April 2020 (UTC)[reply]
100% running in circle. Please just read Wikidata:Verifiability or en:Wikipedia:Five_pillars before expressing your idea on how wikipedia and wikidata should be run. Matthew hk (talk) 22:45, 16 April 2020 (UTC)[reply]
Why can't you answer a simple question without a personal attack or accusation? The message wasn't to you. Your the one accusing me of canvassing. That's exactly what your doing by responding to everything. Also, your the one expressing your idea on how Wikidata should be run by repeatedly asserting that a proposal for a policy has to be followed and that anyone who doesn't is a vandal. This whole thing is 100% founded on your assertions. Wikidata:Verifiability says nothing about how people need to provide sources to translate labels. Which is what your asserting. And en:Wikipedia:Five_pillars isn't relevant because it's a separate project, doesn't apply to this situation, and "Wikipedia has no firm rules" would contradict you that this is an issue anyway. Btw, which is also why I don't think the title of a Wikipedia article should dictate a label in Wikidata. Do you have anything on topic to say or an actual response to my questions? Or is all you have to say personal attacks? --Adamant1 (talk) 00:09, 17 April 2020 (UTC)[reply]
  • Well, there are several ways to comprehend the foreign word, like, transcription, transliteration and sometimes loan translation. The preferred one generally depends on context and the field in which the translation happens. I think, in general, one should prefer the name from corresponding reliable sources in this language to be used in the language section. Only if there are no translations in reliable sources at all one may consider to make "self-translation". As for the main title in case when several options appear in reliable sources, one should stick to the one most recognisable (usually one in English wiki). If there is close match, it should be decided by consensus on case by case basis.
P. S. It's funny how my username is similar with the user who is the cause for this RfC. Adamant.pwn (talk) 03:49, 19 July 2020 (UTC)[reply]

The Belgian railway lines' items seem likely, someone keeps on naming them to be "Line X (Infrabel)" even they know that Englishman will never call so. --Liuxinyu970226 (talk) 02:29, 9 September 2020 (UTC)[reply]