Editing MarioWiki talk:Japanese

Is this a Writing Guideline or not? It is linked to from the policy page yet it doesn't belong to the category.--{{User|Knife}} 17:24, 8 December 2011 (EST)

:The page has been under construction and was being developed largely by {{user|Twentytwofiftyseven}}. But as he seldom appears in the Wiki now, the page is unfinished. You should talk to him about this topic. {{User:Coincollector/sig}}
::I'm not sure if there was much that needed to be added anyway. However, I've been wondering if it might be better to make this page just "MarioWiki:Japanese" and reorganize it a bit, since it already has aspects of policy that go beyond plain romanization. On a policy note, I also think we should just use plain Hepburn and allow long え and い to be romanized as ''ei'' and ''ii'' as well as ''ē'' and ''ī'' - that's what all the other wikis do, iirc. It might also be a good idea to just let proper names/titles be capitalized in foreignname, since that's how most people try to enter them anyway, and from what I've seen, Wikipedia, Bulbapedia and Zelda Wiki all capitalize names even if they're not used in the article's title. - {{User:Walkazo/sig}} 00:03, 9 December 2011 (EST)

== English capitals ==

If an English letter appears in Japanese text, should it be written literally in the romanized text (''A'', ''B'', ''C'', ...) or rendered in Japanese pronunciation (''ē'', ''bī'', ''shī'', ...)? [[User:A gossip-loving Toad|A gossip-loving Toad]] ([[User talk:A gossip-loving Toad|talk]]) 03:59, 4 February 2016 (EST)

:I think it should be written literally, and same with numbers: romanizing them just looks a bit ''too'' earnest and excessive, imho (and can make things hard to understand, since it could seem like the letter/number was skipped, especially if the romanization's done poorly, with no spaces separating the letter/number). On the other hand, I ''would'' advise using the hover-over text to include the kana that would have been used in place of the English letter/number, going by the same logic as showing the kana for kanji - I've even seen furigana used this way in manga (including ''[[Super Mario Kun]]''). The same thing could be done for the romanization itself too, for that matter, so then it'd be best of both worlds: all the info, but none of the artificialness of writing letters. - {{User:Walkazo/sig}} 12:05, 4 February 2016 (EST)

== furigana template ==

Is there a template for furigana marking? A template not only makes code cleaner, it can also make furigana available to those who don't have a mouse. [[User:A gossip-loving Toad|A gossip-loving Toad]] ([[User talk:A gossip-loving Toad|talk]]) 02:34, 13 February 2016 (EST)

== Rendaku ==

The Japanese language has a strange rule such that
:''kodomo'' + ''heya'' = ''[[Gate (Super Mario Galaxy)#Names in other languages|kodomobeya]]''
similar to how "furi" and "kana" make "furigana". Should such things be taken into account if Nintendo does not provide a pronunciation of the kanji in a name? {{User:A gossip-loving Toad/sig}}  21:46, 8 May 2016 (EDT)
: As I continued learning Japanese I found that this needn't standardisation. Shame on me. {{User:A gossip-loving Toad/sig}}  00:13, 31 May 2016 (EDT)

== Official romanizations ==
Are official romanizations which are given by Nintendo allowed? --{{User:FanOfYoshi/sig}} 12:31, 4 November 2018 (EST)

== Kanjis ==

Certain games do not use the "Hiragana above Kanjis" thing, such as [[Super Mario Maker 2]], which complicates reading for non-kanji reading people in Japan. How would they expect everyone to read Kanjis? --{{User:FanOfYoshi/sig}} 10:52, July 3, 2019 (EDT)

== Fake Dakuten R letters ==

I've found out that the "r" letters shown with the Dakuten are fake. If you try to copy-paste them, you can see that it is a fake. If there was an "l" sound, it'd have been a Handakuten. Does the "l" sound even exist in Japanese? --{{User:FanOfYoshi/sig}} 03:41, July 13, 2019 (EDT)
:From what I've seen, they can ''technically'' be used to represent non-Japanese words with L sounds. They're basically never used, though. {{User:Niiue/sig}} 04:04, July 13, 2019 (EDT)

== "So"-like letter ==

I've found a letter that looks like a "so", but looks distinct from it. I've found it [[:File:Grayguy.png|here]]. It has a Handakuten, but it isn't that that makes it looks different. It's at the end of the last sentence. --{{User:FanOfYoshi/sig}} 07:06, August 3, 2019 (EDT)
:That's just another way of writing ''so'' (or in this case, ''zo''). {{User:Niiue/sig}} 07:08, August 3, 2019 (EDT)

== Reliable translator ==

Yes i'm aware that i've been posting multiples messages in a row in htis, but what is the most reliable translator? I use Yandex translator, but i believe it has the same reliability as Google Translate. Also, i've seen names in Hiragana using this ー , see Balloon Teresa. Also, when it comes to loanwords, are they ever written in Hiragana? --{{User:FanOfYoshi/sig}} 05:57, August 20, 2019 (EDT)
:Late reply, but...
::1. Online translators generally aren't very accurate, I'd recommend cross-referencing Japanese words with a dictionary in case something was lost.
::2. The ''chōonpu'' is sometimes used in hiragana for stylistic reasons (such as 「らーめん」), though this is unusual. In the case of ''Barūn Teresa'' however, the word 「ばるーん」 is simply a transliteration of the English word "balloon", which would normally be written in katakana as 「バルーン」. For reasons unknown (probably a stylistic choice), they wrote it in hiragana instead but kept the ''chōonpu'', instead of changing the spelling to 「ばるうん」 or something.
::3. I've seen it before (see above), probably just as a stylistic choice.
:{{User:Niiue/sig}} 15:42, October 6, 2019 (EDT)

== C vs. T ==

:''Note that っち should be ''cchi'', never ''tchi'', as it is in other romanization systems.''

But ''why''? Not only is it, as noted, more commonly spelled with a "t", but it's more obvious to casuals how it's supposed to be pronounced if it's spelled with a "t". <span style="font-family:Mario Party 2/3 Textbox">[[User:RickTommy|RickTommy]] ([[User talk:RickTommy|talk]])</span> 22:05, September 15, 2019 (EDT)

: I think the rule was created to simplify the transcription rules (so that っ always doubles the next letter). --[[User:Dine2017|Dine2017]] ([[User talk:Dine2017|talk]]) 23:56, February 11, 2022 (EST)
:: By the way, here are some romanized words found on the wiki in violation of this rule: <tt>Atchimuitehoi, Atchitchi, Batchiri, Batchī, Chitchai, Dotchi, Etchi, Hotchikisu, Kotchi, Kyatchi, Matchi, Matcho, Ritchi, Suitchi, Suketchi, Sukuratchi, Tatchi, Uotchi, Wotchi, suketchibukku</tt>. --[[User:Dine2017|Dine2017]] ([[User talk:Dine2017|talk]]) 05:53, August 13, 2022 (EDT)

:::I'm here a bit later because I wanted to fix "[[Jump Up, Super Star!]]" which was using both ''-tchi'' and ''-cchi''. Because both were present and this page specifies Hepburn romanization, I assumed that the above quote was a mistake— Hepburn uses ''-tchi''. So saying "other romanization systems" when the prior section says the wiki uses Hepburn is contradictory and confusing.

:::In any case, I'd ''strongly'' advocate for ''-tchi'', as it is consistent with the other rules to use Hepburn romanization and the other standards that prioritize pronunciation along the same lines— such as that は, へ, and を as particles are written as they are pronounced (''wa'', ''e'', and ''o'' instead of ''ha'', ''he'' and ''wo''), and a long o as おう is written as ''ō'' to avoid confusion with separate syllable ''o'' and ''u''. Also, as pointed out, it's more intuitive for those without this knowledge of Japanese. [[User:Chubby Bub|Chubby Bub]] ([[User talk:Chubby Bub|talk]]) 02:48, April 11, 2023 (EDT)
::::Yeah, doesn't make sense to not use this form if we're already using Hepburn and it's more common and helpful. I've always thought the double "c" was standard since that's what I type to get the characters, and I'm not sure what the reasoning behind this rule in the policy was (maybe what Dine said, but it's still inconsistent; and "other romanisation systems" might be an error). I'll change that now. {{User:Mario jc/sig}} 09:35, April 11, 2023 (EDT)

== An edge case: スゥー ==

The page says: "''Sometimes a small kana might be used elongate the preceding vowel for stylistic reasons (e.g. for names, etc.), but should still be romanized the same way as if a ー symbol was used.''" What if both a small kana and a ー symbol are used? [[Swooper Dooper|マスク・ド・'''スゥー'''パー]] is an example (and it's currently romanized as ''Masuku do '''Swū'''pā''). --[[User:Dine2017|Dine2017]] ([[User talk:Dine2017|talk]]) 10:11, July 26, 2022 (EDT)
:Two more examples: [[Dōryī]] and [[Kānyī]]. --[[User:Dine2017|Dine2017]] ([[User talk:Dine2017|talk]]) 10:31, July 26, 2022 (EDT)
::Since it's still elongated, it probably should be romanized the same way. "Swūpā" I think is still appropriate even though the elongation is longer than normal. {{User:Bazooka Mario/sig}} 16:55, July 29, 2022 (EDT)

== Are there any standards for other non-Latin transliterations? ==

I'm sorry if I'm writing in the wrong place now, that's the most relevant talk page I've found. So, I've seen that transliterations of Cyrillic languages (such as Russian) are inconsistent (using different translit methods) and sometimes plain wrong (I've seen Cyrillic Р transliterated as Latin P, although AFAIK all methods use Latin R instead). It's possible that similar problems might exist with other non-Latin languages, such as Korean or Chinese. I know all these languages are, in case of this wiki, less significant since most non-English-language things about Mario are in Japanese. But these languages are commonly seen, for example, in articles devoted to names of characters, items, etc. in different languages. So, are there any guidelines for that? [[User:SuperJosipBroz|SuperJosipBroz]] ([[User talk:SuperJosipBroz|talk]]) 15:08, January 25, 2023 (EST)
:There are no official guidelines as far as I know. The status quo: Chinese text is almost always transcribed using Pinyin, although non-native speakers of Mandarin occasionally make mistakes. Korean transcription exhibits less consistency: either Revised Romanization of Korean or a custom transcription. I'm not knowledgeable about Russian. 
:One issue concerning Pinyin is whether to transcribe 马力欧 and 瑪利歐 as ''Mǎlìōu'' or ''Mǎlì'ōu''. The official Pinyin orthography says an apostrophe should be used before ''a, o, e'' where syllable boundaries may be confused, and convention dictates ''Mǎlì'ōu'' accordingly. But some editors favor ''Mǎlìōu'', possibly because the last two tone marks naturally distinguish the two syllables, dispensing with the need for the apostrophe. It's an unclarified issue in official Pinyin rules.
:Another is that Traditional Chinese may also be pronounced in Cantonese. We don't seem to have any Cantonese transcription on the wiki yet.
:I agree that Cyrillic Р should not be transliterated as Latin P. --[[User:Dine2017|Dine2017]] ([[User talk:Dine2017|talk]]) 22:31, January 26, 2023 (EST)
::Thanks! I guess I won't touch the Cyrillic transliterations unless they have obvious errors. [[User:SuperJosipBroz|SuperJosipBroz]] ([[User talk:SuperJosipBroz|talk]]) 17:12, January 28, 2023 (EST)

==Proposal for new section about correct rendering for certain characters==
From what I've heard, although I live nowhere close to northeast Asia, a handful of so-called "Hanzi unification" Unicode letters are somewhat controversial there, and also cause pretty large problems on Mario Wiki on ninja-related pages.

The problem is most visible on Windows (I tested with Windows 11 23H3), where the default fonts used in web browsers in general render the letters in their Simplified Chinese forms by default. I could test in Android too, but only on request.

As such, a section could, and in my eyes personally probably should, be added about something like ''"In Japanese text, certain characters (known as Unicode hanzi unification characters) are rendered wrong depending on the OS and the installed OS fonts, and instead default to their Simplified Chinese characters. This includes, but is not limited to,'' 窓空今忍骨直. ''To render them correctly, use ''<code><nowiki><span lang=ja style="font-weight:580">character</span></nowiki></code>'', which will result in'' <span lang=ja style="font-weight:580">窓空今忍骨直</span>''. "font-weight" is optional as its effect on other OS' default browser font settings is not yet known."'' [[User:DandelionSprout|DandelionSprout]] ([[User talk:DandelionSprout|talk]]) 16:41, July 14, 2024 (EDT)

: I agree, but manually adding language tags would be a daunting job. It would be ideal if browsers can automatically determine the language of untagged CJK text (and apply appropriate fonts) using heuristics. Otherwise, we can insert some of the tags automatically using Lua scripts in some templates -- {{tem|foreign names}} is a good candidate. --[[User:Dine2017|Dine2017]] ([[User talk:Dine2017|talk]]) 11:21, July 16, 2024 (EDT)
::It appears [[User:Porplemontage]] was up to the task yesterday (https://www.mariowiki.com/index.php?title=Template%3AForeign_names&type=revision&diff=4304656&oldid=4167302), but seemingly by mistake implemented the "lang" element to the romanizations as well, resulting in Windows in particular (and Chrome on Windows even more so) using the CJK fonts' Latin letters instead of e.g. Segoe UI. For Japanese and Korean romanization the results look "Conspicuous but fully readable", Simplified Chinese has too big text, while Traditional Chinese in Chrome looks all sorts of strange (but looks perfectly good in Firefox) due to trying to supplement missing characters in Jheng Hei with Times New Roman. (Firefox also doesn't seem to need "font-weight", but that's a different situation.) [[User:DandelionSprout|DandelionSprout]] ([[User talk:DandelionSprout|talk]]) 08:42, July 22, 2024 (EDT)
@@ Line 23: / Line 23: @@
 == Official romanizations ==
 Are official romanizations which are given by Nintendo allowed? --{{User:FanOfYoshi/sig}} 12:31, 4 November 2018 (EST)
-: Not in |JapR=, but they can be introduced in |JapN=; see [[List of Toadette names in other languages]] for an example. I think the reason is that official romanizations usually have fancy spellings (''Kinopi'''c'''o'' v.s. ''Nokono'''k'''o''), while we want our romanizations to be consistent to aid those learning the language. --[[User:Dine2017|Dine2017]] ([[User talk:Dine2017|talk]]) 14:22, October 6, 2024 (EDT)
 == Kanjis ==
 Certain games do not use the "Hiragana above Kanjis" thing, such as [[Super Mario Maker 2]], which complicates reading for non-kanji reading people in Japan. How would they expect everyone to read Kanjis? --{{User:FanOfYoshi/sig}} 10:52, July 3, 2019 (EDT)
-: If the game does not specify the pronunciation of kanji, then we can only "do as the Japanese do". This requires proficiency in the Japanese language, and is a frequent source of errors (of which I have [https://www.mariowiki.com/index.php?title=Midas_River&type=revision&diff=4131943&oldid=4127329 corrected some]).
-: Sometimes the kanji in a name can be read in two ways and both make sense. For example, ゴツゴツ山 can be read as both ''Gotsugotsu-san'' and ''Gotsugotsu-yama''. Then there's the problem of communicating whether the reading specified in romaji is official or conjectured. I have proposed distinguishing the two using furigana: "ゴツゴツ{{ruby|山|やま}} ''Gotsugotsu-yama''" means the reading is official; "ゴツゴツ山 ''Gotsugotsu-yama''" means the reading is conjectured by editors. --[[User:Dine2017|Dine2017]] ([[User talk:Dine2017|talk]]) 14:34, October 6, 2024 (EDT)
 == Fake Dakuten R letters ==
@@ Line 88: / Line 85: @@
 : I agree, but manually adding language tags would be a daunting job. It would be ideal if browsers can automatically determine the language of untagged CJK text (and apply appropriate fonts) using heuristics. Otherwise, we can insert some of the tags automatically using Lua scripts in some templates -- {{tem|foreign names}} is a good candidate. --[[User:Dine2017|Dine2017]] ([[User talk:Dine2017|talk]]) 11:21, July 16, 2024 (EDT)
 ::It appears [[User:Porplemontage]] was up to the task yesterday (https://www.mariowiki.com/index.php?title=Template%3AForeign_names&type=revision&diff=4304656&oldid=4167302), but seemingly by mistake implemented the "lang" element to the romanizations as well, resulting in Windows in particular (and Chrome on Windows even more so) using the CJK fonts' Latin letters instead of e.g. Segoe UI. For Japanese and Korean romanization the results look "Conspicuous but fully readable", Simplified Chinese has too big text, while Traditional Chinese in Chrome looks all sorts of strange (but looks perfectly good in Firefox) due to trying to supplement missing characters in Jheng Hei with Times New Roman. (Firefox also doesn't seem to need "font-weight", but that's a different situation.) [[User:DandelionSprout|DandelionSprout]] ([[User talk:DandelionSprout|talk]]) 08:42, July 22, 2024 (EDT)
-::: I also noticed that Chrome fallbacks to Times New Roman (perhaps due to [https://github.com/chromium/chromium/blob/main/third_party/blink/renderer/platform/fonts/win/font_fallback_win.cc this]), but I haven't been able to come up with a good solution.
-::: My current idea is to assemble a font [https://developer.mozilla.org/en-US/docs/Web/CSS/@font-face/unicode-range in CSS] with Latin letters from a Latin font and CJK characters from a CJK font. (The effect of simply setting <code>font-family: sans-serif, Yu Gothic</code> is still unclear to me. On some computers it works as intended, on others it doesn't.) --[[User:Dine2017|Dine2017]] ([[User talk:Dine2017|talk]]) 10:57, July 22, 2024 (EDT)
-:::: There was the far-outer fringe option of applying something like this to the cells based on my work on https://github.com/DandelionSprout/adfilt/blob/master/JapaneseTextSoftener.txt, which'd be something like <code>style="font-family: "Segoe UI", "Calibri", "Verdana", "Ubuntu", "Noto Sans CJK JP", "Noto Sans CJK", "Hiragino Maru Gothic Pro W4", "ヒラギノ丸ゴ W4", "Hiragino Maru Gothic Pro W3", "ヒラギノ丸ゴ W3", "Ume Gothic", "Ume UI Gothic", "Yu Gothic", "Yu Gothic UI", "Meiryo", "メイリオ", "BabelStone Han", "Nishiki-teki", "FRBTaiwaneseKana", (...)"</code>  However, considering https://www.mariowiki.com/index.php?title=Template%3AForeign_names&type=revision&diff=4306671&oldid=4304656, I'll give my greatest thanks and appreciations to Porplemontage once again. That fellow is a genius when it comes to wiki coding. I myself am halfway decent at wiki coding and I therefore appreciate all the more when someone surpass me by far and wide. [[User:DandelionSprout|DandelionSprout]] ([[User talk:DandelionSprout|talk]]) 02:20, July 23, 2024 (EDT)