Wikipedia talk:Manual of Style/Japan-related articles

WP:VG/GL#Non-English games

Archives of this discussion can be found at /VGGL and /VGGL2

Un-deprecation of other punctuation marks demarcating aspects of titles of media

The last discussion got archived because someone thought that the page was getting too long.

Basically, I believe that we should end the practice of removing tildes, hyphens, wave dashes, etc., from article titles, as the English-language media does not have any preference as to how to format these items, and as all reliable sources (which are of course Japanese in nature), retain these items.

Why shouldn't we, as an encyclopedia, retain these punctuation marks as well? It is not a trademark quirk. It is not necessarily a relict of Japanese printing. It is a method of marking a different aspect of the song's title, and really gets in the way if there is a song which has both a tilde and a set of parentheses in the title.

This practice is particularly obtrusive when the title of the original subjectis already parsed in grammatically correct English. It does not make sense to have to change the title to eliminate these punctuation marks and have to explain the original "stylization" of the title in the lede when we could just use the title as it appears in all reliable sources anyway.—Ryūlóng (竜龙) 08:34, 9 December 2010 (UTC)

I'm pretty sure that the entirety of Titles of books and other media needs an overhaul. If something is written as "B-E-S-T" or "B·E·S·T" and it is read as "Bee E Ess Tee", why should we remove the hyphens or middle dots?—Ryūlóng (竜龙) 08:49, 9 December 2010 (UTC)

As I said, I'm neutral on this aspect. I would suggest an RfC on it because from what I remember of the discussion there was some serious concerns.陣内_Jinnai 16:51, 9 December 2010 (UTC)

One of the main concerns was that the practice simply isn't English, but then again the subjects being discussed and in most cases the reliable sources being used for these articles aren't English anyway. My examples for the shortcomings of this aspect of the style guide is WBX (W-Boiled Extreme), which in all reliable sources is called "W-B-X ～W-Boiled Extreme～" (except on iTunes where it is "W-B-X ~W-Boiled Extreme~".

By removing this aspect of the style guide, it can make it easier for us to deal with song titles such as "On The Painted Desert - Rampant Colors", "Round ZERO～BLADE BRAVE", or songs like "～それから～".—Ryūlóng (竜龙) 19:15, 9 December 2010 (UTC)

On the English Wikipedia, standardized rules are employed for dealing with special characters and capitalization. Manual of Style (trademarks) says "Avoid using special characters that are not pronounced, are included purely for decoration, or simply substitute for English words (e.g., ♥ used for "love"). In the article about a trademark, it is acceptable to use decorative characters the first time the trademark appears, but thereafter, an alternative that follows the standard rules of punctuation should be used" and "Using all caps is preferred if the letters are pronounced individually, even if they don't stand for anything. For instance, use SAT for the (U.S.) standardized test."

Meaning: The stylized form should be included in the article on the subject, but in normal prose text, standard English formatting and capitalization is to be used. Prime Blue (talk) 19:11, 10 December 2010 (UTC)

The ～ is not decorative, though. It is merely the way chosen to represent subtitles or whatever other part of the song such items are relegated to rather than a colon or parentheses or any other puncutation mark, and in several instances parentheses are also used in the title of a song, so simply throwing out the ～ is unhelpful.—Ryūlóng (竜龙) 19:35, 10 December 2010 (UTC)

I guess that way you can dismantle pretty much every guideline. Good day. Prime Blue (talk) 19:47, 10 December 2010 (UTC)

PB, the point about them not being decorative is valid; it is a legitimate form of punctuation. That said, I do understand the problems this would create. Better would be to bring this up at WP:MOS as WP:PUNCT is the appropriate place for this and it is silent on this issue.陣内_Jinnai 21:29, 10 December 2010 (UTC)

But it's still an issue on this particular manual of style.—Ryūlóng (竜龙) 21:36, 10 December 2010 (UTC)

Jinnai, as I posted: "Avoid using special characters that are not pronounced, are included purely for decoration, or simply substitute for English words (e.g., ♥ used for "love")." But I am sure with the amount of pettyfoggery these guidelines are read when it comes to opposing them, it won't really make a difference. Prime Blue (talk) 09:41, 11 December 2010 (UTC)

Exactly, ♥ is a special character as are a number of others. I don't believe that is the case for tildes in Japanese for how they are in use. I believe the practice has them being used as standard punctuation for titles in the same way semicolons and dashes are used in titles. There is nothing on dealing with nonstandard English punctuation in the MOS. That's why I am neutral. It shouldn't be outright rejected as a special character like the heart symbol because it isn't. On the other hand, its only really used with Japanese and only for titles and taglines.陣内_Jinnai 05:27, 12 December 2010 (UTC)

I think for what this proposal suggests (just use what's used outside of Wikipedia), the guideline is pretty clear-cut: "Follow standard English text formatting and capitalization rules, regardless of the preference of trademark owners." One can argue about hearts, then tildes come up. One can argue about tildes, then interpuncts come up. And so on. It's an endless argument, but pretty much irrelevant to the purpose of keeping formatting consistent across the encyclopedia. Taking into consideration that the stylization is mentioned in the respective articles anyway, I don't think this would have any use. Prime Blue (talk) 17:09, 12 December 2010 (UTC)

A heart being used for the word "Love" is one thing. Tildes and interpuncts, both of which are punctuation marks in English, are another. This isn't a stylization. It's how they're intended to be written out. It's better than forcing articles to use a different name and then need to explain that the original name is a "stylization" when it is just the name.—Ryūlóng (竜龙) 19:01, 12 December 2010 (UTC)

Tildes still are far from being standard English formatting for subtitles, though – that would rather be a colon or a hyphen. Though I have to admit I don't quite understand what you are proposing. Where do you draw the line of what is acceptable and what isn't? I think it would help if you elaborated on that. Prime Blue (talk) 19:15, 12 December 2010 (UTC)

MOSTM

• All tildes and hyphens used in the titles of Japanese media are not to be removed or replaced by colons or parentheses unless an official English title of the media exists and does not utilize these items. Examples are W-B-X ~W-Boiled Extreme~ and ja:On The Painted Desert - Rampant Colors.
• Interpuncts (or similar items used to separate parts of the names of media) are also to be retained unless they are used to stand in for a word break in an English translation. Example is Che.r.ry, and feasibly m.c.A·T.
  • When the name is written entirely in Japanese such as 涙・抱きしめて, the name could be parsed as "Namida·Dakishimete" or "Namida · Dakishimete" or possibly "Namida - Dakishimete".
• Any other characters such as stars or hearts should not be used in the article title, particularly if they are not pronounced in some form in the name. However, if the name is written entirely in non-Japanese characters, the symbol should be used when referring to the title in the article. An example would be the Queen & Elizabeth song "Love♡Wars"; if it were made into an article it would be at Love Wars (Queen & Elizabeth song) but the name would be parsed as "Love♡Wars".

This is the general idea I have in mind. Not sure on the interpuncts in Japanese part, though.—Ryūlóng (竜龙) 19:42, 12 December 2010 (UTC)

If tildes and hyphens were considered standard English formatting for subtitles, then they would be acceptable – otherwise, WP:MOSTM forbids them. The Chicago Manual of Style suggests that colons are normally used to mark a subtitle, though they also mention em dashes (which seems to be a very rare case, though). For cases like "～それから～", WP:MOSTM forbids the tildes anyway as, there, they are clearly decorative. All in all, the hyphen looks debatable (possible to be replaced by an en or em dash, as hyphens are not generally used that way), while the tildes should be flat-out converted to a single colon.

The second point is again covered by WP:MOSTM and also by WP:ABBR. "Che.r.ry" is to be converted to "Cherry" and "m.c.A·T" to "MCAT". Same goes for "W-B-X" which needs to be "WBX". For the interpunct example "涙・抱きしめて", preserving the interpunct as is (or as "·") would be forbidden as it is not standard English formatting, but converting it to an en or em dash would again be up for discussion.

WP:MOSTM is also to be used on the "Love♡Wars" example (again decorative), which needs to be rendered "Love Wars" unless the "♡" is pronounced somehow, in which case it would be written "Love Love Wars" or something along the lines. Prime Blue (talk) 10:06, 13 December 2010 (UTC)

Are you talking about "Avoid using special characters that are not pronounced, are included purely for decoration, or simply substitute for English words"? That's the closest I can see and if that's the case, then it clearly doesn't fall within that as it is not "subsistitue for English words" nor is it a "included purely for decoration". Finally, as a tilde is on the standard 101-keyboard and it isn't a "special character" like a ♥. If you claim it is, I can start claiming pretty much any character outside of A-Z and 0-9 is a special character.陣内_Jinnai 16:34, 13 December 2010 (UTC)

I have already explained in great detail (several quotes and explanations!) how WP:MOSTM dictates the usage of standard English formatting, which – as far as I am aware – tildes for subtitles are not. Prime Blue (talk) 18:01, 13 December 2010 (UTC)

You keep bringing up WP:MOSTM. This isn't a trademark. It's their name. We don't follow the Chicago Manual of Style. That's why we have all these other manuals of style to deal with these little quirks. And we do not have to follow every MOS on the site. We can clearly suggest that there be an exception for these song titles to show that they are Japanese and not have to spend another fifty words explaining that the original title is only a stylization when it is the actual title.

The Manuals of Style on Wikipedia are ultimately way too restrictive in that all of a sudden people are not allowed to use periods (or in the case of the W-B-X song hyphens) to demarcate initialisms. Then why the hell do we have articles on American artists such as B.o.B and N.E.R.D? This was one of my problems with the move request for the band "m.o.v.e" to its current title Move (Japanese band). And by the standard set by k.d. lang, "m.c.A·T" (or "m.c.A.T"), angela (with {{lowercase}}), and other bands who had to be renamed because of all of these conflicting manuals of style should be allowed, but because they're Japanese and we have style police who do nothing but move and edit articles so they are in line with the manuals of style, they all are not allowed to be at the various names they should be.

The Interpunct exists in English, so it should be a perfectly viable substitution for the nakaguro. The tilde exists in English, so it should be allowed, especially if the rest of the name is in English lettering. And finally, there should not be such a disparity in how ignoring specific aspects of the manual of style is allowed for Western musicians, but completely forbidden for Eastern musicians, just because they are Japanese and not British or American.

And also, it is not going to harm Wikipedia to leave the heart in the name "Love Wars" when it is discussed in the body of the article, rather than having to explain over and over again that the title originally includes said heart.—Ryūlóng (竜龙) 19:31, 13 December 2010 (UTC)

MOSTM has no juridiction here because its not part of the trademark name. Its standard punctuation.陣内_Jinnai 19:39, 13 December 2010 (UTC)

Yes, but Prime Blue is arguing that because the tilde, hyphen, interpunct, etc. are not standard English punctuation (and that there are various other guidelines that seem to forbid us from even using them), that it's not allowed.—Ryūlóng (竜龙) 19:54, 13 December 2010 (UTC)

Disregarding that you are aggressively exaggerating here (50 words to explain a stylization? Employing The Chicago Manual of Style on Wikipedia?), you can keep trying to dance around the guideline by introducing special cases and semantic intricacies, but it does not change its spirit. "It exists" does not equal to "it is a standard way of using it". Not to mention that "it exists" and "it does no harm" (the latter is not even the case, giving editors a free pass to write work titles in a stylized form wherever they deem necessary causes a lot of harm – this is an encyclopedia, not a trademark library or database for stylization curiosities) is a fairly weak rationale for a change that would, in the long run, affect thousands of articles. That's what the Manual of Style is about: Keeping formatting consistent across Wikipedia. And per WP:MOSTM, you mention the stylization once in the article on the subject, otherwise use standard English formatting across the encyclopedia. And that makes a lot sense, because else Wikipedia would be drowning in hearts and kawaii flashy neon signs.

If you think these proposed exceptions find support, you can always start a Request for Comment to clarify the status of WP:MOSTM.

As for B.o.B and N.E.R.D.: These are not in line with WP:ABBR, so by all means, if you wonder "why the hell" we have articles with those names, move them. Americans sure don't deserve to have full stops in their band names on Wikipedia if the Japanese do not either! Prime Blue (talk) 20:32, 13 December 2010 (UTC)

Fine. The tilde, hyphen, et al., are the standard methods of writing out the names of Japanese songs in the English language. If you feel the tilde/wave dash/whatever it is to be wrong, maybe we should go with how various online retailers parse the names. I know that the Japanese iTunes Store (when it is parsed in English) uses hyphens rather than the other dashes, as do these various retailers (even though I don't particularly agree with it).—Ryūlóng (竜龙) 20:41, 13 December 2010 (UTC)

Oh, and as a note I missed out on before where you say "tildes should be flat-out converted to a single colon", even this guideline currently proscribes that if it is a song title, these items should be modified to open and close parentheses/brackets. However, this discussion is currently on the merits of throwing out that particular part of the MOS.—Ryūlóng (竜龙) 20:44, 13 December 2010 (UTC)

[User:Prime Blue|Prime Blue]], you are misinterpreting MOSTM is. Special characters do not include the basic keyboard characters. Thus no one here is trying to "dance around the guideline"; instead you are trying to broaden a guideline to include something it was never menat to be used with. The tildes are not part of the trademark anymore than a colon is the part of most titles with subtitles attached to them when we want to describe them in prose (such as Star Trek: The Next Generation). That is what the tilde is used for and thus MOSTM, which is specifically only about the trademark's official name, has no jurisdiction here. WP:MOS would, but it is completely silent here.陣内_Jinnai 03:47, 15 December 2010 (UTC)

Policy

(edit conflict)I'm not sure if someone pointed this out yet, but the Article titles policy states:

Do not use non-language characters: Non-language characters such as "♥", as sometimes found in advertisements or logos, should never be used in titles.

I have added this to WP:MOS-JA#Article titles in order to be clear on policy. ···日本穣^? · 投稿 · Talk to Nihonjoe · Join WikiProject Japan! 09:52, 31 December 2010 (UTC)

The heart is not planned on being used in the title. However, you have summarily outlawed the wave dash which was under discussion in this thread as being shifted to the tilde (~) to demarcate the subtitles (or hyphens such as in "On The Painted Desert - Rampant Colors") example. It would also solve the issue of what to do with songs that begin and end with the wave-dash/full width tilde or simply only have one such tilde in the middle of the song (see ja:Round ZERO〜BLADE BRAVE). I think it would benefit the reader to not have to go through the extra step of explaining a "stylization" when it is not a stylization and just one of the various ways the subtitle is marked in Japanese media.

Also "WBX (W-Boiled Extreme)" is still the worst possible title for that page. If the title is written entirely in English, even if it is a Japanese song, it should not have to be renamed on the English Wikipedia.—Ryūlóng (竜龙) 11:16, 31 December 2010 (UTC)

Well, the nami-dash is a non-language character, and it should never be used in the title per the Article titles policy. Language characters (in the case of English) would be the 26 letters of the alphabet as well as standard English punctuation. The nami-dash is not any of those. ···日本穣^? · 投稿 · Talk to Nihonjoe · Join WikiProject Japan! 03:19, 2 January 2011 (UTC)

The only place I see it as the nami-dash is on ja.wiki. Everywhere else uses a full or half width tilde. Surely we can utilize the tilde rather than throwing whatever the symbol is in favor of unverifiable names.—Ryūlóng (竜龙) 04:22, 2 January 2011 (UTC)

Odd thing with current examples

How come most of the examples under the "Titles of books and other media" section involve Kumi Koda?—Ryūlóng (竜龙) 09:49, 31 December 2010 (UTC)

I have rewritten the section to where it does not use five examples of Kumi Koda's music. I am using the following examples now:

• ja:LØVE (中島美嘉のアルバム) --> Love (Mika Nakashima album)
• ja:i spy i spy --> I Spy I Spy
• ja:CHE.R.RY --> Cherry (song)
• ja:m・a・z・e --> Maze (Kumi Koda song)
• ja:BEST〜first things〜 --> Best: First Things
• ja:I miss you 〜時を越えて〜 --> I Miss You (Toki o Koete)

This is still using mostly female musicians though.—Ryūlóng (竜龙) 03:46, 4 January 2011 (UTC)

Wave dash

It has been brought up multiple times in this discussion that the full width tilde/wave-dash is not a non-language character but a punctuation entity, if at least in Japanese orthography. It should not currently be listed in a group of characters to be avoided in article titles, as we are discussing it here. It is better to have actual characters which we know should not be used (the star and the heart) rather than one that we could decide has some encylopedic merit.

The only place the wave dash is ever used in the ways that are currently described on this page is on the Japanese Wikipedia. Everyone else (record labels, retail outlets, etc.) use the full width tilde. I have never seen a wave dash anywhere other than at ja.wiki.—Ryūlóng (竜龙) 02:35, 4 January 2011 (UTC)

The term "non-language character" is quite bizarre and seems to have been invented at WP:TITLE. I've proposed changing it to "symbol", which is a technical term clearly defined at Unicode symbols, and explicitly extending the overall rule to cover non-Latin punctuation. U+301C WAVE DASH is arguably a bit of both, as it's in Unicode's "CJK Symbols and Punctuation" block. Jpatokal (talk) 09:26, 4 January 2011 (UTC)

Yes, but it can still be approximated with a piece of Latin punctuation.

Any of these "dashes" should no longer be deprecated, but merely approximated with their QWERTY keyboard counterparts: the hyphen and the tilde (e.g. "Round Zero ~ Blade Brave" for "Round ZERO～BLADE BRAVE", "On The Painted Desert - Rampant Colors", "W-B-X ~W-Boiled Extreme~" for "W-B-X ～W-Boiled Extreme～", "Splash Gold -Natsu no Kiseki-" for "SPLASH GOLD-夏の奇蹟-").

"Nakaguro-style dots" should be approximated with the interpunct (·) if it acts as a break between pronunciations (musician "m.c.A·T", song "Namida · Dakishimete" for "涙・抱きしめて", "A·A·Aiyaiya·A·A!" for "あ・あ・あいやいや・あ・あ!").

Anything else (Hearts, stars, male symbols) should be replaced by parentheses should they be used to separate parts of the song title or not be retained (e.g. "Age Age Every Night" for "アゲ♂アゲ♂EVERY☆騎士"). If the song's title is otherwise in English and contains these other characters, the article title should not retain this symbol per WP:AT, but the symbol should be used when referring to the title of the media in prose (e.g. "Love♡Wars" in text, "Love Wars (Queen & Elizabeth song)" as the page title) and not say "originally stylized as" in the first sentence. It is unnecessary prose to say "originally stylized as" when it's just the original name of the song before we tore it apart.—Ryūlóng (竜龙) 09:41, 4 January 2011 (UTC)

Tildes and, even worse, interpuncts are not a part of standard English punctuation, and neither can hyphens be used for anything other than joining words or separating syllables in a word. Using them willy-nilly is bizarre and liable to misinterpretation. Consider "Splash Gold -Natsu no Kiseki-": the average English speaker would read that as "Splash Gold-Natsu no Kiseki", with "Gold-Natsu" a compound word (金玉?) and the hyphen at the end a typo.

English has a perfectly good standard way of separating subtitles: the colon. It also has an excellent replacement for the nakaguro, namely the space. Jpatokal (talk) 12:18, 4 January 2011 (UTC)

One, these are songs that possibly do not have subtitles but just different parts of the title separated by these characters. And the average English reader would not mistake "Splash Gold -Natsu no Kiseki-" as "Splash Gold-Natsu no Kiseki" because there is a space in the first one there. I think we can give English speakers more credit here. Tildes and interpuncts at least exist in English puncuation. Just because they're not standard does not mean that they're not used at all. The project utilized tildes and hyphens and whatever other symbols prior to Hoary's original editing of the page. We can make exceptions Jpatokal. I don't see why you can't understand that.—Ryūlóng (竜龙) 19:40, 4 January 2011 (UTC)

The tilde is never used for punctuation in any language. The wave dash is, but not in English. And the interpunct is simply a Japanese representation of a space: how does "A·A·Aiyaiya" differ from "A A Aiyaiya" in pronunciation? Jpatokal (talk) 21:41, 4 January 2011 (UTC)

It does not. And it is a poor example, I realize that now.—Ryūlóng (竜龙) 22:37, 4 January 2011 (UTC)

I have to admit that I have problems seeing why an exception should be made. The only case where a wave dash or wave dashes could come in handy, I think, is when a work has two subtitles (unless we already have a style guideline on what to do then, but I'm not aware of its existence), though I'd still be in favor of using more standard English methods of separating the second subtitle then (i.e. a semicolon or an en dash). Prime Blue (talk) 20:45, 4 January 2011 (UTC)

Again, the "subtitle" thing only appears in album titles or larger pieces of artistic work. In songs, it is very often the case that it will have another portion of the song's title with parentheses in it, usually if it is being designated as an instrumental or karaoke version or remix of one of the other tracks. So rather than going through the entire list of brackets usable in this case, we can just turn the wave dash into a tilde in all cases and not have to worry about having to make exceptions for certain cases along the line.

For example, we have Namida (Kokoro Abaite). There is currently nothing to suggest that the song's title (not full title) is not "Namida Kokoro Abaite" but the "Kokoro Abaite" part has been separated with the wave dash/full width tilde, resulting in what could be an English language name of "Namida ~Kokoro Abaite~". It's not like (I Can't Get No) Satisfaction or other similar songs in English, where the song is referred to only as "Satisfaction". In the Japanese press they don't call the song "NAMIDA", but by the whole title every time. There's a completely different system of naming in Japan and referring to these other parts of the songs as "subtitles" is just an easy comparison to make.—Ryūlóng (竜龙) 20:49, 4 January 2011 (UTC)

Sounds like a good reason to move it to Namida: Kokoro Abaite then.Jpatokal (talk) 21:41, 4 January 2011 (UTC)

That is not how song titles are ever formatted.—Ryūlóng (竜龙) 22:37, 4 January 2011 (UTC)

Got reference? Jpatokal (talk) 02:24, 5 January 2011 (UTC)

I don't need a reference. It's general practice that song titles do not include colons in them in order to differentiate between a title and subtitle. Look at (I Can't Get No) Satisfaction, Pleasure (Pleasure), Eh, Eh (Nothing Else I Can Say), etc.—Ryūlóng (竜龙) 02:46, 5 January 2011 (UTC)

It is also general practice that song titles in English do not include Japanese punctuation. I find the colon to be the lesser of two evils. Jpatokal (talk) 10:14, 5 January 2011 (UTC)

The colon is not what is used in song titles at all. While it may separate the title from the subtitle in other pieces of media, it is not what is used in English, Japanese, French, Chinese, or any other language to demarcate two different parts of a songs' title. And I am tired of going in fucking circles arguing this with you of all people.—Ryūlóng (竜龙) 10:36, 5 January 2011 (UTC)

So "Kokoro Abaite" is the subtitle of "Namida ~Kokoro Abaite~". Why shouldn't it be written that way, seeing as the full name of the song is referred to in all Japanese language reliable sources as such, when English language sources only mention the title, rather than the title and subtitle? Whenever someone refers to (I Can't Get No) Satisfaction it's just "Satisfaction". However when it's "NAMIDA～ココロアバイテ～" it's always "NAMIDA～ココロアバイテ～" and not "NAMIDA" or "ココロアバイテ" on their own.—Ryūlóng (竜龙) 10:43, 5 January 2011 (UTC)

MOS:TM 2

In a discussion concerning other aspects of article titles at WT:MOS#Hyphens in article titles, it has been pointed out that WP:MOSTM (and WP:ABBR) do not apply to the titles of compositions. So it is clear that the only issue is a stylistic one here (and possibly with WP:MUSIC or WP:MOS-ALBUM). Unpronouncable characters like hearts, stars, and horseshoes aside, the wave dash/swung dash/full width tilde is a character that by no other general manual of style is explicitly forbidden from use in the title of any piece of media. And as it stands, this manual of style conflicts with WP:MOSTM anyway in that it states "editors should choose among styles already in use (not invent new ones)".—Ryūlóng (竜龙) 10:37, 9 January 2011 (UTC)

Okay, this got archived. Can we get some new input with this new discovery?—Ryūlóng (竜龙) 08:33, 10 February 2011 (UTC)

Issue of revised Hepburn

Current MOS-JP adopts the "revised Hepburn", but there are some issues;

1. "Revised Hepburn" is a minority - it is so difficult or impossible for we to find the macron spelling such as Bishōjo, Ichirō (Ozawa), Hokkaidō etc. in modern reliable sources.

2. It's a hassle to enter a character with macron such as "ō".

3. What is called "revised Hepburn" is very old dictionary.

3.1. Extended katakana such as she(シェ) and ti(ティ) are not included in "revised Hepburn".

3.2. It is a little difficult to understand genuine "revised Hepburn" rightly. We can see some interpretive article about "revised Hepburn", but there are some variations; the current MOS-JP includes some elements such as aa in non-revised Hepburn.

3.3. It seems to be a little difficult to distinguish between long vowel and short vowels in translation Japanese term into romaji (see link).

So I think "revised Hepburn" should be replace with "folk Hepburn" such as "Tokyo", "kana spelling" such as "Toukyou" (or "modified Hepburn" such as "Tookyoo"). Of course, they are not perfect. But I think they are better than "revised Hepburn".--Mujaki (talk) 16:43, 17 December 2010 (UTC)

Do you have references where it is shown that revised Hepburn is not the first option in other reference works? The article Hepburn romanization doesn't have a whole lot of sources to verify that revised Hepburn is the primary option. Reading the article, revised Hepburn is used by the Library of Congress, so I believe that is the reason why it was favored over modified Hepburn or any of the other variations which appear to be more widely used. It appears to me that other reference works seem to be favoring the macronless variants but I can't find a specific study where that is mentioned, it's only my perception. Fan pages and other Internet material appear to favor wāpuro style, but they aren't reliable sources. Mainstream media, on the other hand, does not use revised Hepburn at all. But even if it's a hassle, I think it's a plus that the general public is able to read the approximately correct pronunciation like it's presented in the article Tokyo, since the alternative, the International Phonetic Alphabet, is even more complicated than macron usage. Having said that I also think that, for the body text of articles, we should use the romanization used by other reference works. But before taking other actions, we should first verify that revised Hepburn is no longer favored by scholarly sources and reference works. Also, if, as you mentioned, revised Hepburn is anachronic, we should confirm that it is no longer the standard romanization for the English-speaking world. Jfgslo (talk) 01:36, 18 December 2010 (UTC)

"Folk hepburn" isn't a real thing and "Toukyou" or "Tookyoo" would never be used, anyway. Also, leaving macrons in for certain items is fine, as redirects are cheap and there's always a box below the "Save page", etc. buttons that allows you to enter however many Latin letters with đĩāçṝįťıċš one could ever wish for. Wāpuro and kana spelling are used solely by fan media.

Unless your entire argument is on article titles. The revised Hepburn system is fine when we're using it as the third parameter in the family of {{nihongo}} templates. However, as the manual of style currently states (at least for subjects of biographies), the revised Hepburn form of the name should be the last possible choice.—Ryūlóng (竜龙) 01:49, 18 December 2010 (UTC)

Pg 179-180 - there lists that the original hepburn is the most commonly used (latest print is 1992 though) I will have to check for RSes on Wapuro usage as i don't think anyone has really done that yet. However, I doubt with pcs not having macron letters, revised hepburn would have increased in usage since 1992.陣内_Jinnai 17:58, 30 December 2010 (UTC)

[1] Some more info on romanization usgae. Hepburn, specifically original, is used most commonly. Modified Hepburn was abolished as the standard in 1994 in place of the older form. Modified is still in wide use, because it was taught for some time, but its no longer taught.陣内_Jinnai 17:28, 7 January 2011 (UTC)

info on Wapuro usage. Beyond that, I can't really find much for English usage except to say that anime/manga translations tend to favor wapuro over modified hepburn at least (going by official translations of names and titles here). Very few titles use modified hepburn and then only for the titles themselves (not the names).陣内_Jinnai 17:56, 7 January 2011 (UTC)

I have seen instances where the revised Hepburn system (macrons) is used in the original Japanese production. I saw them in promotional materials for the Engine Sentai Go-onger show and also it's on all of their wardrobe.—Ryūlóng (竜龙) 19:41, 7 January 2011 (UTC)

• English (here) (Japanese characters rōmaji) or rōmaji (here) (Japanese characters)

The first part (topic) is generally used in body and title, and articles are normally titled using the name which is most frequently used to refer to the subject of the article in English-language reliable sources. But macron style is not common.

Jfgslo: I see some sources - "駅名とローマ字表記と,オ段長音表記の国際化" written by Taro Takahasi in "国文学 解釈と鑑賞" January, 1997 issue published by gyosei, Comprehensive Database of Japanese Name Variants by The CJK Dictionary Institute and so on.

I think rōmaji can follow revised Hepburn, but revised Hepburn is not suitable for the topic due to minority.

Editors (transtraters) in external source commonly write with consistent style. Library of Congress translates "東京" into "Tōkyō", "弘法大師" into "Kōbō Daishi" and so on with few exception (Latin name etc). Encyclopædia Britannica adopt "Kyōtō", "Dōshisha University" etc., though there are some exceptions such as "Tokyo", "Hokkaido" and so on. Columbia Encyclopedia etc. follows macronless style.

However thare are mix of common usage, official name and translation per MOS-JP in enwp. And someone often persists in MOS-JP in case of moving proposal into common usage. But "revised Hepburn" (macron style) is minor; thorefore "revised Hepburn" for topic should be replaced with more common Hepburn manner.

Ryūlóng: Well... "Toukyou" and "Tookyoo" are examples for "東京" in each style (see article Hepburn romanization). "Folk Hepburn" is one of names for style ignorring long vowel such as "Tokyo", and it is also called "English style" etc. But the name is not important because it is a way for classification in this section. Incidentally, "Tookyoo" (or "To^Kyo^") is valid if diacritical mark such as macron is not able to be used in writting.

I think you have a preconception about Hepburn -e.g. Guide by Agency for Cultural Affairs, Guide by Ministry of Foreign Affairs.

And official transration you said in other section is mostly romanized name per Hepburn family, though there are exceptions such as SION (Kunre-siki (or Nihon-siki)), Naoki Shinjyo (mixed Hepbuen-shiki (jo) and Kunrei-siki (zyo)) Yasuhiro Nightow (original). e.g. "Ichirou Mizuki" is permitted in passport Hepburn and kana spelling.

Jinnai: They mean Hepburn style including traditional, revised and other variations is the most popiler in romanization styles (Hepburn, Kunrei, Nihon, JSL etc.). It is verified by many sources. And wapuro style is usually handled as one of Hepburn style family. Incidentally, some manners in original Hepburn (1867) changed or went out of use. e.g. "h'" ("h'to") in original Hepburn bacome "hi" ("hito") in second version or later. And "Japanese school children now learn Hepburn when they first begin to learn the English alphabet in junior high school" is wrong; Japanese children learn Kunrei-siki style and Hepburn style at 4th grade in elementary school.[2]--Mujaki (talk) 05:13, 8 January 2011 (UTC)

Mujaki, I did not say that revised Hepburn is always used in the original Japanese productions. I only mentioned one instance where I came across the revised Hepburn system in one particular television program. While it is good that you are trying to reform our practices here, please remember that we are using a system that appears in English language academia on the Japanese language. We are not looking to use a system that is primarily used in Japan, but a system that is easily conveyed and the pronunciation can be easily inferred and is used in English language sources. Kunreisiki omits the actual pronunciation of the し, ち, つ, and ふ kana, and the other Hepburn systems make it difficult to differentiate between long O's and o-u.—Ryūlóng (竜龙) 05:35, 8 January 2011 (UTC)

If they consider wapuro to be a branch of Hepburn then it may be quite hard to find statistics for use in English. The most I can say is that the only times I usually see it are title names, in some academic texts (although those rarely follow the macron rules consistently) and occasionally a person's name. For titles of works, the characters in them, do not use macrons and only a handful of titles use macrons. Therefore I'd say macron usage is in the minority for English-speaking world. Beyond that, I can't really say. As mentioned, there hasn't been a comprehensive study for English world on which form of Hepburn romaization is the most common.陣内_Jinnai 16:32, 9 January 2011 (UTC)

What is "general guidelines" in MOS-JP based on

What is general guidelines in MOS-JP based on? Long vowel and Syllabic "n" in MOS-JP are diferrence from revised Hepburn style family such as standard style (Hyōjun-shiki Rōmaji by Rōmaji Hirome Kai in 1908; "revised Hepburn" in article Hepburn romanization) and BS 4812:1972.

• Long vowels
  • Long vowels, including "ああ" and "ええ" in kana/kanji term, are generally written with macrons in revised Hepburn style family. - e.g. "kāsan"(secondary) for "母さん", "nēsan" for "姉さん".
  • Long vowels are also written with circumflex in standard style. - e.g. "kâsan"(primary) or "kāsan"(secondary) for "母さん", "nêsan" or "nēsan" for "姉さん".
  • Long vowel "i" in kana/kanji term is also writen as "ii" in standard style. - e.g. "nîsan", "nīsan" or "niisan" for "兄さん".
  • Long vowel "i" in kana/kanji term is always writen as "ii" in BS and ANSI. - e.g. "niisan" for "兄さん".
  • Long vowel "e" writen as "えい" in kana is writen as "ei". - e.g. "Tokei" for "時計".
• Syllabic "n" (ん)
  • "n" is written as "m" or "n" before "b", "m" and "p" in standard style. - e.g. "Namba" or "Nanba" for "難波".
  • "n" is always written as "n" even if it is before "b", "m" and "p" in BS and ANSI. - e.g. "Nanba" for "難波".

Incidentally, "Bonin (Islands)" in "Determining common usage" section is not Japanese term, but English word of Japanese origin (Bunin (Jima or Tō; 無人(島)) in Edo period). "ボニン(諸島)" is a loanword in Japanese.--Mujaki (talk) 05:13, 8 January 2011 (UTC)

Would you mind making new sections instead of making these sections longer with different discussions in them in the future?—Ryūlóng (竜龙) 05:41, 8 January 2011 (UTC)

Changes to Hepburn romanization

Mujaki has extensively rewritten Hepburn romanization to fit his rather idiosyncratic ideas of what the various flavors of Hepburn are. I think the changes go ahead established practice (eg. traditional Hepburn is "desu", not "desz"!) and will try to clean them up later today, but other opinions and assistance are welcome, since there is a lot of confusion over the labels and what they mean. (See also the talk page.) Jpatokal (talk) 21:24, 11 January 2011 (UTC)

Perhaps it should just be blanked reverted. Like what I'm about to do. It's simple to go back to the previous version and clean it up, rather than having Mujaki's fringe point of view on Hepburn as the current focus.—Ryūlóng (竜龙) 21:31, 11 January 2011 (UTC)

As usual, you're being too rash -- Mujaki has brought up some valid points and they should be integrated, not reverted wholesale. Jpatokal (talk) 23:50, 12 January 2011 (UTC)

They can be reintegrated after the fact.—Ryūlóng (竜龙) 23:51, 12 January 2011 (UTC)

Please cite new sources and don't remove {{fact}} tag without any source, if revert (remove sources).

I based the edit on Hyōjun-shiki Rōmaji along the introduction because the theory exists and it is verifiable with cited some sources (dictionary, book, etc.). And I supplemented other views per WP:NPOV. At least, the explanations reverted by Ryulong are not consistent in the article, and there are a little source.--Mujaki (talk) 16:12, 12 January 2011 (UTC)

Capitalization of titles, and a new guideline

In this very recent edit, Nihonjoe adds a general guideline telling people to use the rules of English capitalization for titles.

In recent educated use, there are, simply, two different approaches to capitalizing titles in English. With an admirable lack of pretentiousness, the Chicago Manual of Style (14th ed [the latest I have on me], §7.1) calls them the up and the down styles; they're exemplified by, respectively:

• The Chicago Manual of Style
• The Chicago manual of style

Each has its minor merits and minor drawbacks. Certain stylesheets demand the one for some purposes; others demand the other.

Nihonjoe and his new guideline clearly want us to use the up style. But while the down style has its minor complications (e.g. some stylesheets say that the first letter of a subtitle must be capitalized, others that it needn't be), the up style has rather more, often involving questions of parts of speech and the counting of letters (or syllables). (And there's more besides: A book that happens to be in front of me is titled How Languages are Learned, showing that, at least in book design, OUP's up style downgrades the copula.)

To my mind, one or other of quotation marks and italics is adequate to set off titles. I'm reluctant to extend the up style, with its requirement for parsing, to Japanese. Take the example that's now given of good use: Otoko wa Tsurai yo: this clashes with the very guideline that's being pointed to:

. . . capitalize:

* the first and last word; [. . .]

* prepositions that are [. . .]

** the first or last word of the title (e.g., "Walk On");

Yes, the last word should be capitalized. Why is this lowercase? Because it's a "particle" -- syntactically, a curious grab-bag of stuff whose list includes postpositions (made) and postposition-casemarker pairs (made ni), but (rightly) not strings such as no mae ni, whose close English analogue is I all lowercased in certain variants of the up style. But not in the WP variant:

. . . capitalize: [. . .]

* prepositions that are [. . .]

**the first word in a compound preposition (e.g., "Time Out of Mind", "Get Off of My Cloud");

Yes, the ni in made ni is a case particle, but (controversially) the of of "time out of mind" is merely a case marker too; arguably, all of this would imply capitalization of the first half: Made ni. To which the answer might be: No, made ni is a listeme; it's a single particle that just happens to be spelled with a space in the middle. (Which would be wrong, but that's another matter.)

Et cetera et cetera. If you simply point people to the English instructions and say "do likewise", you'll soon run into snags. So you either allow for "common sense" (discretion, guesswork, slack) in application, or you write a set of guidelines, which people may or may not then read and follow.

Now, if others want to develop these rules and apply them, I have no particular objection. But I'm not going to develop them, and I'm not going to apply them -- not while an alternative as simple as the down style is also available. It's a style I've used for years in "my" articles here, and nobody had complained about lack of consistency, difficulty of comprehension, until just the other day.

I recommend the down style. But, unlike the writer of this guideline, I don't particularly want to push anyone to use it. And if I did suddenly want to make such a push, I'd invite comments first. -- Hoary (talk) 07:39, 9 January 2011 (UTC)

(quoting what I just posted in answer to your comments on my talk page) Ironically enough, the main discussion I can see was initiated by you back in 2006. While you initially started the discussion talking about foreign language titled using all caps, part of the discussion included my referencing the same Wikipedia:Manual of Style (capital letters) referenced now in the MOS-JA (search for "normal capitalization"). It was also discussed briefly here, here, and here.

So, this isn't anything new, and it wasn't done without comments (as you can clearly see above), but rather just clarifying what has already been discussed and in practice for a long time (except, of course, with "your" articles, as you pointed out—I will note that "your" articles aren't the only ones I've corrected over the years to meet what I thought was already mentioned in the MOS-JA). ···日本穣^? · 投稿 · Talk to Nihonjoe · Join WikiProject Japan! 08:42, 9 January 2011 (UTC)

For starters, that much earlier discussion that I started was (as you point out) about WACKY STUFF LIKE THIS and not about More Tolerable Stuff Like This. Well, what do you think about the syntactic complications of the up style, Joe? Are you planning to flesh out the rules? (E.g. is ni yotte a compound particle, a pair of particles, a particle and a verb?) -- Hoary (talk) 10:10, 9 January 2011 (UTC)

I've never seen wa, e, o, no, ni, ga, na, ja, ka, kana, desu, yo, etc. ever capitalized. If it's not a noun, verb, or adjective (adverbs in my experience are just adjectives appended with ni anyway), it shouldn't be capitalized in a romanization and as a practice isn't anyway.—Ryūlóng (竜龙) 09:09, 9 January 2011 (UTC)

Well, there are words such as yukkuri. But here you're appealing to Japanese-specific convention or to what might be termed common sense. I don't disagree with you, and indeed common sense (as it seems to me, though clearly not to Nihonjoe) would extend to skipping far more capitals; basically, everything other than names. Easy to understand, easy to implement. -- Hoary (talk) 10:10, 9 January 2011 (UTC)

Yukkuri would be subject to WP:MOSCAPS where anything longer than five (or some other number) letters is capitalized. And I don't think you're interpretting Nihonjoe's stance correctly.—Ryūlóng (竜龙) 10:28, 9 January 2011 (UTC)

Time permitting, I'll make a fresh attempt at interpreting it tomorrow. -- Hoary (talk) 11:59, 9 January 2011 (UTC)

"Desu" is a verb, so would be capitalized, and "ni yotte" would be a particle and a verb for our purposes. ···日本穣^? · 投稿 · Talk to Nihonjoe · Join WikiProject Japan! 19:02, 9 January 2011 (UTC)

My mistake.—Ryūlóng (竜龙) 23:29, 9 January 2011 (UTC)

That underlines Hoary's point pretty nicely. If obsessive pedants like us, whose idea of a good time is hanging out on MOS-JA debating capitalization rules, can't get the "up" rule right for Japanese, how can we expect normal users to? Jpatokal (talk) 01:54, 13 January 2011 (UTC)

That's what page moving is for.—Ryūlóng (竜龙) 01:59, 13 January 2011 (UTC)

Redirection? But this works equally well in any direction. -- Hoary (talk) 02:41, 13 January 2011 (UTC)

Well, once we determine what should and should not be capitalized in titles outside of nouns, all we have to do is explain it clearly on this page and point to it whenever we have to enforce it. That's pretty much how any manual of style works.—Ryūlóng (竜龙) 02:53, 13 January 2011 (UTC)

Yes indeed. How's this: "Capitalize the first letter of the title, subtitle, or any proper name; use lowercase for everything else." That's the down style. Shorter, simpler, and I think more likely to be read and understood than a description of the up style. -- Hoary (talk) 10:43, 13 January 2011 (UTC)

But that would mean that a slew of words that look like they should be capitalized (to the eyes of an English language reader) are not. The "up" style is what is used most commonly on this project.—Ryūlóng (竜龙) 19:41, 13 January 2011 (UTC)

(undent)

Is there a reason not to use the "down" style? To my eye things like "Otoko wa Tsurai yo" just look weird, since it makes "Tsurai" feel like a proper noun. Jpatokal (talk) 21:41, 9 January 2011 (UTC)

But "Otoko wa Tsurai yo" is a proper noun as it is a film title. Also the "down" style isn't used by anything. I don't think I've ever seen any piece of media (other than subtitles of scientific papers) that only capitalizes the first word and any subsequent proper nouns.—Ryūlóng (竜龙) 22:29, 9 January 2011 (UTC)

A quick Google indicates a fairly even split between the two styles, with perhaps a slight advantage to the "down" style -- hard to tell since Google doesn't provide comparative counts, and the "up" style gets a boost from Wikipedia and its mirrors.

And re: proper-nouniness, "Otoko wa Tsurai yo" could feasibly be parsed as "The man is [Mr.] Tsurai". "Otoko wa tsurai yo", on the other hand, is unequivocally "[being] a man is tsurai". Jpatokal (talk) 23:25, 9 January 2011 (UTC)

Well the down style also probably gets a boost from a lot of normal sentances in everday speech that are properly translated without capitalizing them. That is a legitimate phrase people use in everyday conversation and we wouldn't capitalize every major word in normal speech.陣内_Jinnai 23:33, 9 January 2011 (UTC)

(edit conflict) I've seen the down style used when the title is a complete question per English grammar (a subject and verb).陣内_Jinnai 23:28, 9 January 2011 (UTC)

I don't see a single reference to anything other than the movie in the search above; besides, any everyday conversation regarding the tsurainess of otokohood would in Japanese, not romaji. Jpatokal (talk) 23:37, 9 January 2011 (UTC)

Well desu can be with or without a capitalization as it is a common verb. In addition to articles and prepositions in English, we don't always capitalize the "to be" verbs because they are short and very common. I would say desu is classified as that and I know I've seen titles where it wasn't capitalized.陣内_Jinnai 04:02, 12 January 2011 (UTC)

Hoary's whole point was that, if we adopt the "down" style, we sidestep this problem entirely. Jpatokal (talk) 23:42, 12 January 2011 (UTC)

I understand. I don't think we should adopt either style since they both seem to be used about the same. For the up style we should not common exceptions and for the down, we should restate the obvious (because it never hurts to be clear when your talking about general terms like up/down), that it doesn't include proper nouns. As I said with the google results, those can be twisted either way so its best to use them as is or not use them at all.陣内_Jinnai 23:47, 12 January 2011 (UTC)

I'm sorry, but although I have read this comment of yours twice, I do not understand what you are recommending. (Me, I am recommending the down style. This comes in a variety of trivially different flavors and we can decide among these later.) -- Hoary (talk) 00:15, 13 January 2011 (UTC)

I mean to support neither becauce evidence of usage doesn't support one over the other.陣内_Jinnai 15:27, 13 January 2011 (UTC)

Yes, this was indeed my point. Incidentally, my hunch is that forms of BE evade capitalization in certain varieties of the "up" style not because they're short ("was" is no shorter than "cut") but instead because of this or that additional specific rule for BE. But if in Japanese we lump the copula together with "particles" and if we also get into letter-counting (as seemingly recommended above), then perhaps we'll end up with the combination of non-capitalized "da" and capitalized "Deshita". How about ある [and いる]? "Wagahai wa Neko de Aru"? "Uchi no Nyōbō nya Hige ga Aru"? (And is nya in The List of Particles?) Yes, because "Aru" is the final word. But in "Go-en ga A/aru Yo", "Yo" is the final "word", so "a/Aru" no longer is. Uh, unless you want to say that no, "yo" is not a word but instead a clitic. However, the eyes of the huge, grammar-uninterested majority of editors will have glazed over by this point. -- Hoary (talk) 00:15, 13 January 2011 (UTC)

What is this "BE" that you are throwing about? Also, neither "Aru" nor "Iru" are particles so they would be capitalized. "Yo" would not in your example.—Ryūlóng (竜龙) 02:53, 13 January 2011 (UTC)

Apologies for the throwing around; I hope that you dodged each one as it flew toward you. By "BE" I meant "the verb 'to be'" in any of its forms. (I put it in all-caps as a simple substitute for all-small-caps, which I'd have used for the verb as a lexeme.) I believe that WP's rule for the capitalization of English titles calls for capitalization of the last word, regardless of which part of speech it is; if you don't want Japanese to do this, your style sheet will have to explain. ¶ I don't understand this list of "Japanese particles", which includes such miscellanea as what I'd consider closed-class suffixes (the mono of tabemono; closed-class because I don't think that you can say for example sutemono) -- and I can't then appeal to grammatical particle as that article is more confused. -- Hoary (talk) 10:43, 13 January 2011 (UTC)

Well, the easiest way to do things now would be to leave any romanicized singular kana (or dual kana if one of the vowels is merely being lengthened as in the rare naa and saa cases) uncapitalized. Another thing currently in practice that is easy to explain.—Ryūlóng (竜龙) 19:41, 13 January 2011 (UTC)

So we'd have "Kore wa Hon Desu", but no caps at all for "Uo no yō na e da"? Or "Uo no Yō-na e da", perhaps? Jpatokal (talk) 23:18, 13 January 2011 (UTC)

"Yō" would be capitalized because it is not a particle. Although I do not think a hyphen would be used. I mentioned that "naa" and "saa" are the only multi-kana cases that should probably not be capitalized because they're usually な and さ, except when it is being emphasized with an あ or ぁ. That would be "Uo no Yō na e da". Any of these that consist of only one kana would not be capitalized.—Ryūlóng (竜龙) 23:32, 13 January 2011 (UTC)

But surely e (絵), a noun, should be capitalized? Why is the copula "Desu" capitalized but the copula "da" is not, even though they mean the same thing? Is "Yō" capitalized when spelled 様 and not capitalized when spelled よう? Do you realize that your "easy to explain" rules keep changing whenever someone points out exceptions, and that down this way lies madness? Jpatokal (talk) 01:29, 14 January 2011 (UTC)

If it is a single hiragana character, it is not capitalized. If it is a kanji that consists of only one mora, it is capitalized. And "yō" is capitalized if it is 様 or よう. And the rules I have put forward are simple, but every exception you think there is is because apparently my rules aren't simple enough.—Ryūlóng (竜龙) 02:50, 14 January 2011 (UTC)

In this case, "E" would be capitalized because it is a noun. In titles, all nouns are capitalized, regardless of length. ···日本穣^? · 投稿 · Talk to Nihonjoe · Join WikiProject Japan! 05:47, 15 January 2011 (UTC)

See? Even you two can't agree on what the rules are supposed to be! Jpatokal (talk) 21:56, 16 January 2011 (UTC)

If e is 絵, it should be capitalized. If it is へ, it should not be capitalized. I think my proposal covered that.—Ryūlóng (竜龙) 21:59, 16 January 2011 (UTC)

ゐ/ヰ and ゑ/ヱ

Yesterday, I modified Hepburn romanization, Hiragana, and Katakana to modify the treatment of the above two kana. I added the letter W to each page. Today, I was reverted on all three pages by User:Unnecessary stuff because my edits were unsourced, despite the fact that we clearly refer to both as the wi and we kana, rather than deferring to the supposed Hepburn romanization that does not pronounce the W, yet we still refer to を/ヲ as wo.

As these are archaic characters is it really required that we only refer to these characters as i and e in tables when we include the IPA pronunciations and a small footnote pointing out that the w is not pronounced in these kana?—Ryūlóng (竜龙) 19:28, 14 January 2011 (UTC)

We don't pronounce the w in wo, yet it's still part of the w syllables. As are the two archaic characters. The w should be included. However, I understand that this could lead to people saying, well, then why not say tu, or si, or ti? But the thing is, those aren't ambiguous. There are no other si, or shi, etc. characters on the chart, and wi is fundamentally different from i, and it would confuse our readers, I think, to pick the more accurate pronunciation (of an archaic character) rather than to elaborate on its etymology by including the w. And I seem to recall in my class actually pronouncing the w from time to time, at least perhaps when learning it. --Golbez (talk) 19:38, 14 January 2011 (UTC)

In addition, they are among the w kana in the Gojūon. And the shi, tsu, and chi characters are utilized as such in the extensions of katakana to approximate shounds like in "shame", "change", and "tsar". The primary issue is that in Hepburn romanization, the w is not even bothered with and apparently this is the form that appears in books on Hepburn. It's better to have a footnote that explains the practice than it is to simply ignore that these are in the W group and they don't have W's when they are referred to.—Ryūlóng (竜龙) 19:45, 14 January 2011 (UTC)

Yes, they are indeed in w- row and you don't need to source or prove that. But romanization is a different issue. Just because they belong to w- row doesn't mean they must have a w; as you know, Hepburn romanization is not 100% systemic.

I have looked up Hepburn's works, and not a single one of them romanized them as wi and we. Yet I actually provided the source few months ago (that romanizes ゐ and ゑ as i and e, respectively), you are still pushing your point of view. For the particle を, Hepburn actually romanized it as wo on his work and that is why I'm keeping it as it is. If any of the future versions of Hepburn romanization uses o, I will agree with romanizing the particle を as o.

If confusion is the matter, we can call them as wi kana, and we kana, only as an individual letter; if we say wi kana combination and we kana combination, then they would be ウィ and ウェ. If someone does not understand the difference between kana and kana combination, then we could refer them as w-row i kana and w-row e kana. Also, what about ji and zu? Aren't they confusing as well because there are じ/ぢ and ず/づ? --Unnecessary stuff (talk) 20:21, 14 January 2011 (UTC)

Yes, the ji and zu kana are also confusing, but the latter of each of those pairs rarely ever appear. And yes, we know that most reliable sources romanize ゐ as i and ゑ as e, but this is for uniformity in a table. If we are already including the w for を when in practice it is always romanized as o, the thing is that we are still pointing out that they are not used in modern usage, and we should refer to the kana name if we already have the IPA pronuniation, particularly in tables just for katakana and hiragana rather than the table of Hepburn romanizations.

I am not pushing any point of view. I am merely trying to make the tables easy or anyone to understand, and by having i and e in a row with wa and wo (and wu on katakana) makes no sense, despite what reliable sources tell us about the romanization scheme we use on this project.—Ryūlóng (竜龙) 20:30, 14 January 2011 (UTC)

"Rarely appear" doesn't mean they're not in use though. I'm pretty sure that ji and zu would be much less confusing than i and e, because ゐ and ゑ are not in use anymore but ぢ and づ are still in use.

Yes, wa wi (wu) we wo would make more sense, but I'd like to say that you don't really need uniformity, and it's not always guaranteed. If that's the matter, then we should use Nihon-shiki romanization that guarantees perfect uniformity. And if we are going to use Hepburn, we should understand that there is no perfect uniformity and give up on having one.

If we are going to use one system, then we should stick to that one only; mixing two or more systems would only be confusing and misleading. Romanizing wi and we while having ta chi tsu te to would be mixing two systems, and that would be more confusing and misleading. --Unnecessary stuff (talk) 21:03, 14 January 2011 (UTC)

But we're already mixing up the systems by leaving を as wo. In nearly all systems, を is o, yet according to your reading of the latest edition of the Hepburn books, を was wo and o, but ゐ was never wi and ゑ was never we. We're already pointing out that these are no longer in use and my recent edits of Hepburn romanization point out that in practice that they are i and e, but named wi and we for uniformity with the other related kana. ぢ and づ are already lacking uniformity because their non-dakuten forms are slightly different from the rest of the T kana. Leaving the name rather than the romanization for two obsolete characters is not going to harm the project and actually provides more information to the reader.—Ryūlóng (竜龙) 21:09, 14 January 2011 (UTC)

I totally agree with the last part, above - that leaving the name helps, and gives better info. When I was personally learning the alphabet, I wondered a lot about the 'missing letters'. Having them in the appropriate locations in the table makes it make a great deal more sense.  Chzz  ►  21:48, 14 January 2011 (UTC)

Leaving を as wo is not mixing up systems. For を, Hepburn used o when it was not a particle and used wo when it was a particle. Since after the kana spelling reform, を is only used as a particle and so wo is the only form that remains. If a proper noun uses を, then it would be romanized o because it's not used as a particle.

You seem to be concerning about uniformity a lot. But it's okay even if we don't have one. The Gojūon itself is already lacking uniformity, and we don't need to try fixing that. --Unnecessary stuff (talk) 06:37, 15 January 2011 (UTC)

But in Hepburn, を is o when it is a particle.—Ryūlóng (竜龙) 06:59, 15 January 2011 (UTC)

I suggest we return to the status quo we had a while back: the table lists them as "(w)o, (w)i, and (w)e", a footnote explains that the kana are obsolete, Hepburn used o/i/e, and the wo/wi/we forms are also used for disambiguation. Jpatokal (talk) 22:01, 16 January 2011 (UTC)

I think that would work at Hepburn romanization, but I believe the wi/we/wo forms should still be used at hiragana and katakana.—Ryūlóng (竜龙) 22:10, 16 January 2011 (UTC)

Please cite source for your view per WP:IRS, Ryulong. Then we can disscuss easily on each source.

Well...

• Unnecessary stuff's edit is verifiable with his/her cited source - ゐ/ヰ and ゑ/ヱ in historical kana orthography are written i and e(ye) in some Hepburn style (Hepburn's dictionary(ye), Romajikwai(e), Japanese passport standard(e), Japanese railway standard(e) etc.).
• ゐ/ヰ and ゑ/ヱ are undefined in some Hepburn style (Kenkyusha's japanese English dictionary, Kojien, etc.).
• ゐ/ヰ and ゑ/ヱ in wa column are undefined, but katakana ヰ and ヱ for loanwords are defined as wi and we in some Hepburn style (ANSI Z39.11-1972, BS 4812:1972 etc.). In case of this, ヰ(wi) and ヱ(we) should be written in "extend katakana" section.
  • There is a way that all kanas are included in the same table with notes (ja:Wikipedia(i,e), de:Wikipedia(wi,we), RÔMAJIBUN no TEHODOKI(wi,we) etc.).

--Mujaki (talk) 16:18, 18 January 2011 (UTC)

My personal view on how the tables should be formatted does not require reliable sources to back them up. If these two kana are universally known as "wi" and "we", we should not ignore that just because an archaic version of the Hepburn system ignores the W.—Ryūlóng (竜龙) 19:02, 18 January 2011 (UTC)

Um... I reverted your edits per WP:V. Jimmy Wales said the following:

There seems to be a terrible bias among some editors that some sort of random speculative "I heard it somewhere" pseudo information is to betagged with a "needs a cite" tag. Wrong. It should be removed, aggressively, unless it can be sourced. This is true of all information, but it is particularly true of negative information about living persons.[3]

This edit is perhaps OR; Wu is undefined in almost Hepburn style or refered "ウゥ" in Hyōjun-shiki.--Mujaki (talk) 07:40, 23 January 2011 (UTC)

Quoting Jimbo isn't really policy. And none of the table on Hepburn romanization is really sourced to anything. It can be verified that ゐ/ヰ and ゑ/ヱ are wi and we and are rarely used at all in modern Japanese and are referred to as wi and we and not i and e and that ヴ is used for V, ヷ and its similar symbols are never used in modern Japanese (WP:COMMONSENSE). If we have を/ヲ as "wo" in the table, then there's no reason why the obsolete characters cannot be "wi" and "we", considering the W is dropped from all three in Hepburn for whatever reason. We do not need to give the archaic out of use Hepburn romanization for these characters when we can simply make an exception considering they're rarely ever used in the language. And the issue of wu is unrelated to this discussion.—Ryūlóng (竜龙) 07:54, 23 January 2011 (UTC)

Sorry for the late reply.

For を, Hepburn used o when it was not a particle and used wo when it was a particle.

This is based on the 3rd and later editions of Hepburn's dictionaries. If in doubt, analyze his works.

"But in Hepburn, を is o when it is a particle."

If so, source this. Then I will agree to を = o when it's a particle.

Hepburn's works are not archaic. And even if they are, just because they are archaic doesn't mean they should be ignored.

You keep saying that you don't need a source, but yes, you do need a source. I have no idea why my sourced edits have to be reverted by you who do not even have a source and only believes in uniformity. I already said that you don't need uniformity, and it's not always guaranteed. If you think uniformity is more important than a source, then please explain why.

BTW, this source is primarily based on Nihon-shiki and Hepburn in parentheses. Or it could be based on Wikipedia's Hepburn romanization page. --Unnecessary stuff (talk) 03:39, 24 January 2011 (UTC)

Um, Hepburn romanization and this manual of style have always had pointed out that を = o in every instance and never wo as a particle. And it is clear that American and British Hepburn standards (as of 1972) say を is o as a particle and wo as a kana, ヰ is wi, and ヱ is we. So do not blindly revert me on every single page and remove content that is sourced and believe that your form is the only possible correct one. And this publication from this past December seems to agree that these two kana include the W in romanization.—Ryūlóng (竜龙) 03:53, 24 January 2011 (UTC)

But how do we know halcat.com is a reliable source? Unless we have access to the actual printed copy, we can't know that it's an accurate transcription.

Also, 外来語の表記につかわれる特殊なカタカナ means "special katakana (combinations) that are used for writing foreign words," which corresponds to Mujaki's "ゐ/ヰ and ゑ/ヱ in wa column are undefined, but katakana ヰ and ヱ for loanwords are defined as wi and we in some Hepburn style. In case of this, ヰ (wi) and ヱ (we) should be written in 'extend katakana' section."

That publication is extended Hepburn system, but that can't really be used as a source because it gives different romanizations for certain kana, and it is based on historical kana orthography, not on pronunciation. --Unnecessary stuff (talk) 05:01, 24 January 2011 (UTC)

Halcat publishes various other forms of romanizations, so it is very unlikely that they are making anything up. And the Extended Katakana publication still shows that ヰ is not i and ヱ is not e, and that can be applied to their hiragana forms. It's still pointed out all over that the pronunciation is identical to i and e. So rather than go against the practice in every other system and omit the w for the ゐ and ゑ, why not go with said systems?—Ryūlóng (竜龙) 05:16, 24 January 2011 (UTC)

They would not be making things up, but those transcriptions might contain errors. As I pointed out, it has si in the text and shi in tables. And if we are following the Extended Hepburn system, then we should romanize じ as zhi, not as ji. --Unnecessary stuff (talk) 05:31, 24 January 2011 (UTC)

The "extended Hepburn" system is only used for the extended foreign language transcriptions.—Ryūlóng (竜龙) 05:48, 24 January 2011 (UTC)

Names versus transcriptions

I am surprised to find myself agreeing to a large extent with User:Ryulong.

If we want to be scientifically accurate – and we should –, we need to clearly distinguish between transcription (or transliteration) of phrases and words, which sometimes consist of a single character, on the one hand and talking about characters (letters, symbols, glyphs, graphemes, syllabograms, …) on the other, especially in articles on scripts and writing systems (i.e. scripts used with a certain language). We must also distinguish (prototypicall) pronunciation hints from character names and transcriptions, especially since the phoneme-grapheme correspondence is weaker in all actual writing systems than most laypeople think.

In many scripts the symbols have well-known, often language-dependent names (e.g. Alpha / alfa, Beta / vita, …). Two distinct symbols within a script never have the same name, except when they are case pairs or are other, non-semantic glyphic variants of each other. For the syllabograms of the kana script[s], too, there can only be one (at most) I, Ji etc. Still, several kana syllabograms (monographs or digraphs) can be transcribed (depending on the transliteration system used) into the roman alphabet or the English writing system with the same letter sequence, e.g. 〈i〉 or 〈ji〉, and likewise several (phonic) syllables, e.g. /i/ and /ji/ or /uː/ and /wu/, can be transcribed with the same (complex) syllabogram, likewise one spoken syllable can be transcribed with two (or more) different syllabograms, e.g. ジ or ヂ for /dʒi/, depending on (usually morphologic) context. And we have not even actually touched words, e.g. jin or gin, yet.

I prefer symbol (character, letter, …) names to be written with an uppercase initial letter as is common for other proper names in contemporary English orthography, but this convention is not strictly followed among linguists. Phonologic information is usually enclosed in slashes when it is a phrase or word (/wɜːɹd/) and in square brackets when it is a phoneme or (allo)phone ([iː]). Lexemes (word paradigms) in the roman script (and often the greek and cyrillic script, too) are italicized (word), whereas those in other scripts are usually just kept in their native unmarked typographic style; word forms, on the other hand, are sometimes also put in italics, but just as often (and when the difference matters) they are put inside matching quotation marks, just like short quoted phrases, but usually simple ones (‘words’), not double ones (“words”), although these are also used to provide the semantics of a word. Foreign glyphs and graphemes are sometimes kept as is, just like words, but at least where metalanguage matches object language angular brackets are employed (〈x〉, ‹x› or <x>).

I digressed into linguistic typography, sorry for that. To conclude, kana syllabogram lemmas in articles and tables should be based on gojūon rows and columns, i.e. conform to the Kunrei system. Where Japanese words and phrases are transcribed a revised Hepburn system is fine for the English Wikipedia. Phonemic information employing IPA has no place in articles on scripts, and it should be avoided in articles on writing systems. — Christoph Päper 14:13, 8 February 2011 (UTC)

This was resolved when it was discovered that wi and i and we and e are used interchangably for these glyphs in the Hepburn system, which is what the issue revolved around.—Ryūlóng (竜龙) 18:47, 8 February 2011 (UTC)

WP:VG/GL mediation

The previous discussions on this point are archived at Wikipedia talk:Manual of Style (Japan-related articles)/VGGL and Wikipedia talk:Manual of Style (Japan-related articles)/VGGL2.

• This new discussion is an attempt to reach a decision on a point, which has run for over 6 months. The disputed point is: how to represent in English Wikipedia articles the Japanese names of Japanese-made videogames whose Japanese names, or parts of them, are English words whose pronunciation has been rendered in the Japanese syllabic writing system called katakana? Such rendering is usually inaccurate because of the limitations of the katakana spelling system, which was developed for Japanese phonetics. An example is "Final Fantasy", rendered in katakana as ファイナルファンタジー, which is literally "fainaru fantazī".

  The question seems to be: in the English Wikipedia page, should we display these "back from katakana to Roman" re-transcriptions of English words, or should we only display the original English spellings of those words?

  The current policy stated in Wikipedia:Manual of Style (Japan-related articles)#Romanization says "use the English spelling".

  One method used sometimes is to use the correct English spellings everywhere in the text, but to put the literal re-romanization in the arguments of the {{nihongo}} template after the first mention of the English name of the game/etc, e.g.:

  Final Fantasy VII (ファイナルファンタジーVII, Fainaru Fantajī Sebun)

  {{Nihongo|'''''Final Fantasy VII'''''|ファイナルファンタジーVII|Fainaru Fantajī Sebun}}

  This method allows access to the literal back-transcription for those who want to see it, without cluttering the text.

  (Note: "nihongo" is Japanese for "the Japanese language".)

  • Please keep discussion concise and on-topic and avoid repeating. Anthony Appleyard (talk) 23:24, 24 January 2011 (UTC)

• Anthony, the dispute was whether or not to include the Hepburn romanization in the third parameter of {{nihongo}} in the lead of the article if the Japanese title of the game was English language words rendered in katakana. This guideline currently suggests that the romanization should be included whereas the users of WP:VG believed that including the romanization of text that is inherently English was superfluous/redundant. This was mostly spawned by my repeated replacement of the "Fainaru Fantajī" text across several articles where they had been removed. It seems that for the most part the practice has been going with "Use romaji" considering the actions of users otherwise uninvolved. Most discussion as of late was "there's no one to mediate so let's just do whatever" as I haven't seen any other disputes.—Ryūlóng (竜龙) 00:28, 25 January 2011 (UTC)
• The argument is not that it should never be done, but that it be required. The VG guideline says that this is optional and this guideline says it should always be there. That is the crux of the problem. The VG guidelines say to leave it up on a case-by-case basis in these circumstance. Since its not necessary to know how to pronounce a English word written in katakana (since it's suppose to represent the English word as best their script can) it shouldn't be forced. 陣内_Jinnai 01:05, 25 January 2011 (UTC)
• I think Jinnai's nailed it. The only thing I'd add is that there seemed to be a move toward agreement that romaji could always be used provided that it be used as hovertext through the {{nihongotip}} template. This compromise was rejected for reasons of access, but I think this was a mistake as WP:ACCESS actually doesn't bar this at all. To interpret it thus is to corrupt its original intent. I argue that it's time to revisit this compromise proposal. -Thibbs (talk) 15:26, 25 January 2011 (UTC)
• The "issues to be mediated" section still sums the problem up pretty well. The crux of the issue is that there is no strict site-wide guideline on including/excluding romanizations for katakana words of English (or rather non-Japanese) origin, along with the sentence "Phonetic transcriptions are, as a rule, not considered to be significantly different and thus do not warrant the inclusion of Japanese titles" from the video game guidelines. Either that has to be removed, or people have to abide by it, or there has to be a compromise. The problem was that some users from both sides were adamantly resisting on their point of view, which made it impossible to enforce either extreme – and compromises were made impossible by additional guidelines. Good luck with unraveling this situation as it's currently at an impasse. Worse even. I'd say it's located in Impasseville, Cul-de-Sac 13. Ahem. In any case, I think the involved users should be renotified as I don't think everyone followed the mediation talk page. With any luck, some might have even had a change of heart, making a true consensus easier. Prime Blue (talk) 17:22, 25 January 2011 (UTC)
• Sometimes back-transcriptions from kana can look very different from the English original, for example Thunderbird → サンダーバード → Sandābādo. Anthony Appleyard (talk) 23:34, 25 January 2011 (UTC)
  • But sometimes they can look very similar - like Nintendo DS → ニンテンドーDS → Nintendō Dī Esu. I think this illustrates the fact that a certain amount of personal discretion is required by the editor to apply WP:VG's "Phonetic transcriptions" rule. Using a nihongotip template would resolve that issue. -Thibbs (talk) 01:29, 26 January 2011 (UTC)
    • That is to be expected with this example: Nintendo is a Japanese word: 任天堂 in kanji = "responsibility/office, heaven/sky, hall". Anthony Appleyard (talk) 09:55, 26 January 2011 (UTC)
      • Not according to WP:MOSJA... I don't see any romaji. That's the basic issue - whether inclusion of romaji should be mandatory or optional (i.e. only used if important to the context of the article). -Thibbs (talk) 13:22, 26 January 2011 (UTC)
        • Well, including the romanization for an expression with kanji was never disputed – that has always been mandatory. Prime Blue (talk) 14:21, 26 January 2011 (UTC)
    • I thought that the proposed nihongotip template was not allowable under WP:ACCESS. Tooltips are also fairly unprofessional looking. Just because you have "Nintendo" and "Nintendō" (neither of which are words of English origin) does not mean the latter should be obscured.—Ryūlóng (竜龙) 01:34, 26 January 2011 (UTC)
      • I disagree that WP:ACCESS bars the use of nihongotip. Like the use of color in articles, the use of Romaji doesn't convey "important information" as the term is used in the guideline. As with color, Romaji can be used to enhance an article, but it still represents a mechanical transformation and so no unique information is being presented (i.e. it is not "important information"), only redundant information in a different form. The fact that hovertext is incompatible with (some) TTS programs shouldn't bar its use on articles unless it conveys "important information" as the term is used in WP:ACCESS. -Thibbs (talk) 01:55, 26 January 2011 (UTC)
        • The method by which to read non-English text in a standard way would be "important information". Also this discussion is being entirely commented on by proponents of the WP:VG/GL treatment.—Ryūlóng (竜龙) 02:47, 26 January 2011 (UTC)
          • Only when its clear that the text is not trying to convey an English word. As for Nintendo, while its not an English word, its a company name registered that way in English so that point is moot.陣内_Jinnai 03:26, 26 January 2011 (UTC)
          • The method by which romaji is derived from kana is not at issue here. That inclusion of romaji itself is the question. Since the method is a fully standardized method that can be performed mechanically by anyone - seeing or blind - it is redundant information and thus not "important information." As I said earlier it can enhance an article, but it shouldn't be used as a means to bar a very workable compromise. I think an old adage sums it up best: The perfect shouldn't be the enemy of the good. -Thibbs (talk) 03:33, 26 January 2011 (UTC)
            Also for the record, I am not affiliated with WP:VG, and I am not a proponent of WP:VG's "Phonetic transcriptions" rule. I am using common sense and my opinion is that the compromise (which makes romaji-inclusion within a nihongotip template mandatory) is ideal. -Thibbs (talk) 03:40, 26 January 2011 (UTC)
            • It just seems that the application of the tooltip would not convey the information as clearly as the basic {{nihongo}} templates do. Every other language receives the "English (Not English, Romanization)" treatment. There's nothing about video game articles that it seems that they deserve separate treatment just because one group of editors feels the information is redundant when others do not.—Ryūlóng (竜龙) 04:32, 26 January 2011 (UTC)

• Whether romaji is redundant or not isn't simply an opinion. It's a fact. An analogy can be drawn to the prime factorization of numbers. Just as romaji is derived from kana by an exact and mechanical method without any new information, each number can be broken down into its prime factors by an exact and mechanical method without any new information. Like romaji, prime factorization can be helpful or it can be unnecessary. Whereas we leave it to WP:WPMATH to determine when prime factoring is helpful in the context of the article, I take it you are arguing that WP:MOSJA supersedes WP:VG in this area. By way of a compromise, the use of the nihongotip template as a mandatory alternative to the nihongo template would mean that romaji would appear on every article just as mandated by WP:MOSJA. I see no downside. If you are still concerned that visually-impaired users may lack the competence to perform the standardized determination for romaji, and that it is important within the context of the article to know for example that Super Mario USA → スーパーマリオUSA → Sūpā Mario Yū Esu Ē, then I suggest we consult someone from WP:WPACCESS rather than speaking for them. -Thibbs (talk) 13:22, 26 January 2011 (UTC)
  • When I asked if it would pose a problem at an FA nomination, there were no strong opinions from the delegates. But asking members of a project that would actually be affected by it is a much better idea to determine if the tooltips could not be accepted at all. Prime Blue (talk) 14:21, 26 January 2011 (UTC)
• By the way: I notified all the remaining users that seemed to be unaware of the mediation. Prime Blue (talk) 14:21, 26 January 2011 (UTC)
• One thing affecting whether romaji is redundant or not, is that most people outside Japan cannot read kana, and most people outside Japan or Chinese communities cannot read kanji. Anthony Appleyard (talk) 14:40, 26 January 2011 (UTC)
  • It's always redundant insofar as it can always be created by any user using a standardized method. Although 「ブラブラブラ」 isn't a real word in Japanese (at least as far as I know since I just made it up), it can still be rendered in romaji (as "buraburabura") by using an unchanging standardized method that anyone can apply regardless of ACCESS-level. That's why the prime factorization analogy works. A non-sentient computer can generate a prime factorization of any number based on set rules that could be applied by anyone. All it needs is the number itself. The same is true for romaji which can be generated based on set rules by anyone given nothing more than the raw kana. It could potentially enhance an article, but in cases where this "enhancement" is as much an enhancement as the addition of color (eg. Sūpā Mario Yū Esu Ē), it seems obvious that it should be allowed to go into hovertext.
    It's also worth pointing out that there are no equivalent examples of this mandatory transliteration occurring in any other-language wikipedias as far as I can see. Thus for example the ja.wiki article on the American iPod product shows the name of the device rendered in katakana and in English letters, but there is no phonetic transliteration of the English apart from the kana itself (which is actually rendered as "aipoddo" in romaji). -Thibbs (talk) 15:02, 26 January 2011 (UTC)
  • Referring to the last example, kana is the equivalent of romaji for all intents and purposes in Japanese as it shows how to pronounce the word in it's most basic form as long as the person has a basic grasp of Japanese pronunciation. I should note that I've already indicated in some other discussion somewhere that I'm fine with having the romaji be in tooltips as an option if that will make WPVG happy, even though I still disagree with their insistence that including the romaji is redundant and that it "clutters" the article unnecessarily. ···日本穣^? · 投稿 · Talk to Nihonjoe · Join WikiProject Japan! 16:49, 26 January 2011 (UTC)
    • Yes but romaji/kana isn't the equivalent of English. We're not talking about providing a phonetic transcription of English, but rather a transliteration of foreign characters for pronunciation purposes. My point is that the ja.wiki (and as far as I know all other-language wikipedias) doesn't share our requirement that a standardized transliteration should be provided for every article on a topic with an originally English-language name. For us the requirement seems to be "Our-language → Their-language → Transliteration" whereas for all other wikipedias the requirement is merely "Our-language → Their-language." As for romaji being redundant, I only make this point to illustrate that for ACCESS purposes it is similar to color. Although it can be useful, it is not "important information" as WP:ACCESS uses the term. -Thibbs (talk) 17:33, 26 January 2011 (UTC)
      • I'm sorry, but I think you misunderstood me. For the Japanese, kana is used to transliterate kanji pronunciation (as well as transcribe words and parts of words not written with kanji), and therefore serves the same purpose in Japanese as does romaji in English: it shows you how to pronounce the word in Japanese. Kanji does not do that, and unless people can read kana (and unless their system is also set up to display the kana), the kana won't do that either here on the English Wikipedia (what other language wikis do is mostly irrelevant to the English-language Wikipedia as every language has the right and ability to choose how to do whatever they want to do). Thus, the requirement for the transliteration/romaji, which provides that pronunciation information. As I stated above (and elsewhere), I'm fine with having the option to put the romaji into a mouseover/tooltip on video game articles since they (which is a weird thing to say since I also edit video game articles a little) seem so bent on completely ignoring MOS-JA on this point. The main two arguments I've seen put forward are that the romaji "takes up too much room" and that it is "redundant". The first argument is only valid in a few, scattered cases, so is mostly baseless. The second could be said about any pronunciation information, so they may as well be arguing to remove all IPA information, too. That's just absurd, though, so that argument is also baseless. I do think your proposal below is a good one, though. ···日本穣^? · 投稿 · Talk to Nihonjoe · Join WikiProject Japan! 05:24, 27 January 2011 (UTC)
        • A few points: 1) I don't see anybody here arguing that MOSJA should be completely ignored and that romaji should be done away with on video game articles. The argument is that it should be allowed to be placed in a tooltip in some cases as an alternative based on editorial consensus. This would apply to all articles, not just video game articles. 2) The IPA phonetic notation is quite different from the standard romaji transliteration of kana. There are multiple ways to pronounce the English word "bow," for instance, and IPA phonetics cannot be applied systematically to this word in the same way that romaji can be to kana. The raw word "bow" itself cannot be the basis of a proper IPA rendering, so external information is required. This is in stark contrast to the regular and systematic way that the rules of romaji are applied to kana based on nothing more than the raw japanese characters. 3) My mention of what other language wikipedias were doing was meant to be persuasive, not controlling. Obviously we can do whatever we want with our own wikipedia, but considering that the entirety of the other wikipedias uses a system that is different than ours, it might give us pause to consider if what we are doing is actually as centrally necessary as we might consider it. -Thibbs (talk) 23:53, 27 January 2011 (UTC)
• If the reader has access to that standardized method, which does not include people who cannot read kana. Luckily I wrote a Windows program that can transcribe various Unicode alphabets to Roman (so far, Indian alphabets and Armenian and kana). Anthony Appleyard (talk) 16:40, 26 January 2011 (UTC)
• I think that the argument that WP:ACCESS bars the compromise reached earlier uses spurious reasoning. My ideal solution would be the following:
  

Proposal:

1 - WP:VG should drop its optional "Phonetic transcriptions" rule and accept WP:MOSJA's mandate that all articles on topics that have Japanese names should contain romaji as in general it is information that enhances articles.

2 - The default is to include romaji using the {{nihongo}} template.

3 - If editors agree by consensus that the transliterated form of the word is substantially identical to that of the English term (e.g. "Sūpā Mario Yū Esu Ē" compared to "Super Mario USA") then the {{nihongotip}} template may be used as an alternative.

Any thoughts on this? -Thibbs (talk) 17:42, 26 January 2011 (UTC)

I simply feel that the 3rd aspect of the proposal would only be dealt with by the WP:VG crowd. However, I believe that we should copy the Japanese Wikipedia's practice of when there is identical hiragana or katakana in their transcription from kanji into hiragana on their various pages. So instead of "Sūpā Mario Yū Esu Ē", it would be "Sūpā - Yū Esu Ē" as "Mario" is going to be the same in both languages. Unless this makes things slightly more confusing.—Ryūlóng (竜龙) 19:25, 26 January 2011 (UTC)

• It makes things more confusing. "Mario" should be in both versions, in the example that you give. Anthony Appleyard (talk) 20:39, 26 January 2011 (UTC)
• No, the 3rd aspect would not be limited to WP:VG, but would of course be available across all projects. As we had brought up in prior discussions, for non-video-game terms like "elevator → エレベーター → erebētā", editors could decide by consensus to put the romaji into hovertext. (Note: Not the best example, I know, but the principle is the same) -Thibbs (talk) 21:20, 26 January 2011 (UTC)

Thibbs, would this mean that an article such as Nintendo DS would have to have the transcription "Nintendō Dī Esu", which is essentially a mangled Japanese attempt to say "Nintendo DS"? I really feel that these transcriptions are confusing to readers who aren't familiar with what they are, which is why I've been so opposed to them all along. It's fine when the name of the game or system is completely different, but it's silly to have when the Japanese are pretty much borrowing mangled English over to the names of their games. Nomader ^(Talk) 21:04, 26 January 2011 (UTC)

What it means is that the term "Nintendo DS" would be required to be written in one of the following two ways:

1)Nintendo DS (ニンテンドーDS, Nintendō Dī Esu) - The default way unless editors object

or

2)Template:Nihongotip - The alternate way that would be available for editors based on consensus.

-Thibbs (talk) 21:20, 26 January 2011 (UTC)

Nomader, it's not really a "mangled Japanese attempt" when "Nintendo" is their word to begin with. The "Dī Esu" bit is the only part that you have some sort of argument that it could be change to just "DS" in the romaji field.—Ryūlóng (竜龙) 23:12, 26 January 2011 (UTC)

Thanks for clarifying, Thibbs. I'd have absolutely no objection to the second version. Can we hear from WP:WPACCESS about this to see if it would work? As I recall, that was the problem with it last time. To clarify, Ryulong, I feel if both of the versions sound the same and mean the same thing, I'm not really sure if we need to list the Japanese romanization next to it, especially when the pronunciation of the Japanese version is an attempt to sound like the English version. And sorry if that came out a bit negative Ryulong, my bad. Nomader ^(Talk) 01:38, 27 January 2011 (UTC)

But in the case of the DS it's the other way around. "Nintendo" is the English approximation of 任天堂. So even if the name is parsed in katakana, "Nintendo DS" is a poor example for this discussion.—Ryūlóng (竜龙) 01:49, 27 January 2011 (UTC)

As I said, I don't think its necessary in every article, but I'd be okay with the hovered text compromise. Final Fantasy is a better example of when it becomes redundant imo. In this case it is clear both words are suppose to be English words written as closely as kana system allows.陣内_Jinnai 03:44, 27 January 2011 (UTC)

Good point, Ryulong-- I wasn't aware of that, although other things such as Final Fantasy are better examples of what I have a problem with. I think the tooltip workaround would be the best route and I'd completely support it if it's usable. Nomader ^(Talk) 04:43, 27 January 2011 (UTC)

@Jinnai: Such a treatment I believe could work for the Final Fantasy articles, so long as at least one of them (primarily the main franchise article) retains the visible romaji. However, I feel that visible romaji would be useful on articles with otherwise ideosyncratic pronunciations compared to the English name (e.g. "UFO" being "Yūfō" and not "Yū Efu Ō", Final Fantasy Versus XIII having "Verusasu" rather than "B/Vāsasu") or having names that don't follow how English grammar would otherwise have it (Kingdom Hearts 358/2 Days's "Three-Five-Eight Days Over 2" being an issue in my book even if the current title is ridiculously long).—Ryūlóng (竜龙) 06:04, 27 January 2011 (UTC)

• I agree. Whenever there are idiosyncratic mismatches or otherwise notable differences then the Romaji should be visible. In all of the examples listed above I agree that visible romaji would be best. The {{nihongotip}} template should be reserved only as an alternative when the English and romaji are substantially identical. -Thibbs (talk) 12:12, 27 January 2011 (UTC)

Almost don't want to say anything, but..
I'm fine with the tool tip. While I originally thought that some Romaji was redundant, it eventually became apparent to me that the Romaji could be confusing to those unfamiliar with it. So my main concern became the absence of proper clarification. Like some label indicating that the text seen was Japanese and a romanization (like {{Leonese}}, {{Hebrewterm}}, and {{Lang-ja}}), as well as a help page that explains the pronunciation. I think the Romaji can be confusing because it does not follow English language pronunciation, but will surely be read that way. (Guyinblack25 ^talk 19:39, 27 January 2011 (UTC))

I thought someone was suppose to update the help page to include romaji pronunciations, although (not to get off the subject) it isn't clear modified hepburn is the most common. Still whatever form of hepburn is used, even wapuro, they follow enough basic rules that we could come up with something.陣内_Jinnai 20:26, 27 January 2011 (UTC)

Help:Japanese already exists, but I think some one took a stab at creating something else. I don't quite remember who or where though. Also, I think we should stick with the Hepburn described on this MoS page to keep things standardized. (Guyinblack25 ^talk 20:48, 27 January 2011 (UTC))

Are you thinking about this one? User:Joren/Help:Japanese. I'd created that as a sketch of how the page could be made more user-friendly to those coming from the question-mark link on the nihongo templates; my concern was that the existing Help:Japanese presents them with a wall of text that doesn't really answer the questions that likely would bring them to that page. My hack 'n slash version is not really ready to be the help page yet (some people had pointed out that the information originally in Help:Japanese was wrong/out of date to begin with) but perhaps others could take a stab at modifying it and making something workable out of it. Feel free to edit my user page version of it if you have the energy...

-- Joren (talk) 06:08, 4 February 2011 (UTC)

Request for comment on Nihongotip Compromise at WP:ACCESS

I've requested commentary from the editors at WP:ACCESS who in my estimation are most likely to be attuned to any possible problems the the Nihongotip Compromise might pose. The request I made can be found here. Please add anything that you may feel I left out. -Thibbs (talk) 00:43, 28 January 2011 (UTC)

The compromise which uses {{nihongotip}} in the example "Template:Nihongotip" does cause accessibility concerns.

It uses a tooltip, which indeed violate WP:ACCESS. WP:ACCESS prohibits the use of tooltips, except for the {{abbr}} template. The abbr template uses the HTML semantic tag <abbr>, which is used to provide the long form of an abbreviation or an acronym. The abbr template can be used solely for the purpose of explaining abbreviations. Any other uses - like in {{nihongotip}} - will produce unexpected and confusing results in screen readers, applications that reuse Wikipedia's content, etc. For example, screen readers will read: "スーパーマリオUSA", abbreviation: Sūpā Mario Yū Esu Ē". Yours, Dodoïste (talk) 21:15, 29 January 2011 (UTC)

Hm... That's a pity. Out of curiosity, when you say "screen readers will read: 'スーパーマリオUSA', abbreviation: Sūpā Mario Yū Esu Ē" I would be interested to know how the screen reader would pronounce 「スーパーマリオUSA」. If screen-readers provide the proper pronunciation anyway then there might still be no need for the romaji in the same limited cases that nihongotip would have been used in. Of course if they just said "Japanese characters" or something similar then I think this compromise is effectively dead. -Thibbs (talk) 23:38, 29 January 2011 (UTC)

It depends if the corresponding add-ons are installed in the screen reader. I've never tried myself, but I know that JAWS can support many unicode characters and languages if the corresponding modules are installed. Japanese is a popular language, so I suppose it can be supported by JAWS. I also know there is a screen reader made specifically for reading japanese. However, most screen readers will not be able to read japanese, but might be able to read the romaji (if the user wants it).

I am only providing explanations and details about WP:ACCESS, as requested. I do not have an opinion on whether we should keep or delete the romanji. Yours, Dodoïste (talk) 01:16, 30 January 2011 (UTC)

I would agree with Dodoïste that the tooltip will cause an annoyance for those using screen readers, by announcing the tooltip as an abbreviation (which probably wouldn't make much sense to a blind reader), so Compromise part 2 is a non-starter.

As for your other point, JAWS is a popular screen reader globally, and can use many different optional language packs. Since the web designer can never be sure whether the screen reader will hear the Japanese text, we always have to make sure alternative text is available.

But looking at your Guideline 1, a non-Japanese speaker won't have the installed the language pack so will probably hear "Link Super Mario USA, USA, Sūpā Mario Yū Esu Ē, Link Help:Installing Japanese character sets" or something similar – perhaps '?' for each unicode character. Whereas anyone who has installed the Japanese language pack (presumably a Japanese speaker), will hear the Japanese characters (which I presume are a phonetic transliteration in these cases). Frankly, I'm not sure what audience Guideline 1 is aimed at. However that isn't an accessibility issue, so take my thoughts as you will. Purely from the viewpoint of anyone using a screen reader, Guideline 2 will impart the information to either non-Japanese speakers or Japanese speakers, with the least amount of repetition, so is a good choice. Hope that helps. --RexxS (talk) 23:56, 30 January 2011 (UTC)

Breakpoint

• I tried using {{nihongotip}}, using the test form {{nihongotip|aaaaa|bbbbb|ccccc}}; it showed "Template:Nihongotip"; "ccccc" appeared on a small hover window, but only if I put the mouse in exactly the right place within about 3 screen pixels vertically, and not at all on some attempts, and many readers likely may not work out how to make the hover window appear, or may not know that the dotted underline means "here is a hover window".

  Please avoid letting the discussion get wordy and repetitious; I suggest "let the standard be to let the editor choose between omitting the romaji and putting the romaji in the 3rd argument of {{nihongo}}". Anthony Appleyard (talk) 23:06, 27 January 2011 (UTC)

  Hmm... I think consensus was forming around the idea of using nihongotip as a mandatory alternative (I count 5 of the 6 respondents in favor).
  Anthony, I think your suggestion runs against the current consensus at WP:MOSJA that romaji is helpful and should therefore be included on every page. Since WP:VG already mandates using it unless it is a phonetic transcription, I think a better compromise is to mandate its inclusion on every page but allow it to be placed in a tooltip as a mandatory alternative based on editorial consensus. -Thibbs (talk) 23:58, 27 January 2011 (UTC)

  And we're slowly discovering that the tooltip doesn't work as well as it has been planned.—Ryūlóng (竜龙) 02:01, 28 January 2011 (UTC)

  Please explain. -Thibbs (talk) 02:03, 28 January 2011 (UTC)

  Anthony seems to explain it well enough.—Ryūlóng (竜龙) 02:08, 28 January 2011 (UTC)

• We're talking about something that I believe is not "important information" per WP:ACCESS. The point is that it doesn't matter if some users have difficulties with tooltips. It's not important information and it can be derived simply and consistently by any editor regardless of ACCESS-level and without any additional information. It's the same as with the use of color to convey redundant information. Color and tooltips are both allowed. Just because WP:ACCESS mandates equal accessibility to colorblind users, this doesn't mean that color can never be used to convey information. What Anthony suggests is to make the use of romaji optional. This is completely contrary to your earlier position. Would you not agree that the use of tooltips to convey romaji in limited circumstances is closer to your vision than Anthony's suggestion? -Thibbs (talk) 02:18, 28 January 2011 (UTC)
  • But there's a difference between using colors and using a hovering text box which may not work as well as we previously thought.—Ryūlóng (竜龙) 02:29, 28 January 2011 (UTC)
    • Only on a semantic level. If the tooltip is used inappropriately, such hiding hepburn romanization of original Japanese words then that's another thing.陣内_Jinnai 02:33, 28 January 2011 (UTC)

• Deciding on a case-by-case basis should be the last resort as having no guideline on what to do would not solve the problem, but merely postpone it and spread it over to article talk pages. Either drop, include, or use the middle way. That said, {{Nihongo tip}} has to be used with Japanese text in the second field, not Roman letters. The other issues are legit concerns (albeit strange ones as I've never had any problems with tooltips on Wikipedia), but as said before, WP:ACCESS can provide a good opinion on if this template could be used at all or not. Prime Blue (talk) 02:38, 28 January 2011 (UTC)
• OK, likely including the romaji should be either compulsory or strongly advised, unless the romaji is spelt exactly the same as the "English" form. Anthony Appleyard (talk) 06:52, 28 January 2011 (UTC)
  • What about a situation like "Super Mario USA → スーパーマリオUSA → Sūpā Mario Yū Esu Ē"? Technically the spelling is not "Exactly the same," but WP:VG views it as "not significantly different" because it is only a phonetic transcription of kana rendering English. In that case under your suggestion would the use of romaji be optional, strongly advised, or compulsory? -Thibbs (talk) 11:27, 28 January 2011 (UTC)
• (Not quite the same. 'S' is "es" in English and "esu" in Japanese.) OK, strongly advised or compulsory always. Anthony Appleyard (talk) 11:53, 28 January 2011 (UTC)
  • Sometimes the same English word can be rendered in Japanese in more than one way. For example, "dollar" got into Japanese as "doru" from British English speakers and "dara" from USA English speakers. And "super" heard from a strongly rhotic English-speaker may on occasion be rendered in kana as "sūparu" or similar. Anthony Appleyard (talk) 11:58, 28 January 2011 (UTC)
    • Oh we should definitely include kana always. I think everyone here agrees with that. But the topic of this discussion is romaji. We've been talking about the fact that the transliteration of kana into romaji follows a rigid standard where there are no idiosyncrasies or irregularities. That being that case, I think it's fair to say that romaji is redundant as long as the reader is presented with the kana. It's like using color as secondary means of conveying information. It can be helpful sometimes - like when it's important to the context of the article that the kana rendering reflects the intonation of a strongly rhotic English-speaker - but when it is not important to the context of the article then I think it is fair to give editors the option of putting it into a tooltip (my first preference) or to simply delete it (your original proposal and my second choice). -Thibbs (talk) 12:15, 28 January 2011 (UTC)
• "romaji is redundant as long as the reader is presented with the kana": But most people outside Japan cannot read kana. Anthony Appleyard (talk) 15:12, 28 January 2011 (UTC)
  • And that's why the nihongotip alternative would only be available when the romaji was substantially identical to the English. Another solution that was brought up in a past discussion was to include a note linking the reader to the romaji article similar to the "^?" mark currently adorning words in the nihongo template. Do you think that would work better? -Thibbs (talk) 15:19, 28 January 2011 (UTC)
• OK so I'm sorry to be so thick-headed here, but I have a question about your most recent proposal, "strongly advised or compulsory always."
  Doesn't "compulsory" preclude "strongly advised"? Do you mean that the rule should be 1) that the use of romaji is always strongly advised, or 2) that the use of romaji is always compulsory? -Thibbs (talk) 12:17, 28 January 2011 (UTC)
• A small link (say from a '?') to the romaji article would be useful.

  This query arose in connection with videogames because videogame name Japanese is unusual in having a big content of English words written in katakana. In other fields where the names are all or mostly of Japanese origin, presence of a romaji transcription is currently variable: e.g. there is a romaji transcription in page Kawasaki, Kanagawa (about the city) but not in page Hirohito. As regards "compulsory" versus "strongly advised", the WP:IAR rule exists for when circumstances and cases arise which were not foreseen by those who made the rule. I put in the "strongly advised" alternative because I was unwilling to appear dictatorial by saying "compulsory". Anthony Appleyard (talk) 16:23, 28 January 2011 (UTC)

OK I agree that IAR may apply to this case although my general impression of it is that realistically it tends to carry very little weight. It sounds to me then, that by using "strongly advised" under an IAR rationale to get around the mandatory inclusion of romaji you've described the original policy endorsed by WP:VG/GL in different words. As I've said earlier, I would agree to this but only if the Nihongotip Compromise fails. So far there seem to be no objections to it at WP:WPACCESS. -Thibbs (talk) 16:54, 28 January 2011 (UTC)

The reason for the difference between the usage of {{nihongo}} in Kawasaki, Kanagawa and Hirohito is because "Hirohito (裕仁, Hirohito)" would be redundant (the romanization is exactly the same as the English), whereas "Kawasaki (川崎市, Kawasaki-shi)" has the addition of the 市 in Japanese, so romanization needs to be included. This is different from romanization of katakana which represents English because the Japanese pronunciation of that katakana is almost always quite different than the English pronunciation. ···日本穣^? · 投稿 · Talk to Nihonjoe · Join WikiProject Japan! 17:14, 28 January 2011 (UTC)

• My experiences with {{nihongotip}} are stated just below the "Breakpoint" section header. I have had similar bad behaviour with hover help mini-windows on controls on various Windows applications. Anthony Appleyard (talk) 23:26, 28 January 2011 (UTC)
  • First of all, your experience seems to be rare. 5 of 6 editors above have endorsed the tooltip compromise and nobody from WP:WPACCESS has taken issue with it yet. But secondly and more importantly, I still think it is consistent with WP:ACCESS. It doesn't matter if you can't see what the romaji transcription is provided that the romaji transcription is substantially identical to the English which I presume you can see. This compromise is not supposed to replace the {{nihongo}} template but only to provide an alternative to users who agree by consensus that the romaji is so similar to the English that it is not important to the context of the article. -Thibbs (talk) 02:25, 29 January 2011 (UTC)
• But many readers do not know that dotted underline means "here is a hover window". Anthony Appleyard (talk) 06:45, 29 January 2011 (UTC)
  • If tooltips are used, the situation will be explained at the new help page reached from the "?" after the Japanese text in the template. Users who do not know that they are confronted with a tooltip but wonder what the characters are will naturally hover the mouse over the text as it appears differently, or click on the question mark to see what it means. But it is useless to argue whether tooltips are viable or not if no one actually affected by them has given their opinion on the matter yet. We are going in circles. Prime Blue (talk) 09:54, 29 January 2011 (UTC)

Breakpoint 2

• This discussion seems to be leading to:
  • If the usual English form is "aaaaa", and the Japanese form is "bbbbb", and the literal romaji form is "ccccc", use "{{nihongo|aaaaa|bbbbb|ccccc}}" "aaaaa (bbbbb, ccccc)" usually, but "{{nihongotip|aaaaa|bbbbb|ccccc}}" "Template:Nihongotip" if "aaaaa" and "ccccc" are identical.

  Can that be assumed? Anthony Appleyard (talk) 11:26, 3 February 2011 (UTC)

• From what the folks at WP:WPACCESS have said, the tooltips option is non-viable since screen-readers would read "Template:Nihongotip" as "aaaaa (bbbbb) abbreviation: ccccc". And this would confuse users because the fact of the matter is that the romaji (ccccc) is not an abbreviation.
  On the other hand, the WP:WPACCESS people said that the kana ("bbbbb") would be pronounced correctly by screenreader programs. Since the algorithm that would be used to read and pronounce the kana can be no worse than a romaji approximation, there's still room for WP:VG's argument in terms of its adherence to WP:ACCESS. In fact, there's even been some suggestion by WP:WPACCESS that the romaji would be confusingly duplicative for screenreader users in situations where the English was highly similar.
  So... I think we're really back to square one again. I'd love to hear some compromises suggested by the critics of the Nihongotip idea. -Thibbs (talk) 13:07, 3 February 2011 (UTC)
• {{Nihongo tip}} is not an option anymore due to the concerns of WikiProject Accessibility. Also, if the first and third fields ("a" and "c") are absolutely identical, the third field is not used at all (on any project) as it is truly redundant then. What I noticed in this discussion is that the WPVG participants seem to be a lot less insistent on the exclusion of the third field, so I'll just suggest to drop the sentence "Phonetic transcriptions are, as a rule, not considered to be significantly different and thus do not warrant the inclusion of Japanese titles." from WP:VG/JP. A project style guideline that is not used consistently and that editors can violate without sanctions is irrelevant. Prime Blue (talk) 13:35, 3 February 2011 (UTC)
  • Just to play devil's advocate a bit here, I think the rationale behind the less than absolute rule used by WPVG is that in some cases the Hepburn romanization of a game's title may simply not be relevant to the topic. Clearly MOSJA believes romaji to be relevant to every article about any product of Japanese origin, however I think there is a fair argument that WPVG would be the authorities on what is and isn't relevant to video game articles. So the question would then become one of whether the article was more a "Japanese product" article or more a "video game" article.
    If WPVG were to decide as a WikiProject to adopt MOSJA's stance that the Hepburn romanization of the titles of all video-game-related products of Japanese origin are always relevant then that's one thing, but I don't know if there could be any binding decision on the topic without project-wide consensus from WPVG. I'm unfamiliar with WPVG's methods, but is there some way that such a suggestion could be brought before WPVG as a whole? -Thibbs (talk) 14:34, 3 February 2011 (UTC)
    • It's not that they think it's not relevant. They believe that if you have "Super Metroid", having "Sūpā Metoroido" is redundant.—Ryūlóng (竜龙) 20:08, 3 February 2011 (UTC)
      • Well my argument (from above) was that unlike translations or phonetic renderings, the romanization of kana produces redundant information inasmuch as it is the exact same information in two different forms with direct 1:1 parity similar to "2+2" and "4". Simple deductive evidence is available in the fact that romaji never needs to be sourced at Wikipedia. Why is this? Because it is mechanically-derived and thus redundant information. It is not new information that an editor is introducing (which would have to be sourced), but redundant information (which doesn't require sourcing). But I don't think that's what WPVG is saying about romaji that is extremely similar to English anyway...
        No, I'd say they consider it to lack "significant difference." What makes something significant if not its relevance to the article? -Thibbs (talk) 23:03, 3 February 2011 (UTC)
    • (edit conflict)Bring it up on the main talk page, but I'll give fair warning that if you propose to simply submit to this guideline, it doesn't stand a good chance of flying because the comprise failed; indeed if you present the findings by WP:ACCESS that it can correctly pronounce them similar to what romaji is, i think there will be even less support for supporting this guideline which doesn't allow for such exceptions.陣内_Jinnai 20:14, 3 February 2011 (UTC)
• But I and many do not have a screenreader and never learned to read kana. (I can decode kana because I wrote a Windows application that can translate kana (and Armenian and Indian) to the Roman alphabet.) We need the romaji to see how much it differs from the usual English spelling :: I thought that that was agreed on. Anthony Appleyard (talk) 22:31, 3 February 2011 (UTC)

• • Only if there is a signifigant difference. If its the difference of lack of a macron or 1-2 sylables, it doesn't. I also would say that the help page could easily include a katakana/romaji chart based on the one from katakana and amount to the same thing.
  • No other place in the English Wikipedia is such a thing required without exception for other languages that use English loanwords and as such I'd argue against its adoption as required at WT:VG. I didn't really agree with the idea that its required, but rather was holding my piece to see how the ACESS issue played out.
  • The bottom line is you don't need to see how Final Fantasy's katakan, ファイナルファンタジー, is pronounced because its clear that its the Japanese writing system trying to get as close to the English one as possible similar to how we translate words from Russian, Greek, Chinese and yes, Japanese. If you think its really nessasary then soundlink like that in anime is far better.陣内_Jinnai 22:51, 3 February 2011 (UTC)
    • Although I was more in favor of a compromise between WPVG and MOSJA's guidelines, this now seems to me to be impossible. If we are forced to go with one or the other guideline then I agree with Jinnai that MOSJA's absolutist rule seems less useful than WPVG's rule which allows for some exceptions. The payoff for following MOSJA's guideline is certainty/consistency and the payoff for WPVG's rule is pragmatism/adaptability. As Anthony pointed out above, this whole topic area seems prone to evocations of IAR if a "without exceptions" rule is in place, but if a limited exception is codified in both guidelines then I think we'd sacrifice a little bit of consistency for a more pragmatic set of guidelines and a lot fewer arguments. This trade-off is one I'd gladly take. -Thibbs (talk) 23:17, 3 February 2011 (UTC)
      • "Absolutist" is a bit of an extreme adjective. I think that the romaji should be included on at least one page if it's a large franchise (e.g. Final Fantasy, SoulCalibur) without requiring romaji for sequels unless there's a new word introduced (e.g. Final Fantasy Versus XIII, Final Fantasy Crystal Chronicles) or if the sequels aren't numbered (e.g. Kingdom Hearts 358/2 Days [although this case could be dealt with a footnote explaining the intended reading instead of Surī Faibu Eito Deizu Ōbā Tsū).—Ryūlóng (竜龙) 23:55, 3 February 2011 (UTC)
        • As I mentioned, it would be better if the concern about voicing would be to create a hepburn equivalent of {{IPA-ja}} and have the parameter able to use voice clips. If someone speaks how the katakana is suppose to sound then there is no need for romaji for those words. What's more, there's no confusion about r-/l- sounds and the like.陣内_Jinnai 00:09, 4 February 2011 (UTC)
          • I don't see how getting a Japanese speaker to read the names of all of these video games is going to help. And Hepburn romanization serves the same purpose as pinyin on Chinese subjects.—Ryūlóng (竜龙) 00:22, 4 February 2011 (UTC)
            • Indeed, but pinyin isn't used for titles that are taken solely from English words.陣内_Jinnai 00:56, 4 February 2011 (UTC)
              • I am fairly certain that we provide the pinyin for any Chinese, unless it's from a dialect that isn't Mandarin.—Ryūlóng (竜龙) 01:04, 4 February 2011 (UTC)
                • I don't and nor do I see the like for Greek or Russian for words that come in English (indeed for some that orgininate in that language but have common translations/pronunciations in English we don't even include them such as Hercules.陣内_Jinnai 01:10, 4 February 2011 (UTC)
                  • Hercules is the Roman myth. Heracles is full of said information.—Ryūlóng (竜龙) 01:12, 4 February 2011 (UTC)
                    • Fair enough for that example, but pinyin is not universally used. However, no one here was really arguing to eliminate them from native Japanese words, the few loanwords that may use kanji, those with mixed kana/kanji in the title or those that do not sound similar to the original english word. As such, using Chinese or any other east-asian language for comaprison is innaporiate as not many video games have been made nor titles in English; indeed very few asian languages use their language titles to create any title with English characters the way Japanese does.陣内_Jinnai 02:14, 4 February 2011 (UTC)

Chinese topics have infoboxes for that. And you are definitely right that Japanese is the only language that gets this treatment. However, the names are still inherently Japanese in origin. But as I said, the exclusion should only occur if the game is part of a series of games and the name doesn't really change other than numbers.—Ryūlóng (竜龙) 03:15, 4 February 2011 (UTC)

• I'm still not convinced that WikiProjects like WPVG shouldn't be allowed to make the determination themselves for what material is relevant to the articles within their limited purview. If WPVG determines that some romaji in some cases (such as "Nintendō Dī Esu" in the "Nintendo DS" article) isn't relevant to some articles, why do we think MOSJA would be a better judge? What's the argument against modifying MOSJA's doctrinaire "zero exceptions" guideline to allow for a more flexible and pragmatic use throughout Wikipedia? Is it really just based on consistency? And isn't consistency mainly an aesthetic concern? -Thibbs (talk) 04:53, 4 February 2011 (UTC)

Breakpoint 3

• The discussion (which has run for over 6 months) seems to be getting wordy again. Please, again state your suggestions about the rule:
  • If the usual English form is "aaaaa", and the Japanese form is "bbbbb", and the literal romaji form is "ccccc", use "{{nihongo|aaaaa|bbbbb|ccccc}}" "aaaaa (bbbbb, ccccc)" usually, but "{{nihongotip|aaaaa|bbbbb|ccccc}}" "Template:Nihongotip" if "aaaaa" and "ccccc" are identical.
  • Or what?

    Anthony Appleyard (talk) 23:13, 5 February 2011 (UTC)

    • {{Nihongotip}} isn't a viable option anymore according to the WP:ACCESS people.—Ryūlóng (竜龙) 03:21, 6 February 2011 (UTC)
• As mentioned, I think a vocal option would be a viable option, but I'd still not want that or the romanji as getting a Japanese speaker can be difficult. I'd put it as a 3rd option.陣内_Jinnai 05:37, 6 February 2011 (UTC)
• In that case, the rule better be "{{nihongo|aaaaa|bbbbb|ccccc}}" "aaaaa (bbbbb, ccccc) everywhere", or perhaps omit 'ccccc' if it is all native Japanese words and identical with 'aaaaa'. And, please keep discussion concise and avoid repeating. Anthony Appleyard (talk) 10:40, 6 February 2011 (UTC)
  • Not perhaps, the third field is omitted always when it is identical to the first field. This was never questioned and never the intended usage of {{Nihongo tip}}. Prime Blue (talk) 12:11, 6 February 2011 (UTC)
  • The issue has always been what to do with romaji that is almost identical to English. The question is whether to A) leave the already-existing official alternative in the WPVG guideline or to B) mandate that WPVG change its guidelines to match MOSJA's or C) seek a compromise. -Thibbs (talk) 12:43, 6 February 2011 (UTC)
    • And the two positions are thus: WPVG wants to be able to omit some romaji derived from English-word-based-kana. WPVG's concerns are content-relevance and clutter/confusion. MOSJA wants to require the inclusion of this romaji. MOSJA's concerns are inter-article consistency (and perhaps rigidity of rule?) -Thibbs (talk) 14:33, 6 February 2011 (UTC)
      • Inter-article consistency is the only main concern here. Content-relevance is also an issue as a Japanese game will necessarily have a Japanese title. The romaji is completely relevant in that case. ···日本穣^? · 投稿 · Talk to Nihonjoe · Join WikiProject Japan! 21:38, 6 February 2011 (UTC)
        • But isn't MOSJA a style guideline instead of a content guideline? -Thibbs (talk) 22:01, 6 February 2011 (UTC)
          • Indeed all MOS are suppose to be style guidelines.陣内_Jinnai 22:15, 6 February 2011 (UTC)
• Try this?:- "{{nihongo|aaaaa|bbbbb|ccccc}}" "aaaaa (bbbbb, ccccc) everywhere", but omit 'ccccc' if it is all native Japanese words (including old loanwords from Chinese) and identical with 'aaaaa' including presence/absence of macrons." Anthony Appleyard (talk) 09:30, 7 February 2011 (UTC)
  • That would give people the option of writing "Nintendo (任天堂)" instead of "Nintendo (任天堂, Nintendō)." But I think that what we've been talking about is not English loanwords from the original Japanese but rather Japanese loanwords from the original English (i.e. English gairaigo). -Thibbs (talk) 13:58, 7 February 2011 (UTC)
• In "Nintendo (任天堂, Nintendō)" the "aaaaa" and the "ccccc" are not identical, because there is a macron different. Anthony Appleyard (talk) 23:28, 7 February 2011 (UTC)
  • OK then I don't understand your rule. Could you restate it in a more concise form? -Thibbs (talk) 23:37, 7 February 2011 (UTC)
• "{{nihongo|aaaaa|bbbbb|ccccc}}" (which displays as "aaaaa (bbbbb, ccccc)") everywhere, but omit 'ccccc' if it is all Japanese words and identical with 'aaaaa' including diacritics. Anthony Appleyard (talk) 10:11, 8 February 2011 (UTC)
  • So this is an expansion of MOSJA's current rule to make it even more doctrinaire? Why? -Thibbs (talk) 12:47, 8 February 2011 (UTC)
• Wikipedia:Manual of Style (Japan-related articles)#Japanese terms seems to say "use ccccc always", but I have seen examples in articles with the ccccc missing if it is identical with the aaaaa. What changes do you want making to the rule as I suggested it here? Anthony Appleyard (talk) 15:02, 8 February 2011 (UTC)
  • :- Use "{{nihongo|aaaaa|bbbbb|ccccc}}" (which displays as "aaaaa (bbbbb, ccccc)") everywhere, but omit 'ccccc' if it is all Japanese words and identical with 'aaaaa' including diacritics.
    But WPVG conflicts with this and I think that they should come to this consensus themselves. -Thibbs (talk) 15:51, 8 February 2011 (UTC)
• 'WPVG' is correctly WP:VG, i.e. Wikipedia:WikiProject Video games. What part of it or its daughter pages specifically mentions English words written in kana in names of Japanese videogames? I was starting to think that we had got this matter nearly settled at last after it has rhubarbed on for over 6 months. Anthony Appleyard (talk) 21:27, 8 February 2011 (UTC)
  • WP:VG/GL#Non-English games.—Ryūlóng (竜龙) 21:30, 8 February 2011 (UTC)
• This discussion started as only about what to do with katakana transcriptions of English words. The difference between WP:VG/GL#Non-English games and here is where WP:VG/GL#Non-English games covers situations not covered here, including where the English form and the Japanese form are not translations or transcriptions of each other but have different meanings, e.g. the Japanese videogame name 悪魔城ドラキュラ (Akumajō Dracula (Akumajō Dorakyura)), which means "evil demon castle Dracula", is rendered into English as Castlevania. In those cases, WP:VG/GL#Non-English games adds a 4th argument: "{{nihongo|aaaaa|bbbbb|ccccc|ddddd}}" = "aaaaa (bbbbb, ccccc, ddddd)", where 'ddddd' is for the literal English translation (rather than transcription) of the Japanese. Anthony Appleyard (talk) 23:36, 8 February 2011 (UTC)
  • This is the contentious line: "Phonetic transcriptions are, as a rule, not considered to be significantly different and thus do not warrant the inclusion of Japanese titles." -Thibbs (talk) 23:46, 8 February 2011 (UTC)
• That rule was likely intended to apply to small differences in transcription of native Japanese words, e.g. whether to transcribe ジ (hiragana) / じ (katakana) as "zi" or as "ji", or っち / ッチ as "tchi" or as "cchi", not to big differences resulting from trying to spell English words in kana. It seems that also, we will have to decide which of these two sets of rules overrules the other. Anthony Appleyard (talk) 06:25, 9 February 2011 (UTC)
  • No, it was intended such that "Fainaru Fantajii" or "Souru Kyaribaa" or "Suupaa Mario" or "Kingudamu Haatsu" are not to be put on video game pages.—Ryūlóng (竜龙) 06:45, 9 February 2011 (UTC)
    • I think the original intention was that. I don't know how its applied today with regard to Japanese words. There is an argument to be made that if the translation uses a form of Hepburn other than modified, we shouldn't try to "correct" it with a second pronucitation that would be just one character different, such as Nintendo.
    • I will note that there was also been an argument made when I PRed Hayao Miyazaki that we should be using IPA tag for pronunciation.陣内_Jinnai 07:18, 9 February 2011 (UTC)
  • Which rule overrules the other? They both overrule each other. If you are asking which should be the controlling rule, I think we must look to the scope and purpose of the rules. MOSJA is a style guideline aimed at words of Japanese origin. WPVG is a style and content guideline aimed at video games. So is "Fainaru Fantajii" more like a game or is it more like a Japanese word? Obviously it's both, but which one is it more like? -Thibbs (talk) 12:05, 9 February 2011 (UTC)
    • I also think that it's rather unhelpful to regard the problem as one of "whose rule can be imposed on the other." Rule by fiat has never been as effective as rule by consensus. -Thibbs (talk) 12:17, 9 February 2011 (UTC)
      • I think Ryūlóng's argument is WikiProject guidelines are by default subordinate to the guidelines like MOS similar to how this guideline is is subordinate to the main MOS guieline which is subordinate to policy.陣内_Jinnai 17:22, 9 February 2011 (UTC)

I don't follow the analogy. WikiProject VideoGames isn't a daughter project of the MOS. Anyway my understanding was that WikiProjects were designed to provide focal points of activity where local consensus could be achieved with regard to specific topics. WPVG's requirement to include a section on "Gameplay," for instance, would only apply to video game articles but it would be required for all video game articles as by the broad consensus of WPVG. I think that's one of the WikiProjects' greatest uses here at Wikipedia. If WPVG has come to a consensus that would indicate that romaji is not always helpful to an understanding of the topic then I respect their decision. I happen to disagree with their stance on this point, but rather than overruling a fairly sizeable local consensus with a style guideline on a topic that is arguably only tangentially related, I think we should work together to craft a carefully-worded and convincing argument that we could present to WPVG to make the case that inter-article consistency is more helpful overall than the avoidance of a little clutter, confusion, and redundancy. The idea would be to form a new consensus at WPVG. Perhaps it may even be a good idea to try to convince WPVG that its guidelines should be subsumed under MOS just as with film, music, anime, etc. Then we would have a legitimate argument based on guideline hierarchy. -Thibbs (talk) 19:22, 9 February 2011 (UTC)

The issue is that WikiProject level style guides (e.g. WP:VG/GL) should be subordinate to general style guides (e.g. WP:MOS-JA) which are then subordinate to sitewide policy (WP:AT).—Ryūlóng (竜龙) 19:52, 9 February 2011 (UTC)

Are you saying that you would prefer to force WPVG to adhere to a "superior doctrine" rather than to change its established consensus by persuading it that inter-article consistency (the basis underlying the "superior doctrine") is valuable? -Thibbs (talk) 20:10, 9 February 2011 (UTC)

I am saying that specific topic area guidelines should not necessarily go against the general practice of the rest of the project and more general guidelines. So maybe I am saying that WP:VG should not make up its own rules concerning the treatment of other languages when the entirety of the project has their own little rules that WP:VG may or may not be ignoring.—Ryūlóng (竜龙) 20:35, 9 February 2011 (UTC)

Sadly it's already too late. This isn't a hypothetical new rule, but an already existing rule that WP:VG made up quite some time ago. From the current point where we find ourselves, and in deference to their presumptive expertise when it comes to guidelines concerning video games, I would be in favor of trying to persuade WP:VG to change their rules voluntarily through consensus. -Thibbs (talk) 21:31, 9 February 2011 (UTC)

And why shouldn't the push for the consensus be with the more general guidelines and practices of non-video game articles?—Ryūlóng (竜龙) 21:45, 9 February 2011 (UTC)

If you re-read my last few comments I think you'll find that we are arguing for the same thing. -Thibbs (talk) 21:56, 9 February 2011 (UTC)

The question is though really how one bases consensus. There are more articles that fall under the scope of WP:VG/GL than this article and based on the activity level at WT:VG there it is certainly one of the larger wikiprojects in terms of activity. In addition, the guideline could be seen as similar to WP:MOS-AM, but just in a different Wikipedia space.陣内_Jinnai 21:49, 9 February 2011 (UTC)

From MOSJA's perspective, though, considering how many other broad topics (anime, music, film, etc) have made their guidelines part of the MOS, perhaps WPVG should move toward this goal as well. -Thibbs (talk) 21:56, 9 February 2011 (UTC)

But the treatment of Japanese text should still defer to this guideline, rather than if there's going to be a WP:MOS-VG that has its own guideline for Japanese text.—Ryūlóng (竜龙) 23:44, 9 February 2011 (UTC)

But this guideline should also give room for reasonable exceptions; even the policy does so.陣内_Jinnai 23:46, 9 February 2011 (UTC)

What purpose is there to have an exception?—Ryūlóng (竜龙) 23:49, 9 February 2011 (UTC)

Look, these issues would be bound come up if WPVG's guideline were to be brought under the MOS. I think there are strong arguments in favor of making the style portions of the video game guideline agree with the Japanese language's style guideline. I'm not really certain that the content of video game articles should be dictated by a parallel language style guideline, but the arguments underlying the other MOS guidelines should obviously be taken into account by WPVG should they decide to go this route. My hope is that by consensus they would agree to adopt MOSJA's stance. -Thibbs (talk) 00:13, 10 February 2011 (UTC)

I agree they would be brought up, but I don't they would be because of this guidelines no-exception rule. It's suppose to be a guideline, not a dictatorial document.陣内_Jinnai 00:27, 10 February 2011 (UTC)

But the articles don't gain anything by omitting a Japanese text reading.—Ryūlóng (竜龙) 01:17, 10 February 2011 (UTC)

As far as I understand it, the argument is that it can sometimes cut down on clutter, confusion, and redundancy. There may be other reasons such as the degree of relevance. -Thibbs (talk) 01:22, 10 February 2011 (UTC)

Breakpoint 4

• This discussion (which started as only about what to do with katakana transcriptions of English words) has yet again got wordy and digressiony. Please help to get this matter resolved: it has dragged on since 2 July 2010, i,e. for over 7 months. OK, here I summarize:-
  • Suggested rule: :- Use "{{nihongo|aaaaa|bbbbb|ccccc}}" (which displays as "aaaaa (bbbbb, ccccc)") everywhere, but omit 'ccccc' if it is identical with 'aaaaa' including any diacritics.
  • A new suggestion: insert in Wikipedia talk:Manual of Style (Japan-related articles)#Names and product titles a new subsection "===Names of videogames=== For names of Japanese videogames see Wikipedia:WikiProject Video games/Article guidelines#Japanese titles.".
  • OK, it is established that the text "Phonetic transcriptions are, as a rule, not considered to be significantly different and thus do not warrant the inclusion of Japanese titles." in page Wikipedia:WikiProject Video games#Non-English games contradicts the current suggestion. Please take that as known and not needing to be again re-stated. OK, if this query about English in katakana in Wikipedia:Manual of Style (Japan-related articles) is decided here, Wikipedia:WikiProject Video games#Non-English games may have to be edited to match, unless the new suggestion here is followed. I have put in Wikipedia talk:WikiProject Video games#How to display videogame names which have an English version and a Japanese version? a link to this discussion.
  • Again, please keep discussion concise and to the point and avoid repeating points. Anthony Appleyard (talk) 06:47, 10 February 2011 (UTC)
• First of all, your current suggestion does not work as the ā, ē, ī, ō, and ū are necessary for proper romanization of Japanese, even though ā, ē, and ī only appear in loan words. And this case will never show up in video games because the use of foreign words transcribed into katakana will never be a 1:1 (ignoring diacritics) conversion due to the nature of the syllabary.
  • And I do not think that video games require their own section on a manual of style. The only reason music dominates the "Names of companies, products, and organizations" is because we don't really cover much else that would have such titles.—Ryūlóng (竜龙) 08:31, 10 February 2011 (UTC)
• OK, "..., but omit 'ccccc' if it is identical with 'aaaaa', and if any diacritics in 'aaaaa' and 'ccccc' all match", or similar.
  • The section Wikipedia:WikiProject Video games/Article guidelines#Japanese titles has been there for some time, and I was merely suggesting putting a link to it in Wikipedia:Manual of Style (Japan-related articles).
  • (Information: 'nihongo' can have 4th and 5th arguments: "{{nihongo|aaaaa|bbbbb|ccccc|ddddd|eeeee}}" = "aaaaa (bbbbb, ccccc, ddddd) eeeee", where 'ddddd' seems to be for an English translation (not transcription) of 'bbbbb', if that translation is markedly different from 'aaaaa'.)
  • Anthony Appleyard (talk) 09:34, 10 February 2011 (UTC)
• I like Anthony's suggestion to try to have a section on video game titles in the MOS, but I'd put it here. Just as WP:ALBUMCAPS delegated decisions on capitalization, we should ask WP:VG if it would also be willing to delegate this issue to MOSJA. The handling of music titles is also completely inconsistent with MOSJA - the article for every single musical example listed at my earlier link omits romaji and usually even omits kana. -Thibbs (talk) 13:03, 10 February 2011 (UTC)

  "Titles of media" regards article titles. Not the romanization of content within the articles.—Ryūlóng (竜龙) 18:45, 10 February 2011 (UTC)

• Having both sections Wikipedia:Manual of Style (Japan-related articles)#Videogames and Wikipedia:WikiProject Video games/Article guidelines#Japanese titles would cause content forking, unless one of them is merely a link pointing to the other. (I take it that "WP:VG", spoken of as a living agent, means "the people who have edited the page Wikipedia talk:WikiProject Video games/Article guidelines" or similar.) As the matter about Japanese videogame names is already at Wikipedia:WikiProject Video games/Article guidelines#Non-English games and likely has been there for a long time, it may as well stay there, to avoid confusion, and this page would have a link pointing to it. The {{nihongo}} template call examples in section Wikipedia:WikiProject Video games/Article guidelines#Non-English games all seem to use the romaji parameter already, so there seems to be less for us to argue about than may be thought, as this topic has already run for over 7 months. All that is needed seems to be for someone to put in here a pointer to Wikipedia:WikiProject Video games/Article guidelines#Japanese titles and to put in Wikipedia:WikiProject Video games/Article guidelines#Japanese titles an extra short paragraph:-
  • In 'nihongo' template calls for videogame names, always insert the third parameter (the romaji parameter); except that it can optionally be omitted if it is identical with the first parameter including all diacritics matching.
    • Anthony Appleyard (talk) 13:56, 10 February 2011 (UTC)
      • That won't work as it doesn't allow for WP:COMMON SENSE to be used when things are similar and doubt it would fly if brought up at WT:VG.陣内_Jinnai 16:00, 10 February 2011 (UTC)

Wow, it seems absolutely nothing has changed here in 7 months except that the new mediator is either missing the point or is very biased. Either way, this would only "settle" the issue to the extent that WP:VG editors accept being completely blown off. As I see it, these are the cases:

1. The Japanese title is a transliteration of the English, e.g. "Final Fantasy VII" → "Fainaru Fantajī Sebun".
2. The English title is a direct transliteration of the Japanese, possibly with minor differences in diacritics or specific letter choice, e.g. "Katamari Damashii" → "Katamari Damacy".
3. The Japanese title is a transliteration of some other English text, e.g. "Super Mario Bros. 2" → "Super Mario USA" → "Sūpā Mario Yū Esu Ē".
4. The English and Japanese titles are translations of each other, e.g. "The Legend of Zelda: Phantom Hourglass" ↔ "Zeruda no Densetsu Mugen no Sunadokei". (Although "Zelda" vs "Zeruda" could make this example case 7)
5. The English and Japanese titles are translations of something similar, e.g. "The Legend of Zelda: The Minish Cap" ↔ "Zeruda no Densetsu: Fushigi no Bōshi" ("The Mysterious Hat"). (Although "Zelda" vs "Zeruda" could make this example case 7)
6. The English and Japanese titles are something different, e.g. "The Legend of Zelda: A Link to the Past" ↔ "Zeruda no Densetsu: Kamigami no Toraifōsu" ("The Triforce of the Gods"). (Although "Zelda" vs "Zeruda" and "Triforce" vs "Toraifōsu" could make this example case 7)
   • Note that that particular article is even odder, as it includes both "Zelda no Densetsu: Kamigami no Triforce" and "Zeruda no Densetsu: Kamigami no Toraifōsu".
7. A mixture of the above, e.g. "The Legend of Zelda: Twilight Princess" → "Zeruda no Densetsu Towairaito Purinsesu" (which could be "Zelda no Densetsu Twilight Princess").

The proposal here is to allow omitting the romaji in only a subset of case 2 where there are no differences, which is probably an infinitesimal fraction of games that have any common English name. The current WP:VG practice (AFAIK) is that the romaji is optional in cases 1 and 2, and for case 3 the plain English text may used instead. I don't think anyone is really objecting to romaji in cases 4, 5, or 6. Case 7 may go word by word as the previous cases. What exactly does it contribute to the article to tell English speakers that "Final Fantasy" spoken by a Japanese speaker comes out something like "Fainaru Fantajī", which the English speaker will probably read with the wrong vowels anyway? Anomie⚔ 16:58, 10 February 2011 (UTC)

The argument is that a romanization or IPA phoneticism occurs on nearly every article the original spelling of whose topic uses non-roman letters. So it's essentially based solely on broadening inter-article consistency. I think that this argument isn't terrible and I think WPVG should seriously consider adopting it. I see no sense in rejecting it out of hand, but if the consensus opposes it then so be it. Perhaps MOSJA could then consider codifying an exception to their "mandatory romaji" rule. -Thibbs (talk) 17:14, 10 February 2011 (UTC)

My main reason for excluding it is that it will confuse the layman rather than educate. If the nihongo template can display something like "aaaaa (Japanese: bbbbb, Romanization: ccccc)", then I'm willing to accept all romanization. The added terms should link to pages in the "Help" namespace that provide some guidelines for proper pronunciation. I believe the pages already exist, but I think it would be good idea to include IPA pronunciations on them for good measure.(Guyinblack25 ^talk 17:39, 10 February 2011 (UTC))

Joren has already started work on an improved help page that would be linked from the "?" superscript like so: "^?" (Here I've linked Joren's help page). This help page fully covers how to pronounce the kana and it includes links for how to give it a Hepburn romanization. Maybe this problem is as simple as improving the "?" help menu. -Thibbs (talk) 18:20, 10 February 2011 (UTC)

My issue with the link is that it too discreet. I understand that was probably the intent, but I think that it negatively impacts the role it's intend to fill. (Guyinblack25 ^talk 18:31, 10 February 2011 (UTC))

(edit conflict)I'm glad that I wasn't the only one reading this and thinking that the mediator seemed biased. I've mostly been disengaging from this discussion because I probably forgot what the original argument was and I think that a visit to Japan may have changed my appreciation for Romaji since I don't know katakana. I'm reminded of an author in an English newsletter that I read who lamented the realization that you cannot just pronounce English words with a supposedly Japanese accept and expect it to be understood. The pronunciation of any translated text would be important when speaking to someone in Japanese about the title and the katakana is more useless for readers without it.

If a correctly recall, my main concern has been that the use of {{nihongo}} in prose after the lede tends to disturb the flow of reading, especially for long titles. While we are focussed on Japanese text, I venture to say that I would see this problem for any language that needs romanization. While I had felt the tooltip option was an appropriate way to convey the information without an extended parenthetical that many would ignore, I understand that speaking "abbreviation" before it would not be ideal. Unfortunately, I would also think that a screen reader would pronounce the romaji incorrectly (not knowing that it has specific pronunciation) or that the romaji would be irrelevant if the screen reader read kana. I like hearing the discussion about IPA because as this is an English encyclopedia that expects its reader to learn IPA, I don't like the idea of forcing readers to learn a new system; to this end, I agree that at least the IPA should be on the romaji explanation page (and the link to the romaji page should also be specified near the top of Help:Installing Japanese character sets so that readers could easily determine pronunciations). —Ost (talk) 18:41, 10 February 2011 (UTC)

In Anthony's defense, I think it's not so much that he is trying to champion MOSJA's argument as it is that he just doesn't understand the central issues yet. I think we should assume he's neutral. But also I wanted to mention one thing that we neglected to look at in the tooltip discussions. Although screen-readers may insert the word "Abbreviation:", I'm not sure how bad that really is. Apparently tooltips are fine and dandy when it comes to the {{R-phrase}} template. This template uses hovertext but the hovering messages are not actually abbreviations for the underlying R-phrases. R1 means "Explosive when dry" but it's not an abbreviation for that expression; it's an abbreviation for "Risk #1". If R-Phrase hovertext is allowed under an argument by analogy to abbreviations then surely the same rationale would hold for kana conversions to romaji. While romaji isn't exactly an abbreviated version, it is an alternate version just as R1 is an alternative way of saying "Explosive when dry." -Thibbs (talk) 19:00, 10 February 2011 (UTC)

スーパーマリオUSA

And also, if the "Abbreviation:" problem is just a formatting issue, then what about an alternative formatting like the following:

This provides hovertext, but is it read as "Abbreviation:" by screen-readers? (Obviously it would have to be modified for inline use by someone who knew what they were doing.) -Thibbs (talk) 19:11, 10 February 2011 (UTC)

Thanks for clarification of Anthony's position and why it could be confused; I'd been keeping quiet to agf about it. I also like the clever approach to getting hovertext; it shows that there is another way to get the tooltip, although I personally don't know what part of table code produces them. —Ost (talk) 20:52, 10 February 2011 (UTC)

Well, the main issue with adding "Japanese:" and "Romanization:" is that it would bring ire to the people who feel that the {{nihongo}} template causes the lead to have too much in it. Also, we would have to impliment a new parameter into the existing template to remove links to the above pages. However, I think that this could be implimented in a new template along the lines of "nihongo lead" like we have {{nihongo title}} currently (which does bold and italics), making it bold and including links to the language pages as the various lang-XX templates do already, but that really has nothing to do with the issue at hand as to whether or not WP:VG should adopt a style for dealing with romaji that all other pages and projects do when it comes to non-Latin text.—Ryūlóng (竜龙) 18:38, 10 February 2011 (UTC)

{edit conflict)From what he says its a bit more because its not clear to the layman what romanization is and seeing it out-of-context would confuse them while adding unnessasary detail. That's why he suggests romanization, although if we go that route, it should mirror the other templates and list the type of romanization ie,"aaaaa (Japanese: bbbbb, modified Hepburn: ccccc)" similar to Chinese. While we might only use 1 type of romanziation, the average person won't know what romanization is let alone which one we are using. I still think though requiring it on everything (rather than strongly advising it) is a direct violation of the policy WP:COMMON. If it was reworded to wording that was strongly advised rather than require I'd be more willing to let MOSJA handle it if it did so appropriately, ie allowed for reasonable exceptions and didn't use it as an excuse to have none.陣内_Jinnai 18:43, 10 February 2011 (UTC)

That's why we can have one (new) template that says "AAAAA (Japanese: BBBBB, Hepburn: CCCCC^?)" and use it in the lede (or modify {{nihongo title}} and remove the italic forcing coding), and retain the original {{nihongo}} for uses in prose/lists.—Ryūlóng (竜龙) 18:53, 10 February 2011 (UTC)

I agree with your take on it, Jinnai. And I think WPVG would be likely to accept this. -Thibbs (talk) 19:02, 10 February 2011 (UTC)

Ditto. I think this is probably the biggest step we've made in reaching a resolution. To my knowledge, the issue with the VG guideline applied mainly with the lead. Inclusion of the some clarification would fix my concerns, and I assume other VG members. Modifying the title template is probably the best practice as it would quickly apply to current uses and minimize any work with the implementation. (Guyinblack25 ^talk 19:17, 10 February 2011 (UTC))

@Thibbs - if it goes along with a rewording of this guideline to have more flexability I could see them accepting it.

@Ryulóng - it should still say modified Hepburn, unless we are going to start allowing any form of hepburn; it wouldn't be clear to everyone what form we'd be using otherwise.陣内_Jinnai 19:32, 10 February 2011 (UTC)

We use revised.—Ryūlóng (竜龙) 19:37, 10 February 2011 (UTC)

Breakpoint 5

• OK, the disputed topics seem to be:
  1. Japanese words which are English written as katakana :: include a literal romaji back-transcription?
  2. Native Japanese words: include a romaji transcription if that differs a bit from the usual English form?
  3. What proportion of readers want/need to see the romaji, anyway?
  • Other topics seem to be:
    • New template that expands e.g. "{{nihongo|[[Thunderbird]]|サンダーバード|Sandābādo}}" into not "Thunderbird (サンダーバード, Sandābādo)" but "Thunderbird (Japanese: サンダーバード, Hepburn: Sandābādo)" or similar. That looks useful.
    • Whether (to avoid content forking) to put the new rules in Wikipedia:WikiProject Video games/Article guidelines#Japanese titles, linked to from Wikipedia:Manual of Style (Japan-related articles)#Videogames.
    • Please append, concisely, any other undecided topics.
    • I am aware that there are persistent differences of opinion here; please help to get this matter settled at last. Anthony Appleyard (talk) 23:58, 10 February 2011 (UTC)

There does not need to be a special heading here for video games because the only issue is that of whether or not to include a romanization. What can be said at WP:VG/GL is that "To be consistent with articles on other subjects, the romanization of the Japanese title of a video game should be included in the lead paragraph. For more information on how to romanize Japanese, see WP:MOS-JA" and then WP:VG/GL can have the various methods by which to deal with the names like they do now (even though I have modified Oracle of Ages to where their "JPN" formatting is now not necessary).—Ryūlóng (竜龙) 00:41, 11 February 2011 (UTC)

The primary problem I have with literal back-transcriptions is that, without any context, they look like vandalized pronunciation guides. They go where the IPA would, but often look nothing like how the word is said in English (which is the entire point of romanization in the first place). I think even adding some context to explain that it's the "Japanese" pronunciation would not make up for the confusion caused by having the romanization there in the first place. Nifboy (talk) 01:11, 11 February 2011 (UTC)

"wu"

Would somebody please comment about "wu" for "ウー" at Talk:Hepburn romanization#About my edit on extended katakana.--Mujaki (talk) 17:55, 2 February 2011 (UTC)

Ugh. If ウー is being used to approximate the "wu" sound in various languages, then it probably isn't meant to be read as ū in those situations. How is that so fucking hard to comprehend Mujaki?—Ryūlóng (竜龙) 19:59, 2 February 2011 (UTC)

Stay civil, please. (Especially when you are demonstrably in the wrong, as in this case.) Jpatokal (talk) 21:54, 7 February 2011 (UTC)

I'm not demonstrably wrong if I have proof that ウー is used for the wu phoneme.—Ryūlóng (竜龙) 22:05, 7 February 2011 (UTC)

Hepburn is a system of romanization, so you would need to provide proof that "wu" is an accepted Latin transliteration of the Japanese kana "ウー". Jpatokal (talk) 10:58, 10 February 2011 (UTC)

I've removed ウー = wu from everywhere on the project that I am aware of its usage.—Ryūlóng (竜龙) 18:40, 10 February 2011 (UTC)

• This is merely another case of kana being imperfect at rendering foreign names, since e.g. nobody in Japan ever invented a kana equivalent of the Indian virama as a vowel-canceller. Anthony Appleyard (talk) 23:45, 8 February 2011 (UTC)

  No. I was suggesting that in specific cases, ウー which is normally the long U sound, could be romanized into the WU sound in words that we know are read with the W sound in other languages. However, Mujaki cannot understand the concept of exceptions it seems.—Ryūlóng (竜龙) 20:37, 9 February 2011 (UTC)