Wikipedia talk:Manual of Style/Japan-related articles/Archive 22

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Archive 21 Archive 22 Archive 23


℃-ute vs. Cute?

Shouldn't the title of the band "℃-ute" be "Cute"? moocowsrule 01:25, 22 December 2008 (UTC)

Or at least "C-ute"? Rules for using symbols probably apply for every possible article on Wikipedia. So ask on Wikipedia:Manual of Style talk page. --staka (T) 22:21, 4 January 2009 (UTC)
It's a Japanese band, so I thought I'd bring it up on the Japan related Manual of Style, rather than the other one. moocowsruletalk to moo 00:34, 12 January 2009 (UTC)

Ryokū vs. Ryokuu

Basically romanization of KANJIs should be done one by one.

Saitō Ryokū (斎藤緑雨) should be moved to Saitō Ryokuu. ”緑雨” is not "Ryokū" but "RYoku u": "uu" in his name is not a long vowell "ū" but two short vowells "u+u". See Talk:Saitō Ryokū.--miya (talk) 00:26, 4 January 2009 (UTC)

I thought we agreed to that principle—vowels that span multiple kanji are not indicated by macrons— but I can't see it in the MoS. Was it formerly there? Fg2 (talk) 02:29, 4 January 2009 (UTC)
I believe it was agreed on, but no one ever added anything. Also I note that it's more general than just spanning kanji: 湖 mizuumi should probably not have a macron either, unless I'm mispronouncing it. (I mentioned the term "morpheme boundary" last time, but that would admit okāsan and onēsan; judging strictly on aesthetic grounds, I have no problems with the former, but the latter still seems weird — and I doubt anyone would be happy with onīsan.)  –Aponar Kestrel (talk) 04:22, 9 January 2009 (UTC)
湖 should be "Mizūmi". The "okaasan" "oniisan" "oneesan" is covered in the Manual of Style. moocowsruletalk to moo 05:12, 25 February 2009 (UTC)
It would surprise me if 湖 were mizūmi. I would have guessed that the word came from mizu + umi, and that it was in Japanese before the arrival of the kanji. So I'd treat it as two words, despite being represented by a single character. Would that be incorrect? Fg2 (talk) 07:17, 25 February 2009 (UTC)
Fg2 is right. it's mizuumi. It is definitely a combination of mizu and umi. Maybe the original meaning is where there is a lot of fresh water like sea. Oda Mari (talk) 07:48, 25 February 2009 (UTC)
Indeed, it is not a long vowel. Saitō Ryokuu is correct. This is リョクウ rather than *リョクー. While the kanji may be a hint and it is sufficient in this case, there are exceptional cases. Romanization is based on the actual pronunciation. Some dictionaries such as 日本国語大辞典 and 新明解日本語アクセント辞典 list the pronunciation in addition to the spelling. Both of which contain entries for this case as well for verification if needed. Bendono (talk) 02:46, 4 January 2009 (UTC)
Yes, the "uu" is correct. I hadn't noticed that before. ···日本穣? · Talk to Nihonjoe 20:17, 4 January 2009 (UTC)
Agreed, "uu". ¶ My agreement aside, is it this subtlety that Quasirandom misinterprets here (the consensus about macrons in romanji [sic] article names seems to be in flux) or is he referring to something else? (In the same discussion, also note Move to Yuko Ito if kept; we don't need untypeable diacriticals in the English Wikipedia by one Smerdis of Tlön [sic].) -- Hoary (talk) 03:59, 10 January 2009 (UTC)

Body text rule apply for titles

All rules in the body text section applies for titles as well, correct? By titles I mean, "Wikipedia:Manual of Style (Japan-related articles)" and such.. I don't think it states here. --staka (T) 22:19, 4 January 2009 (UTC)

Generally, yes. ···日本穣? · Talk to Nihonjoe 05:14, 9 January 2009 (UTC)

Differencing "おお" from "おう"

The manual never really describes the difference between "おお" from "おう" (oo and ou). How are they to be romanized? Are they both to be romanized as "ō"? moocowsruletalk to moo 00:38, 12 January 2009 (UTC)

They are both romanized as "ō" unless the two parts are different kanji, which is rare. ···日本穣? · Talk to Nihonjoe 01:24, 12 January 2009 (UTC)

Designated city naming

Back in 2006, there was a long discussion which resulted in most designated cities shedding their ", Prefecture" tails. Recently, the newest designated city was moved back to Hamamatsu, Shizuoka from its shortened name, which led me to search the MOS for our rules; but, either they were never added, or, they were deleted sometime in the last 2.5 years. So, should the "place names" section be updated to match the old consensus (and, Hamamatsu moved back?) or, is there a new consensus?? Neier (talk) 13:01, 25 January 2009 (UTC)

I don't see Hamamatsu in the list of exceptions to the standard rule. I don't see any reason why it shouldn't be added, though. ···日本穣? · Talk to Nihonjoe 06:11, 26 January 2009 (UTC)
Hamamatsu became "designated" in 2007, after our discussion. Okayama, Okayama will gain that status this year; but, since the city and prefecture are the same name, there is no need to do the move. Neier (talk) 11:37, 26 January 2009 (UTC)

Japanese counters

How do you romanize Japanese words with counters (i.e. 三人, 10年)? Do you use a dash between the word and the counter, or just romanize it as one word? Or does the romanization depend on the counter you are using (if it's a common counter or not)? I really think MOS:JP should address this to solve any problems of articles with counters on their name (i.e. Morning Musume Tanjō 10 Nen Kinentai). Eugh jei Kaorin 22:47, 28 January 2009 (UTC)

IMHO, they are one word. Or I take them as one word in a native speaker's sense. [1], [2], [3] and [4]. Oda Mari (talk) 05:26, 29 January 2009 (UTC)


Any opinions on what the title of Shekwasha should be, perhaps after a glance at ja:シークヮーサー#表記? It seems to me that we're not at the ideal title now, but I don't really have a grasp on how we treat that type of thing. I remember a similar question about goya a long time ago, but all of that appears to have been excised from the Bitter melon article. Dekimasuよ! 12:25, 25 February 2009 (UTC)

Not a topic that I am too familiar with, so I do not have any strong opinions either way. However, in academic papers and technical translations I have often used scientific names as they are most precise and also neutral. Also, WP:NC(flora) makes the same recommendation. Even if Shkwasha or some variation of that is somehow the common name, the recommendation is to redirect it to the scientific name. Thus I'd suggest Citrus depressa Hayata. I suggest taking it up at Wikipedia talk:WikiProject Plants for a better response. Bendono (talk) 13:06, 25 February 2009 (UTC)
Thanks for the help. The naming convention is good enough for me, so I've gone ahead and moved the page to Citrus depressa. I'll create a redirect from the longer title you mentioned. Dekimasuよ! 15:31, 25 February 2009 (UTC)

Casing in romanization

What part of WP:MOS-JA supports this edit: [5]? It is a romanization, nothing more, and capitalizing certain words as if they are proper nouns makes little sense. Bendono (talk) 08:21, 5 March 2009 (UTC)

Asuka is a proper noun, though, and should be capitalized. When it comes to other things, like capitalizing Kabuki or Karate by way of emphasizing it or denoting it as a foreign word, I am for the most part opposed to it, and have in fact spent some time lately decapitalizing instances of "Karate" all over Wikipedia. But I don't think special rules need apply within the romanization brackets that don't apply elsewhere; proper nouns should be capitalized, and Asuka is a proper noun. LordAmeth (talk) 11:43, 5 March 2009 (UTC)
I agree that Asuka is a proper noun. And it is appropriately capitalized in the English text. However, that is not what I am referring to. I am asking about the romanization given in the {{Nihongo}} template. It makes little sense to apply any special grammatical or style guidelines such as casing there. If you do, then while 飛鳥 is a proper noun, then what about 飛鳥時代? Similar logic could be applied to capitalizing jidai as well. There are a number of other cases (no pun intended) as well. Rather than waste time on such issues, I would opt to always make it lower case and simply be done with it. Bendono (talk) 12:25, 5 March 2009 (UTC)
"to always make it lower case and simply be done with it" may be a little too unsubtle. While I agree that when romanisation is used as a purely technical device it is inappropriate to apply style/grammar guidelines, consider the situation where the romanised text is the primary focus e.g. Bashō's bibliography. There may be other comparable scenarios that need separate consideration.--Yumegusa (talk) 13:43, 5 March 2009 (UTC)
None of those are in the Nihongo template and do not act merely as a romanization but as real titles. So really outside the scope of this topic. Bendono (talk) 13:50, 5 March 2009 (UTC)
I understand where you're coming from, but I don't feel that any special rules need to apply to things specifically within that template vs outside of it. Even in that particular spot, it serves not solely as a pronunciation guide, not solely as romanization, serving a purely linguistic purpose, but also serves to provide an alternate title or name for the subject of the article, and thus should reflect a standard of capitalization. Whether we write Asukajidai, Asuka jidai, Asuka Jidai etc is of course a matter of debate; there can be no single definitive answer about whether or not jidai ought to be capitalized. But, again, for proper nouns such as Asuka, Tōkyō, etc, I see no reason to make them lowercase, ever, no matter what template they're in, no matter what context. That's my two cents. LordAmeth (talk) 16:44, 5 March 2009 (UTC)
As LordAmeth wrote, that edit (and all the others I made in that series) were because the romanization should follow the same capitalization rules as for the English words. WP:MOS-JA talks about it in the Capitalization section. ···日本穣? · Talk to Nihonjoe 04:22, 6 March 2009 (UTC)

Searching the archives

I've added a search field in the archive box to allow searching of just the archives of this page. Enjoy! ···日本穣? · Talk to Nihonjoe 20:17, 8 March 2009 (UTC)

Thanks x 2! This and the one at WikiProject:Japan will save a lot of work searching. Fg2 (talk) 20:35, 8 March 2009 (UTC)

Linking dates

Wikipedia:Linking#Chronological items has just been updated with the results of a major poll on whether or not to link dates. Fg2 (talk) 10:40, 14 April 2009 (UTC)

Need some help on the Rampo Edogawa article.

Rampo Edogawa has been changed to Edogawa Rampo near the end of last month. I don't know why it keeps getting changed back and forth between either one of these names, but I have now changed the name back to "Rampo Edogawa", with a note about its romanization. Is this change acceptable? Thank you in advance. Sjones23 (talk - contributions) 00:38, 20 April 2009 (UTC)

The change is correct. Oda Mari (talk) 05:31, 20 April 2009 (UTC)

Article names for mountains

There's a near-consensus discussion at Wikipedia talk:WikiProject Japan#Mount O on how to name Japanese mountains -- please chip in! Jpatokal (talk) 02:49, 29 April 2009 (UTC)

Why is the discussion there and not here, where is belongs? Bendono (talk) 03:06, 29 April 2009 (UTC)
If you go read the discussion, you'll see it initially was just a quick question of how to name a specific article. As more people frequent that talk page, it only makes sense that someone would ask a question there first. ···日本穣? · Talk to Nihonjoe 03:23, 29 April 2009 (UTC)

Discussion from WT:JA

Moved here for convenience.

A new article describes Mount O (御岳 O-take). In contrast, we already had Mount Ontake (御嶽山 Ontake-san), rather than Mount On. Is "Mount O" the best name for the article on 御岳? Is "Mount Ontake" the best name for the article on 御嶽山? Fg2 (talk) 04:40, 25 April 2009 (UTC)

When I search "御岳山", I get Mitakesan... And the article says it's in Kagoshima, but also says it's a Hokkaido stub... there's something very wrong with this article. (talk) 06:24, 25 April 2009 (UTC)
Depends on which 御岳, 御嶽, or 御岳山 it is. (talk) 06:49, 25 April 2009 (UTC)
The point is how to deal with the tautological mountain names of xxx岳/嶽 with 山 and the names with xxx岳/嶽 without 山, isn't it? BTW, the precise name of the volcano in Sakurajima peninsula is Ontake/御岳 too, even Japanese usually call it Sakurajima. There are many mountains of these kinds. See the ja links above. The matter makes me headache. I'm not sure...but both of the names might be best. Oda Mari (talk) 07:02, 25 April 2009 (UTC)
Although ja:御岳 calls it On-take, it seems from short search that 御岳 on Nakanoshima (Kagoshima) is called in three ways in Japanese: O-take, O-dake and On-take. --Sushiya (talk) 07:09, 25 April 2009 (UTC)
This local site says it's Otake. Oda Mari (talk) 07:33, 25 April 2009 (UTC)
And this page by Kagoshima prefecture calles it おんたけ (Ontake). The name of the article should be, as long as the style is concerned, in the same way as the other Ontake, like "Mount XXXXtake (Kagoshima)". "On", "O" or "Mi" is an honorific prefix added to take, so the name really means "The Mountain". There are alikes in Japan, for example 江の川, whose name is "The River - River". --Mantokun (talk) 07:49, 25 April 2009 (UTC)
国土地理院 also says it's Ontake. (Click Kagoshima in the map.) Then the volcano name of Nakanoshima is Ontake? Oda Mari (talk) 08:51, 25 April 2009 (UTC)
Incredible... Each island in the Tokara islands has its own peak named 御岳 and they're all called differently.... At any rate, the article name for the one in question be named "Ontake" after GSI preference on WP, I suppose. --Mantokun (talk) 09:35, 25 April 2009 (UTC)
Someone should build a WP:SETINDEX/{{SIA}} and WP:CJKV dab page for it then. (talk) 13:44, 26 April 2009 (UTC)
If Japanese name is Ontake, article title should be "Mount On" in accordance with the naming convention inferred from the examples of Mount Tsurugi (Toyama) (Tsurugi-dake), Mount Yake (Yake-dake), Mount Yari (Yari-ga-take), etc. --Sushiya (talk) 12:33, 25 April 2009 (UTC)
And what is wrong with Otake or Ontake? If we consider -take, -dake as part of the proper noun which forms the name, the controvery and confusion can be reduced. As with Ōshima, we do not have "Ō Island". For Shimbashi, we do not have "Shin Bridge". For Yamashiro Province, we do not have "Mount Shiro Province". In naming of mountain articles in German (Matterhorn, not "Matter Peak") and French (Mont Blanc), not "Mount Blanc" or "Blanc Mountain"), we have many similar examples. Given the variety of methods by which the Japanese name is transliterated depending on English-language source, would it not be better to stay with the original name as the article title? --MChew (talk) 03:53, 26 April 2009 (UTC)
Of course exceptions are allowed if you have some ground to break the de facto (or documented? I'm not sure) naming convention. We already have exceptional forms like Asahidake, Kurumayama, Daisen (mountain) and Pinneshiri. --Sushiya (talk) 08:35, 26 April 2009 (UTC)
Hi folks, we have had this discussion repeatedly. MChew has always taken the stance that we should not translate these parts of the proper nouns. I have over time taken the opposite end of the argument. For example, we translate Kaikyō in Tsushima Strait. So with Otake I made it Mount O. The past discussions have run along the line that we translate these nominal parts. There different people have created different exceptions, like the well-known English name is not translated, or if the name ends up being too short, with no clear-cut definition of what too short is. I choose a methodology that leaves as few exceptions as possible, just so I do not have to think too much. Do I apply it universally? No, otherwise I would have made Nakanoshima into Naka Island. Sometimes it is hard with names like Furanonishidake. See past discussions here and here.
As for the reading, I based the reading on the Quarternary Volcanoes of Japan article for GSI. There they give it the reading 'O'.imars (talk) 11:22, 26 April 2009 (UTC)
If we are looking for a method with the least exception, the only way I can think of is to stick strictly to local names. However, let's put that aside a moment because there's a reasonably stated guideline here: Wikipedia:Naming conventions (geographic names).
Although some changes were made to the guideline over time since 2008, the basics as to how to treat non-english names have not changed largely, with its essence excerpted as follows:
  • The title: When a widely accepted English name, in a modern context, exists for a place, we should use it. This often will be a local name, or one of them; but not always. (August 2008 - further elaborated but, in essence, still the same today)
  • If no name can be shown to be widely accepted in English, use the local official name. (April 2008 -)
I may have overlooked a J-specific rule in which case I would like to be pointed to it, but otherwise, the above seems to be where we should fall back upon as reference.
Then, 御岳 is "Ontake" or "Otake", unless there's a clear explanation provided as to why it should be named otherwise. (Note: How the name is called locally is yet another issue and has to be treated apart from whether the article should be named with an English name or not.) A reason why people tend not to be too irritated by kaikyō being translated into English as "straits" is likely because such is a relatively newer notion, or term, that have only become prominent in modern times, or earliest in pre-modern times beyond the 17th century. In other words, it is a term too new for people to look at the term as an inseparable, integral part of a name; names of moutains and rivers, are often much older to be allowed to de-composed. I have already stated in my previous post that "御" in the case of "御岳" is a prefix. Regardless of how it is pronounced, its meaning is "THE mountain". There was not a mountain called "O" or "On"; if there was not THE mountain there was nothing, thus calling it "Mount O" is just nonsense. Nevertheless, if there is a proof that it became to be called "Mount O" or "Mount On" in English, then let's just take it.
Examples as Mount Tsurugi, Mount Yake or Mount Yari are some of those whose naming is yet subject to review than being referred to as any sort of "standard" naming examples.
I have included a list of search results here related to this matter on my User page (just because it is so lengthy but I am ready to copy it over here if a portal admin so requests). --Mantokun (talk) 15:01, 26 April 2009 (UTC)
I conceede that Mantokun has raised some good points. I am rethinking my position. One question though, how do we interpret widely accepted in English in the guideline? imars (talk) 20:14, 26 April 2009 (UTC)
Tsushima Straits is quite okay under Wikipedia:Naming conventions (geographic names), as it is an internationally recognized name of an international waterway. Likewise Mount Fuji is the commonly used name for that mountain. However, very few Japanese geographic names are not widely known outside of Japan. --MChew (talk) 01:48, 27 April 2009 (UTC)

I agree that "Mount O" sounds bizarre and is not the most commonly used form. I would suggest the same compromise that we have for shrines and temples: translate -san into "Mount", but keep -yama, -dake as is. Jpatokal (talk) 03:17, 27 April 2009 (UTC)

I think this is a good compromise. ···日本穣? · Talk to Nihonjoe 07:03, 27 April 2009 (UTC)
The Japanese Wikipedia has one article on Oyama (小山), one on Oyama (雄山) (with the disambiguation page listing one more), and two articles on Ōyama. There's a disambiguation page for two mountains named 御山, which might be Oyama or Miyama. So there could be lots of Oyamas and Ōyamas. But that's a matter for disambiguation (e.g. Ōyama (X Prefecture)) and not naming the mountain. I prefer Oyama to Mount O. Fg2 (talk) 11:53, 27 April 2009 (UTC)
This sounds good to me too. Dekimasuよ! 14:31, 27 April 2009 (UTC)
In Jpatokal's proposal, the -dake part sounds reasonable, but I don't agree to the -yama part, assuming titles like Mount Asama, Mount Kumotori etc. Unlike -dake, -yama is merely an alternative reading of the kanji 山 so that distinctive treatment based on Japanese readings would cause confusion. --Sushiya (talk) 15:07, 27 April 2009 (UTC)
Mmm. Do we need to qualify by syllable count instead then? Monosyllabic names (excluding suffix) retain the Japanese form (Oyama, Ondake, Gassan), longer ones are English (Mount Fuji, Mount Asama, Mount Haguro)...? Jpatokal (talk) 03:14, 28 April 2009 (UTC)
A modified version of Jpatokal's original proposal sounds like a reasonable compromise to me, (although I understand Sushiya's concerns) and it would still leave us with "Mount Ga" instead of Gassan. On the other hand, going purely by syllable count would leave us with "Mount Yatsu", or even worse, "Mount Yatsuga", instead of Yatsugatake. Perhaps keeping -yama, -dake, -san and monosyllabic -yama names "as is", and changing multisyllabic -yama to "Mount" would work? I would not object as much to "Mount Asama" as I would to a "Mount Ga". I believe that Sushiya's point is that the final kanji 山 could be read as either -san, or -yama, depending on the mountain. Perhaps those cases could be addressed case-by-case depending on the prevalent local name?--MChew (talk) 00:44, 29 April 2009 (UTC)

OK, I think MChew's version is the best yet. Draft for MOS-J: Jpatokal (talk) 02:44, 29 April 2009 (UTC)

Use the Japanese name for all mountains suffixed with -dake/take (岳), and monosyllabic names suffixed with -san or -yama (山). For longer -san/yama names, use the form "Mount X" without the Japanese suffix. If there are multiple mountains with the same name, disambiguate by prefecture.
Examples: Ondake, Yatsugatake, Gassan, Oyama but Mount Fuji, Mount Asama, Mount Tsurugi (Tokushima)
I strongly disagree that there is any consensus on this issue. Yatsugatake is a redirect, Gassan does not even refer to the mountain, Oyama is a disambiguation page. Why is -yama translated and -dake not? What is the reasoning for that? Either translate both or neither? imars (talk) 07:27, 29 April 2009 (UTC)
This is only a proposal — I'm not claiming that there is consensus nor that current names match this. (FWIW, you'll find 月山 at Mount Gassan.)
And there's a good reason to treat -dake/take separately: unlike -san/yama, it often takes the particle ga, and unlike -san/yama, it's also used to mean "peak" as well as "mountain". This is impossible to smoothly translate into English, and makes shoehorning such names into "Mount X" format awkward at best: for example, 八ヶ岳 is actually "Eight Peaks", as opposed to a single "Mount Yatsu" (or, worse yet, "Mount Yatsuga"). Jpatokal (talk) 08:36, 29 April 2009 (UTC)
Yatsugatake is perhaps a bad example, as it does not really represent a single mountain, but rather a mountain range. You see the same thing repeated again and again: 八甲田山 should not be Mount Hakkōda, but Hakkōda Mountains. Nor should 大雪山 be Mount Daisetsu, but Daisetsuzan Volcanic Group. I see nothing wrong with 松田岳 Mount Matsuda, 比布岳 Mount Pippu, or 石狩岳 Mount Ishikari so long as Mount Asama, Mount Fuji, and Mount Kumotori are acceptable. imars (talk) 13:19, 29 April 2009 (UTC)
Sigh. So, what do you propose instead? Treat -dake/take just like -san/yama, and add a specific exception for the case where -san/dake/take represents multiple mountains? Jpatokal (talk) 03:17, 30 April 2009 (UTC)
On the other hand, Mount Warusawa is not actually a mountain as implied by the article title, but simply is one of three named protuberances on ja:荒川岳. imars illustates the danger of simply translating -san/-yama as Mount, as it could be a single mountain, or an entire range. Likewise -dake/-take can be a mountain, or it could be a single peak on a mountain. Which brings me back to my original position that -dake/-take, -sam/-yama should be considered intregal parts of a proper name, and not divided out. If it helps drive a compromise, I do not see a problem with the redundant Mount XXXdake or Mount XXXsan, if imars is adamant that the use of the word "Mount" is essential to the understanding of non-Japanese speaking readers. I would prefer a redundant "Mount Ontake", which preserves the native form proper name over a truncated version. --MChew (talk) 03:35, 30 April 2009 (UTC)
I am beginning to weaken my position on forcing Mount. I clearly see the error in Mount O. If we continue down the path of adding Mount, then it should be consistent no matter if it is -yama, -san, -zan, -dake, -take, -mine, -hō or whatever. Mantokun raised the point that we have a general standard to use the widely-accepted English name. Mount Fuji is clear. If none exists then use the local name. I am writing a lot of articles about relatively obscure mountains as far as the English-speaking world goes. Let's take an example: 幌尻岳 Mount Poroshiri, the tallest mountain in the Hidaka range of Hokkaido. Literally this is Big Mountain Mountain. Poroshiri is big mountain in Ainu. Dake we already know. A quick Google search gives Mount Poroshiri 201, Mount Poroshiridake 0, Mt. Poroshiridake 4, Poroshiridake -Mt. -Mount 126, 幌尻岳 36,900. Is this enough of a sampling to determine if there is a widely accepted English name? If it is, Mount Poroshiri seems to come out slightly ahead of Poroshiridake, but this is hardly a scientific count. By excluding Mt. and Mount from the Poroshiridake search I may have influenced the result. imars (talk) 07:08, 30 April 2009 (UTC)
I would oppose redundant "Mounts", we don't do this for any other articles and we don't need to start doing it now. So either "Mount X" or "X-san", but not "Mount X-san". Jpatokal (talk) 12:56, 30 April 2009 (UTC)
Not exactly true: Mount Kabutoyama, Mount Nakayama, Mount Nishigatake, Mount Miyama, Mount Bunagatake, and Mount Hinokizuka Okumine. OK, I could only find 5 out of a few hundred. ;-) imars (talk) 15:05, 30 April 2009 (UTC)
Allow me to correct my wording: we don't need to start enshrining such naming in policy. And, interestingly enough, none of the examples you note have -san as an ending. (I'll add one for you though: Mount Gassan.) Jpatokal (talk) 06:24, 1 May 2009 (UTC)
Back to my question, Encyclopedia Britannica writes Mount Poroshiri. So it looks to me like they tend to translate these endings. Additionally all the river articles in List of rivers of Japan translate -k/gawa 川, X River, with the exception of Kinokawa River and Mukogawa River, which choose the redundant X-gawa River. Lakes are also mostly translated. imars (talk) 15:22, 30 April 2009 (UTC)
I agree, and I think we should mostly follow suit, but there are some names for which this would be absurd ("Mount On", "Mount Mi", "Mount Gas"), so the question remains where to draw the line. Jpatokal (talk) 06:24, 1 May 2009 (UTC)
I agree that "Mount Gas" is absurd, as it doesn't represent the Japanese name (although it is what you get by separating "san" from "Gassan," it invites interpreting the name as gasu). And short names like "Mount On" and "Mount Mi," especially those with name components that seem to modify "Mount" rather than to be independent names followed by san, are the sort of thing that prompted me to ask this question in the first place. But it's not necessarily shortness; I can see the thinking behind Mount Tate (which looks like the one-syllable English name Tate) even if it was alarming the first time I encountered the article. Still, I think this is getting close to a good solution. Fg2 (talk) 08:12, 1 May 2009 (UTC)

Problems with the "romanization" section of the manual?

This isn't the first time I bring some of this up, but I hope I'll get a bit more feedback this time around...

(and I hope you'll forgive my poor English skills, too)

Like the title says, I'm not too sure about some aspects of the "romanization" section of the manual...

For one thing, it starts off by saying that "revised Hepburn romanization (described below) should be used in all cases". But the "description below" obviously is a very summarized one, and even the article this sentence links to doesn't appear to offer (or link to) any exhaustive description of this "revised Hepburn romanization". I'm still fairly new to Wikipedia, but... wouldn't a source be nice, here?

Especially considering the same manual also advises the use of "modified Hepburn" in several instances. There's apparently some general confusion regarding those two variants, as amusingly noted in the Hepburn romanization article (emphasis mine):

  1. The second is revised Hepburn, a revised version of traditional Hepburn, in which the rendering of syllabic n as m before certain consonants is no longer used. This is used by the Library of Congress. (Revised Hepburn may be referred to as modified Hepburn.)
  2. The third is modified Hepburn, which builds on revised Hepburn to further modify traditional Hepburn. This version is consistent in its treatment of long vowels (always doubling the vowel) and syllabic n (always n-bar). It has been adopted by some major dictionaries (e.g. the Pocket Kenkyusha Japanese Dictionary published by Oxford University Press), but is still mainly the preserve of linguists. (The term modified Hepburn may also be used to refer to revised Hepburn.)

... Well, that's obviously very convenient.

So what is it we're supposed to be using anyway? Revised? Modified? Where do you go when the short description provided by the manual doesn't suffice?

The article says at one point that "in 1972, a revised version of Hepburn was codified as ANSI standard Z39.11-1972". So I looked around and found this. It's only a summary, apparently, but it's a lot more detailed than the one provided by the manual, so I guess that's always something.

According to the Hepburn romanization article, in "modified Hepburn", long vowels are always indicated by doubling the vowel. ANSI Z39.11-1972 disagrees and says macrons are the future of mankind, so it would appear it really is "revised Hepburn" indeed, not a mislabeled modified Hepburn.

Now, the manual also advocates the usage of macrons, which, at first glance, seems to indicate that it's really referring to revised Hepburn / ANSI Z39.11-1972 when it comes to romanization. But it says that only long "u" and "o" sounds should be written with macrons, and "all other long vowels are written without macrons: ああ → aa, いい → ii, and ええ → ee." That actually goes against ANSI Z39.11-1972, which states that long "a" and "e" sounds should also be written with macrons. So this is getting confusing already. Why does the manual say that "aa" and "ee" should be used? Based on what? Is there another (apparently slightly different) revised Hepburn out there? And if so... source?

When I brought this up a while ago, I was told that maybe the manual says to write "aa" because more often than not, "ああ" won't really be a long "a" sound, but rather two juxtaposed "あ" (and I'm guessing the same could be said about "ee"). Now, long "a" sounds certainly aren't common, and I guess there indeed is a good chance most "ああ" you'll find will simply be two juxtaposed "あ"... but that's still a rather unfortunate way to put things, for a manual, don't you think? After all, we are talking about long vowels, here, no matter how rare some of them may be.

Then, there is the matter of those two points:

  1. For transliterations from kanji and kana, long o and u are written with macrons as ō and ū respectively. If you have difficulty typing these characters with your IME, you can click on the special characters below the Wikipedia edit box, or see Help:Macrons for instructions on setting up your computer to input them directly from the keyboard. You can also enter the HTML entity ō for ō, and ū for ū. All other long vowels are written without macrons: ああ → aa, いい → ii, and ええ → ee. Apostrophes and hyphens are not needed to distinguish i)(i from ii.
  2. For transliterations from katakana, use the English spelling if available (i.e., Thunderbird (サンダーバード Sandābādo) instead of Sandābādo). If an English spelling is not available, but a spelling from another language of origin exists, use it (i.e., Homard (ja:オマール Omāru) rather than Omāru, and Zha cai (ja:ザーサイ Zāsai) rather than Zāsai). Otherwise, macrons should be used for all long vowels indicated with ー, including "a", "e", and "i".

Right now, the manual makes a distinction between "transliterations from kanji and kana" and "transliterations from katakana". Naturally, katakana are kana, so this is some unfortunate wording right here. From what I can tell, the intent was actually to distinguish words of Japanese origin and words of foreign (well, "non-Japanese") origin. If that is indeed the intent, and considering words of Japanese origin can be written in katakana (... and words of foreign origin are sometimes written in hiragana... there are some jokers out there), I think it would make a lot more sense to... well, explicitely have the first point be about "transliterations of words of Japanese origin" (a case could be made for "transliterations of words of Japanese or Chinese origin", but that would only concern on'yomi and thus might muddle the issue quite a bit), and the second one be about "transliterations of words of foreign origin".

Finally, am I the only one a bit perplexed by the way the second point is worded? "If an English spelling is not available, but a spelling from another language of origin exists [...]"? Yes, if we're talking about a word of non-Japanese origin that is not (amazingly enough!) of English origin either, it's a pretty safe bet that it won't have an English spelling and we'll have to try and look elsewhere... I'm thinking a slightly less anglocentric take would probably be more logical in this context (and perhaps even be shorter in the end).

Soooo... what do you think? Is it just me, or is there some room for improvement here? Erigu (talk) 06:15, 21 May 2009 (UTC)

This is the English language Wikipedia, so of course it's going to be more anglocentric. However, I fail to see how the MOS-JA telling you to use the spelling from another language is "anglocentric". There are plenty of words not of English origin which have English spellings, and some which may even be different than the original language spelling (Brazil/Brasil springs to mind, and I'm sure there are plenty of others). As this is the English language WIkipedia, English language spellings are preferred if they exist; otherwise, use the spelling of the language of origin. In cases, such as with Japanese, where the language of origin does not use Latin letters, then the appropriate romanization scheme should be used. ···日本穣? · Talk to Nihonjoe 15:51, 21 May 2009 (UTC)
This is the English language Wikipedia, so of course it's going to be more anglocentric.
In this particular context, anglocentrism makes no sense. Words of foreign (non-Japanese) origin may be of any origin, and it makes no sense to make a distinction between those of English origin and those of other origins if, in the end, the exact same principle should be applied anyway: you're supposed to use the spelling of the original language.
There are plenty of words not of English origin which have English spellings, and some which may even be different than the original language spelling (Brazil/Brasil springs to mind, and I'm sure there are plenty of others).
I wouldn't say "Brazil" is the English spelling of "Brasil". "Brasil" is the Portuguese name, and "Brazil" is the English one. They just happen to be very similar. Or would you say "Londres" is the French spelling of "London", by the same logic?
And I believe Japanese dictionaries specify the origin of gairaigo (which would tell us which spelling to use). My dictionary, for example, says that "burajiru" comes from the English "Brazil". So "Brazil" it would be.
In cases, such as with Japanese, where the language of origin does not use Latin letters, then the appropriate romanization scheme should be used.
And that's a precision that should probably be added to the manual, actually. While it does provide an example from Chinese, it doesn't explain why the alphabet spelling "Zha cai" is used. Erigu (talk) 16:11, 21 May 2009 (UTC)

Romanization of "ou" in words of foreign origin

I've recently been arguing with others about the romanization of "ソウル" (from the English word "soul").

My opponents argue (well, argued, as somebody decided the discussion was over) that "ソウル" should be romanized as "sōru". I'd personally use "souru".

Are you referring to Wikipedia talk:WikiProject Japan#Katakana question? There was no such "decision", it was agreed that the correct spelling depends on the pronunciation. Jpatokal (talk) 09:45, 21 May 2009 (UTC)
No, I was referring to this. But I guess I should have expected the announcement of this new Pocket Monsters game would generate similar discussions. Erigu (talk) 10:12, 21 May 2009 (UTC)

My opponents' argument seems to be (feel free to barge in if I get something wrong) that we should simply follow policy, i.e. check the manual of style. And the manual of style says this:

  1. For transliterations from kanji and kana, long o and u are written with macrons as ō and ū respectively. If you have difficulty typing these characters with your IME, you can click on the special characters below the Wikipedia edit box, or see Help:Macrons for instructions on setting up your computer to input them directly from the keyboard. You can also enter the HTML entity ō for ō, and ū for ū. All other long vowels are written without macrons: ああ → aa, いい → ii, and ええ → ee. Apostrophes and hyphens are not needed to distinguish i)(i from ii.
  2. For transliterations from katakana, use the English spelling if available (i.e., Thunderbird (サンダーバード Sandābādo) instead of Sandābādo). If an English spelling is not available, but a spelling from another language of origin exists, use it (i.e., Homard (ja:オマール Omāru) rather than Omāru, and Zha cai (ja:ザーサイ Zāsai) rather than Zāsai). Otherwise, macrons should be used for all long vowels indicated with ー, including "a", "e", and "i".

According to the first point, long "o" sounds should be written with macrons, and the second point (about "words in katakana", but like I explained in that other section I created just above, I suspect it's really about "words of foreign origin") doesn't say anything about that one way or the other. And this would mean we should just apply what's said in the first point, as there is (apparently) no reason for an exception to be made.

I disagree (obviously enough) and here's my reasoning:

Are we even dealing with a long "o" sound, here? Maybe most Japanese people would pronounce it like a long "o" sound, but then again the same would probably pronounce "ヴァ" like "ba", and both kana and revised Hepburn romaji make the distinction, here...

So should "so / u" in a word of foreign origin really be considered a ("true") long "so" sound (and romanized as such)?

To me, the "u" being used to indicate a long "o" sound is something that's only relevant to the actual Japanese language, by which I mean "words of Japanese origin" (on'yomi included, naturally).

Not only that, but "ou" isn't necessarily a long "o" sound either, in words of Japanese origin. There are quite a few examples of words or names that contain "ou" that are really juxtaposed "o / u". In fact, I'd argue that "ou" standing for a long "o" sound or "ei" standing for a long "e" sound are actually irregularities, not "the rule". Extremely common irregularities (born from centuries of evolving pronunciations and kana orthographies), obviously, but irregularities all the same. Normally, "u" stands for "u" and "i" stands for "i".

So when I'm told that there is no apparent reason to make an "exception" for "ソウル" (a word that isn't even of Japanese origin in the first place), that doesn't make sense to me.

I think that, when dealing with words of foreign origin, juxtaposition should be assumed (after all, for the long vowels, we have the chōonpu... that's pretty much the whole purpose of the thing). I see very little reason to bring up irregularities that are intimately tied to the long evolution of (the actual) Japanese language, here. And even less to see those irregularities as "the rule".

"Ball" is generally rendered as "ボール" in kana, whereas "bowl" ends up as "ボウル". Why? If it's all the same anyway, why the (extremely) common distinction? I think there's information, here, and very little reason (if at all) to lose it by transliterating both words as "bōru".

But I was told I wasn't making any sense... What do you think?

(obviously, the whole matter could be resolved pretty quickly if somebody knew of an exhaustive description of the revised Hepburn romanization that would address this particular matter, but...?) Erigu (talk) 06:18, 21 May 2009 (UTC)

I only briefly skimmed your rather long comment, but the word ソウル, meaning "soul", is transliterated as souru, not *sōru. The reason being because the /o/ here is short, not long. While vowel + u is often realized as a long vowel, that is not the case here, which is two short vowels /o/ and /u/. Although Japanese spelling is often quite rational and transparent, there are a few rough spots that need special care. This is a fairly common word, but in difficult or uncertain situations, there are dictionaries (such as 日本国語大辞典) that, in addition to the spelling, will also give the pronunciation as well. Regards, Bendono (talk) 07:59, 21 May 2009 (UTC)
Thank you for your reply (and sorry about the length of my initial comment).
So it would appear we agree about "ソウル" being transliterated as "souru". Right now, several articles (Soul Edge, Soulcalibur, The SoulTaker, Bleach (manga), Soul Eater (manga), Persona: Trinity Soul, Dragon Soul and possibly more) transliterate that word as "sōru", and my attempts to change the spelling to "souru" were reverted by an admin who pointed me to the manual of style, and explained the spelling "sōru" was dictated by Wikipedia policy and there was nothing I could do to change that. Nihonjoe joined the discussion and agreed with him.
I don't know. Erigu (talk) 09:34, 21 May 2009 (UTC)
Whack 'em with a trout and point them to this discussion. Jpatokal (talk) 09:45, 21 May 2009 (UTC)
I'm still confused by the information Oda Mari provided at Wikipedia talk:WikiProject Japan#Katakana question. She mentioned that Rubber Soul is ソウル whereas rubber sole is ソール. I have no reason to doubt this. Do Japanese people pronounce them differently? If they do, it makes it easier to believe that pronunciation causes the difference. But if they pronounce them the same (as I do, being a native speaker of English), I still wonder about the cause of this difference. Not that we need an explanation to make a rule; I simply wonder. Fg2 (talk) 11:31, 21 May 2009 (UTC)
How to convert English into kana is one question, but this is about the reverse: how to convert to kana into romaji. I think it's pretty self-evident that ソウル is meant to be pronounced "souru", because if not, it would be ソール. Jpatokal (talk) 13:35, 21 May 2009 (UTC)
At least I pronounce them differently. Oda Mari (talk) 15:02, 21 May 2009 (UTC)
I pronounce them the same in English and differently in Japanese. That seems to be consistent with my experience as well as dictionary references. Bendono (talk) 15:30, 21 May 2009 (UTC)

I don't see how "ソウル" and "ソール" would be pronounced differently at all in Japanese, when the words they represent (generally) in English ("soul" and "sole") are homophones. Just because the two words have been written differently does not mean that they are pronounced differently. The word "Soul" in Japanese has for the longest time been written as "ソウル" for whatever reason, and even if you separate the "so" and the "u" you are still producing a long o sound.—Ryūlóng (竜龙) 20:19, 21 May 2009 (UTC)

I don't see how "ソウル" and "ソール" would be pronounced differently at all in Japanese, when the words they represent (generally) in English ("soul" and "sole") are homophones.
Then again, "コンピュータ" and "コンピューター" represent the same English word ("computer"), and it seems obvious to me that you would pronounce them differently in Japanese.
As for "ソウル" and "ソール", one might wonder what came first: the kana spelling or the Japanese pronunciation? Maybe (that's really just a theory formulated on the spot) the two words were spelled differently in kana (to avoid any confusion?), and the pronunciation difference (in Japanese) mentioned by other users above was actually caused by that spelling difference.
Just because the two words have been written differently does not mean that they are pronounced differently.
But I could also argue that just because two words are written differently in kana and pronounced the same in Japanese, that doesn't necessarily mean they're to be transliterated the same according to revised Hepburn. Just above, I mentioned "ヴァ" being pronounced by most like "ba" but transliterated as "va". Erigu (talk) 20:58, 21 May 2009 (UTC)
The vu kana is completely different. For many many years, there was no way to transliterate the foreign V sound in Japanese. The B kana were used for a long time until someone decided to make "ヴ" which is transliterated as "vu". Not everyone uses this and it's still difficult in the Japanese phonemic system to add a new sound after several thousand years without that sound existing which is why "ヴァ" is transliterated as "va" but consistently pronounced as "ba". Again, you are not convincing me of anything. "ソウル" and "ソール" are homophonous. You have yet to convince me otherwise. And I do not want to see my response picked apart as you usually do. I want a straight linear response.—Ryūlóng (竜龙) 21:21, 21 May 2009 (UTC)
And the same goes for "ボール" and "ボウル" in Japanese, even though the same does not hold true for English. And the only difference between "コンピュータ" and "コンピューター" is that you pronounce the end of the latter with a long a. The choice between the two is just a stylistic choice.—Ryūlóng (竜龙) 21:24, 21 May 2009 (UTC)
For your first point, I could argue back that, just like it's difficult for Japanese people to adapt to a new sound, it might not be that surprising to see them pronounce "ou" as a long "o" sound simply by habit... But I could also point out that we have a bunch of examples, in words of Japanese origin, of "ei" that are pronounced by some exactly like "ē". Different kana spelling, same pronunciation, different revised Hepburn romanization.
My point being that I don't think you can argue there's a rule of thumb that says that, regardless of the Japanese spelling, "if it's pronounced the same, it should be transliterated the same in romaji".
As for your second point, yes, the only difference between "コンピュータ" and "コンピューター" is that you pronounce the end of the latter with a long "a". So? My point was that there was a pronunciation difference despite both words being based on the exact same English word (and I'm glad to see we do agree about that). So your arguing that "ソウル" and "ソール" should logically be pronounced the same simply because they're based on English homophones... Well, it's doesn't quite add up.
(I "pick apart" your responses in order to address each point individually, and I'm not sure what's so wrong with that. Unless you're implying I'm twisting your words, in which case I'd really like you to explain when/where/how exactly.) Erigu (talk) 21:37, 21 May 2009 (UTC)
I am not saying that the two words should be pronounced the same just because the words are homophonous in English. I'm saying the two words should be pronounced the same because they are homophonous in the modern Japanese phonetic system. I don't know why the Hepburn system choses to use "ei" and "ee" instead of "ē" but "oo" and "ou" become "ō" and "uu" becomes "ū", but the homophonous nature of "ou" and "oo" and "o-" to be "ō" is pretty much set into the Japanese phonemic system. And I cannot find one instance other than the anecdotal instances here, where "ソウル" and "ソール" are pronounced differently. I am currently trying to find a native speaker to get his or her opinion on the matter, rather than those of us who have learned it in an English speaking culture.—Ryūlóng (竜龙) 21:47, 21 May 2009 (UTC)
I am not saying that the two words should be pronounced the same just because the words are homophonous in English. I'm saying the two words should be pronounced the same because they are homophonous in the modern Japanese phonetic system.
If your point was just that the two words should be pronounced the same in Japanese because they're homophonous in Japanese (tautology alert! ^^;), why bring up the fact they're homophones in English at all?
And you've been saying that they're homophonous in Japanese from the beginning, but I'm still not sure about that. Apparently, some outright disagree.
And again, the thing about "ou" standing for a long "o" sound is (often) true in words of Japanese origin, but you have yet to prove it's (always?) true for words of foreign origin. I explained above why I don't think it makes sense to bring up "ou" -> long "o" sound in this particular context. Erigu (talk) 21:56, 21 May 2009 (UTC)
Why wouldn't the words be homophonous if they're not Japanese in origin? If the phoneme is the same in native words, why do you believe the phoneme wouldn't be the same in gairaigo? Just because of the difference in how the word is written? I have other studiers of Japanese and native speakers who are telling me that the words are homophonous.—Ryūlóng (竜龙) 22:05, 21 May 2009 (UTC)
If the phoneme is the same in native words, why do you believe the phoneme wouldn't be the same in gairaigo?
I explained all that in my initial post. Erigu (talk) 22:28, 21 May 2009 (UTC)
Give me the Cliff's Notes.—Ryūlóng (竜龙) 22:33, 21 May 2009 (UTC)
Come on, it's not that long... And considering we've been debating this for a while, I'd appreciate it if you could take the time to actually read my reasoning eventually... Erigu (talk) 22:50, 21 May 2009 (UTC)

(outdent)コンピュータ vs コンピューター is a different kind of matter. it's a matter of notation, not of pronunciation. It is said that コンピュータ is started to be used by scientists especially when they type the word, using word processor/computer just because it's too much of a bother. So please put the matter aside. "ソウル" and "ソール" are not homophones in Japanese. As a native speaker, I totally agree with User:Erigu. Oda Mari (talk) 04:37, 22 May 2009 (UTC)

Well, that's strange because I've spoken with other native speakers who pronounce the two the same way so they are homophones.—Ryūlóng (竜龙) 05:00, 22 May 2009 (UTC)
コンピュータ vs コンピューター is a different kind of matter. it's a matter of notation, not of pronunciation.
I'd say it's both... Those two spellings aren't pronounced the same, right?
But if your point is that the "コンピュータ"/"コンピューター" difference (and the reason behind that difference) has nothing to do with the "ソウル"/"ソール" difference... Sure, I agree. I only brought the "コンピュータ"/"コンピューター" thing up in response to Ryulong's argument that "ソウル" and "ソール" had to be homophones in Japanese because "soul" and "sole" are homophones in English. It wasn't my intention to claim that both cases were extremely similar or anything like that. Erigu (talk) 05:11, 22 May 2009 (UTC)
To Ryulong. Ask the native speaker how does s/he pronounce Seoul/ソウル. Thank you.
To Erigu. As for "コンピュータ"/"コンピューター" , the pronunciation is the same. Oda Mari (talk) 05:21, 22 May 2009 (UTC)
Seriously?? 0_o;
Well, I guess you learn something new every day... Erigu (talk) 05:24, 22 May 2009 (UTC)
I was looking around for more about this "コンピュータ"/"コンピューター" thing (I'm still flabbergasted, sorry), and I happened to stumble upon this.
This part seems quite relevant to the debate at hand:
〔例〕エネルギー オーバーコート グループ ゲーム ショー テーブル パーティー ウェールズ(地) ポーランド(地) ローマ(地) ゲーテ(人) ニュートン(人)
注1 長音符号の代わりに母音字を添えて書く慣用もある。
〔例〕バレエ(舞踊) ミイラ
注2 「エー」「オー」と書かず,「エイ」「オウ」と書くような慣用のある場合は,それによる。
〔例〕エイト ペイント レイアウト スペイン(地) ケインズ(人) サラダボウル ボウリング(球技)
注3 英語の語末の-er,-or,-arなどに当たるものは,原則としてア列の長音とし長音符号「一」を用いて書き表す。ただし,慣用に応じて「一」を省くことができる。
〔例〕エレベーター ギター コンピューター マフラー エレベータ コンピュータ スリッパ
(I have to admit I hadn't thought of "バレエ" and "ミイラ"...)
So according to the second note, it is a long "e" sound, in "スペイン"? I'm... surprised (but I'm thinking I won't be alone in this). "ボウル" is listed, too.
While it doesn't say that "エイ" and "オウ" necessarily stand for long "e" and "o" sounds in words of foreign origins, it does call them long vowels in at least some cases... Huh. Erigu (talk) 05:50, 22 May 2009 (UTC)
You mean kana aren't perfect for rendering foreign words and that there's some confusion about how to pronounce them? Who woulda thunk it?
I think we can make this real easy-like: for words in katakana, always follow the spelling. If the katakana spelling オウ is used, we render it ou. If the katakana spelling オー is used, we render it as ō. End of story. Jpatokal (talk) 06:16, 22 May 2009 (UTC)
That's absolutely what I'd go for, personally. Erigu (talk) 08:00, 22 May 2009 (UTC)
But this is not what the revised Hepburn (or whatever the MOS uses) says should be used for "オウ". The current Hepburn says it should be "ō" just like "オー" would be.—Ryūlóng (竜龙) 04:22, 23 May 2009 (UTC)
You're just being dense, here...
According to revised Hepburn, long "o" sounds should be written as "ō". But is that really supposed to be a long "o" sound in "ソウル"? Erigu (talk) 04:48, 23 May 2009 (UTC)

But there is another exception above. See #7 Ryokū vs. Ryokuu. And please compare the pronunciation and the pitch accent of soul Seoul and insole. Soul→0:54 and 1:09. Seoul→0:02 and 1:33. Insole→0:24. Oda Mari (talk) 05:12, 23 May 2009 (UTC)

No exceptions. For words in katakana, always follow the spelling. Very, very easy.
Seriously, it's not our job to start examining pronunciations and pitch accents and trying to determine if something is a long vowel or two vowels. That's subjective WP:OR, varies between speakers, and will just lead to endless pointless debates. The kana, on the other hand, are clear and unambiguous. Jpatokal (talk) 07:29, 23 May 2009 (UTC)
Well, I'd say "for words of foreign origin" rather than just "for words in katakana".
But yeah, simply following the spelling sounds like the most sensible option to me. Macrons would only be used when there's no ambiguity (basically, when there are chōonpu). Erigu (talk) 08:18, 23 May 2009 (UTC)

Section break

"Ou" in nearly every single other instance (except for very very VERY few exceptions like "Inoue") is a long "o". That is what gives us the long o when it is kanji such as "王" and in the given name "一郎". This isn't me being dense. This is me being consistent with what is definitely known in all Japanese speakers. Just because the word isn't Japanese in origin does not mean that it is not pronounced as if it is. "Souru" is going to be read as "Sōru", even if it was written with a chōonpu or if it written in katakana to be "so-o-ru" instead of "so-u-ru" or "so-(chōonpu)-ru". I've honestly never run across this problem with anyone else in any Japanese setting. I've asked multiple other native speakers and non-native speakers, and they are all telling me things that you all here are not. If these were written in hiragana, I assume there wouldn't even be an argument. I'd be willing to bring this up to discussion with WP:JAPAN rather than here, because this is merely a manual of style and not at all indicative of full knowledge of Japan and its culture. From what I know and can read from current guidelines as well as the Hepburn dictionary's forward (where they use katakana) "ソウ [is written as] sō". You three (Erigu, Jpatokal, Oda Mari) are the only ones I have ever seen argue anything along these lines.—Ryūlóng (竜龙) 02:17, 24 May 2009 (UTC)

If these were written in hiragana, I assume there wouldn't even be an argument.
Katakana or hiragana, it doesn't matter. What matters is that we're talking about a word of foreign origin. With words of Japanese origin, you can tell fairly easily the difference between "ou" and "ō", between "o / u" and an actual long "o" sound. With words of foreign origin? Not so much.
From what I know and can read from current guidelines as well as the Hepburn dictionary's forward (where they use katakana) "ソウ [is written as] sō".
Yes, they use katakana. But they're not talking about words of foreign origin (hell, they even explain stuff like "a-fu becomes au or ō").
(besides, that's not revised Hepburn)
This has been going on for a while, and you apparently still don't really know what we're talking about... And when I suggest you try and read the initial post, all I get from you is "Give me the Cliff's Notes"...
If you're not interested in this debate, why take part in it? If you are, please take the time to read what the others have to say... Erigu (talk) 07:42, 24 May 2009 (UTC)
I spurred this debate and have been reading what has been going on. I still fail to see how just because a word is of foregin origin, it means that it does not follow the normal phonemic and phonetic rules of the language it is now in. So far, the arguments have not been convincing to me. And in words of Japanese origin, what normally happens with words with an "o" kana (ko, mo, so, do, ro, etc.) followed by the "u" kana, that it is parsed as a long O, resulting in the romanization to the ō. I am sure that if you see other words such as "藻類" or "走塁", you'd romanize them as "sōrui". Being written in katakana or being from a different language does not prevent it from being pronounced a different way. This is why everything pronounced in that way is listed on the Japanese Wikipedia under ソール. Apparently, the Japanese pronounce Sol, Sól, soul, sole, and Seoul the same way (and I know I do the same in English). And all of those ways involve the long O, which is parsed in the revised Hepburn of Wikipedia as ō.—Ryūlóng (竜龙) 07:59, 24 May 2009 (UTC)
Also, with one of the main dictionaries I utilize, WWWJDIC which has audio recordings of some words, the only difference between "ソール" and "ソウル" are the accent, but there is still a long O being spoken.—Ryūlóng (竜龙) 08:10, 24 May 2009 (UTC)
But there is another example for ou in Japanese word. 格子/lattice and 子牛/calf are both こうし in hiragana, but the pronunciation is different. 格子 is kōshi, 子牛 is koushi. Oda Mari (talk) 08:28, 24 May 2009 (UTC)
That's different because the kanji separate the o sound from the u sound in 子牛. The website I listed above pronounces the two differently. However, they still pronounce ソール and ソウル identically.—Ryūlóng (竜龙) 08:55, 24 May 2009 (UTC)
I am sure that if you see other words such as "藻類" or "走塁", you'd romanize them as "sōrui".
Because there's no ambiguity, there. Sometimes, there is. We've seen "緑雨" (りょくう) above, Oda Mari just produced a couple of other examples, and how would you romanize/pronounce stuff like "桃生" (ものう) or "黒河内岳" (くろごうちだけ)? At least, here, we have the kanji to try and understand what's going on and whether or not those are supposed to be long "o" sounds.
But how do you do that with words of foreign origin? Erigu (talk) 08:38, 24 May 2009 (UTC)
Those words are in the same family as "子牛" in that kanji exists to separate the pronunciation. But with ソール and ソウル, the pronunciation is identical. I haven't found one source that says that the latter is pronounced differently from the former. I know that you are insinuating that there is no way to discern this for words of foreign origin, but all you do is look at the word in the original language, and in this case, the two are pronounced identically.—Ryūlóng (竜龙) 08:55, 24 May 2009 (UTC)
And in the dictionary I linked, "ボウル" and "ボール" are pronounced identically, as well. I can't discern "bouru" from "booru", and neither should Hepburn romanization.—Ryūlóng (竜龙) 08:59, 24 May 2009 (UTC)
Those words are in the same family as "子牛" in that kanji exists to separate the pronunciation.
And yet, I believe that was actually a long "o" sound, in "黒河内岳"...
But yeah, the kanji do help. That was my point, even.
with ソール and ソウル, the pronunciation is identical.
Well, clearly, some disagree...
I know that you are insinuating that there is no way to discern this for words of foreign origin, but all you do is look at the word in the original language
The thing I quoted above said that it was really a long "e" sound, in "スペイン". Would you agree? Same thing for "エイト" or "ペイント"? Would you say it could just as well have been a chōonpu instead, and they only used "イ" because that's one way to indicate long "e" sounds in Japanese? No particular reason for that "イ" to be there, beyond that? Erigu (talk) 09:07, 24 May 2009 (UTC)
No, the "ou" in 黒河内岳 (Kurogouchidake) is not a long "o". That's why it would be written "Kurogouchidake" on Wikipedia. The kanji in the name are Kuro-go-uchi-dake. It is not a long "o" any more than the "ou" in the name of Narumi Kakinouchi. As for スペイン, it is not a long "e" sound. "ペイ" (not the long "e" sound) and "ペエ" (which would be the long "e" sound") are absolutely not pronounced the same, though they are very similar and someone who doesn't know how to discern the difference would have a hard time on first listen. The "エイ" is pronounced similar' to the "ay" in "hay", and the second is pronounced similar to the "e" in "met", only held for slightly longer. But that's neither here nor there when it comes to the long "o" sound. ···日本穣? · Talk to Nihonjoe 06:50, 26 May 2009 (UTC)
No, the "ou" in 黒河内岳 (Kurogouchidake) is not a long "o". That's why it would be written "Kurogouchidake" on Wikipedia. The kanji in the name are Kuro-go-uchi-dake.
And since when is "河" pronounced "go", exactly?
I may be wrong about this, but I believe this is just another case of "au" ending up being pronounced "ō" and the kana orthography eventually catching up with that pronunciation shift (a bit like "ikamu" eventually became "ikō", or "o hayaku" became "o hayō"): "Kurogawauchi" became "Kurochi" over time. And that would be why "河内" is written "Kōchi" on Wikipedia.
As for the "ei" thing... So you do pronounce the "i" in "sensei", and the guys over at the Japanese Wikipedia are just plain wrong? Hmm.
(and I do think it has much to do with the "ou" situation, actually) Erigu (talk) 11:21, 26 May 2009 (UTC)
Apparently since it's the name for a mountain. We aren't here to argue why they are using a particular kanji and pronouncing it in an uncommon way. They pronounce it that way, and that's that. And it's not "Kurogawauchi", as the 内 is said "uchi". You can't break up the kanji pronunciation the way you are trying to do here. The same goes for "Inoue": "ue" is the pronunciation of the final kanji in that name, so you can't go attaching half its pronunciation to something else. I've fixed Kouchi, Hiroshima to be correct per WP:MOS-JA. As for the pronunciation of "ei", please drop it as that's not part of this issue. It has nothing to do with anything we're talking about here. ···日本穣? · Talk to Nihonjoe 18:23, 26 May 2009 (UTC)
it's not "Kurogawauchi"
I was simply emphasizing which part of the name got its pronunciation altered over time. If you break it up according to the kanji, it's obviously "Kuro - gawa - uchi - dake", originally.
I've fixed Kouchi, Hiroshima to be correct per WP:MOS-JA.
Well, thanks for that: the article is now inaccurate... 'Guess I'll have to go and fix it back...
It is "Kōchi", it is a long "o" sound. It was originally "Kawauchi" ("Kawa - uchi"), but the pronunciation shifted to "Kōchi", and the kana spelling simply reflects that. See this and this, for example...
Sorry, but you simply don't know what you're talking about, here... Erigu (talk) 19:05, 26 May 2009 (UTC)
Please keep civil here. There is no reason to start making personal attacks, and your attitude (at least to me) appears to be one of "you are all wrong, and I'm the only one who is right". That is not a productive angle from which to approach discussions, especially since you obviously haven't been here since the beginning of discussions which resulted in WP:MOS-JA. The romanization is not necessarily a reflection of the pronunciation (though in most cases it is). It was determined quite a while ago that any romanization across kanji boundaries was not a long "o" sound; therefore it should be "Kouchi", not "Kōchi", because the first kanji is pronounced "Ko". Do not revert changes such as that without discussion again. ···日本穣? · Talk to Nihonjoe 23:47, 26 May 2009 (UTC)
Please keep civil here. There is no reason to start making personal attacks, and your attitude (at least to me) appears to be one of "you are all wrong, and I'm the only one who is right".
That sounds a lot like your own attitude so far. You're obviously working under the assumption that you're right and I'm wrong, and you apparently haven't been paying much attention to the explanations I've been giving you, here or in the initial discussion (the one you unilaterally closed and archived).
It was determined quite a while ago that any romanization across kanji boundaries was not a long "o" sound; therefore it should be "Kouchi", not "Kōchi", because the first kanji is pronounced "Ko".
Here, it is a long "o" sound, like I explained (twice). The first kanji isn't pronounced "ko".
Do not revert changes such as that without discussion again.
Please take the time to actually check your facts before editing articles like you just did. Erigu (talk) 00:24, 27 May 2009 (UTC)
How about レインボー? Sometimes it's レインボウ but never レーンボー nor レーンボウ. Sea lane is always シー・レーン. The pronunciations are distinctive. Oda Mari (talk) 09:42, 24 May 2009 (UTC)
You're wrong, Erigu. A long "e" is not the same as "ei", ever. The long "e" is said similar to the "e" in "met", but held slightly longer, and the "ei" is said similar to the "ay" in "hay". Not even close to the same sound. However, a long "o" (either "oo" or "ou" is always pronounced the same (please note that the "ou" in words such as "Inoue" is not a long "o", so that's not an issue here). ···日本穣? · Talk to Nihonjoe 15:40, 24 May 2009 (UTC)
A long "e" is not the same as "ei", ever. The long "e" is said similar to the "e" in "met", but held slightly longer, and the "ei" is said similar to the "ay" in "hay". Not even close to the same sound.
What was the thing I just quoted above even talking about, then?
And what about this:
Or this:
Things don't seem nearly as clear-cut as you're saying... Or do you actually pronounce the "i" in "sensei"?
please note that the "ou" in words such as "Inoue" is not a long "o", so that's not an issue here
That's just the crux of this debate ("ou" or long "o" sound?), but if you say "that's not an issue", OK, then. ^^; Erigu (talk) 22:09, 24 May 2009 (UTC)

Jesus H. Christ on a pogo stick. Why does the pronunciation matter? Especially in the case of artifically created words like "Soulcalibur", which you can't find in a dictionary, it's literally impossible for us to say how the creators intended it to be pronounced. The only solution possible without WP:OR is to transcribe the kana, one for one, into romaji. Ryulong, do you have a problem with this, and if yes, what is it? Jpatokal (talk) 00:34, 25 May 2009 (UTC)

Yes, there is a problem with that, because that would introduce a whole level of non-uniformity into the manual of style.
Quite the opposite: the rule "romanize as it is written" is extremely uniform and has no exceptions at all(*). Follow the katakana: オー is ō, オウ is ou. Jpatokal (talk) 08:56, 25 May 2009 (UTC)
(*) Well, almost. In the rare cases where katakana is used to write words normally written in hiragana/kanji, such as the scientific names of plants, the spelling rules of hiragana/kanji should be used instead. Eg. イチョウ (銀杏) is still ichō and not ichou.
Which is why I think it's important to have the distinction be between "words of Japanese origin" (that would include on'yomi, such as "イチョウ") and "words of foreign origin", rather than between "words in kanji and kana" and "words in katakana", like it is right now in the manual of style (as I mentioned already in another section above). Erigu (talk) 05:58, 26 May 2009 (UTC)
If it is clear that any long O sound, such as that in ソウル or 格子 or ボール is written in romaji as "ō", which is the current practice anyway. The exceptions that have been brought up where an o kana followed by u such as 子牛 or 井上 which clearly are not long O sounds are simply exceptions and should not be expected everywhere you get オ, コ, ゴ, ソ, ゾ, ト, ド, ホ, ボ, ポ, モ, ヨ, or ロ followed by an ウ. There are exceptions, and I understand that, but we should not change the MoS to favor those exceptions where they might not exist. From what we already know, "ソウル" is a transliteration of the English word "soul". The only other place I have seen the proposed "souru" is where the Hepburn system isn't used and that's over at Wiktionary (I'll have to see their Manual of Style, but that spelling has been there for a while). This Manual of Style currently does not treat it the way Wiktionary does. And I know that if it was changed, there'd be a hell of a lot of work to do on our end.—Ryūlóng (竜龙) 01:23, 25 May 2009 (UTC)
any long O sound, such as that in ソウル
Source? Erigu (talk) 02:44, 25 May 2009 (UTC)
How the heck do you pronounce the word "Soul"? "So ul" or like "Sole"? There's no source for that. It's common sense. I've even been speaking with a native speaker who says that in order for "ソウル" to have the O and U separated, that it'd have to have its roots in a word that would normally be pronounced as "so uru", and none exists. As I have been saying, you are practically the only person to suggest that the word is pronounced any differently between languages.—Ryūlóng (竜龙) 03:38, 25 May 2009 (UTC)
You are laboring under the mistaken assumption that Japanese spelling or pronunciation systematically correlate to English spelling or pronunciation. Tell me, how would you pronounce or romanize ミシン, which is the Japanese rendering of "(sewing) machine"? Jpatokal (talk) 08:56, 25 May 2009 (UTC)
That's "mishin". But that still does not mean anything.
But we should be romanizing it according to its English roots as "masheen", no? Or are you saying that mishin is a Japanese word and not necessarily correlated to its English roots anymore? Jpatokal (talk) 09:16, 25 May 2009 (UTC)
"Souru" still has a long O in it. And everyone else with knowledge of the Japanese language is telling me that it does, unlike you all here. ソウル should be romanized as sōru, because that is how it is pronounced in Japanese. There's no separation of ソ from ウ as Oda Mari and Erigu suggest exists. It is simply a long O that does not use a chōonpu.—Ryūlóng (竜龙) 09:01, 25 May 2009 (UTC)
I'm afraid your imaginary friends do not count as sources. Compare WWWJDIC's pronunciations of ソール and ソウル; they're quite clearly different. Jpatokal (talk) 09:16, 25 May 2009 (UTC)
There is no different pronunciation. One just has a different accent, but they still say "sōru". And I have users of the Japanese language Wikipedia saying that they are pronounced identically. Not "imaginary friends" as you are calling them.—Ryūlóng (竜龙) 09:21, 25 May 2009 (UTC)
Err, it's the same person reading both. Pay attention to the vowel: in ソー it's steady and long, in ソウ you can hear the transition from "o" to "u". (This would be quite obvious if you graphed them, but ripping out the waveforms from that Flash app appears to be nontrivial.) Jpatokal (talk) 09:42, 25 May 2009 (UTC)
Ryulong is correct: the pronunciation is the same, but the accent is different. He wasn't talking about an accent like a Southern accent, but rather how the emphasis is put on words. This is the same as 橋 (hashi = bridge) and 箸 (hashi = chopsticks). They are pronounced the same, but have slightly different accents placed on the syllables. ···日本穣? · Talk to Nihonjoe 06:33, 26 May 2009 (UTC)
Romanizing it as written tells nothing about the pronunciation of the word, but is instead a transliteration, and will only affect the extremely small instances of "oo", "ou", and "uu" in katakana, which if they were kanji or hiragana, they would be "ō" or "ū" without any exceptions. There should be no special rule just because "ソウル" was pointed out to be "sōru" instead of "souru" for Oda Mari and Erigu.—Ryūlóng (竜龙) 09:13, 25 May 2009 (UTC)
"Romanizing it as written tells nothing about the pronunciation of the word, but is instead a transliteration" ← yes, that would be exactly what I've been trying to say here. No special rules needed: transliterate the kana and let the reader worry about pronunciation. Jpatokal (talk) 09:16, 25 May 2009 (UTC)
That doesn't serve any purpose then.—Ryūlóng (竜龙) 09:21, 25 May 2009 (UTC)
On the contrary, it serves the purpose of providing the romanization for people who can't read kana or kanji. Jpatokal (talk) 09:42, 25 May 2009 (UTC)
I agree with Jpatokal. And I want to be bold and ignore Hepburn romanization and MoS ja with this matter. Oda Mari (talk) 14:49, 25 May 2009 (UTC)
Well, considering Hepburn romanization only tells us to write long "o" sounds with macrons and we're actually questioning whether or not this is a long "o" sound, we wouldn't even be ignoring it, just working around a loophole. Erigu (talk) 15:10, 25 May 2009 (UTC)
if they were kanji or hiragana, they would be "ō" or "ū" without any exceptions.
You and I both know there are "exceptions". Erigu (talk) 15:13, 25 May 2009 (UTC)
Alas, Mr.Hepburn didn't think so. But there are plenty of exceptions in any language. Languages are not that systematic. Oda Mari (talk) 15:56, 25 May 2009 (UTC)
I still disagree, and I would like to see someone other than the five of us (Erigu, Oda Mari, Jpatokal, Nihonjoe, and myself) add something. We all know our opinions by now.—Ryūlóng (竜龙) 21:02, 25 May 2009 (UTC)
There would also be Bendono, who was the first to reply. Erigu (talk) 00:36, 26 May 2009 (UTC)
This discussion still needs new blood.—Ryūlóng (竜龙) 01:46, 26 May 2009 (UTC)
I asked for comment/opinion at WP Japan. Oda Mari (talk) 02:01, 26 May 2009 (UTC)

Section Break #2

I saw Jpatokal made two changes to the article that Nihonjoe and I have removed. Only you three (Jpatokal, Erigu, Oda Mari) seem to believe that ソウル is not pronounced with a long O, whether it represents the word "Soul" or the name of the city "Seoul". As far as I can tell, you three are still in the minority, but are a vocal minority in this idea about the pronunciation of these three katakana in succession. Everything else I have been finding says that ou is a long O. The instances where "ou" are not a long O are rare and have been described in detail throughout this argument. Before any changes are made to the MOS, I would like input from other active members of WP:JAPAN.—Ryūlóng (竜龙) 01:08, 27 May 2009 (UTC)

Actually, it appears that at the moment you are in a minority of one (1), since Erigu and Oda Mari have both endorse the above proposal, and you're the only one who has explicitly opposed it.
More importantly, though, you are misrepresenting my view. I'm saying that the only sensible, non-WP:OR way out of this mess is to follow the katakana and romanize it as it is spelled. To illustrate why your approach is hopeless, let me repeat my earlier extremely basic question for you: how are we supposed to find out the correct pronunciation for a made-up word like "Soulcalibur", which is missing from all dictionaries and other reliable sources? Jpatokal (talk) 01:24, 27 May 2009 (UTC)
It's "SoulCalibur" which comes from "Soul" and "(Ex)Calibur" which are "ソウル" and "キャリバー". Both words exist in dictionaries. And from what I can tell from the dictionaries is that the words should be written in Hepburn romanization rules as "sōru" and "kyaribā". I have not seen any good argument to say that "ソウル" is not pronounced the same as "ソール", because that's what basic Japanese studying has told me. In the attack of my argument with the audio recording of the two words on the dictionary I put forward, the only difference was how the word was accented, and both recordings said "soh'ru". I've yet to see some definitive evidence that the word is pronounced "so-u-ru" because as far as I can tell, when you pronounce them anyway, the ウ still lengthens the pronunciation of the ソ part, which is why the phoneme produces the long O anyway, and why you have to type any of those O kana followed by a u to produce those sounds in an IME editor. I'm sure if you input "こうし" into one of those, it will bring up the word for "crystal" before the word for "baby cow" as was argued by Oda Mari. As far as I'm concerned, and from what I can tell from Nihonjoe's replies on this page, there is no consensus to change the MOS and there is certainly no consensus that "ソウル" should not be "sōru" and should be "souru".—Ryūlóng (竜龙) 01:36, 27 May 2009 (UTC)
I agree that it would be fair to assume the "soul" part would be pronounced like the word "ソウル" by itself... But the problem remains: while any of us can tell how we would be pronouncing that word, or how people around us are pronouncing that word, we've seen that there is disagreement on this particular matter. So the question shouldn't be "how is "ソウル" pronounced?", but "how should "ソウル" be pronounced?"
Now, if you can think of a way to answer that particular question beyond the shadow of a doubt, I'd be glad to hear it, but short of getting ahold of whoever decided the English word "soul" should be spelled as "ソウル" rather than simply as "ソール", and asking them why they did that and whether or not that implies a different pronunciation (maybe Ryulong and Nihonjoe would argue that the decision to use "オウ" over "オー" in some words of foreign origins is a purely random one, but based on my experience, I'd personally say there's a pattern, here), I'm not sure how you would want to do that.
... And of course, there would also be the matter of all the other words of foreign origin that would present that same ambiguity...
So, yes, I think simply "following the kana" is the sensible option when dealing with words of foreign origin, like Jpatokal suggests. And while I guess it may result in some intended long "o" sounds being transliterated as "ou", I don't think many of us have had a problem with revised Hepburn using "n" even when it's really pronounced more like "m", or keeping the spelling "ei" in all cases, even when a lot of Japanese people would pronounce that kana string as a long "e" sound... Erigu (talk) 01:53, 27 May 2009 (UTC)
To be fair, Nihonjoe also opposes that proposal. But to be even fairer, Bendono apparently agreed with it, so... Erigu (talk) 01:55, 27 May 2009 (UTC)
I've been very clear as to how "ソウル" should be pronounced. I have a hell of a lot of evidence that shows that it is pronounced identically to "ソール" just as it would if it were written as "ソオル" (but there's no word like that that I can find). I think that the spelling in katakana was an attempt to emulate the original English spelling. "Soul" became "Souru" and "Bowl" became "Bouru". How "ball" became "booru" was because there's no way to emulate the English pronunciation of that word, because the phoneme does not exist in Japanese. "Sole" became "Sooru" with the chōonpu because there was no other way to emulate the original English spelling. All of these words rhyme or are homophones. I know the choice between the use of the chōonpu and simply using an "ou" combination is at times random. I remember trying to romanize the title of Engine Sentai Go-onger before actual spellings in Latin letters came about. Some chose "Gouonger" from one of its base words 轟音 which is romanized outside of Hepburn as gouon, while I knew that it was using a chōonpu and tried to avoid using "Gooonger" because that would be how I would romanize it outside of Wikipedia. This means that it is pretty clear to some that "オウ" is homophonous to "オー" in almost every single instance of it being used. I can understand that there are instances where it isn't, but in "ソウル", it is not an exception and the spelling choice is simply one to emulate the source language.—Ryūlóng (竜龙) 02:11, 27 May 2009 (UTC)
I've been very clear as to how "ソウル" should be pronounced.
You've been very confident as to how "ソウル" should be pronounced.
I think that the spelling in katakana was an attempt to emulate the original English spelling.
So you and I both agree that wasn't an entirely random choice, then? Well, that's something. ^_^;
I know the choice between the use of the chōonpu and simply using an "ou" combination is at times random.
Well, I certainly wouldn't say I'm floored with the consistency of the whole thing, far from it... But I think there's a pattern, a tendency, at least. From what I've seen, when you see a word of English origin with "エイ", chances are you'll find a "i" or a "y" there in the original word. And when it's "オウ", you'll generally find a "ou" or a "ow" in the original English word.
Now, as you've said, the reason might simply be that they were trying to emulate the original English spelling. And maybe it really is just that, and that oddity doesn't mean anything as far as the pronunciation is concerned. But it sure would be nice to have a confirmation of that.
I remember trying to romanize the title of Engine Sentai Go-onger before actual spellings in Latin letters came about. Some chose "Gouonger" from one of its base words 轟音 which is romanized outside of Hepburn as gouon, while I knew that it was using a chōonpu and tried to avoid using "Gooonger" because that would be how I would romanize it outside of Wikipedia. This means that it is pretty clear to some that "オウ" is homophonous to "オー" in almost every single instance of it being used.
Keyword being "almost"? ^^;
And like you said, these people were only theorizing the "Gouon" spelling based on the (non-Hepburn) romanization of the Japanese word "轟音"... and there's no doubt we're dealing with a long "o" sound, in "轟音" (and no question "ou" is often used to indicate long "o" sounds in words of Japanese origin). Erigu (talk) 02:33, 27 May 2009 (UTC)
I thought I've been clear when it isn't homophonous, like where you have the words of Japanese origin. However, in these words of English origin, even when the u is used instead of the chōonpu, it's still clear that it's a long O. Why else would the Japanese Wikipedia have both choices on their disambiguation page?—Ryūlóng (竜龙) 02:43, 27 May 2009 (UTC)
in these words of English origin, even when the u is used instead of the chōonpu, it's still clear that it's a long O.
Not to me. And you simply stating it's clear doesn't make it any clearer, sorry. ^^;
Why else would the Japanese Wikipedia have both choices on their disambiguation page?
For the same reason the same page gives us two possible kana spellings for "Sól": "ソール" and "ソル"? There's an alternate kana spelling (and in the case of "ソール" for the English word "soul", a fairly uncommon one). That doesn't imply both kana spellings are pronounced the same. "ソール" and "ソル" most certainly aren't... Erigu (talk) 03:22, 27 May 2009 (UTC)
From what I can tell on that page, "ソール" and "ソル" are only alternate spellings (and pronunciations) of Sol and Sól, and not for the other words on that page. Again, you're just finding flaws that don't really exist in my arguments. For "soul(ソウル/ソール)" and "sole(ソール/ソウル)" they appear to use the same spellings for both words, but one is just more prevalent than the other. The fact that these are coupled together (ignoring everything else on the page) shows that they're homophones.—Ryūlóng (竜龙) 03:28, 27 May 2009 (UTC)
From what I can tell on that page, "ソール" and "ソル" are only alternate spellings (and pronunciations) of Sol and Sól, and not for the other words on that page.
I never said otherwise.
I was just saying that "ソウル" and "ソール" may very well both be listed next to "soul" for the same reason "ソール" and "ソル" are listed next to "Sól", i.e. they're alternate spellings. You're apparently arguing that, by listing "ソウル" and "ソール" next to "soul", the page is implying those two kana spellings are pronounced the same... but by the same logic, that would mean "ソール" and "ソル", the two kana spellings listed for "Sól" are homophonous, and I believe we can agree that's not the case. Therefore, your argument is flawed, and your assumption that the page implies that "ソウル" and "ソール" are homophones is a baseless one.
The fact that these are coupled together (ignoring everything else on the page) shows that they're homophones.
"Conveniently ignoring everything else on the page", you mean? ^^; Erigu (talk) 03:38, 27 May 2009 (UTC)
Stop twisting what I've been saying. "ソウル" and "ソール" are both alternate spellings of "Soul" and "Sole". One is just used more than the other for each word. This means that they are homophones of each other in both languages. I am not saying anything for how they have different spellings and pronunciations for the other words listed on that disambiguation page, because they obviously are pronounced differently because one uses a chōonpu and the other does not. But that has no merit on the difference between "ソウル" and "ソール". Sol and Sól are from Latin and Norse. The pronunciations are very likely to be different in pronounced in the original language and the subsequent Japanese or English. I know that the first word "Sol" I'd pronounce differently if it were in Sol Invictus or Sol as in the sun. I'm not sure how to pronounce Old Norse, but I'm sure that both of the pronunciations they have make sense. After all, it's a disambiguation page for that particular combination of katakana.
What is definitely clear about that disambiguation page their inclusion of "Soul" and "Sole" (and "Seoul") on the page. That means that they're homophonous with the title of the disambiguation page which is "ソール" and they only give the alternate readings of "ソウル" for those words.—Ryūlóng (竜龙) 03:55, 27 May 2009 (UTC)

Section Break #3

I have been away for a few days and find this wonderful mess. I can not hope to respond to it all, but here are a few general comments:

  1. The MOS:JP is a manual of style, not a guide on romanization.
  2. The MOS:JP is consistent on the relevant issues. Specifically, macron for long vowels, nothing for short vowels.
  3. This is a content dispute. Does ソウル have a short or long o? There has been much discussion but no resolution. Again, please refer to points #1 and #2 again.
  4. Why does it even matter? In the relevant article (Soulcalibur), just list the kana without any romanization. The English Soulcalibur, along with the JA interwiki link is sufficient. While it may often be helpful to romanize Japanese text for topics that sufficiently differ from the English terms, there are no obligations or requirements to always list it. I do not think that it is very useful or beneficial in this case. Especially considering that there is conflict, omission is neutral for all.
  5. I hope that I do not see this nominated one day for WP:LAME.

Regards, Bendono (talk) 03:48, 27 May 2009 (UTC)

I'll agree that this is a content dispute that started here, and was deferred here. Nihonjoe and I both argued that the manual of style dictated that the romanization of the katakana ソウル would be sōru, and it evolved into this. I'm not totally sure if omission is the best idea, regardless of conflict.—Ryūlóng (竜龙) 03:55, 27 May 2009 (UTC)
Sorry, I don't find points 1 and 2 very helpful. This is the discussion page for MOS:JP, and I think we can all agree that we have found a (minor) problem in the current rules: namely, the MOS:JP provides no guidance for determining whether certain character sequences are long vowels or not. The guidelines should be amended in some way to resolve this issue, the debate is over how that should be done.
I do agree that, for the specific case of Soulcalibur, re-romanizing the kana of an English word is fairly pointless, and it might be worth debating if the MOS:JP should be amended to recommend against such romanizations. But that's a separate issue, really, and unfortunately this problem is not limited to that one word alone. Jpatokal (talk) 02:53, 28 May 2009 (UTC)
There is no problem in determining whether certain character sequences are long vowels are not. "ソウル" is most definitely "sōru" from all of the evidence I've put forward in the previous sections, which includes linguistic analysis and a dictionary that has the pronunciations which are identical except for the inflection on the word.—Ryūlóng (竜龙) 03:11, 28 May 2009 (UTC)
If there was no problem, we would not be having this debate. But come, tell me: can you suggest a reliable, sourced, non-OR method for determining whether オウ is or is not a long vowel in any given word? Jpatokal (talk) 06:54, 28 May 2009 (UTC)
There is no problem in determining whether certain character sequences are long vowels are not.
No, no problem at all, as we've seen above with "Kōchi"... Erigu (talk) 06:59, 28 May 2009 (UTC)
The only instances where it is not is when the o and u are separated across two kanji (or a kanji and a kana). It is always a long vowel when the word is solely hiragana or katakana. I have yet to see one of these cases be brought up. That is why the Hepburn romanization system calls for the categorical lumping of "ou", "oo", and "o-" as "ō". Words like "だろう", "そう", "インチョウ", and "ソウル" all have the "ou" in them and are pronounced with long Os. Words like "王", "象", "狼", and "大" have "ou" or "oo" and are pronounced with a long O. Words like "追う", "負う", and "仔牛" have "ou" but these are separated very clearly and have been beaten like a dead horse in this discussion not to have long Os despite having the "ou". The argument that "ソウル" falls under this last set of words is itself one of "OR" as you have been referring to it. I've provided audio recordings of the word alongside a recording of "ソール". In my ears and in the ears of nearly everyone else except the group on this page (who still has not changed), the only difference in the recordings is the inflection on the word, and there is no discernable syllabic shift where the "ウ" is emphasized as being separate, as it is in words like "追う", "負う", and "仔牛" and the surname "井上". I cannot understand why you three/four cannot grasp this part of my argument and instead continue to say that I am wrong and there has to be another conclusion. Most of the instances where katakana is used over hiragana and kanji (anyway) are for the popular culture articles (musical performances, television programs, films, video games), and when they make up words that utilize katakana, there is at least a clear pronunciation in the source material. For the source of this dispute, the Japanese video game uses an American voice actor to read the title of the game in most instances (however, I recall that SCII has the characters read the name of the game, and switching the language settings on US game version might be able to shed some light).—Ryūlóng (竜龙) 07:19, 28 May 2009 (UTC)
Ryulong, seriously, could you please stop ranting and give focused responses? I asked you if you can suggest an "reliable, sourced, non-OR method" for determining vowel length, not "please repeat everything you've said earlier". Jpatokal (talk) 07:53, 28 May 2009 (UTC)
I've given online dictionaries and the Hepburn dictionary. Both of those deem how the words are pronounced, in addition to what I've been saying and expounding based on what these dictionaries say. The only way to determine whether or not something is a long vowel is based on how the Japanese pronounce it, which exists in the source material for popular culture terms, which is also which I brought up as these discrepancies will only appear in katakana.—Ryūlóng (竜龙) 08:12, 28 May 2009 (UTC)
An example is this video where I cannot discern any difference between the pronunciation of the long vowels in ゴールド and ソウル (at the 1 minute mark).—Ryūlóng (竜龙) 08:15, 28 May 2009 (UTC)
...and I have trouble distinguishing between some tones in Cantonese, but that doesn't mean they're not there. User:Bendono earlier noted that dictionaries like 日本国語大辞典 and 新明解日本語アクセント辞典 include transcriptions of the pronunciation, presumably as IPA or equivalent, which would be great -- alas, neither appears to be available online. Jpatokal (talk) 00:15, 29 May 2009 (UTC)
I'm not sure the manual should explain how to determine whether certain kana sequences are long vowels or not... As we've seen, there's a bunch of "odd" cases, so a summary will be incomplete anyway (and possibly misleading), and an exhaustive explanation would turn the whole thing into a Japanese lesson... Well, in my opinion, anyway.
I think the manual should assume the reader is knowledgeable enough to tell the difference by himself/herself and simply specify which romanization system is to be used on Wikipedia (again, a link to an exhaustive explanation of revised Hepburn would probably be a good idea) and how long vowels should be rendered (macron? double vowel? other?). ... Which is pretty much what it does in its current form, in fact.
(not that I don't have issues with that section of the manual, but once again, I didn't get much feedback...) Erigu (talk) 06:59, 28 May 2009 (UTC)

Section Break #4

So, I raised this question on sci.lang.japan, and got an answer from none other than Jim "WWWJDIC" Breen himself. Here's what he has to say:

ソウル and ソール are written differently because the writers are reflecting different pronunciations. To romanize them the same way is plain wrong.

And "Sean" chimes in:

If the disputants are native English speakers, perhaps the source of their confusion is the tendency in English (at least the Englishes I am familiar with) to diphthongalize /o/, so that many English speakers would end up pronouncing ソール as if it had been written ソウル. So, in your flame war, taunt the people who say that ソウル and ソール are the same thusly,"You idiots are speaking with a Canadian accent!"

I rest my case. Jpatokal (talk) 00:23, 29 May 2009 (UTC)

I consulted someone else and they also said that ソウル and ソール aren't pronounced identically. However, this is also a question about Romanization. So you are using that in your argument when native words such as 藻類 and 走塁 which are written as そうるい, which in Hepburn would both be sōrui which is pronounced like "ソウル" (as far as I can tell) but with the i at the end. Why don't you ask Mr. (Dr.?) Breen how he believes the word "ソウル" would be parsed in the Hepburn romanization system.—Ryūlóng (竜龙) 00:35, 29 May 2009 (UTC)
As you can see below (and on s.l.j), we now have Jim's answer. You might also want to read what the s.l.j thread above has to say about super-duper experts... Jpatokal (talk) 09:09, 30 May 2009 (UTC)
So you are using that in your argument when native words such as 藻類 and 走塁 which are written as そうるい, which in Hepburn would both be sōrui which is pronounced like "ソウル" (as far as I can tell) but with the i at the end.
Those are obviously long "o" sounds... Erigu (talk) 04:27, 29 May 2009 (UTC)
And they're written in the same way as ソウル(イ). Why wouldn't ソウル have a long O like 藻類 and 走塁 have?—Ryūlóng (竜龙) 04:34, 29 May 2009 (UTC)
... Just in case you aren't kidding with this, "オウ" isn't always a long "o" sound, and can be "o / u" instead. So just because you have two words with "ソウ", that doesn't necessarily mean the "ソウ" will be pronounced the same way. Erigu (talk) 05:08, 29 May 2009 (UTC)
Can you show me a word in Japanese where "そう" is "so)(u" and not "sou"? I can't think of any that aren't verbs like 添う or 沿う.—Ryūlóng (竜龙) 05:09, 29 May 2009 (UTC)
Off the top of my head, "細腕". Erigu (talk) 05:15, 29 May 2009 (UTC)
Okay, that's separated across kanji. However, in the pronunciation of that word, the "so)(u" is clearly annunciated. I don't hear it in the pronunciation of "ソウル". And now I hear the difference in the pronunciation of "ソール".—Ryūlóng (竜龙) 05:20, 29 May 2009 (UTC)
How about 木曽馬? The pronunciation of U is not as clear as 細腕 and as vague as soul. Oda Mari (talk) 14:47, 29 May 2009 (UTC)
My dictionaries are telling me that 木曽馬 is two words. "Kiso Uma".—Ryūlóng (竜龙) 22:54, 29 May 2009 (UTC)
Even though the origin of the word is two words, unlike Shetland pony, it is more like thoroughbred and considered and pronounced as one word. Do you think 民主主義, 脳外科, 白米, 弁当箱, etc. are also two words? Out of curiosity, Ryulong, exactly how did you ask your native Japanese friends about the pronunciation of ソウル and ソール? Oda Mari (talk) 04:34, 30 May 2009 (UTC)
For romanization, (compound) words such as 木曽馬 do not contain a long "o" as the "ou" is across kanji boundaries. "木曽" is "kiso" and "馬" is "uma". I don't know that it was ever specifically codified in the MOS-JA, but I know there was discussion in the past regarding this and it was unanimously (or very nearly so) decided that a long "o" is never across kanji boundaries for the purpose of romanization here. So, whether it's considered one word or two is irrelevant as it would be romanized "kisouma" or "kiso uma". ···日本穣? · Talk to Nihonjoe 19:53, 30 May 2009 (UTC)
For romanization, (compound) words such as 木曽馬 do not contain a long "o" as the "ou" is across kanji boundaries.
1) I don't believe anybody was arguing that there was a long "o" sound in "木曽馬"... ^^;
2) You say "for romanization" as if there were other contexts where this could be considered a long "o" sound?
3) The long "o" sound in "河内" still mocks your "kanji boundaries"! Seriously, it's not a long "o" sound in "木曽馬" because it's "kiso - uma" and that's it... Erigu (talk) 20:43, 30 May 2009 (UTC)
Please excuse me while I wipe all the sarcasm you're dripping everywhere...seriously, could you try to show at least a little respect toward others in this discussion? Your attitude throughout the entire discussion has been one of "I can't believe how stupid you all are." You've regularly mock the opinions of and information presented by others. Please stop.
As for 河内, there will ALWAYS be some exceptions due to local unusual pronunciations (and MOS-JA specifically allows for exceptions). You can see this page I, LordAmeth, and Fg2 translated a while ago for hundreds of examples. I only brought up 木曽馬 because people were discussing it for some reason. ···日本穣? · Talk to Nihonjoe 20:57, 30 May 2009 (UTC)
Your attitude throughout the entire discussion has been one of "I can't believe how stupid you all are."
If you say so...
for 河内, there will ALWAYS be some exceptions
It's the same kind of "exception" that gave us "arigatō", so it's not all that uncommon either...
I only brought up 木曽馬 because people were discussing it for some reason.
The reason is right there: Ryulong asked for examples of Japanese words with the kana string "ソウ" not pronounced with a long "o" sound, and Oda Mari offered "木曽馬" as an example. So when you then commented that it wasn't a long "o" sound, all I could think was "well... yeah, that's the point?" And when you explain that it's not a long "o" sound "for the purpose of romanization" (for what purpose would it be a long "o" sound?) and that this kind of thing is decided based on "kanji boundaries", that sounds like some weird logic to me.
But apparently, pointing that out is horribly rude and sarcastic. Erigu (talk) 21:35, 30 May 2009 (UTC)
I asked them if they were pronounced the same. They agreed, although some said there were some different inflections. Also, the first and last of the kanji sequences you gave are showing up in my dictionary as two words (minshu shugi, nōgeka, hakumai, and bentō bako).—Ryūlóng (竜龙) 04:38, 30 May 2009 (UTC)
Native Japanese think the words above as one word. It seems to me the concept of it is different between English and Japanese. And what dictionary are you using? Oda Mari (talk) 04:42, 30 May 2009 (UTC)
I use WWWJDIC and the J-Talk Kanji converter usually. The latter shows those as two words. The former didn't have the word for that horse breed. But this is tangential. From what I can tell, ソウル should still be sōru.—Ryūlóng (竜龙) 04:48, 30 May 2009 (UTC)
I'd romanize ソウル as "souru" and ソール as "sōru" in Hepburn. (And my name is Jim.) JimBreen (talk) 08:29, 30 May 2009 (UTC)
I have now checked this with two native speakers of Japanese (both teachers of Japanese as a foreign language) and a Japanese-speaking Korean. All agree that ソウル and ソール are not pronounced the same way, and that to romanize ソウル as "sōru" is an error as it obliterates that difference in pronunciation. JimBreen (talk) 01:11, 31 May 2009 (UTC)
The problem is that we can't base this on original research. Do you have a textbook or something similar which indicates this? None of the "I checked with a native speaker" information given by anyone here can be used for this. ···日本穣? · Talk to Nihonjoe 02:06, 31 May 2009 (UTC)
Again, you're mixing two things here. The pronunciation is debatable, and that why I keep saying we don't really need to get into this mess. What is clear is that according to Hepburn rules, ソウル should by default be romanized "souru": look up ソ, ウ, ル in any kana conversion table and that's what you'll get. Do you agree? Now, maybe there's an exception for this particular word... but that claim, that this word does not follow the normal Hepburn rules, is the one that requires proof. Jpatokal (talk) 00:53, 1 June 2009 (UTC)
This page is written about 十パーセント. The pronunciation is jyū and ji and both are accepted. But when writing, it should be じっ. Some pronounce ソウル and ソール differently and some don't. But as the notation in Japanese is different, romanization should be different too, as Japatokal and Erigu wrote above. Oda Mari (talk) 05:17, 31 May 2009 (UTC)

Purpose of Romanization on Wiki

Interesting discussion. I'm not sure how, but this ended up on my watchlist some time ago, and lo and behold it's been going crazy lately! Nice work guys :) One thing that occurs to me though... even supposing the pronounciation IS the same, between ソウル and ソール, would that still mean they should be written the same? I saw the following comments above:

  • Romanizing it as written tells nothing about the pronunciation of the word, but is instead a transliteration
  • The only solution possible without WP:OR is to transcribe the kana, one for one, into romaji

So it sounds like there is an unresolved question wanting to be asked here, which is: What is the purpose of the Romanization system here? Is it to help English speakers pronounce Japanese words like Japanese people do in real life? Or is it to help English speakers know the original syllables with which the Japanese word is written? Which is more important here? Joren (talk) 04:05, 31 May 2009 (UTC)

I would say the first item noted in your short list (which corresponds to the second one noted in your last paragraph). If you don't know Japanese, romanization isn't going to help you pronounce the word anyway as there is some small amount of training which is required to learn pronunciation and such. ···日本穣? · Talk to Nihonjoe 04:20, 31 May 2009 (UTC)
I think there's a bit of schizophrenia at work, here... If the article is to be believed (and again, more sources wouldn't hurt), traditional Hepburn had the "n" turn into "m" to reflect the pronunciation shift in some cases, but that was dropped by revised Hepburn... and on the other hand, revised Hepburn kept "ei" in all cases, whereas modified Hepburn only kept it when both "e" and "i" could be heard individually.
So I don't think there's a clear answer to that question. Sometimes, Hepburn romanization focuses on actual pronunciation to make it easier for foreigners ("fu" instead of "hu"), but at other times, it goes "ah, well! that's not really how it would be pronounced, but let's be consistent with the Japanese spelling, here"... Well, maybe the former was the original intent of traditional Hepburn, but... Erigu (talk) 07:39, 31 May 2009 (UTC)
The primary purpose of Wikipedia's romanization system, I think, is to minimize the number of revert wars concerning variant romanizations. By that metric, I'd say it's doing a very good job: sure, once or twice a year we have — as Bendono so delightfully puts it — a "wonderful mess" like this one, but it's almost entirely confined to the WP:MOS-JA talk page, not scattered incoherently across ten thousand discrete pages.  — Aponar Kestrel (talk) 21:53, 1 June 2009 (UTC)


I just saw Erigu perform the change from "Sōru" to "Souru" on the Soulcalibur pages and I've undone it because we never really came to any sort of sufficient conclusion here. I understand that the pronunciation of ソウル and ソール are not identical, but we never finalized how the two words (or simply kana combinations for any o kana followed by a u kana or any o kana followed by the choonpu) should be written in the Hepburn romanization system. There's been nothing to prove that the ソ and ウ are pronounced separately such as when it is in the unconjugated verb form like in 添う or mildly dipthong...ized as the sound exists in words such as 想 and 僧. I am unconvinced that the former case is what is happening with the words, and the choice of spelling in katakana is (or was) a stylistic one and should not be treated any differently just because the word exists in Japanese solely as a loan word in katakana.—Ryūlóng (竜龙) 03:43, 15 June 2009 (UTC)

Oh, for Pete's sake...[6][7][8][9][10][11]
"Nothing was ever sufficiently determined about this"? Doesn't that go both ways? Besides, we've discussed this extensively, and only Nihonjoe and you were in disagreement... Erigu (talk) 03:41, 15 June 2009 (UTC)
Yes it does. That means there is still no consensus to do either thing. I've also moved your comment down in the discussion for chronological stuff.—Ryūlóng (竜龙) 04:23, 15 June 2009 (UTC)
There's been nothing to prove that the ソ and ウ are pronounced separately
And there's been nothing to prove that they're supposed to be pronounced "sō". In fact, you said just above "I understand that the pronunciation of ソウル and ソール are not identical". Erigu (talk) 03:47, 15 June 2009 (UTC)
"Sō" is not a pronunciation. It's a romanization. And while it is not a romanization, it is still dependent on pronunciation. That fact is evident as to how there are words that should be romanized a particular way because of the pronunciation and etymology. In this case, I'm arguing that 想 is homophonous with the ソウ in ソウル. Both use the same moras. And I'd like someone to say something else before you continue to reply to me as you have.—Ryūlóng (竜龙) 04:19, 15 June 2009 (UTC)
"Sō" is not a pronunciation. It's a romanization.
It's the revised Hepburn romanization of the long "so" sound. So what I meant by "pronounced "sō"" should be quite obvious, really. Erigu (talk) 04:38, 15 June 2009 (UTC)
Very well then. Still, there's been no finalized consensus here on what to do with the word "Soul".—Ryūlóng (竜龙) 04:49, 15 June 2009 (UTC)
'Guess they're all tired of discussing this (and I can't blame them).
So we have no consensus, but we need a romaji spelling. And since we apparently all agreed "souru" was the sensible solution except for Nihonjoe and you... Erigu (talk) 04:00, 22 June 2009 (UTC)
Because Nihonjoe and I feel that it is wrong. I've yet to see something to convince me otherwise. The "Sou" in "Souru" is the same as the "Sou" in various Japanese words, both of which should be "sō" in the Hepburn system that this Wikipedia uses.—Ryūlóng (竜龙) 04:53, 22 June 2009 (UTC)
Because Nihonjoe and I feel that it is wrong.
Thanks for clarifying that you disagree because you don't agree. The fact remains that you're clearly the minority, so, like I said above...
The "Sou" in "Souru" is the same as the "Sou" in various Japanese words, both of which should be "sō" in the Hepburn system that this Wikipedia uses.
Why, it's always a pleasure to go back to square one.
Didn't you say just above "I understand that the pronunciation of ソウル and ソール are not identical"?
Do you agree that the chōonpu in "ソール" indicates a long sound?
Do you know that the macrons used in revised Hepburn romanization indicate long sounds ("From Greek μακρόv (makrón) meaning "long"")?
So, if "ソウル" isn't pronounced with a long sound... why should we use a macron? Erigu (talk) 06:38, 22 June 2009 (UTC)
No one has proven that last point of yours. We've only had someone say that ソウル and ソール are not identical, which I agree with now. However, that does not mean that ソウル is not pronounced the same as words that are written with the moras of "そうら", "そうり", "そうる", "そうれ", or "そうろ". Being that "ソウ" and "そう" are pronounced identically, and the latter would always be written as "sō" in Hepburn. The fact that the item is a loan word from another language and is written in katakana in a particular way does not mean it is parsed in Hepburn any differently.
There is no emphasis on the U in "Souru". I've shown an instance of this earlier in the thread in a video link (somewhere). Therefore, it does not follow the rules that various words in Japanese that do have one of the O-voweled kana followed by a U kana pronounced separately, solely due to the fact that the word is comprised of two kanji where the O is at the end of one and the U at the beginning of the other. It instead follows the rules where the two kana are in the same kanji, such as 総理, 壮麗, 草案, 創意, 相応, 騒音, 草花, 総轄, etc.—Ryūlóng (竜龙) 08:36, 22 June 2009 (UTC)
No one has proven that last point of yours. We've only had someone say that ソウル and ソール are not identical, which I agree with now.
---> Do you agree that the chōonpu in "ソール" indicates a long sound? <---
If so (and I sure hope so), and if you also agree that "ソウル" and "ソール" aren't pronounced the same... I mean, come on. Work with me, here.
Being that "ソウ" and "そう" are pronounced identically
The pronunciation of "ソウ"/"そう" "kinda" depends on the context. And that's something you should really, really, really know at this point of the discussion. Erigu (talk) 08:55, 22 June 2009 (UTC)

Oh, fer chrissake. We've had a professor of Japanese at the Tokyo University of Foreign Studies, a member of the board of the Japanese Studies Centre and the guy who compiled WWWJDIC, unequivocally state that ソウル should be romanized "souru"... and you're telling us we should ignore him and trust your ja-1 original research instead?

And I'll repeat my last comment, because you never replied to it:

What is clear is that according to Hepburn rules, ソウル should by default be romanized "souru": look up ソ, ウ, ル in any kana conversion table and that's what you'll get. Do you agree? Now, maybe there's an exception for this particular word... but that claim, that this word does not follow the normal Hepburn rules, is the one that requires proof. Jpatokal (talk) 08:49, 22 June 2009 (UTC)

I've seen the word freaking pronounced. It's not "so-u-ru". And the "exception" would be that "ソウル" is to be romanized as "souru". The current rules state "For transliterations from kanji and kana, long o and u are written with macrons as ō and ū respectively." No one has succinctly proved that the word does not contain a long o simply because it is written as "ソウ" and not "ソー" or "そう".—Ryūlóng (竜龙) 09:02, 22 June 2009 (UTC)
And this bullshit over my "ja-1" thing is getting tiresome. I only keep it at that because I do not know the verb rules. I can identify kanji, I can identify other words, I just don't know how to chain them into a full sentence.—Ryūlóng (竜龙) 09:03, 22 June 2009 (UTC)
And this bullshit over my "ja-1" thing is getting tiresome. I only keep it at that because I do not know the verb rules.
Well, that's... Er... Are you sure you're really qualified to keep this discussion going? Erigu (talk) 09:07, 22 June 2009 (UTC)
What the hell are you talking about? I don't know how to conjugate Japanese verbs. That does not mean I don't know how to read (apart from understand) the language. And that has no bearing on how I know how a word is pronounced, particularly when it is in one of the simplest writing systems of the language.—Ryūlóng (竜龙) 09:12, 22 June 2009 (UTC)
Well, I suspected you didn't know all that much about the language based on some of your comments so far, but... Anyway, that explains a lot (not your strange confidence though).
(and you just changed your "ja-1" to "ja-2"? seriously?) Erigu (talk) 09:24, 22 June 2009 (UTC)
So you are telling us that we should ignore the professor of Japanese and trust your ja-1 original research instead?
And for the umpteenth time, this is the discussion page for the MOS-JA, and I'm proposing that the wording be clarified to note that the kana sequence ソウ is romanized "ou", whether or not it's a long vowel. Even you have now admitted that ソウル and ソール are not pronounced the same -- why do you want to erase this distinction in the romanization? Jpatokal (talk) 09:25, 22 June 2009 (UTC)
I'm proposing that the wording be clarified to note that the kana sequence ソウ is romanized "ou"
I realize I'm repeating myself (then again, at this point... heh) but I really think the distinction should be made between words of Japanese origin (on'yomi included) and words of foreign origin, rather than between words in hiragana and words in katakana. Or we'd have stuff like チョウゲンボウ romanized as "chougenbou". Erigu (talk) 09:34, 22 June 2009 (UTC)
I can't seem to make any headway in this. I give up. Modify the manual of style to say that katakana readings of foreign words are romanized as a direct transliteration of the katakana, which apparently only applies to "souru" and "bouru" (if there are any pages about bowling in Japan on the English Wikipedia). I'll perform the changes on the articles I'm aware of once the MoS is changed.—Ryūlóng (竜龙) 09:36, 22 June 2009 (UTC)
Modify the manual of style to say that katakana readings of foreign words are romanized as a direct transliteration of the katakana
I don't think there's any reason to mention katakana specifically. "Kana" will do. Erigu (talk) 09:39, 22 June 2009 (UTC)
How often would hiragana ever be used to write out words of foreign origin?—Ryūlóng (竜龙) 09:41, 22 June 2009 (UTC)
It happens. That's reason enough. Erigu (talk) 09:43, 22 June 2009 (UTC)
As in?—Ryūlóng (竜龙) 09:45, 22 June 2009 (UTC)
Off the top of my head, Magical Tarurūto-kun. Erigu (talk) 09:47, 22 June 2009 (UTC)
Well that is a rarity. But it's not really affected by this change. It even has a chōonpu in the made-up word portion of the title.—Ryūlóng (竜龙) 10:02, 22 June 2009 (UTC)
Well that is a rarity.
it's not really affected by this change.
I didn't say it was. You asked for an example of a foreign word written in hiragana.
It even has a chōonpu in the made-up word portion of the title.
I'm not sure how that has anything to do with anything, really... Erigu (talk) 10:11, 22 June 2009 (UTC)
Why must you always do that to comments?—Ryūlóng (竜龙) 10:13, 22 June 2009 (UTC)
Did I twist your words? Erigu (talk) 10:15, 22 June 2009 (UTC)
Haha, I have to agree with Ryulong here. We can see what's written right above your comments, so you really don't have to quote parts of them every time you reply. It's kind of distracting. Dekimasuよ! 10:17, 22 June 2009 (UTC)
Well, that's... pretty much the opposite of what I'm going for. Sorry? Erigu (talk) 10:26, 22 June 2009 (UTC)
たばこ (sometimes katakana, sometimes kanji)、てんぷら (or 天ぷら)、ramen or かすてら (usually katakana, but not always)... there are varying degrees. I'm not agreeing or disagreeing with anyone here. Dekimasuよ! 09:57, 22 June 2009 (UTC)
Those seem to be rarities. Tempura is a word that is entirely in kanji most of the time, and ramen is always in katakana and written with the chōonpu.—Ryūlóng (竜龙) 10:02, 22 June 2009 (UTC)
I didn't know I was supposed to be avoiding rarities. But it's actually less common to see tempura in kanji than in hiragana (or 天+hiragana), regardless of the level of restaurant. And like I said, not always when it comes to things like ramen. Lots of people do it, although it is probably a stylistic choice. [12] Dekimasuよ! 10:09, 22 June 2009 (UTC)

Infobox City Japan vs. Infobox Settlement

I thought I would inject a different debate. I was adding a few infoboxes to Japanese cities, when I noticed some problems with the presentation (in IE 6.0 to be specific). So I looked at the talk page of the template to find an answer. While there, someone raised the question why we need the infobox for Japanese cities. Why don't we use the generic Infobox Settlement?


  • Standarization (no need to maintain a separate temple)
  • Solves the display issues
  • Support for multiple symbols. (In this case the emblem and the flag.) And there are a number of other features that this infobox supports


  • We lose the address for the govenment office. One could argue whether we really need that in the infobox, but it is a nice feature.
  • The settlement infobox is a rather large template. It is more complex to use.

As a means of comparrison, I used the infobox settlement on my sandbox page. Compare that with the infobox at Bihoro, Hokkaidō.

imars (talk) 12:32, 29 May 2009 (UTC)

We could use the "blank_name_sec1", etc., to enter the address and anything else missing. Looks like there can be two additional sections, and the template could always be modified to add more. ···日本穣? · Talk to Nihonjoe 20:35, 29 May 2009 (UTC)

Nihongo2 (for marking up Japanese script)

I've moved the following from Wikipedia talk:WikiProject Japan; it seems better placed here. -- Hoary (talk) 05:17, 1 June 2009 (UTC)

I had a nasty surprise today while attempting to remove gratuitous links to Help:Japanese from Tankōbon (an article that's less about tankōbon than about manga). Converting from Nihongo to Nihongo2 did nothing.

Template:Nihongo2 has recently been -- to put my opinion as politely as possible -- dehanced. Yes, a little question-mark link to Help:Japanese has been added to it. The effect on a lot of articles I've edited is grotesque. Unfortunately it wasn't on my watchlist, so I didn't notice the change, and the lack of discussion at Template talk:Nihongo2 suggests to me that it wasn't on the watchlist of more than a handful of other people either.

I haven't dared look at the other Nihongo templates. Perhaps they too have been similarly ... dehanced. Perhaps the only way to prevent Japanese text from being dehanced is to leave it untemplated (as here), but this defeats the original purpose of Nihongo2: to add <span class="t_nihongo_kanji"><span lang="ja" xml:lang="ja"> for the benefit of any software that may use it.

I'm inclined just to revert, but in WP one should never go for a simple fix when a laborious argument can be started instead, so here goes. Your comments either here or there, please. -- Hoary (talk) 02:05, 1 June 2009 (UTC)

I seem to recall strenuously opposing that hideous little question mark some time ago during what I believe was a vote on the matter, but I can't remember when and where and, as you can see, I failed to convince enough other editors. Exploding Boy (talk) 02:10, 1 June 2009 (UTC)
Yes, I hazily remember that. Perhaps the proposal was that WP's readers (presumed to be dimwitted, incurious, inobservant, or some combination thereof) needed more help; why a discreet little question mark when a bloody obvious "Help" or similar could appear instead? ¶ I'm not so happy about the question mark but I raise no objection to a single, even conspicuous appearance in a given article. Bad:
Natsume Kinnosuke (夏目金之助?), better known by his pen name Natsume Sōseki (夏目 漱石?, February 9, 1867 – December 9, 1916) [blah blah blah]
Better: One link to Help:Japanese, thanks to judicious use of a Nihongo2 template that does not link to it. Which is impossible now that Nihongo2 has been pimped with a link. -- Hoary (talk) 02:26, 1 June 2009 (UTC)
Earlier discussions: here, here, here, and here (chronological order). If anyone would like to move this new discussion to talk:MoS (ja), please feel free to do so. Wherever the new discussion is, it's surely better in one place (and it's already in two). -- Hoary (talk) 03:24, 1 June 2009 (UTC)
I have created {{nihongo4}} as a temporary fix for an article. Unfortunately, as the template uses {{nihongo2}}, the intended effect of suppressing the question mark doesn't work any more. I think these templates should all be consolidated into one template, with the default behaviour, i.e. without any parameters specified (except for the displayed text), showing just the Japanese text, without any question marks etc.  Cs32en  03:36, 1 June 2009 (UTC)
That may be a good idea for the medium or long term. But even if it's welcomed and implemented, nihongo2 is going to survive by the thousand for the short and even medium term. I suggest that we limit this discussion to nihongo2. As soon as it's settled one way or another, do please bring up the notion of a single supertemplate. -- Hoary (talk) 03:41, 1 June 2009 (UTC)

Comments welcome; please add them below. -- Hoary (talk) 05:17, 1 June 2009 (UTC)

I fixed Template:Nihongo2 back to how it's supposed to be (without the question mark) and I deleted Template:Nihongo4 as a duplicate. ···日本穣? · Talk to Nihonjoe 06:02, 1 June 2009 (UTC)
Well done, Joe, and most appropriately signed! -- Hoary (talk) 08:22, 1 June 2009 (UTC)

I would be happy to get rid of all the question marks. Don't all major operating systems have Japanese fonts by default these days? --Apoc2400 (talk) 08:41, 1 June 2009 (UTC)

Yes, I think that both Windows and OS X install Japanese fonts by default. So does Ubuntu 8.04 (and I presume so do its successors). So that's most people catered for. But I'd be surprised if WinCE (or whatever it's called these days), Symbian and so forth do this. Anyway, I think one link to a help file can be, well, helpful; and the question mark doesn't seem a bad way of doing this. -- Hoary (talk) 09:07, 1 June 2009 (UTC)
But Help:Japanese is about more than font support - it's also about pronunciation and other things. Though I would welcome a workable alternative to the question mark. Unfortunately, the last discussion on the subject died right when it looked like changes were going to be made. --Eruhildo (talk) 20:30, 3 June 2009 (UTC)
Perhaps it would be good to create a new help page specifically about font support. ···日本穣? · Talk to Nihonjoe 22:48, 3 June 2009 (UTC)
Yes, if some other kind person were to do this: I shan't have time to be of much help. -- Hoary (talk) 00:30, 4 June 2009 (UTC)
I forgot to mention it here, but I created Help:Installing Japanese character sets based on the content that was formerly at Help:Japanese. ···日本穣? · 投稿 · Talk to Nihonjoe 20:04, 22 July 2009 (UTC)

Draft template Nihongo core

I've created a draft template {{Nihongo core}} that might be useful to consolidate the existing nihongo templates (i.e. they would be short-cuts to specific styles of {{Nihongo core}} after consolidation). The version for long titles is not yet worked into the template. Please let me know what you think about it.  Cs32en  03:00, 4 June 2009 (UTC)

→ The discussion on {{Nihongo core}} has been moved to the Talk page of the template.

Trouble brewing at Hiragana

Please come participate in the discussion at Talk:Hiragana#Table and wording. An anon IP and a new editor are fighting over the presentation of information in the article. Thanks! ···日本穣? · 投稿 · Talk to Nihonjoe 23:23, 6 July 2009

Naka Castle?

I have proposed that the article Naka Castle be moved to either Nakagusuku Castle, Nakagusuku (castle) or something else that preserves the placename Nakagusuku. I invite your comments at Talk:Naka_Castle#Requested move. LordAmeth (talk) 14:10, 25 July 2009 (UTC)

Preserving original formatting of Roman characters in titles of Japanese media

Regarding the section on "Titles of books and other media", I understand that the policy is to follow conventional English formatting guidelines when displaying titles, even if the original title uses Roman characters in an unconventional way (for example, using all caps or all lowercase). While this is sensible, I am concerned that by not displaying the original formatting at all, we are losing information about the works in question, and the specific formatting used may be relevant to understanding the author's intention behind writing the title in a given way. As such, I think it makes sense to treat titles that use unconventional formatting of Roman characters as though they were titles originally in Japanese. One would then not omit the original formatting altogether, but only use it once to indicate the original formatting, in the same way one would write the original title if it were in Japanese. The {{Nihongo}} tag could suit this purpose, though the '?' help link seems a bit out of place. An example of this would be the title of Buono!'s seventh single, originally written "MY BOY". One could display it as

  • "My Boy" (MY BOY?)

in places where displaying the original Japanese spelling of a title would be appropriate (the header of an article or a track listing, for example), and write "My Boy" elsewhere. I have seen this used in a few places, but for most cases in which Roman characters are used unconventionally, the original formatting has been omitted (sometimes after editing to conform to capitalization guidelines), which makes for a slightly less informative article. We preserve the original formatting of Japanese titles with regard to whether they use kanji or hiragana or katakana when we display them in Japanese, so why not preserve the original use of Roman characters as well? Kirarin☆Snow ☃ 02:39, 22 July 2009 (UTC)

Because typographical gimmickry (or, if you prefer, results of idiosyncratic views of the significance of capitalization) such as this are endemic in Japan and common elsewhere too. We distinguish between hiragana and katakana because neither SCREAMS for attention over the other, and where appropriate we also distinguish between traditionally complex and recently simplified hanzi/kanji, because such distinctions don't give anyone a typographic megaphone, because they help in database searches and so forth, and because they don't give a green light to people outside Japan wanting more attention for themselves. If "My Boy" had been written in red, would you want {{Nihongo|"My Boy"|<span style="color:#ff0000">My Boy</span>}}? Shall we distinguish between hankaku and zenkaku lettering? How about double underlines, smiley marks and the like? But all that's a bit hypothetical; rather, let's stick to the actual example: What might the full capitalization of "MY BOY" tell anyone about the song? -- Hoary (talk) 02:55, 22 July 2009 (UTC)
It seems to me that idiosyncratic uses of capitalization are generally consistent throughout a work (e.g. front of CD disc, album cover, spine, etc.), while things such as color or typeface aren't. I was also under the impression that differences in capitalization were preserved in databases, but a quick search through the JASRAC database suggests that all Roman character titles are shown in all caps, regardless of the original capitalization, so I'm not so sure anymore. Still, the representation in a database doesn't necessarily reflect the author's intentions. As for the "MY BOY" example, to me the choice of all caps lends the title a more emphatic but serious flavor (without the mildly comic effect of, say, an exclamation point: "My Boy!"), which is appropriate to the tone of the song. And supposing the title were "My Boy!", should we omit the exclamation point? What about "My! Boy"? I've seen titles with unconventional placement of exclamation points, so it's not clear what differentiates idiosyncratic punctuation from idiosyncratic capitalization with regard to this policy (if indeed there is a difference).
To go with another example, Buono!'s sixth single is originally written "co・no・mi・chi". What is the proper way to represent this title in Wikipedia? "Cono Michi"? And should we not even mention the original punctuation and spelling altogether? In this case, the unusual spelling of a Japanese title suggests that there could be some purpose other than typographical decoration (I'm not saying there is one, but I don't think that's a question for Wikipedia editors to decide).
You do raise a good point with the hankaku/zenkaku question. Leaving aside the question of half-width and full-width forms of Roman letters (they seem, at least from my limited experience, to some extent interchangeable in official representations of titles, so perhaps this would be along the lines of color or typeface), how should we treat typographical flourishes in kana? A title could distinguish between hankaku and zenkaku forms, or use small versions of kana when regular-size versions are conventional (for example, S/mileage's single "ぁまのじゃく", which has a small 'ぁ' instead of the conventional full-size 'あ'). Do we leave the title the way it was originally written, or display it in a more conventional form (as "あまのじゃく")? And if we do we leave it as "ぁまのじゃく", what makes this different from capitalization of Roman characters? Kirarin☆Snow ☃ (talk) 04:18, 22 July 2009 (UTC)
It's fine to use the formatting (other than colors, as mentioned by Hoary) used in the original when the title is first mentioned at the beginning of the opening paragraph. The article title, infobox, and any subsequent uses should be using standard capitalization and formatting. As for your question about punctuation, if the title has an exclamation point, question mark, or other punctuation, it's fine to use that in the title wherever it appears. ···日本穣? · 投稿 · Talk to Nihonjoe 06:26, 22 July 2009 (UTC)

Preserve special characters in album title?

Singer Ayumi Hamasaki has an album properly titled A Song for ××. It's (commonly) mistakenly written A Song for XX, and currently, Hamasaki's article writes it the latter way. Without getting too much into the meaning of the title, the x-like characters, batsu, can mean "fill in the blank", and this is the meaning the singer intended (they're not supposed to be x's.) (In fact, the title is pronounced "a song for", not "a song for ex ex" or "a song for batsu batsu".) The MoS-Ja advocates the standardization of titles according to proper English formatting, but I'm not sure if this title falls under its guidelines. So: A Song for ×× or A Song for XX? Ink Runner (talk) 19:49, 22 July 2009 (UTC)

Considering we have this article, I guess using "×" should be OK? Besides, that's not the "X" letter... Erigu (talk) 05:20, 30 July 2009 (UTC)
Only if the appropriate redirects are created. ···日本穣? · 投稿 · Talk to Nihonjoe 06:50, 30 July 2009 (UTC)
Yet we have the article at XxxHolic for a manga titled ×××HOLiC. Akata (talk) 09:19, 14 August 2009 (UTC)
I moved the page back too a A Song for XX. Per, MOS:TM: Avoid using special characters that are not pronounced. I guess in the article it can be mentioned that the album title is pronounced as A Song for. MS (Talk|Contributions) 12:07, 14 August 2009 (UTC)
Are album titles considered trademarks, though? (This may be a stupid question, but law has never been my forte.) Ink Runner (talk) 18:50, 14 August 2009 (UTC)
Not usually as multiple artists/groups can have an album with the same title. ···日本穣? · 投稿 · Talk to Nihonjoe 17:16, 20 August 2009 (UTC)

Naming conventions

Shrines in Japan come in a Myriad of formalities that distinguish one from another. The current policy states to use "shrine" in place of -jinja -jingu and -myojin. However the suffix title does differentiate between several types of shrines and their status accordingly. Many of the imperial shrines are -jingu shrines as an example. The suffix is needed to make the subtle distinction and is part of a formal name. I have used this example before but it it like calling "The Vatican" the vatican church, or St Anthonys Cathedral the Anthony church. We need to change the policy and use formal names that a shrine wishes to be called and not shorten it. Takashi Ueki (talk) 19:06, 6 August 2009 (UTC)

As all of the suffixes are generally translated as "Shrine" (or "Grand Shrine" in the case of -taisha), I don't think I agree; Wikipedia is not written for a specialist audience, and "Shrine" is a much clearer way (in common English usage) to state the title of many of these than "Jingu", etc. As for the argument that these are official names, the English Wikipedia relies on the common names for article titles. "Heian Shrine" appears to be more common than "Heian Jingu" according to Google (I personally use the second); "Ise Shrine" outnumbers "Ise Jingu"; "Sumiyoshi Shrine" outnumbers "Sumiyoshi Taisha"; "Hachiman Shrine" outnumbers "Hachiman Gu". All of these show equal numbers whether the suffix is capitalized or attached to the name. On the other hand, "Izumo Taisha" appears to outnumber "Izumo Shrine". Cf. ja:Category:大聖堂 and ja:大聖堂; none appear to be called "cathedrals" on the Japanese Wiki, nor does ja:ウェストミンスター大聖堂 call it anything other than a "church" or the Japanese word for cathedral in the text as far as I can see. Dekimasuよ! 19:29, 6 August 2009 (UTC)
I generally agree with everything Dekimasu states. Also, the specific Japanese names are found in the lead in the nihongo template for any readers that want to know. --TorsodogTalk 19:36, 6 August 2009 (UTC)
However, the convention would not hold up to a specific Shrine like Ise Jingu as it is called just Jingu. That is a formal name and not a suffix. The name is correct as a formal title. Also there is a distinct difference as the formal name is a trademark by most Shrines for one, and the preferred name by the Shrine. Separating Tiasha only proves the point. There is no specialist reader needed to understand the need to allow for self naming conventions among formal titles. They are not all just shrines, they belong to different subsets of shrines and that is relevant to religious self determination inside Shinto belief. It is similar to saying that the naming conventions between Catholic and Protestant clergy is irrelevant because they are all Christian (lets just call them all priest). Also I fail to see the relevance of using Google as a determinant method for naming conventions. Eg: Tao vs Dao. All of the conventions to date show that Tao is more commomly used due to age of mistranslation, however Dao is the Library of Congress method of modern convention. Pin Yin is more accurate. I go with the modern convention. It is the same in this case. Google can be manipulated and popularity does not make correct convention. Plus the internet is notoriously bad at getting information academically correct and book sources would be needed to support that supposition for these naming conventions. I am going to make contact with one of the experts that I resource about Japanese language conventions, especially for this purpose.

It deserves the formal title and I completely disagree that naming conventions can simply be thrown out because it is more convenient for the western reader. Takashi Ueki (talk) 20:39, 6 August 2009 (UTC)

I also agree with Dekimasu. As Torsodog states the specific translation appears directly after the name in the nihongo template. Not only that but you are free to create Ise Jingu as a redirect to the article. The point is that the English reader can find the article about the topic. That is our target audience, the English reader. For the non-English reader there are other Wikis. imars (talk) 08:29, 7 August 2009 (UTC)
I also agree with Dekimasu. I jumped on my local university website and looked up one of the academic databases used for research and did a search for both Ise Shrine and Ise Jingu. Ise Shrine came up with 4 times as many hits. Also, having lived in Ise for 4 years, the most common translation into English I saw there for the shrine was "Ise Grand Shrine" (on road signs and such). To me, this argument is about what is 'more correct' based on the official Japanese name (Jingu) as opposed to the more accepted name in standard English use (Shrine). It would be nice to use the proper Japanese name 'Jingu', but this is English wikipedia and we should use the English name here. Ka-ru (talk) 13:52, 7 August 2009 (UTC)
I did the same thing as Ka-ru and also checked within the article to see if Jingu was being used as the common name. Ise Shrine is the academic standard, so as a publication based on consensus of sources Wikipedia must employ this naming convention. Shii (tock) 14:09, 7 August 2009 (UTC)
Do remember that there is an exception in the WP convention on temple and shrine naming that allows for a formal name when distinguishing between shrines. This change is based on that convention. Ise Jingu is different than every other Ise branch shrine and ther are about 2000 of them. The head shrine is meant to be named formally where every other branch shrine can be called Ise Shrine. The convention then states to make a disambiguation page to clear up the matter. Ths looks appropriate to me. If everyone would be happier calling it Ise Grand Shrine that is fine as well, but it as all other head shrines need to be diffferentiated. Takashi Ueki (talk) 17:34, 7 August 2009 (UTC)
I agree 100% with Takashi Ueki. Abandoning the distinction between the various types of shrine to put them all in one category doesn't make sense to me. Why junk useful information? If those names are used and are seen as having a meaning in Japanese, should we not use them in English too? A cathedral and a chapel are both churches, but that doesn't mean the distinction between the two can/should be lost. Saying that Wikipedia is not for specialists, so we shouldn't use words that are too technical is paternalistic: it assumes people are not capable of and uninterested in learning. In any case if Wikipedia is not for specialists, it's supposed to be at least correct and teach something. Rather than using redirects from jingu, -gu, -sha and so forth to shrine, how about doing the opposite? About common names: it surely makes sense to use "Marilyn Monroe" rather than "Norma Jean Baker". That's the name she was known under. But surely this rule wasn't meant to actually validate and perpetuate errors. Urashima Tarō (talk) 23:34, 7 August 2009 (UTC)
Because "Jinja", "Jingu", "Myojin" etc have no meaning in English, and because the best possible English translation for each is "Shrine". No information is being lost, those Wikipedia readers who understand Japanese can see the Japanese name right at the beginning of the article.
However, this discussion belongs on Talk:Ise Jingu, not here. Jpatokal (talk) 09:38, 9 August 2009 (UTC)

Yes, perhaps it would be good to use the official name the shrine calls itself. Shii (tock) 14:43, 8 August 2009 (UTC)

Once upon a time, The Love Boat visited Tokyo. One of the attractions they mentioned was the Meiji Jing Shrine. This illustrates the difficulty of relying on what's known in English, since that show was viewed by an audience of millions (and has, no doubt, been rerun ad nauseum). It sounds absurd, but after all, since means "shrine," it's a valid translation. ("Jing" is my transcription from memory. The script might have had "Jin.") A difficulty in using the names shrines (or other establishments or people) call themselves is that they follow a variety of romanization conventions. The purpose of a style manual is to provide a standard for this publication. So unless something is extremely widely known in English, I'd advocate using Wikipedia style. I remain open-minded about the possibility of distinguishing the different kinds of shrines if the community feels it's best. Although I'm happy with the status quo. Fg2 (talk) 09:55, 9 August 2009 (UTC)

Please comment on this issue

Talk:RKKY_interaction -- Shocklord (talk) 08:19, 7 August 2009 (UTC) Should 芳田奎 be romanized as Yoshida Kei or Yosida Kei ?

Place names

There is a section Place names in this article, however there is no description about what kind of romanized name should be used. I am discussing this issue at Talk:Daikanyama#Requested move, but I came to a decision to bring it to this page. I would like to discuss a general criteria not only a solution of this specific name.

In above mentioned article, the name of the place in Tokyo is disputed, 代官山 "Daikanyama" or "Daikan'yama". Please read above link for detailed discussion.

Basically my proposal is as follows;

  1. The romanized name used by published map in Japan should be used for Wilipedia. (for example Google Map) In this case, apostrophe is probably never used.
  2. If the government(either state, prefectural or municipal) indicates the romanaized name in their official document or web site, it should be used for Wikipedia.
  3. If the romanized name of a place is widely used within Japan, it should be used for Wikipedia, even if it is not widely known around world.

If there is any conflict between above, another reasonable criteria should be employed.

Please let me know your opinion. ― Phoenix7777 (talk) 07:30, 9 August 2009 (UTC)

  • I vote for either Daikanyama or Daikan-yama. While I am normally a proponent of a standard system of romanization, I don't think apostrophes are as big of a deal as macrons or the "correct" romanization of shi and chi as not being si and ti. I don't have any specific sources on me, but I don't believe I have ever seen "Daikan'yama" before, with the apostrophe. When it comes to commonly known placenames - and Daikanyama is certainly commonly known to anyone who's ridden the trains around Tokyo enough - I think it's just fine to go with a common version. LordAmeth (talk) 12:40, 9 August 2009 (UTC)

Please note that there is an ongoing discussion occurring at Talk:Daikanyama#Requested move, so discussion relating specifically to and about the name "Daikanyama" should occur there. Note that I've moved one comment specifically about "Daikanyama" to the Talk:Daikanyama#Requested move page. Phoenix7777 is (or, at least, should be) discussing the overall romanization guidance which is currently offered in this guideline. He feels that it shoudl be changed to a system which he is apparrently more comfortable with.
V = I * R (talk) 15:54, 9 August 2009 (UTC)
I restored a post by User:LordAmeth. Please continue a discussion of a specific name of Daikanyama as well as general rule. It is not permitted to move/delete other person's descriptions on talk page without permission of author.
Yes, certainly this discussion began at page Daikanyama. However its audience is very small, so I brought the discussion here where there are many persons with knowledge of this topic. ― Phoenix7777 (talk) 21:49, 9 August 2009 (UTC)

To make this discussion a bit more general, I'd suggest that the syllable にゃ nya is so rare in Japanese -- as far as I can see from WWWJDIC, the only kanji with that reading is 若, which in turn is only used in the compound 般若 hannya -- that apostrophes for it are unnecessary, at least in article titles. But for avoidance of doubt, the apostrophe can still be indicated as as separate romanization, like this: "Daikanyama (代官山 Daikan'yama?) is..." Jpatokal (talk) 07:27, 10 August 2009 (UTC)

Train and Subway stations

There is a section Train and Subway stations in this article, however there is no description about what kind of romanized name should be used.

My proposal is "Station name should be the romanized name Railway companies publish".

Please let me know your opinion. ― Phoenix7777 (talk) 07:39, 9 August 2009 (UTC)

On the other hand, I think it a very dangerous precedent to set to go by what railway companies publish. I don't know of any specific examples right now immediately offhand, but I would not put it past railroad companies in Japan to make basically every "mistake" in the book, from ignoring macrons to writing shi and chi as si and ti, to mis-romanizing words that are English anyway.
There are a great many rail companies in Japan; they don't all agree on a single standard system, and I wouldn't be surprised if there are even inconsistencies within any given rail company as to romanization. In short, I don't think they can be trusted to get it right.
I would absolutely support the idea of using the name officially used by rail companies when it comes to naming and translation issues - e.g. Naha-kūkō Station, as is clearly printed on the giant sign at the station itself, rather than "Naha Airport Station", or Ōsakajōkōen Station and not "Osaka Castle Park Station" even though we have an article at Osaka Castle Park - but I cannot support that notion as it applies to correct romanization. The name given a station is a very purposeful and intentional thing, and so the official name of a station should be honored. However, I firmly believe that the romanization of that name is not nearly as purposeful and intentional, and represents a haphazard application of poorly understood standards and logic. Where a rail company leaves off a macron, this is not with purposeful intent to misrepresent the Japanese name, it is merely a slip-up, where someone doesn't know better or doesn't care, and we should not honor that as if it were an official, purposeful, creative choice. LordAmeth (talk) 12:52, 9 August 2009 (UTC)
I agree with LordAmeth. Another thing I noticed: what is the policy in hyphenation ? I notice there is Kita-Senri Station, Kita Kurihama Station and Kitahiroshima Station. The pages without hyphen are redirected to pages with hyphen. Just some examples. Shocklord (talk) 16:36, 9 August 2009 (UTC)
I also agree with LordAmeth. The guideline for hyphens is that they should be avoided. However, in the case of train station names, they are often used as we use the official name of the station used by the company which operates it. I agree that the romanization following the kanji should be standardized, and there should also be redirects for other likely romanizations or spellings. ···日本穣? · 投稿 · Talk to Nihonjoe 18:07, 10 August 2009 (UTC)
  • First I would like to point out that si and ti are based on the Japanese standard. See Romanization of Japanese#Chart of romanizations. Although this Romanization is not accepted today even in Japan. Yes, it is certainly beautiful if those names were KitaSenri Station, KitaKurihama Station and KitaHiroshima Station. However, it is not productive to complain about current condition. We should not be arrogant to change the existent name, only because it does not conform to the Wiki standard or it is not beautiful. The only choice we can take is to use the current romanization used officially by railway company without adding and reducing any apostrophe or macron. And use of such a name is in reality de facto standard in en:Wiki now. So I propose again the following description be added to WP:MOS-JP#Train and Subway stations, similar wording as in "Names of companies, products, and organizations".
"Honor the current romanization used officially by Railway companies." ―― Phoenix7777 (talk) 02:11, 13 August 2009 (UTC)
Here is my counterproposal:
"Use transliteration of official Japanese name in revised Hepburn, unless it is apparent that the railway company prefers different name or spelling."
Reasons: Firstly, Japanese railway stations hardly have "official" romanized name that have to be followed under the MOS. Romanized names are mere transliteration of the official names that are made by railway companies as a guide for foreigners or just for decoration of webpages, with no intention to establish official names.
Secondly, even if the spellings by railway companies should be considered as "official," we would face difficulties to determine which spelling is the official because of companies' indecisive attitude to the "official" names. For example JR West uses macronned form on the station signages (example: Ōtsu Station) but does not on the official website (example: Otsu Station). Such a situation would easily cause a serious confusion in naming if we declare that the "official" romanization is absolute. --Sushiya (talk) 13:58, 13 August 2009 (UTC)
My position is even in your first case, a name written in station sign should be used for a name for en:Wiki. Could you show me some examples of this case? I may be convinced to agree your opinion.
As for the second case, it shows that it is not so easy to find the fact. Please see this page [13] by JR West or Google Map[14]. This is because macron is often omitted in website. This can be solved by adding a caution like "the station name in website may sometimes be written without macron(s)".
My concern for your counter proposal is that it often becomes a dispute as to whether "it is apparent that the railway company prefers different name or spelling" ―― Phoenix7777 (talk) 22:40, 13 August 2009 (UTC)
My first point is that railway companies do not consider the romanized names so important as the official names of the stations. One evidence of this is that even station signages at one station have different names in romaji. For example Hanamaki Airport, Hanamaki-Kūkō and Hanamaki-Kuko. See also the inconsistency in macron usage already mentioned above. --Sushiya (talk) 02:05, 15 August 2009 (UTC)
I think in cases like that we should go with the standard name which would be correct under the general WP:MOS-JA guidelines. As Phoenix7777 indicates, you are listing an extreme case which is unlikely to occur all that often (I never once saw signage that diverse in romanization in one station in my many years living in Japan). ···日本穣? · 投稿 · Talk to Nihonjoe 04:47, 15 August 2009 (UTC)
It often occurs when a company changes its romanization method but leaves old signages unchanged. I've seen some in stations of Tokyo Metro, which recently cease to use macrons. cf. ja:神保町駅#歴史. --Sushiya (talk) 07:11, 15 August 2009 (UTC)
I understand your concern, however you are talking about an extreme case. So I support User:Nihonjo's clarification as a confirmation of current de facto naming method. And in your extreme case, editor should follow the description written in lede of WP:MOS-JP#Names "If you are unsure of how to write a name after reading the information below, please post your question on the Talk page." ―― Phoenix7777 (talk) 02:42, 15 August 2009 (UTC)
I am talking about the nature of the romanized names inferred from the cases, not the naming procedures under your proposed rule. --Sushiya (talk) 03:02, 15 August 2009 (UTC)
Okay, now you're just making things confusing. This isn't someplace to academically discuss romanization inconsistencies just for the heck of it. All we care about is what we are going to do here, on enwiki. We aren't here to infer anything. Your questions have been answered already (unless you've come up with new ones). ···日本穣? · 投稿 · Talk to Nihonjoe 07:49, 15 August 2009 (UTC)

Frankly speaking I don't believe the naming based on station signages would cause so many problems. However, as a part of Wikipedia, I am unconfortable to declare that the station signage must be the basis for the article naming because signages are not a kind of reliable sourses and to check it for naming is a kind of original research. --Sushiya (talk) 03:13, 15 August 2009 (UTC)

It's no more original research than looking on the official sites of the train companies. The signs are officially approved and produced, and they are there where everyone can see them. If I had to choose one over the others, I would probably pick the sign above the entrance to the station. ···日本穣? · 投稿 · Talk to Nihonjoe 04:47, 15 August 2009 (UTC)
I agree with Sushiya. Signage cannot be taken as a reliable source, because the train companies themselves do not follow any consistent style or logic in their romanizations, and because, as Sushiya writes, Japanese railway stations hardly have "official" romanized name ... Romanized names are mere transliteration of the official names that are made by railway companies ... with no intention to establish official names. I'm sorry to repeat what's already written here, but I think this is completely the core of the point. Our station naming on en:wiki should reflect a certain degree of consistency and logic in the use of macrons and in spelling, even when that trumps "official" names (e.g. Ōtsu Station instead of Otsu Station; shi & chi instead of si and ti). LordAmeth (talk) 14:19, 15 August 2009 (UTC)
If the Revised Hepburn Romanization take precedence over station sign, then Daikan'yama, Shin'ōsaka and Nishishinjuku must be used instead of Daikan-yama, Shin-Ōsaka and Nishi-shinjuku which are on the station sign. Because usage of hyphen instead of apostrophe is not defined in Revised Hepburn. Furthermore there is no description "after a word shin or nishi, hyphen should be used" in WP:MOS-JP. Someone would never compromise on apostrophe usage, strictly following Revised Hepburn Romanization. See Talk:Junichiro Koizumi#Article title. As for the usage of the name on station sign, correction of apparent misuse of romanization will not be precluded, unless it is not a change from hyphen to apostrophe.
Anyway I came to realized that there is no merit to clarify the guideline, because the naming of the station is properly made by following the first line of WP:MOS-JP "Use common sense in applying it; it will have occasional exceptions." I withdraw my proposal. Thank you all for participating this discussion. ―― Phoenix7777 (talk) 22:31, 15 August 2009 (UTC)
Instead, please participate in Talk:Daikanyama. ―― Phoenix7777 (talk) 23:34, 15 August 2009 (UTC)

Names of fictional characters: Hepburn or "intended" romanization?

The discussion that lead to this one can be found here: Talk:List of One Piece episodes (season 10)#Margaret//Marguerite//Māgaretto

There has been a bit of edit warring at List of One Piece episodes (season 10) about a character that first appears in that season. Her name is written in kana as マーガレット. One side wishes to romanize her name as "Margaret", like the common first name. The other side wants to use "Marguerite", as she and several related characters are apparently named after flowers. My question is: Is going with the "intended" romanization without an official source to back it up acceptable at all? Goodraise 21:30, 11 August 2009 (UTC)

For the record, I now support Marguerite rather than Margaret, as according to the Japanese wikipedia, the spelling for the Marguerite flower is indeed マーガレット. The Splendiferous Gegiford (talk) 21:37, 11 August 2009 (UTC)

If it's clearly established that ALL of the names are flowers (perhaps in an interview with the author or something similar), then that would make sense. I don't see a need to modify the MOS-JA, though. ···日本穣? · 投稿 · Talk to Nihonjoe 00:25, 12 August 2009 (UTC)
So, if it's not "clearly established that ALL of the names are flowers" (as is the case here), you'd use Hepburn? (I assume that with "something similar" you mean a non-primary source.) Goodraise 01:42, 12 August 2009 (UTC)
Given the popularity of the series, it's likely that someone, somewhere, in some article about the series, commented on all those names being flower names. ···日本穣? · 投稿 · Talk to Nihonjoe 09:53, 13 August 2009 (UTC)
I don't share your optimism, but that is beside the point. What do we do until such a source has been found? -- I'm not asking this because of a single character. I've been wondering about this question for quite a while. This incident only provided the incentive I needed to ask. Goodraise 23:42, 13 August 2009 (UTC)
You may want to take a look at the old talk page of Talk:Burdock (Dragon Ball). Dekimasuよ! 08:18, 12 August 2009 (UTC)
Thanks for the link, but the case is a lot different here. Burdock has several official English spellings, as opposed to none. There is no writing system trickery involved here: As Geg pointed out above, the kana is the same for the flower and the character. Also, a decision does not seem to have been made there, so even if the cases were compatible, I'd be left as clueless as I am now. Goodraise 15:42, 12 August 2009 (UTC)
It maybe original research, or it may simply fall into common sense. I have no idea in what character is that, but if someone can suggest a set of characters are all named after flowers( and no other exception could be found in such set), and all of them sound uncommon, I would say that using the flower name is much better than using the normal common English name. —Preceding signed comment added by MythSearchertalk 07:59, 14 August 2009 (UTC)
Use Marguerite with a notation about the original spelling. Alternative, you could ask Funimation what they intend to use as her name citing the problem.Jinnai 20:25, 14 August 2009 (UTC)


ウォー, is it Wō or Uō? --Apoc2400 (talk) 14:54, 13 August 2009 (UTC)

Wō. Uō needs a large ォ, as in ウオー. Jpatokal (talk) 16:00, 13 August 2009 (UTC)
Ok, thanks. As I thought, but I wasn't sure. --Apoc2400 (talk) 17:31, 13 August 2009 (UTC)

Romanization of foreign words rendered in Japanese katakana

As per this discussion I am being point here. I find it very redundant and unnecessary, since the Japanese katakana is actually only telling what the word in front would be pronounced by the Japanese. Of course it may be useful some of the times, like when the pronunciation is very different from the origin or the displayed words. However, there are times when it is only stating the obvious joke in the eyes of westerners about how poorly Japanese render those words and sometimes even more misleading since katakana pronunciation separate each word as oppose to English where people try to link each of them. For example, Aerith in English would be something like Arith (2 distinct vowels), but the katakana is E A Li Su (4 distinct vowels) and the romanzation would be Earisu, where it would be 4 vowels but people will tend to pronounce the first part more like Ear, an i sound instead of the eh sound. —Preceding signed comment added by MythSearchertalk 14:49, 15 August 2009 (UTC)

First of all, there is no "Li" sound in Japanese, so your example is incorrect. That said, the Japanese only needs to be presented once, so I'm not seeing what the big deal is here. If it's a Japanese game where the characters were originally named in katakana, having that where people can easily see how the Japanese pronounced it is important information. It's made even easier by using the {{nihongo}} template. ···日本穣? · 投稿 · Talk to Nihonjoe 18:10, 15 August 2009 (UTC)
From what I heard, the Japanese romanization have no li, but most Japanese do not pronounce ri as in English as well. Also, what does it add to the article for people who do not read Japanese, cannot read Japanese, and do not bother to learn Japanese? Wouldn't including the IPA of the term much more efficient than romanizing the katakana? —Preceding signed comment added by MythSearchertalk 19:09, 15 August 2009 (UTC)
It adds nothing for those people, but the same can be said for any Japanese anywhere on enwiki. The point of having it is to be complete and provide the information for those who do read Japanese and for those interested in it whether or not they can read it. Since it's a Japanese topic, using the Japanese is entirely appropriate. ···日本穣? · 投稿 · Talk to Nihonjoe 19:52, 15 August 2009 (UTC)
I did not mention anything about removing the katakana, if people see the katakana, and is interested in the Japanese pronunciation, they will check it for themselves, if they do not, the romanization do not add anything to it. —Preceding signed comment added by MythSearchertalk 14:50, 16 August 2009 (UTC)
I agree that providing Romanization for Japanese script is redundant only when the romanization is essentially a copy of the English phrase. The example in the VGProj discussion had to do with Kingdom Hearts, which at one point was introduced as: Kingdom Hearts (キングダムハーツ Kingudamu Hātsu?) . The point of contention was that the romanization was redundant, and to most readers just looks like a guide on how to say an English title with a Japanese accent. However, where the romanization IS interesting and important would be in cases where the Japanese title is significantly different from the English - for example, Tetris Attack, released in Japan as Yoshi no Panepon (ヨッシーのパネポン?) .
I think the issue is that this MOS seems to say that we have to provide romanization for every instance of Japanese script so that people can see how to pronounce it. I agree with others from the VGProj, though, that it makes article leads unnecessarily long and doesn't add anything of significant value when it basically looks like an English title, Japanese script, and then the English title again. — KieferSkunk (talk) — 05:40, 17 August 2009 (UTC)
It does add value, even if it's not significant, which no one ever claimed it was. Every last word doesn't have to add significant value, however, and including the rōmaji for the Japanese will not significantly increase the article lead size unless the article lead is significantly smaller than it should be. Rōmaji is not English, even if it looks similar due to using Roman letters. Due to being written in a completely different character set, having the rōmaji allows those who don't read Japanese to have some idea of how it's actually said in the original language in which it was released (which could be considered significant). It also helps to keep the site from being Euro-centric. ···日本穣? · 投稿 · Talk to Nihonjoe 07:36, 17 August 2009 (UTC)
That is exactly what I am saying. I am elaborating the discussion concerning the titles to other parts, like characters that appear in video games. like Vincent romanization as Vinsento is quite redundent in the English wiki(but it is not redundant in the French wiki as the pronunciation of Vincent is different in French). —Preceding signed comment added by MythSearchertalk 06:26, 17 August 2009 (UTC)
The problem with this is the consensus was removing romanization of Japanese game titles; but it was too general in a sense: there was talk of using footnotes if the article lead was unnecessarily long which was not elaborated further, and somehow the removal included character names and plot terms - which wasn't in the reached consensus. Also, I think most who were for removing the romanization didn't see it from the point of view of or didn't take into account those who'd be interested to know how the kana was pronounced (such as moi). I agree with Nihonjoe's view that romaji isn't English - it's the pronunciation of the kana using latin letters - and is thus not redundant. — Blue 08:49, 17 August 2009 (UTC)
If they are interested, it would be very simple to look up the kana and see how they are pronounced. If they do not bother to do that, I do not see why we need to compensate for their laziness and force feed it to them. Japanese is not a totally out of hand language like the tokien elvish, it is rather simple to find web sites to at least have the kana's pronunciation displayed. Saying it is not redundant because someone might be interested sounds like saying someone in the street might be hungry, so you place lunch boxes in every corner thinking hungry people might not want to walk the length of the street to go to a dinning place. —Preceding signed comment added by MythSearchertalk 12:56, 17 August 2009 (UTC)
Eh, you completely missed the point; it's not redundant because romaji isn't English. Plus, your comparisons could also be placed on Wikipedia as a whole; "readers can look up the nearest library and find out about Homunculi, we don't need to "force feed" them by writing an encyclopedic article about it." So, yeah.
I'm for a) restoring the romaji back into articles that have been using kana, and b) should the article intro be rendered needlessly long because of the Japanese titles, then placing them in footnotes should be adequate, but not wholesale removal. — Blue 14:14, 17 August 2009 (UTC)
What have they come to wikipedia for? If they are reading a Japanese related page, they are here for the information for the article specific, and those articles are not here to teach Japanese.(And they are doing a pretty bad job if they are.) Does anyone really look at an article about something and decided they have to know how it is pronounced in Japanese, when the name could be very simply pronounced in English? It seems more like making fun of Japanese pronunciation than actually helping the article. Vincent Valentine is very obvious to English speakers, and informative enough for any regular readers. The rationale used for supporting the romanization could be well be used on adding the IPA of the Japanese characters so readers can know exactly how the words are pronounced. The list can keep on going if more information is always good arguement is used. —Preceding signed comment added by MythSearchertalk 03:29, 18 August 2009 (UTC)
I agree that the romaji transliterations need to be added back. Romaji pronunciation is not English pronunciation as the Japanese do not have the schwa and many other sounds seen in English and other European languages. WhisperToMe (talk) 20:06, 17 August 2009 (UTC)
I don't think we need pronuciation for obviously English words. We don't do that for greek figures like Pericles even though its pronounced in Greek as Periklis. It's because we have a well-known English name we don't need the pronunciation. The same is true for something like Vincent. For words which are radically different when translated, that might be differnet.Jinnai 20:25, 17 August 2009 (UTC)
Vincent is a fictional character from a video game. I'm not sure that's comparable to Pericles, really...
(that being said, I don't think including a note about the original pronunciation of Pericles's name would hurt that article) Erigu (talk) 20:38, 17 August 2009 (UTC)
Those "obviously English" words are not pronounced that way in Japanese. I.E. Vincent versus Binsento. The reason why we give the romaji is to show how it is pronounced in Japanese. Likewise a note about the original pronunciation about Pericles's name is perfectly reasonable. WhisperToMe (talk) 20:55, 17 August 2009 (UTC)
Also, the intended pronunciation of the name "Vincent" is not obvious based on the alphabet spelling alone: it could also be the French name "Vincent". The fact the name is pronounced as "binsento" in Japanese does seem (<- that's me being ridiculously cautious, here) to indicate it's supposed to be pronounced according to English phonetics. I think that's valuable information. There's a bunch of cases where the original pronunciation of a video game character's name has little to do with English phonetics (in that same Final Fantasy series, even).
And while I guess you could argue that somebody who would know enough about Japanese phonetics to understand that based on the romanized spelling would probably also be able to read the kana spelling, the "probably" bothers me... Erigu (talk) 21:41, 17 August 2009 (UTC)
You missed what I said, the name Vincent is obvious in the English wiki to be pronounced like Vincent in the common English speaking world, and thus in this case, this specific character would need the romanji in the French wiki, since it is not the common pronunciation for them. Include them if not common, skip when they are. —Preceding signed comment added by MythSearchertalk 03:29, 18 August 2009 (UTC)
You might then have a better argument imo for the likes of Vincent if we also included the French pronunciation since the idea is it would be unclear what the proper source for Vincent's name is as it obviously isn't Japanese. Same could be said for Edward Elric who has even more evidence of his intended nationality. And yes, other articles that deal with real or fictional individuals who have different nationalities than than English or where they are primarily known from (for real people) or the original work (for fiction) do list such pronunciations if they decide to list them at all. Most do not though if the English equivalent is obvious.Jinnai 22:16, 17 August 2009 (UTC)
Where's the "evidence" of Edward Elric 's "intended nationality"? The character has a name that looks English, that's all.
I'm just saying it would be better not to make assumptions (as "educated" as they may be). Let's just provide the reader with relevant and objective information about the name. I think the romanized spelling of names from Japanese works is both relevant and objective. Anything beyond that, like the assumption that the pronunciation should be obvious for English readers, is subjective territory, in my opinion. Erigu (talk) 23:30, 17 August 2009 (UTC)
In the narrative. It is a fictionalized version of Germany before WWII, although I meant was his intended citizenship (since he could be something other than German).Jinnai 01:13, 18 August 2009 (UTC)
Er... As far as I know, it's a fictional world that's merely inspired from 19th century Europe... Erigu (talk) 01:34, 18 August 2009 (UTC)
A simple question, will you pronounce Vincent Valentine as Vinsento Varentain? Everytime you use the term? Or are you going to prnounce the term simply as its English form Vincent Valentine(VIN-sənt VAL-ən-tien key)? —Preceding signed comment added by MythSearchertalk 03:29, 18 August 2009 (UTC)
The question seems too self-centered for one to answer. One needs to understand that the Japanese pronounce things differently than we do, and we need to express that. WhisperToMe (talk) 04:58, 18 August 2009 (UTC)
Why would I pronounce that name exactly according to the kana? I'm not Japanese. It looks like you're missing the point, really... Erigu (talk) 04:20, 18 August 2009 (UTC)
Your are missing my point. If someone like you who insist of having the romanization have no use for them, why do you think anyone else would want it other than making fun of Japanese pronunciation? I understand there are times when it is a must to have, like Tidus is pronounced as Tida in the Japanese version. However, I do not see the article(s) of the long list of Louis(es) like this one have the pronunciation loo-EE attached to it stating the name should be Loui instead of Louis. The WP:MOS-FR did not have such ideas but they are more likely to cause confusion for readers. WP:MOS-KO also did not require the pronunciation of the Korean words eventhough it is similar to Japanese in this sense. Some Chinese article having that simply because there are so many dialects in China(some argue those are separate languages as well, whatever) and the same words can have different pronunciation. —Preceding signed comment added by MythSearchertalk 04:43, 18 August 2009 (UTC)
When a pronunciation is unusual we often use IPA. As for China, we always include Mandarin and also include any applicable local dialects, depending on the involved regions. For MOS-KO, I believe if it is a "foreign character in a Korean work" we also have the romanization in Revised Romanization and McCune Reischauer. WhisperToMe (talk) 04:58, 18 August 2009 (UTC)
If someone like you who insist of having the romanization have no use for them, why do you think anyone else would want it other than making fun of Japanese pronunciation?
I personally have no use for the romanization because I can read kana. Last I heard, quite a few people can't read kana. And if you're suggesting they should just learn... Well, Bluerfn already covered that one.
Now, do I have a use for the kana spelling? Yes, I do, as I'm interested in the original pronunciation of those names. If I couldn't read kana, I'd still have the romanization, lucky me.
And I can think of reasons to be interested in this other than "making fun of Japanese pronunciation" (now, that part was quite depressing, really). The kana spelling (or the romanization, if you can't read kana) gives us the Japanese pronunciation of the name. In the case of a non-Japanese name, it will naturally be an approximation, but that's still information.
I know you're saying we could just make an exception when the name is "obviously supposed to be pronounced according to English phonetics", but I wish you'd realize the limitations of that approach. What appears "obvious" to you may not be obvious to someone else. Start making exceptions like that, and you'll end up creating a bunch of arguments just for the sake of removing a few characters. I just don't think it's worth complexifying the guidelines. Erigu (talk) 06:06, 18 August 2009 (UTC)
Try to convince MOS-FR to include Loui in Louis related article then, it seems much more appropriate to include than the romanji here. Since most people do not see it this way, I am not going to comment on it anymore, I have no idea how to use the romanji anyway.(I can read the Japanese, but have no knowledge in the áàâäãǎāă type of stuff.) —Preceding signed comment added by MythSearchertalk 06:31, 18 August 2009 (UTC)
I think that exceptions for obvious English words isn't that big of an exception. Can it be helpful? Yes, but in such cases it shouldn't be required, only suggested.Jinnai 21:36, 18 August 2009 (UTC)
As in the original discussion I still think romanization of non-Japanese words is superfluous for the en.wikipedia readers. And if pronunciation should be indicated at all it should be the English one in IPA (as heard in voiceovers or from the developers' mouth). Megata Sanshiro (talk) 21:54, 18 August 2009 (UTC)
I wouldn't make it absolute. There are first and foremost foreign words that aren't of English origin. There may also be justified reasons for using romaji if the name includes either kanji and/or hiragana. My only concern is the idea of it being absolute for words like English words like ドラゴンクエスト.Jinnai 22:02, 18 August 2009 (UTC)
If it includes kanji then it's a Japanese word. Megata Sanshiro (talk) 22:10, 18 August 2009 (UTC)
Not quite true as there are a few exceptions. 珈琲 is coffee and coffee certainly isn't a native Japanese word.Jinnai 23:20, 18 August 2009 (UTC)
The original discussion and the guideline it resulted in were about article intros, and Japanese titles made up of plain English words (and that goes against this manual of style). You're extending the whole thing to entire articles, non-Japanese words and names.
(and I didn't know "rokujūyon" was a common English word...) Erigu (talk) 00:58, 19 August 2009 (UTC)
There wasn't any consensus on the extension as far as I can read, only on article titles which was too absolute - "should not be included" - and too general, not elaborating on better treatment regarding the romanization. Erigu's statement is worth repeating - what's fairly obvious to some may not be obvious to others. I'm fairly sure that the developers for "Dragon Quest" didn't pronounce "ドラゴンクエスト" like an English speaker. I'll just repeat my suggestions on a) including romanization and b) using footnotes if the romanization rendered the article lead too long.
And correct me if I'm wrong, but I do believe MOS-JP trumps any decision regarding the inclusion/exclusion of kana romanizations. I'd like to point out Wikipedia Japan's practice of writing the kana pronunciation in their game article leads, even though we'd expect the Japanese to know their own language already; I know they have a good reason for this because of the numerous meanings to one kanji character, but that's still to be commended, unlike the view that "if they want to read the kana, they can go to a kana article". — Blue 03:31, 19 August 2009 (UTC)
And as a result of this discussion we need to restore all kana readings for "foreign names" in the context discussed here. Other projects should yield to MOS-JA's current practices. WhisperToMe (talk) 02:11, 20 August 2009 (UTC)
Bl- if the developers didn't pronounce ドラゴンクエスト as "Dragon Quest" then its more likely do to the difficulty in phonetics of another lanuggae purposeful.Jinnai 05:16, 20 August 2009 (UTC)
I agree with Jinnai. Also MOS-JA and VG/GL are both guidelines, why would one have to "yield" to the other? WikiProjects are allowed to have their guidelines, which may or may not differ from other WikiProjects's guidelines (but not main Wikipedia policies). As it is, the VG guidelines are pretty clear:
"Japanese titles should only be provided for games of Japanese origin whose official English name differs significantly from its Japanese name."
"Dragon Quest" does not significantly differ from "Doragon Kuesuto", since it is quite exactly the same name except one of them is an approximation of the other. As for "rokujūyon", it's just "64". Megata Sanshiro (talk) 10:33, 20 August 2009 (UTC)
MOS-JA and VG/GL are both guidelines, why would one have to "yield" to the other?
It's only natural to include the original title of a work (movie, TV series, comic book or video game). And if that original title isn't written in our alphabet, it's only natural to include a romanization. Why would video games be an exception?
We have a system that works pretty well and is fairly objective, but you apparently want to complexify the guidelines (for video games specifically) in such a way that editors will have to decide on a case-by-case basis whether or not a title/name/whatever's pronunciation should be "obvious" for an English speaker (as long as they all agree all the time, that shouldn't be a problem, right? it's not like there's any subjectivity involved, here, after all...), and all that just just for the sake of removing a few characters.
That simply sounds like a terrible idea to me. Erigu (talk) 11:44, 20 August 2009 (UTC)
"MOS-JA and VG/GL are both guidelines, why would one have to "yield" to the other? "
1. MOS-JA is an across the board Manual of Style guideline and it overrules local project guidelines like Wikipedia:WikiProject Video games/Article guidelines
2. The Japanese language is not within the expertise of VG/GL, so VG/GL will defer to MOS-JA for matters related to the Japanese language. VG/GL ought to accept MOS-JA's guidelines about romanization of "foreign" names. WhisperToMe (talk) 18:46, 20 August 2009 (UTC)
Allow me to present a new argument in countering the point raised up there. Bluefn, I'm fairly sure that the developers for "Dragon Quest" didn't pronounce "ドラゴンクエスト" like an English speaker Japanese speakers having Japanese accent is not a valid argument of including the info. People keep thinking that the words are pronounced differently and thus it must be included, but they missed the fact that accent in different countries are what going on here. The Japanese have their own accent, since they use their own native language to ease the difficulty in learning a new language. Katakana was not designed for foreign words, it is a simplified method of helping kids and women to learn the pronunciation of words, since in ancient times, there are a lot more hiragana in Japan. (Source: Nippon jin no shiranai nihongo 日本人の知らない日本語 ISBN 978-4-8401-2673-1) It is used as a similar tool for Japanese to learn foreign languages and thus the accent of Japanese will be influenced by it. It does not mean that they made up whole new words or whole new pronunciations for words. Saying they pronounce the words differently and we must include it is only WP:POV and should not be used in wikipedia. If the words are made up, so be it, include the romanji since no one kows what on Earth are they. If it is a simple English word, do not make fun of the accent they carry in Japan, it is only a bad joke. —Preceding signed comment added by MythSearchertalk 11:05, 20 August 2009 (UTC)
It does not mean that they made up whole new words or whole new pronunciations for words. Saying they pronounce the words differently and we must include it is only WP:POV
Editors deciding on a case-by-case basis whether or not the romanization should be included sure sounds a lot like WP:POV to me. Systematically including the romanization, on the other hand...
If it is a simple English word, do not make fun of the accent they carry in Japan
That again? Nobody said anything about making fun of Japanese people, really. Well, apart from you, anyway... Erigu (talk) 11:44, 20 August 2009 (UTC)
Insisting on having the Japanese pronunciation of English words with no reason other than stating they pronounce the word differently is making fun of other countries' accent, whether you intended to do so or not. You are trying to emphasize the difference of accents stating in this reply that "Systematically include everything" is faulty since you keep using the argument of they pronounce it differently above. Editors are making case-by-case judgement all the time, it is called common sense. Calling Dragon Quest Dragon Quest and not Doragon Kuesuto is common sense. I will give you that Aerith Gainsborough might not be common sense, but Vincent Valentine is. And obviously, the consensus claimed here is not quite as it claimed. You have 4 people in this particular discussion supporting your arguments, on the other hand, there are 4 disagreeing your view on it should be required(At least for English words rendered in Japanese). It is so disputed that I can hardly agree on the consensus anymore. It should be removed from the guideline, and if disputed, do not say that it is consensus of the MOS-JP. It is not consensus anymore. —Preceding signed comment added by MythSearchertalk 13:56, 20 August 2009 (UTC)
I will give you that Aerith Gainsborough might not be common sense, but Vincent Valentine is.
"Vincent" is also a French name, and "Valentine" is also both a French name and a French word. And even if those weren't used by other cultures, we're talking about a game set in an imaginary universe, and Square is well known for its strange alphabet spelling/pronunciation combos. The pronunciation of that name isn't obvious.
"But if we omit the romanization, people will just assume the English pronunciation is the correct, and based on the kana, it would seem it is indeed, so where's the problem?"
Problem 1): You're assuming English phonetics should be used based on the kana spelling. While I'd agree with you here, that's still an assumption, as educated as it may be.
Problem 2): People might also just wonder why the romanization is missing.
(and I find the fact you keep saying that "mentioning the romanization of a non-Japanese word is like making fun of the Japanese" a tad sad, really...) Erigu (talk) 22:59, 20 August 2009 (UTC)
Consensus isn't determined by the majority, Myth - that's voting.
Yeah, katakana was made to simplify learning words - BUT, modern Japanese now render katakana's use for foreign words mostly, others include for emphasis, onomatopoeia, among other use. Problem is, I'm sure most of us knew kana enough so that when we see the kana script, we know how it sounds and that supposedly makes it redundant - since we already know. What about others? We're not going to tell them that モーグリ is "Moogle" and be done with it, are we?
And really, who is "making fun of the Japanese accent"? I have to ask that because that's been Myth's argument for some time, and yet I don't see any evidence of people making fun of the Japanese accent - aside from Myth calling inclusion of the romaji "stupid". People making fun of the accents - that can't be helped - whole spectrum of people are out there, ignorant ones included. That shouldn't be a reason to deter us from including the romaji.
"Hey guys, バクナムス is pronounced 'Baknamy'. Why? Because it's obvious. Not sure? Go to the kana article, since it's redundant to provide you with the romaji." I wince at the thought.
The discussion's been long enough as it is, and removing the romaji is just going to open a can of worms, complicating guidelines; and I can be sure when an editor sees the romaji missing, they'll be adding it unawares of a guideline asking to remove romaji - because it's just natural for us to guide whenever possible. Anyone got a plan for a compromise? — Blue 15:18, 20 August 2009 (UTC)
Consensus is about how a group of people reached a common ground and agreed on it. If a majority of people in the group agreed on the same thing, you can say that the group has consensus on it. I am not saying we should vote, but it is very obvious that when then are 8 people participating a discussion, which is the group of people concerning this particular problem/dispute, and they are split into two groups of similar number, we can say that we do not have consensus. This is not voting, but an obvious deduction of why we do not have consensus here. And Moogle is not an English phase, thus is not falling into the area of omitting the romanji proposal here. —Preceding signed comment added by MythSearchertalk 17:29, 20 August 2009 (UTC)
Consensus isn't determined by voting, but it can give a picture. In addition several reasons were brought up and there has been no good reason why for English words the pronunciation should be used. Indeed other language articles don't use it for words common in English. Therefore I'd say on the larger picture consensus is against the current version of MOS-JP. And the argument that it's a dialect very much holds water as Indian English is considered a set of dialects and similar to Japan, English is a second language in India.Jinnai 16:32, 20 August 2009 (UTC)
Except that in a few edits made by Megata, such as in List of Castlevania media, removal of "Akumajō" and "Densetsu" is clearly beyond what the consensus had reached, and last I read, the consensus was regarding "English words in article intros". A much clearer guideline should be derived from this consensus. Plus, I don't like it when two brilliant editors get worked up into edit warring.
Try "Include Hepburn romanization for articles about games developed in Japan, except for those with entirely English titles that are not rendered differently than the Japanese title eg. Dragon Quest. This would mean Drakengard is an exception since the Japanese title is Drag-on Dragoon." How is that? — Blue 16:56, 20 August 2009 (UTC)
(ec) I still don't see how including romanization makes the lead too long, so I don't see why including the romanization is such a big deal. Making an exception just for video game titles seems a little pointless and rules creepy to me. Romanization should be the same across the board so we don't confuse people. If we made this rule, people would be constantly adding the romanization because they thought it was just forgotten, so making an exception like this will only create a pile of unnecessary work in the future, whereas simply including the romanization makes it conform with WP:MOS-JA and prevents unnecessary headaches in the future. Including the romanization does not significantly increase the lead of an article, either. ···日本穣? · 投稿 · Talk to Nihonjoe 17:35, 20 August 2009 (UTC)
The dispute here is about the general guideline on all English words used in Japan that have or eventually will have their own articles. Drag-on Dragoon seems to be a bad example(As bad as Aerith Gainsborough I used up there, so I changed to Vincent Valentine as my major example here) If the objects are common English words/names, do not include the romanji since it is only a different accent of English. The rationale of it causing confusion/frustration/complexity to the guideline is not strong enough, this is only bureaucratic. I can agree on that if it is the consensus of the group, but it is obviously not right now. The other rationale used stating because people might want to know is simply irrelevant to the article and swaying away from the intended topic of the article(when it is only a different accent, not a new word needing people to know how to pronounce it). The reader came to search for Dragon Quest, not the Japanese katakana pronunciation method or the Japanese accent. If they look at the article and decided they want to know the pronunciation of ドラゴンクエスト, in which they became interested in when they are reading the article, they can look for it in other articles, like katakana. It is just like any other links in the article, say, role-playing game. The reader might be interested in what RPG is, thus we link them to the right page, not include a part of the RPG page in the Dragon Quest article telling people what it is. I am sure that including the link to katakana in a nihongo template should be very simple if needed. —Preceding signed comment added by MythSearchertalk 17:29, 20 August 2009 (UTC)
Nihonjoe makes a great point about the rule creep. MOS-JA has jurisdiction over all Japan-related articles, so a local video game project cannot overrule what MOS-JA has. WhisperToMe (talk) 18:46, 20 August 2009 (UTC)
It's not that it holds jurisdiction, or that it blankets everything else, but looking back, since the consensus talks about supposed problems with the Japanese romanization, I unearthed this from WP:UEIA. — Blue 19:32, 20 August 2009 (UTC)
There seems to be a few misconceptions in the discussion:
1/ The VG guidelines apply to video game articles in their entirety, not just article intro. There was a discussion which had the words "article intro" in the name of the talkpage section, but the discussion clearly did not concern just intros. If something applies to an intro, there's no reason it shouldn't apply to similar cases in the rest of the article.
2/ I'm not really seeing Nihonjoe's point. MOS-JA says that if romaji is needed then Hepburn Revised Romanization is the system that should be used (as opposed to JSL romanization, etc.). It doesn't say "Insert the romanization of every kana you mention". It states quite clearly that:
"For transliterations from katakana, use the English spelling if available (i.e., Thunderbird (サンダーバード Sandābādo) instead of Sandābādo)."
We have to use Thunderbird instead of Sandābādo, so let us use Dragon Quest instead of Doragon Kuesuto. Megata Sanshiro (talk) 19:46, 20 August 2009 (UTC)
The body of each article, preferably in its first paragraph, should list all common names by which its subject is widely known. I don't think the romanji is widely known in any English case. Beijing, Peking are both wildly known, but Doragon is definitely out of the question. All of the WP:UEIA is not supporting using the romanji at all, it is talking about using the common names and latin transliteration of non latin words, not latin romanization of non latin words. —Preceding signed comment added by MythSearchertalk 19:50, 20 August 2009 (UTC)
Yes, and following the example you quote, it would be Dragon Quest (ドラゴンクエスト Doragon Kuesuto?). ···日本穣? · 投稿 · Talk to Nihonjoe 20:15, 20 August 2009 (UTC)
Mmh, no, that's precisely what is being discussed here. It's not in any of the guidelines. Megata Sanshiro (talk) 20:35, 20 August 2009 (UTC)
So far, in most removal edits I've seen concerning this discussion - non-English names were also removed of their romanization. That wasn't in the discussion as far I can read in the archive.
And in WP:UEIA there's also the second line, "When the native name is written in a non-Latin alphabet this representation should be included along with Latin alphabet transliteration". Romanization encompasses several transliteration methods, thus UEIA supports romaji usage.
An average reader wouldn't be able to tell which is the English word and which isn't in 葛葉ライドウ 対 アバドン王, without the use of romaji. — Blue 20:53, 20 August 2009 (UTC)
Mmh, no
Mmh, yes. It would be Dragon Quest (ドラゴンクエスト Doragon Kuesuto?), like Nihonjoe said. It's right there in the manual of style, in the part you quoted above even:
"For transliterations from katakana, use the English spelling if available (i.e., Thunderbird (サンダーバード Sandābādo) instead of Sandābādo)."
It doesn't say use "Thunderbird" instead of "Sandābādo". It says use "Thunderbird (サンダーバード Sandābādo?)" instead of "Sandābādo". It doesn't say anything about removing the romanization altogether. Erigu (talk) 22:59, 20 August 2009 (UTC)
Read the guideline again. It does say use "Thunderbird" instead of "Sandābādo". The kana and romaji in parentheses (they're parentheses inside parentheses...) is just for the sake of explanation in this guideline; it's not necessarily what should be in articles. Click on "Edit" there and notice how that part is written with real parentheses rather than the {{nihongo}} template. The actual Thunderbird (train) article doesn't have the romaji either. Nowhere in the guideline is it written "Use romaji for English words written in kana", and this is precisely why we are currently discussing the issue here. Megata Sanshiro (talk) 01:13, 21 August 2009 (UTC)
Read the guideline again. It does say use "Thunderbird" instead of "Sandābādo". The kana and romaji in parentheses (they're parentheses inside parentheses...) is just for the sake of explanation in this guideline
No, it's not. I'm not sure what else to tell you, really.
The actual Thunderbird (train) article doesn't have the romaji either.
It should. And now, it does.
(and that article wasn't using the nihongo template either (fixed)... obviously, these things happen) Erigu (talk) 02:06, 21 August 2009 (UTC)
You are pushing your POV to the articles. The guideline is in dispute right now, it is a bad idea to change the sample articles. —Preceding signed comment added by MythSearchertalk 07:03, 21 August 2009 (UTC)
Like Nihonjoe said... Erigu (talk) 09:25, 21 August 2009 (UTC)
You don't seem to understand, you are trying to push your own POV during a dispute. Yes, Nihonjo agreed that there is a dispute ongoing, and the article in question you edited is one of the dispute targets. So your reply is only further pushing your POV. —Preceding signed comment added by MythSearchertalk 09:54, 21 August 2009 (UTC)
You don't seem to understand. I'm simply following this manual of style. Erigu (talk) 10:18, 21 August 2009 (UTC)
You are an involved party in the dispute, you should be well aware of the dispute. I am telling you your actions are highly similar to using the current guideline to WP:Game the system to push your own WP:POV in the edits. Nihonjoe's editing out the dispute tag is also very questionable, it may be good faith, but for someone involved in the argument should not remove the tag. The guideline states Include in all case and it is obviously disputed and being discussed here. Yet you claimed your edit is only following the guideline is obviously point 4 at WP:WL —Preceding signed comment added by MythSearchertalk 10:33, 21 August 2009 (UTC)
You are an involved party in the dispute, you should be well aware of the dispute.
I know what the dispute is about, too. Which is why I don't see the problem, here: why would the Raichō article be concerned by a new WikiProject Video games guideline? Erigu (talk) 10:45, 21 August 2009 (UTC)
Because the current discussion includes ALL English terms used in Japanese which is rendered in katakana. Thunderbird is one of them. It seems like you are really not clear of what the dispute is, and by that, no wonder you made such an edit. —Preceding signed comment added by MythSearchertalk 12:15, 21 August 2009 (UTC)
MythSearcher, the guideline doesn't actually say "Include romaji in all cases", it says "Romanization: Revised Hepburn romanization should be used in all cases". IMO, this means "If romaji are needed, then use the Revised Hepburn romanization system in all cases" (as opposed to other systems of romanization). The guideline never actually states when romaji are needed, which indeed is why we are discussing the issue here on this talk page. Megata Sanshiro (talk) 12:02, 21 August 2009 (UTC)
If this is so, then Erigu's edit is even more controversal. —Preceding signed comment added by MythSearchertalk 12:15, 21 August 2009 (UTC)
The guideline never actually states when romaji are needed
Always? Obviously (?), you can't include Japanese text in kanji and/or kana without romanizing it for the Japanese-illiterate readers. Erigu (talk) 12:55, 21 August 2009 (UTC)
This is your POV, and what this discussion is all about, is trying to find out if the romanji is redundant or not. You edit and actions are pushing your own POV of it being not redundant, that is doing something that is not up to consensus using the name of following a guideline. The problem is that the guideline in question is the centre of dispute, so your actions are only worsening the situation since it is controversal. —Preceding signed comment added by MythSearchertalk 14:40, 21 August 2009 (UTC)
This is your POV
It's just common sense, really. Erigu (talk) 15:05, 21 August 2009 (UTC)

And I still don't see the point of saying "Thunderbird (サンダーバード Sandābādo?)" other than to point out Japanese have weird dialects. If we do that we might as well start doing "tomato (tə-ˈmā-tō, tə-ˈmä-tō, tə -ˈma-tō)"Jinnai 23:13, 20 August 2009 (UTC)

Japanese have dialects, but that has nothing to do with this discussion. Almost anyone who is speaking in a language other than their native language will pronounce things differently, and pronunciation of English by Japanese people is no different. It is not a different dialect, though. It's simply pronunciation. The point of having Thunderbird (サンダーバード Sandābādo?) instead of Thunderbird (サンダーバード?) is to provide a pronunciation guide for the katakana because it is not pronounced the same as "Thunderbird". Including the rōmaji does not expand the size of the lead sections of articles significantly (unless, as I previously mentioned, the lead is too short and in need of significant expansion). The whole point of articles here is to provide useful, complete, and accurate information to the reader, and for games originally developed in Japan, the Japanese title and the transliteration of that title are an important part of that coverage in the article. ···日本穣? · 投稿 · Talk to Nihonjoe 00:23, 21 August 2009 (UTC)
Like I said, it is merely an accent of the Japanese, not a whole new pronunciation. People not knowing how the kana are pronounced have nothing to do with the article. The article provides how it is written, and if you want to provide the information, link the kana's to the appropriate article like the katakana article, like all the other interwiki links in all the articles. —Preceding signed comment added by MythSearchertalk 07:03, 21 August 2009 (UTC)
it is merely an accent of the Japanese
It's not a matter of "accent" at all...
People not knowing how the kana are pronounced have nothing to do with the article.
Are you saying the English Wikipedia isn't for people who don't know their kana? Erigu (talk) 09:25, 21 August 2009 (UTC)
If the title is suppose to be the English word which was used in translation there is no need for phonetics; we don't do that for other languages with non-roman script. I mean there are always exceptions to this rule, but that can be done by simply making it optional and using common sense to decide when transliterations are necessary and when they aren't when the Japanese is used for an equivalent English word.Jinnai 03:07, 21 August 2009 (UTC)
I am saying putting a link to the katakana article is much better than including all of the pronunciation for the English words. This is the English wikipedia, readers have no problem in pronouncing Thunderbird in English, and if they want to know what the katakana サンダーバード stands for, a link to katakana is more than sufficient and enough like all other interwiki links. Let's have the Thunderbird (train) as an example, I am interested in it, and is reading the article, and I have no idea what limited express is. So I click on the link, and it brings me to the page giving me that precise information. We don't try to explain limited express in the thunderbird article, we have interwiki links for that purpose. —Preceding signed comment added by MythSearchertalk 09:54, 21 August 2009 (UTC)
if they want to know what the katakana サンダーバード stands for, a link to katakana is more than sufficient
And including the romanization is even simpler. Erigu (talk) 10:18, 21 August 2009 (UTC)
And including the romanization is making the page overwhelmed with information that some readers are not looking for. Especially the subject being discussed here is about terms that are originally English terms. —Preceding signed comment added by MythSearchertalk 10:33, 21 August 2009 (UTC)
"Overwhelmed"? It's only a few extra characters... Erigu (talk) 10:45, 21 August 2009 (UTC)
A few? hmmm... I wonder what does a few means. Sandābādo is 9, okay, it is still less than 10, I'll give you that. Doragon Kuesuto, 15, Vinsento Varentain, 18, not so few. Also, this is talking about including everything is overwhelming, Thunderbird (サンダーバード Sandābādo) as oppose to Thunderbird (サンダーバード) is 50% more words on screen in most cases. —Preceding signed comment added by MythSearchertalk 12:23, 21 August 2009 (UTC)
Doragon Kuesuto, 15, Vinsento Varentain, 18, not so few.
That's "few", in my opinion.
Thunderbird (サンダーバード Sandābādo) as oppose to Thunderbird (サンダーバード) is 50% more words on screen in most cases.
For select words, not the entirety of the article. Definitely not "50% more words on the screen"... Erigu (talk) 12:55, 21 August 2009 (UTC)
Ok, screen is not the correct word, but a word is split into 3 equivalent meaning strains, this is the problem here. You need 18 extra characters for a word with only 5 syllable, how few is that? Don't forget the katakana is already taking up 11 kana spaces(23 character spaces) Think more on the wikipedia server load for all these words. —Preceding signed comment added by MythSearchertalk 14:31, 21 August 2009 (UTC)
The server load? Seriously? Erigu (talk) 15:05, 21 August 2009 (UTC)
Seriously, if every single editor thinks like you, that We should include it just because we can and it does not hurt to be included since it is just a few words, we will have articles overwhelmed with tons of redundant data and it will cause the server load to raise. Think about the few thousand articles that can be rid of the few characters, which will also reduce server load when readers loading the pages. Do not underestimate the few characters in articles, a guideline is influencing thousands of articles and many of those articles are being loaded from time to time. It may seem small for one article, but if no precaution is taken, it can get out of hand. You may not have set up your own server and website before, but a few hundred people loading one single page can already cause quite some bandwidth problem if the page is a few bytes longer for small servers. I know wikipedia does not run on small servers, but the bandwidth is still limited and thus there is the good o WP:IINFO policy. I am not stating the IINFO supports the removal of this particular guideline, I am only using it to tell you that we are working on limited resources. —Preceding signed comment added by MythSearchertalk 16:16, 21 August 2009 (UTC)

I don't think this passage from VG/GL has been brought up before:

Japanese titles should only be provided for games of Japanese origin whose official English name differs significantly from its Japanese name. Phonetic transcriptions are, as a rule, not considered to be significantly different and thus do not warrant the inclusion of Japanese titles. However, games known in English-speaking countries by their phonetic Japanese titles (e.g. Katamari Damacy) are an exception to this rule; these games should also have their Japanese titles included for clarity.

Though personally I am not very fond of that guideline, it would mean that even the kana for, say, Kingdom Hearts should not be included – the exception states, however, that it should be given if the English release adopted the Japanese title (consisting of actual Japanese words). More examples of that exception would be Shin Megami Tensei III: Nocturne and Ōkami. Prime Blue (talk) 10:50, 21 August 2009 (UTC)

I did mention the first sentence, but yeah the second sentence, in bold, is even more clearcut. Megata Sanshiro (talk) 12:02, 21 August 2009 (UTC)
Just so we're clear: would you consider the Japanese text (i.e. the title in kanji and/or kana) superfluous, in such cases? Erigu (talk) 13:01, 21 August 2009 (UTC)
Japanese titles including kanji would apply to the aforementioned exception as they would either be significantly different from the English title (e.g. Seiken Densetsu), or would be adopted for the English release along with the Japanese pronunciation (e.g. Ōkami). Furthermore, it seems the issue discussed here is already covered in existing guidelines, though not in VG/GL, but in MOS-JP itself: Give the romanization for any name or term written in kanji or kana when the Japanese pronunciation is different from the English pronunciation.
So, the romanization would have to be dropped from titles like Kingdom Hearts, Seiken Densetsu 2 (since the pronunciation of the number is two/tsū and, thus, does not differ) and Akumajō Dracula (same here with Dracula/Dorakyura), whereas it would have to be kept for titles such as Mario Kart 64 (since the pronunciation of the number is not sixty-four/shikkusutī-fō but rokujūyon and, thus, differs).
The guidelines from my last edit dictate that the kanji and kana would have to be dropped from Kingdom Hearts (though I'm personally not happy with removing it) and, of course, kept for Seiken Densetsu 2 and Akumajō Dracula due to the exceptions. Prime Blue (talk) 15:13, 21 August 2009 (UTC)
So instead of systematically mentioning the original title of Japanese video games and applying this manual of style (by including both the Japanese characters and their revised Hepburn romanization for Japanese-illiterate readers), like we do for Japanese movies or comic books, we'd make an exception when the original title is in English already (in which case, no Japanese characters and no revised Hepburn romanization), or another kind of exception when the original title isn't in English but its revised Hepburn romanization, even if it differs, would be deemed "obvious enough" to be redundant (in which case the Japanese characters would stay but their revised Hepburn romanization wouldn't), and let's not forget the various exceptions to the aforementioned exceptions (Mario Kart 64, Castlevania Legends, Shadow of the Colossus... all exceptions for different reasons).
Sure, it would only save us a few characters here and there, but it's so straightforward. Erigu (talk) 17:24, 21 August 2009 (UTC)
I often find myself referencing Wikipedia when I need to find out how to type a particular term in katakana, and the policy to always include those has made that possible. I can see the romanization being redundant, but we should keep the katakana even with names like Kingdom Hearts - the exact katakana used is often elusive and hard to reverse engineer. Doceirias (talk) 18:44, 21 August 2009 (UTC)
The kana or kanji I don't have a problem with, especially when it's an untranslated work/element, but romaji can be an lot extra. Take FE, School Rumble episode 26 - 突然の「さよなら」... 迷い込んだラビリンス... あなたはだれ? ... 教えて.「すれちがい」「片想い」とどけ, ボクの気持ち.とどけ, ワタシの想い.たぶん一度しかない季節, 青春の1ページ.これが最後のチャンス, 確かめたい... キミの気持ち.伝わる言葉, 伝わらない想い.あの日の告白, 永遠の一日, だけど... いつまでも続いていく, わたしたちの「いま」.そして明日へ... 「スクールランブルフォーエバー」- the romaji now is - Totsuzen no "sayonara"... Mayoikonda rabirinsu... Anata wa dare? ...Oshiete. "Surechigai" "Kataomoi" Todoke, boku no kimochi. Todoke, watashi no omoi. Tabun ichido shikanai kisetsu, seishun no 1 pēji. Kore ga saigo no chansu, tashikametai... Kimi no kimochi. Tsutawaru kotoba, tsutawaranai omoi. Ano hi no kokuhaku, eien no ichinichi, dakedo... Itsumademo tsuduiteiku, watashitachi no "ima". Soshite ashita e... "sukūru ranburu fōebā. Now that episode is an exception as it is believed to be the longest tv (or at least anime) episode in existance (i still need to find someone with a copy of something like Japanese guiness record book), but School Rumble other titles like that, FE 2nd ps2 game School Rumble Ni-Gakki Kyōfu no (?) Natsugasshuku! Yōkan ni Yūrei Arawaru!? Otakara wo Megutte Makkō Shōbu!!! No Maki (スクールランブル二学期 恐怖の(?)夏合宿! 洋館に幽霊現る!? お宝を巡って真っ向勝負!!!の巻?). My point is at this point adding both original and romaji to the pose just clutters things up.Jinnai 19:46, 21 August 2009 (UTC)
You're quoting an extremely rare example. The vast majority of Japanese video game titles (which is the sole topic of this discussion) are quite short, usually only 2-4 words, so you're comparing apples to black holes. I would say that 95% or more of the Japanese titles for video games are going to be somewhere around 2-4 words, and therefore will not cause much change in length of the lead. As for Mythsearchers "Oh no! What about the bandwidth?!?" argument, 18-20 bytes (or even 50 bytes, for that matter) is not going to make any significant difference in server load, so that argument is simply specious. ···日本穣? · 投稿 · Talk to Nihonjoe 20:03, 21 August 2009 (UTC)
Still there is no common sense principle applied here for making execptions. Just a blanket "add romaji to everything, no questions asked, no exceptions made." principle being applied here.Jinnai 21:22, 21 August 2009 (UTC)
I don't consider that a bad thing, personally. Erigu (talk) 00:39, 22 August 2009 (UTC)

I think the existing guidelines for dropping or including the romanizations are pretty clear-cut and easy to follow, and hence should not pose any problems. I understood them the first time I read them, and, judging from Erigu's answer, another user also did. I don't have any objections against adding another guideline to always include the kana writing for Japanese works, though. As for Jinnai's problem with the length of some terms: There's always the possibility to add notes (as a reference). I think that is a good solution for longer Japanese titles where the romanization and translation would only hinder the readability of the article. Prime Blue (talk) 00:22, 22 August 2009 (UTC)

I still believe it serves absolutely no purpose for English words and you'll find other languages do not include the pronuciations as such, though they do include the original writing. It does not add to the average Wikipedian reader to know that ドラゴンクエスト is pronounced Doragon Kuesuto other than to not an accent of English and perhaps make fun of it. We do not promote individual usage of accents on other wikipedia articles and that would be what in essence would occur here. And yes it is dialeectal the same as someone from France coming sepaking English would have a cerain accent.Jinnai 01:17, 22 August 2009 (UTC)
I have suggested to add "always include the kana writing for Japanese works" as another guideline. I guess you just confused that with the Hepburn romanization. Prime Blue (talk) 01:45, 22 August 2009 (UTC)
Using the original kana or kanji (for rare instances it may occur...see above for an example) I have no problem with as it is worthy of mention. My problem is when it extends to romanization of word when those would basically be pronounced as an accented version of the name. When the translation isn't the same it may be apporpritate.Jinnai 01:55, 22 August 2009 (UTC)
It's not a matter of "accent". French text is natively in Latin alphabet, whereas Japanese text isn't. Which is why this manual of style says the latter should be accompanied by a (revised Hepburn) romanization: so the Japanese-illiterate reader would know "what those symbols stand for".
And yes, I understand the "redundancy" argument for titles that are (or stand for) plain English words in the first place, but I'm concerned with the layers of complexity your proposed solution adds to the guidelines, and the arguments it will undoubtly generate in the (many) not-so-clear-cut cases. That's a lot of trouble for very little reward. Erigu (talk) 02:01, 22 August 2009 (UTC)
And, how complex is that? When it is English, do not include the romanization but include the kana to indicate it is originally a Japanese work. is complex? It is not very convincing. —Preceding signed comment added by MythSearchertalk 05:41, 22 August 2009 (UTC)
We've been over several exceptions and oddities already... Erigu (talk) 05:57, 22 August 2009 (UTC)
The simplest way is If greatly different from the English title, include the romanization. If there is no consensus, do not make it a guideline, let individual editor decide whether it is necessary or not. If there is argument, bring it here or relevant project and let consensus decide on a case-by-case basis, or simply respect the original editor and keep the article at that state. —Preceding signed comment added by MythSearchertalk 07:21, 22 August 2009 (UTC)
Again, letting editors decide (or argue) on a case-by-case basis just to save a few characters... Simply not worth the trouble, in my opinion. Erigu (talk) 14:22, 22 August 2009 (UTC)

Again, there already are easily understandable and established guidelines pertaining to the inclusion or omission of Japanese titles and/or romanizations, which I have posted in my first two contributions above. It's hard to see what the prolonged discussion is now about – especially the comments from MythSearcher and Jinnai, since we seem to have the same point of view on the matter. I would like to have a clarification on that by those two users.

As for the argument raised by Erigu, it's not a "proposed solution" but rather existing guidelines. MOS-JP states to omit the romanization if it is not different from the English pronunciation. All exceptions to the rule given so far were easy to classify for me and Erigu. I could see some room for discussion if a very ambiguous example would be pointed out, along with an explanation as to why the case would be ambiguous. Prime Blue (talk) 09:39, 22 August 2009 (UTC)

Well, I am not replying to counter your arguments, but it seems like there are people who do not agree to the idea. Since people try to interpret the guidelines differently, I think it is best to not stick to that in this discussion and make sure everyone agreed on the same thing. Since Erigu made an edit to the thunderbird (train) article and claims s/he is following the guideline, which is different from what you are saying the guideline is. —Preceding signed comment added by MythSearchertalk 10:16, 22 August 2009 (UTC)
There are a few cases for non-video games titles. Again I'll use School Rumble (because I know it's an example). Kenji Harima (播磨 拳児 Harima Kenji?) as one of the main protagonist would by this guideline clearly get a romanization as his name is Japanese. However, Harry Mackenzie (ハリー マッケンジー Harī Makkenjī?) would need the romanization for the reader to understand why his name is signifigant to several of the jokes in the series. At the same time, that case is unusual which is why I support an optional usage for clearly English Harry Mackenzie (ハリー マッケンジー Harī Makkenjī?) words.Jinnai 13:44, 22 August 2009 (UTC)
As for the argument raised by Erigu, it's not a "proposed solution" but rather existing guidelines. MOS-JP states to omit the romanization if it is not different from the English pronunciation.
But that's in a section called "Japanese terms", and the example that follows is "Mount Fuji". I'm thinking this might just be a poor choice of words (not the only one, in my opinion, but I've raised that issue several times in the past and never got much feedback, so...).
All exceptions to the rule given so far were easy to classify for me and Erigu.
I wouldn't say that... Erigu (talk) 14:22, 22 August 2009 (UTC)
As for Erigu's two replies:
1) The Japanese terms guideline states Give the romanization for any name or term written in kanji or kana when the Japanese pronunciation is different from the English pronunciation. The wording any name or term covers work titles, which is also shown in the ancillary template section that states: There is a template (Template:Nihongo) to help standardize the entries for Japanese terms. In the same paragraph, a Japanese work title for a dictionary, the Shin Meikai Kokugo Jiten, is given as an example for the template usage.
2) Two additional examples (Castlevania Legends and Shadow of the Colossus) for the exception were given by Erigu without any problems, so that leaves me thinking that the guideline is very clear. If there are still concerns as to its comprehensibility, additional examples could be added to the manual of style to clarify possible ambiguities.
As for the example given by Jinnai: Disregarding the notability of the misunderstanding scene as I'm not familiar with this anime and how frequently it is used – even if the romanization is removed from Harry Mackenzie, I don't see much of a problem raised as the context with the similar pronunciations is explained explicitly in the article (again, though, I can't tell whether the misunderstanding itself is even notable or to be classified as mere trivia). Prime Blue (talk) 15:07, 22 August 2009 (UTC)
Give the romanization for any name or term written in kanji or kana when the Japanese pronunciation is different from the English pronunciation.
What "English pronunciation"? When is the Japanese pronunciation identical to the "English pronunciation"? Just what exactly are we talking about, here?
That's why I said that might be a poor choice of words... Erigu (talk) 15:35, 22 August 2009 (UTC)
Kingdom Hearts and Seiken Densetsu 2 are examples for an identical pronunciation (the "2" is pronounced English as "two/tsū" in Japanese, thus dropping romanization), whereas Mario Kart 64 and Hoshi no Kirby 64 are two examples with a different pronunciation (as "64" is pronounced rokujūyon and not as the English sixty-four/shikkusutī-fō, thus keeping the romanization). Prime Blue (talk) 16:08, 22 August 2009 (UTC)
I wouldn't call "Kingdom Hearts" and "kingudamu hātsu" "identical pronunciations"...
The "Mount Fuji" example makes me think that particular section is about cases where we're not already using revised Hepburn romanization of the original Japanese word/name in the body of the text. When we are (like here, for example), including the revised Hepburn romanization a second time in the nihongo template would indeed be redundant. And on the contrary, when there's a difference (like for "Mount Fuji", as compared to the original "Fuji-san"), revised Hepburn should be included in the template.
And if I'm right about that, if that's what that section is really about, I think talking about "English/Japanese pronunciations" wasn't such a good idea... Erigu (talk) 16:29, 22 August 2009 (UTC)
I'll bold the important parts of what MOS-JP states: Give the romanization for any name or term written in kanji or kana when the Japanese pronunciation is different from the English pronunciation.
It is talking about the differences between the Japanese and English pronunciations, not the differences of Japanese and English names and terms (and the resulting redundancy of a second romanization in such a case). And it uses the Mount Fuji example as one such case.
And, according to the guideline, Kingdom Hearts is identically pronounced in both languages, as, per your logic, katakana would always be different from the English pronunciation (due to the syllabic concept of Japanese) and its romanization would have to be included. But the MOS-JP guideline says "written in kanji or kana", which explicitly includes katakana. And, again, the issue is expanded on in VG/GL with Phonetic transcriptions are, as a rule, not considered to be significantly different and thus do not warrant the inclusion of Japanese titles. Prime Blue (talk) 00:13, 23 August 2009 (UTC)
It is talking about the differences between the Japanese and English pronunciations
If so, it doesn't make sense, as Japanese and English pronunciations pretty much always differ, basically.
I'd like the opinion of some "veterans" on this matter, if you don't mind... Erigu (talk) 01:26, 23 August 2009 (UTC)
I'm in the opinion that the guideline shouldn't compensate anything for our average readers. It should be rewritten for editors to guide our average readers how to pronounce the kana/kanji pronunciation, and not just for editors who's at least well-versed in kana, which is how they could tell if the kana was explicitly invoking an English term. Phonetics transcriptions are not considered to be significantly different by who, actually? Having the ones who know kana/kanji determine whether the romanization gets included/excluded is not helping to create a comprehensive Japanese-origin article.
As well, when there's the case of the title combining two or more languages at one time ie. デビルサマナー 葛葉ライドウ対超力兵団, of course removing the romanization isn't ideal - considering that even the literal translation is different. The same is also applied to 悪魔城ドラキュラ - removing the romanization completely for these titles would render the characters unreadable, except for those who know kana/kanji and that's withholding information. — Blue 04:53, 23 August 2009 (UTC)
These are examples where the guidelines dictate to include the Japanese title, because they clearly differ from the English release titles. They would thus be rendered something like the following paragraphs in the respective articles (most likely using a reference note for the lengthy Japanese title inclusion in Shin Megami Tensei):
Castlevania, known as Akumajō Dracula (悪魔城ドラキュラ?, lit. "Devil's Castle Dracula") in Japan...
Shin Megami Tensei: Devil Summoner: Raidou Kuzunoha vs. The Soulless Army, known as Devil Summoner: Kuzunoha Raidō tai Chōriki Heidan (デビルサマナー 葛葉ライドウ対超力兵団?, lit. "Raidō Kuzunoha vs. The Super-Powered Army Corps") in Japan...
The Japanese titles for those two have to be included as they are not just "phonetic transcriptions", which VG/GL excludes. The romanization is omitted because the Japanese and English pronunciations do not differ. The katakana transcription and how it is uttered in Japanese is merely of phonetic nature. The kanji pronunciation is still given.
I don't think we should start to even deal with the "average reader" issue now, as it is very debatable if they would be interested in seeing the katakana. If they don't know katakana, why would they want to have it in the article, or why would they assume it means something different than the English title if it is included and not elaborated on, etc. I guess that is also the reason why VG/GL suggests to drop the katakana spelling from an article altogether in case of just a phonetic transcription (e.g. Kingdom Hearts). Prime Blue (talk) 11:48, 23 August 2009 (UTC)

Individual opinions and suggested compromises

The discussion has reached a considerable length now and is not very easy to read for users not engaging in it. Since that issue has been discussed in the past, but people eventually just stopped (because they probably got tired of it), I suggest a summary of everyone's opinion and compromise on the matter.

Prime Blue's point of view:

  • The existing guidelines in MOS-JP cover the romanization issue sufficiently.
  • This guidelines are compromising, as they do not suggest to always drop the romanization but rather classifies inclusion per differences in the Japanese and English pronunciations (especially apparent for example in Biohazard: The Darkside Chronicles and Baiohazādo / Dākusaido Kuronikuruzu, where the romanization would be kept).
  • If there are concerns as to the guidelines' understandability, explicitly mention the katakana case and add more examples for different and identical pronunciations to it (possibly the ones in this discussion).
  • The existing guidelines in VG/GL cover the inclusion of Japanese titles and an exception to this rule (phonetic transcriptions).
  • As a compromising additional guideline to VG/GL, I suggest to always include even the katakana for mere phonetic transcription titles of Japanese works, though to drop the romaniation per the existing MOS-JP guidelines in such a case.
  • An additional guideline in VG/GL or MOS-JP that states that if a Japanese title, its kanji/kana form and possibly its romanization and literal translation are too long and it hinders readability of the article, it is to be included as a note in the references.

Prime Blue (talk) 11:48, 23 August 2009 (UTC)

I suggest to always include even the katakana for mere phonetic transcription titles of Japanese works, though to drop the romaniation per the existing MOS-JP guidelines in such a case.
I still don't think that's what that particular section of the manual was about, but I'd like to hear some "veterans"' opinions... Erigu (talk) 14:30, 23 August 2009 (UTC)
First off, these articles aren't about teaching people to speak Japanese. We do have articles like Kanji that can help with that much better. And saying phoetics differ from the English is a backdoor way of saying there's no exceptions because the way Japanese and English work is so fundimetally different. Other articles, even those without roman alphabets do not have their spellings requir3ed for well known English names/terms.Jinnai 02:14, 24 August 2009 (UTC)
I'm no veteran, but the line Give the romanization for any name or term written in kanji or kana when the Japanese pronunciation is different from the English pronunciation. doesn't ask the editor to drop the romanization when pronunciations are similar. In fact, that line had been something of a contentious issue.
There was a discussion in the archives about the issue, including notions that "Wikipedia isn't the place to teach people Japanese". No absolute decision was reached, except the opinion that it is perfectly acceptable to have something such as Kingdom Hearts (キングダムハーツ Kingudamu Hātsu?), and that the romanization was used in place of the IPA for Japanese, in which the latter was "tricky to place unless you're a trained linguist".
An additional guideline to VG/GL (why is it just Video Games and why is it discussed here, I wonder?): As for the kana, I suggest to always include the non-latin alphabet for the game's title in the article intro; it's important to establish that the games are developed in their native country ie Japan, and they should retain the native title spelling, even if it is written, say, in katakana.
This is a side case, but I suggest that romanization be retained for, save for article titles, all fictional names even if the names use common English names or that there is little to no phonetic differences. The removal of romanization isn't warranted by the guidelines, at least with the romaji retained, the kana characters can be at least somewhat read by anyone, even if the demographic is small.
I agree on PrimeBlue's inclusion for the "footnote". This was something which wasn't included in the guideline edit, even though it was in the WP:VG discussion. — Blue 04:57, 24 August 2009 (UTC)
EDIT: Just to clarify, my stance is still to include romanization for the Japanese titles; it's the simplest approach, much like the whole "use commonly used terms" the Wikipedia's been using.. — Blue 06:55, 24 August 2009 (UTC)
Including the rōmaji in the footnotes is next to useless as it's very unlikely someone is going to click on the footnote link just to read the rōmaji (and the footnote text will likely be far down at the bottom of the article, too, so it won't be helpful to anyone trying to read the Japanese title anyway). I still maintain that something like "Kingdom Hearts (キングダムハーツ Kingudamu Hātsu?)" is hardly intrusive, and that really long titles are a very rare exception, so arguments that the rōmaji makes the lead too long are specious at best. We need to keep the guidelines as simple as possible so that people aren't having to check them all the time to see if there's an exception for this, that, or the other thing. Including the rōmaji is not going to teach anyone Japanese, either, so "we aren't here to teach Japanese/kanji/kana" is also a specious argument. The least work for everyone will be sticking to the "use the {{nihongo}} template, and fill in the English, Japanese, and rōmaji fields completely" guideline for Japanese in prose anywhere on the site. ···日本穣? · 投稿 · Talk to Nihonjoe 06:25, 24 August 2009 (UTC)
As a clarification to Nihonjoe: The footnote would affect the whole long Japanese title, not only the kanji/kana form and/or romanization, a case that is not as rare as one might think. It would look something like in the The Legend of Zelda: Oracle of Seasons and Oracle of Ages article. Also, here and here are the two previous discussions about the "Give the romanization..." guideline Bluerfn mentioned. Prime Blue (talk) 12:07, 24 August 2009 (UTC)

But again, I request everyone to keep the discussion down for a moment and ask Erigu, Jinnai and Nihonjoe (and others still following the discussion) to reply with a summary of their opinion and a possible compromise on the matter, as I and Bluerfn did (preferrably in the format I used above since it is easy to read). Prime Blue (talk) 12:07, 24 August 2009 (UTC)

I think I made it clear enough already, but...

Erigu's point of view:

  • The original Japanese text (by which I mean kanji and/or kana) should be included at least once somewhere for Japanese terms, Japanese names, titles of Japanese works, made-up names and terms from Japanese works. No exceptions, not even for titles that are made up of non-Japanese words, or fictional names that clearly aren't meant to be of Japanese origin.
  • Japanese text should never be left without a revised Hepburn romanization, for the sake of Japanese-illiterate readers. Same thing as above: no exceptions.

... Yes, I suck at compromises. Erigu (talk) 13:12, 24 August 2009 (UTC)

  • Comment I keep my POV, that in most cases, romanization is only necessary when the Japanese is either kanji, non-English words or made up words. If it is common English, I cannot see people confusing the pronunciation. If it is a direct romanji of the kana, it is more redundant to include the romanization. It is not convincing to say that it is hard to make exceptions, if there is no consensus reached, and no new party came in to make a majority !vote and keep the guideline as the consensus, I see no reason to keep the guideline in such controversial terms. Prime Blue is saying the guideline covers everything clearly, but it is obvious that about half of the involved party here think otherwise and keeping the guideline is not going to solve the problem of different people interpret the guideline differently. It there is no consensus, I am afraid that the guideline has to go, to prevent it being used in edit warring. If there is no guideline on this, whenever there is a dispute, solve it case by case or respect the original editor. —Preceding signed comment added by MythSearchertalk 15:28, 24 August 2009 (UTC)

Jinnai's point of view:

  • Keep kana/kanji from the original work for titles and elements within it (VG guidelines also cover all related works).
  • Anytime there is a major difference between the kana/kanji and the Japanese, the romanization should be used. In the case of post-Edo names, always use it (as the naming will be reversed: surname then personal name in Japan per the guidelines).
  • For words, especially English words, common sense should be applied to determine whether the difference would be signifigant enough to warrant romanization. In general, on or two minor sylable differences isn't enough.
    I am specifically not stating only English words as words like katana have become well known in the English-speaking world and a game with English word(s) and Japanese word(s) common to the English language may take stylation liberities and change all the text to hiragana/kanji.Jinnai 21:49, 24 August 2009 (UTC)

Megata's point of view:

  • I essentially agree with Prime Blue and Jinnai. Always include the kana/kanji for Japanese titles and names, and include the romanization only when there is a significant difference between the official title and the romanized title. A few examples (I left out the translations in these for clarity, but obviously they should be indicated when necessary):
  • No significant difference: Kingdom Hearts (キングダムハーツ?), Akumajō Dracula (悪魔城ドラキュラ?), Vincent Valentine (ヴィンセント・ヴァレンタイン?).
  • Significant difference: It's a Wonderful World (すばらしきこのせかい Subarashiki Kono Sekai?), Baten Kaitos: Eternal Wings and the Lost Ocean (バテン・カイトス 終わらない翼と失われた海 Baten Kaitosu: Owaranai Tsubasa to Ushinawareta Umi?), Tidus (ティーダ Tīda?).

Megata Sanshiro (talk) 09:50, 25 August 2009 (UTC)

Since I didn't put it the way everyone else did, I'd settle with this:

Bluer's POV:

  • Utilize the {{nihongo}} template completely: fill in the English, Japanese letter and romaji as recommended below:
  • "Significant difference" is a subject of POV between those who know kana and those who don't, and it is common sense to include a kana pronunciation for readers, even if the demographic of those who care is small. Therefore Hepburn romanization should be included where kana/kanji has been written, especially for fictional character/location names and terms such as Ivalice (イヴァリース Ivarīsu?), Moogles (モーグリ Mōguri?) and Laguna Loire (ラグナ・レウァール Raguna Rewāru?).
  • In the case where the English release title is different than the Japanese release title, including where there is a combination of languages, the {{nihongo}} template should be used for the Japanese release title, for example in "Castlevania" as: "Castlevania, known is Japan as Akumajō Dracula (悪魔城ドラキュラ Akumajō Dorakyura?, lit. "Devil's Castle Dracula")" wherein "Japanese title, kanji/kana title, pronunciation guide, unofficial literal translation".
    • Addendum: I'd like to point out these "literal translations" are also educated guesses by editors at best, and is in the same level as the pronunciation guide - I'd ask whether editors want literal translations to be removed as well? — Blue 12:17, 25 August 2009 (UTC)
I agree overall, and I also think the difference should be made (when it exists in the first place, naturally) between the English and Japanese release titles, indeed. In Megata Sanshiro's examples, the The World Ends with You article makes that distinction, whereas the Valkyrie Profile: Covenant of the Plume one doesn't and uses the nihongo template for the English release title. Some consistency would be nice, in my opinion...
(by the way, Megata Sanshiro, I noticed you were removing the bold font for Japanese titles, and I'm not sure that's helping with the readability of the article intros...) Erigu (talk) 13:09, 25 August 2009 (UTC)
I looked up WP:BOLDTITLE which says "do not boldface foreign names not normally used in English", but those "that use the Roman alphabet should be italicized if they are not bolded". I didn't know that either, so I'll have quite a few corrections to make to my former edits...uh...
As for Bluerfn's "literal translations" remark: VG/GL suggests to always include a translation for the Japanese title and to put it in quotation marks to signify its unofficial status. The translations are frequently assessed and improved upon by many users, so I think there is no problem with that guideline. Prime Blue (talk) 14:42, 25 August 2009 (UTC)
Oops yeah Covenant of the Plume is not a relevant example (I hadn't checked the translation). I wanted to give an example of a game for which there is no significant difference in the root title but one in the subtitle. I changed it to Baten Kaitos: Eternal Wings and the Lost Ocean. As for unbolding secondary titles, see [15]. Megata Sanshiro (talk) 21:07, 25 August 2009 (UTC)

Nihonjoe's opinion and compromise is not there yet, so in the meantime, I would like to ask MythSearcher to restructure his answer to use the format suggested above. Prime Blue (talk) 14:42, 25 August 2009 (UTC)

My POV is exactly what Bluer wrote above. I see no need to make things needlessly complicated. Also, all alternate titles should be bolded, so Japanese titles should be bolded (especially since that's the original title of the item). ···日本穣? · 投稿 · Talk to Nihonjoe 02:40, 26 August 2009 (UTC)

MythSearcher's POV:

  • If the word(s) are of English origin, and is commonly used in the English languages, except for the ones used as puns in the Japanese language(which should be easily sourced), do not include the romanization. Drop the kana as well if it is only showing the Japanese renderization/accent of such words, unless the Japanese title is different from the English title. (Like the old Dragon Quest/Dragon Warrior problem, in this case, include Dragon Quest's katakana but skip the romanji.)
  • If those are made up words, keep everything. So, Ivalice, Moogles and such should have romanji and kanas attached to them.
  • If the word(s) are not English, or it is pronounced in a non-English way.(Like the name Louis is pronounced as Loui in French, and it is of French origin) Then keep the romanji.
  • All cases when the word is in Japanese, include the romanji unless the romanji is exactly the same as the word itself. Like Hitodama, Oni (folklore), Kanabō, etc. and I would suggest removing the romanji in the kitsune article as well, the IPA is sufficient.
  • For all special cases, like when the title is too long, keep the kanas and romanji as a footnote, unless the title itself is completely not English and is not comprehensible by a normal English reader. If people really wanted to know it, they will try to click on something tagged after the word, doubting this simple action is quite irrational.
  • These should not be much different from Jinnai, Megata and Prime Blue's POV. I do not have a strong preference for the kana, so I can agree with keeping them as well.
  • If there is no consensus, the guidelines do not represent a consensus and should be removed.

WhisperToMe's POV:

  • I am in favor of a guideline requiring that the romaji be listed only when the pronunciation significantly differs from the English
  • In practice this will be the case almost 100% of the time: I consider all three "Kingdom Hearts (キングダムハーツ?), Akumajō Dracula (悪魔城ドラキュラ?), Vincent Valentine (ヴィンセント・ヴァレンタイン?)." to have significant differences in pronunciation in English - the Haatsu and Hearts sound very different, and it would illustrate how many sounds simply do not exist in Japanese. Readers need to accept this as being necessary and crucial in understanding the subjects. Even the slight differences need to be documented. Only when the pronunciation is exactly the same, will the romanization be omitted.
  • Having romanizations of "foreign" names from Japanese products in articles is NOT intrusive to the headers. WhisperToMe (talk) 08:16, 2 September 2009 (UTC)

Mediation request

I have read the individual POVs now, and from the look of things, I don't think a clear consensus can be reached here on our own. An almost equal amount of people are arguing for their cause and there seems to be little room for a compromise that will satisfy everyone. I think the best way to resolve the dispute would be to request formal mediation. What are your opinions on this? Prime Blue (talk) 11:55, 27 August 2009 (UTC)

I will not agree on having formal mediation without the other sides agreeing on the request(since it will be totally useless if only people agreeing with each other request one). Actually, I had thought of filing an informal mediation for this before, but since most are not showing any signs of compromising(at least any more) I don't see how mediation can work at all.(i.e. I don't know what to write in the section stating how the mediation cabal can help) Especially when Erigu and Nihonjoe clearly stated they have no will at all to compromise and Bluefn is still talking about what he thinks instead of having plans of compromises. —Preceding signed comment added by MythSearchertalk 17:17, 28 August 2009 (UTC)

Users for mediation (use an asterisk and four tildes in a row to sign):

Further discussion

Actually, the side saying cases should be made separately have tried to compromise a lot. Like I was originally asking to have all romanji removed as long as it got an official English name with little different from the Japanese katakana(like Aerith Gainsborough) and it seems like this is what Megata and Jinnai wished to do earlier as well. I have since ditched that idea on the highly resistive "too many cases" claim and I can understand the unusualness of some of those words. I see no will of compromise from the other side at all. —Preceding signed comment added by MythSearchertalk 14:48, 27 August 2009 (UTC)
I won't speak for the others, but I personally don't wish to compromise, indeed. I'm opposed to the idea of complexifying the guidelines and introducing subjective calls on a case-by-case basis just to save a few characters. Any kind of compromise would do just that. Erigu (talk) 20:51, 27 August 2009 (UTC)
And I believe forcing it on everything needlessly expands the prose (sometimes to unreadability) without adding anything of value in many cases (the few examples have been noted) and further the usage could be seen as promoting dialectical differences which which are not promoted elsewhere, especially for words originally or commonly used in the English language. Moreover I believe it is a violation of WP:MOS for such words to promote one pronunciation over another.Jinnai 23:21, 27 August 2009 (UTC)
I've already said I'm fine with putting excessively huge title information in the footnotes, but I also don't see how it hurts anything to have it in the prose. If people are wanting to find out about the game, they will likely be interested to know the Japanese title, even if it's excessively long. ···日本穣? · 投稿 · Talk to Nihonjoe 01:49, 28 August 2009 (UTC)
I can't see how dropping the romanization is going to improve an article's informative value at all. With the kana, we tell the reader "this is how they spell this Japan-developed game's title in Japan" and with the romaji, we tell the reader "this is a way to pronounce that kana, English-speaker; it may not be exact, but it's your best bet". The romaji may not have been used/promoted elsewhere, but so is the unofficial literal translations. I believe one of the purposes of the romanization is to establish something is of Japan-origin/developed in Japan, which moots the notion that we're promoting one pronunciation over the other. — Blue 02:39, 28 August 2009 (UTC)
And the side refusing to have any compromises refuses to have formal mediation. You guys do understand that a stalemate meaning there is no consensus and the guideline would risk to be removed as a whole, and all the things you are fighting for is not going to be granted. Yes, you have repeated said what you prefer to do, and I am sure that you can also see what others prefer to do, with a certain level of compromises. Stubbornly sticking to your own POV is not going to help you win this argument since there are no consensus at all in the guideline right now. I am sure that everyone understand how you guys do not think the romanji as redundant, other do not think so, and I hope that you can at least understand this and about the guideline being not a consensus anymore and the consequences of not trying to compromise. You might be able to keep you own view, but it is impossible to force others to do so with a guideline that shows no consensus. —Preceding signed comment added by MythSearchertalk 17:11, 28 August 2009 (UTC)
You guys do understand that a stalemate meaning there is no consensus and the guideline would risk to be removed as a whole, and all the things you are fighting for is not going to be granted.
Er... You do understand I actually disagree with that new guideline over at the video game project, right? Erigu (talk) 18:14, 28 August 2009 (UTC)
You do understand we are discussing also about the guideline on the MOS-JP, namely the Revised Hepburn romanization (described below) should be used in all cases, right? —Preceding signed comment added by MythSearchertalk 19:04, 28 August 2009 (UTC)
No, the whole argument stems from that new guideline over at the video game project. Erigu (talk) 19:36, 28 August 2009 (UTC)
It's not "new" being around since late 2006 and even if it were, it could be seen as more accurately reflecting the newer consensus as consensus isn't static. The issues dealt with would impact both guidelines and perhaps others as well.Jinnai 20:12, 28 August 2009 (UTC)
I'm talking about this. Not that there aren't other (and older) video game guidelines I'd disagree with... Erigu (talk) 20:18, 28 August 2009 (UTC)
I thought you didn't agree with this one either: "Phonetic transcriptions are, as a rule, not considered to be significantly different and thus do not warrant the inclusion of Japanese titles."[16] Megata Sanshiro (talk) 21:21, 28 August 2009 (UTC)
Like I said above, "not that there aren't other (and older) video game guidelines I'd disagree with". I disagree with this guideline as well, indeed. And on a bunch of levels, too. Erigu (talk) 21:36, 28 August 2009 (UTC)
I that case it's obviously been shown that through the discussion in this talk page there is no consensus for either extreme. I believe that was why mediation was suggested.Jinnai 02:42, 29 August 2009 (UTC)
And in this case, it is perfectly inline with what I said, Erigu, that we do not agree with the guideline here, on MOS:JP about the romanization, and thus a stalemate only means that there are no consensus and the romanization guideline would be removed to reflect it being not a consensus. Other guidelines may be removed as well, but the one backing you up all along is the romanization one and I do not see how you can keep trying to be stubborn about this since it is not helping your view of having romanization in all cases at all. —Preceding signed comment added by MythSearchertalk 18:20, 29 August 2009 (UTC)
So you don't understand a guideline, and that means it's getting removed? Interesting.
(not that I'd mind this manual getting reworked/reworded here and there) Erigu (talk) 18:36, 29 August 2009 (UTC)
I understand the guideline, but that does not mean that I agree with it. Do you even understand what is being discussed here? There are at least 4 people here who do not agree with what the guideline is saying, and wished to amend it so that redundant information could be removed. You do not agree with these people(along with 3 others) and thus making the discussion coming to a stalemate status, meaning the guideline is not a consensus (anymore). And your stubbornness of not willing to compromise is not going to help your view. I suspect you are not trying to understand anything here. The guideline itself is in dispute, since people do not agree on the terms it is forcing them to do. We are not only discussing about video games, all other related subjects on Japanese languages are also included right now. Your arrogance of the guideline being there is not going to help your view at all if you do not try to communicate and compromise, since guidelines are created by consensus, and if there are no consensus, the guidelines can well be removed. —Preceding signed comment added by MythSearchertalk 19:01, 29 August 2009 (UTC)
Problem is, the information (rōmaji) is not redundant. It serves a very clear purpose in providing a transliteration of the Japanese. Rōmaji is not (and should not ever be considered) redundant to the kanji or kana it represents. ···日本穣? · 投稿 · Talk to Nihonjoe 19:26, 29 August 2009 (UTC)
I think, Nihonjoe, it's because some people wrote the romaji for the kana differently - especially when it was blatantly English like キングダム - they romanized it directly to "Kingdom" instead of kana-by-kana "Kingudamu", which made it appear redundant. Maybe establish how to write romaji? — Blue 06:57, 30 August 2009 (UTC)
Like I said, if it is redundant or not, it is also disputed in this discussion, thus if you are not going to try to compromise, the result is the romanization guideline not being a consensus and would be removed. Like I said, my view point is that the romanji is merely an accent the Japanese people uses, and thus it is redundant on pure English terms. Emphasizing on the little different in accent is redundant. I know you do not agree on it being redundant, but the same thing is happening here that we think it is redundant and this is what the dispute is all about. —Preceding signed comment added by MythSearchertalk 07:38, 30 August 2009 (UTC)
Like I said, my view point is that the romanji is merely an accent the Japanese people uses
Thing is, that point of view is flat out wrong on several levels. Erigu (talk) 10:10, 30 August 2009 (UTC)
Thing is, that POV is yours and have no source backing it up. Japanese people do speak English, just that they might not be good at it and their accent is affected by their katakana. —Preceding signed comment added by MythSearchertalk 16:05, 30 August 2009 (UTC)
No, objectively, you're flat out wrong on several levels. 1) Romaji (or "rōmaji", I guess, but definitely not "romanji") are just Latin alphabet letters, 2) we're actually talking about a systemized romanization, here (revised Hepburn), 3) it's not an "accent" at all. Erigu (talk) 16:52, 30 August 2009 (UTC)
First off, if the discussion continues, I would like to request both parties to keep it civil. Also, I think MythSearcher is using the term "accent" for the syllabic concept of the Japanese language, so that might have been just a misunderstanding. Prime Blue (talk) 17:10, 30 August 2009 (UTC)
The syllabic structure of the Japanese language can't quite be passed off as a simple detail ("merely an accent"). So it would be a pretty big misunderstanding right there. Erigu (talk) 17:18, 30 August 2009 (UTC)
We aren't talking about the Japanese lanuage. We are talking about Japanese pronunciation of common English words which is part of the realm of dialect just as a US pronunciation of common Japanese words is.Jinnai 06:51, 31 August 2009 (UTC)
We are talking about the Japanese language. We're talking about transliterated Japanese characters. Not an "accent" nor a "dialect". Erigu (talk) 07:58, 31 August 2009 (UTC)
Transliterated Japanese characters of non-Japanese words. Megata Sanshiro (talk) 08:44, 31 August 2009 (UTC)
'Doesn't matter. The Japanese characters are there, and they should be transliterated for Japanese-illiterate readers. Erigu (talk) 08:46, 31 August 2009 (UTC)
We don't do that for other languages where the words have a common English counterpart. We do list the original spelling, but we don't go out and give a sylable-by-sylable pronunciation.
I'm wondering though if rather than list in the prose after it the romaji for words we deem need such clarification if we could put the translation of the kanji or kana above similar to furigana for kanji.Jinnai 01:02, 1 September 2009 (UTC)
Could you link to manual of styles that specifically state the romanization shouldn't be included for words of English origin? Erigu (talk) 10:41, 1 September 2009 (UTC)

Suggestions to resolve the dispute

I overlooked another action to be taken before formal mediation can be requested, requesting comments from uninvolved users, that is. While this will not end the dispute, it might add something to the discussion. I posted that on the bottom of this talk page. Furthermore, I think it is unlikely that this dispute will be resolved by discussion alone, so if any of the participants reject formal mediation, I would like to see other resolution ways suggested. Prime Blue (talk) 17:10, 30 August 2009 (UTC)

The only problem I see with the RFC, below, is that it slightly misrepresents the discussion. This discussion is not how to romanize Japanese, but whether to include romanization in addition to the English when an obviously English (or other language using Roman characters) is rendered in katakana (e.g., Kingdom Hearts (キングダムハーツ Kingudamu Hātsu?) vs. Kingdom Hearts (キングダムハーツ?)). WP:MOS-JA is already clear on how to romanize words, so I'd ask that you please clarify the RFC on this point. ···日本穣? · 投稿 · Talk to Nihonjoe 18:38, 30 August 2009 (UTC)
Clarification added, feel free to rephrase. Prime Blue (talk) 19:04, 30 August 2009 (UTC)
And I would like you to clarify the fact that that particular WP:MOS-JA is already clear on how to romanize words, is related to a part of the dispute.(It is not the centre of the dispute, but related, since people are quoting that section to claim romanization is always required.) —Preceding signed comment added by MythSearchertalk 13:57, 31 August 2009 (UTC)
I'm sorry, but your comment makes no sense. Could you try rephrasing it? ···日本穣? · 投稿 · Talk to Nihonjoe 05:53, 3 September 2009 (UTC)

I went Bold, and removed the wording that made the guideline too specific, plus there's no update on this matter for a month now. Its really obvious that there's a lot of cases that one single guideline couldn't easily be applied to. Plus, turning them into rules doesn't really help either. The nihongo template is there, and so far there's no problem using it, so I guess we could use it as it fits, and decide on a case-basis when there's a dispute. Of course, if anyone's still not in agreement, umm..... they can start a new discussion in a new section? So, yeah... — Blue 11:07, 3 October 2009 (UTC)

I waited for the RfC to be closed before I'd make a new comment. I think that discussions for every single case and no clear guidelines are not the best solutions, judging from this talk page here. After all, it doesn't (for example) make the Kingdom Hearts issue go away. But since no other suggestion to resolve this dispute was made, let's just see how the problem is handled in the future. As you said, the discussion could be reopened at any point, should it become necessary. Prime Blue (talk) 17:15, 15 October 2009 (UTC)

I thought it was general practice to add the Hepburn transliteration when the "English" title is extensably different from the Japanese pronunciation of the title. While it is clear that Kingdom Hearts is what is meant by キングダムハーツ, by omitting Kingudamu Hātsu, we do not provide to the reader the proper meaning of the Japanese text within a Japanese context. I can understand not having an identical Hepburn transliteration for Shin Megami Tensei (although I did see that Maniakusu and the like were absent from the already lengthy lead of Shin Megami Tensei III: Nocturne).

In my opinion, the guideline should state that only if the English name is identical to what would be the Hepburn romanization can the Hepburn romanization be omitted from the nihongo template. It doesn't seem right to omit Katamari Damashii from Katamari Damacy and related articles, particularly when one in English would not normally parse the "cy" to be homophonous with the word "she". For articles such as Shin Megami Tensei and Tokugawa Ieyasu, there's no need to have the Hepburn romanization. But for articles such as Tenjho Tenge, Kingdom Hearts, and Final Fantasy, where the English name is similar to but not exactly the same as the Hepburn romanization, both should be included. It seems that WP:VG/GL currently says "look here for rules on Japanese text". So we should either make this clarification or say "Non-Japanese words rendered in katakana do not need Hepburn romanization."—Ryūlóng (竜龙) 03:14, 16 October 2009 (UTC)

Correction should be applied for the description about apostrophe

I was reading bits and pieces of the Daikanyama vs. Daikan'yama debate, and noticed that it was pointed out that this MOS contains the text:

Use of apostrophes should be avoided except in the case of the syllabic "n" followed by a vowel (see "Body text", below).

This sounds like nothing but an error. Those who know the correct way to apply the apostrophe in revised Hepburn romaji should already know that apostrophe is used between syllabic N and a proceeding vowel or y. It is also unnecessary as it already states "Revised Hepburn romanization (described below) should be used in all cases, excepting..." While the inclusion of additional description about the usage of apostrophe was supposed to help reiterate what's already known about revised Hepburn, in turn it's just adding confusion because it sounds like it might be one of the exceptions that are supposed to exist.

The sentence should first distinguish itself either as a new rule or as a clarification of an already stated rule. In the case of the latter, it could be removed entirely as it is redundant and also adds only confusion rather than clarification. The need for the additional emphasis on the usage of apostrophe is not apparent. We have already stated that we're using Revised Hepburn at this point in the text. —Tokek (talk) 14:05, 17 August 2009 (UTC)

The relevant edits were made two years ago: [17], [18], [19]. Also, I really have no idea what "i)(i" is supposed to mean or where the notation comes from. The relevant discussion, according to the timestamp, was: [20]Tokek (talk) 14:16, 17 August 2009 (UTC)

The "i)(i" thing was added because someone got it into their head that an apostrophe was needed between two similar vowels. I agree that it's unnecessary and I've removed it as pointless. The apostrophe bit should remain, though I've clarified it. ···日本穣? · 投稿 · Talk to Nihonjoe 15:40, 17 August 2009 (UTC)
Tokek: There are non-standard versions of Hepburn which apply apostrophes in cases of i's and n and an (i.e. Ei'ichi and Sen'nin); it turns out that this is NOT standard Hepburn. But sometimes one sees it (i.e. Viz Media's version of Rurouni Kenshin) WhisperToMe (talk) 05:32, 20 August 2009 (UTC)

Request for comment

Requesting comment on the inclusion of the romanization of foreign words rendered in Japanese katakana in addition to the foreign words (for example: "Kingudamu Hātsu" in "Kingdom Hearts (キングダムハーツ Kingudamu Hātsu?)"). While it is no substitute for the discussion above, the users involved summarized their opinions and suggested compromises. Prime Blue (talk) 17:10, 30 August 2009 (UTC)

Comment I think we need to remember that this is a guideline. The guideline is to use the nihongo template. Leave it to the individual article authors to decide when it makes sense not to follow the guideline. However, these authors should be prepared to defend their choice when other authors and editors assess their work. That said, I think it makes sense to include the romaji to indicate how a Japanese person would tend to pronounce what are essentially English loan words. Imagine a Japanese title like Strength of Statistics or the Nissan Bluebird. Is that fence-sitting enough? imars (talk) 13:32, 1 September 2009 (UTC)

That's this catch, some people refuses to have articles decide individually and claims that the guideline should supercedes all other projects on Japanese topics. Erigu even went as far as editing an article that did not have romanji during the conflict when it was used as an example and claim that s/he is only following the guideline, without any view of it being a conflict of interest edit within the dispute. —Preceding signed comment added by MythSearchertalk 13:10, 2 September 2009 (UTC)
I'm afraid I still don't quite realize the gravity of my sin... Erigu (talk) 15:38, 2 September 2009 (UTC)
As an example, in a court case, both the prosecution and defense do not go to alter the opposing side's evidence or case samples and claimed that since the other side is incorrect, the evidence or the samples must be incorrect also. In this case, the dispute is about the guideline stating the romanji should be included, and mainly focusing on whether the romanji should be included or not, an example was brought up by the side claiming it is an example that should not include the romanji and it did not include it, you went to change it to include the romanji, claiming it a simple action following the guideline without thinking about your action is controversial due to the fact that the guideline itself is disputed. It is like changing what is written on a statement in court just because it is not favourable to your side.
To be more clear, it is like when you are in the middle of an edit war with someone else, and moved to discuss about what should be in the article, a similar article was brought up by your opponent, and you change that article just because it is against your view.
If someone else who is not involved in this dispute changed it, it would be totally fine, the main problem here is that you are an involving party of the dispute, thus you should refrain from making controversial edits in such cases. It is not a matter of sin, it is a matter of good practise during disputes. I think this pretty much is the spirit behind the WP:3RR policy and the WP:1RR essay. I know the exact refrain from making controversial edits during disputes is not written in words, but to my understanding this is the spirit behind the policy. —Preceding signed comment added by MythSearchertalk 08:57, 3 September 2009 (UTC)
I do hope you could carry on with this comment on user conduct on said user's talk page and let other users leave a comment on the issue at hand, thank you ^^ — Blue 11:02, 3 September 2009 (UTC)


How exactly should these three kana be put into Hepburn? I'm aware that the "ni" is a particle and is generally separated from the rest of the phrase, but is it "you ni" or "yō ni"? This is effectively an issue at Hoshi no Yō ni... and Tsutsumikomu Yō ni... (I had moved the former to the "You ni" variant, but now it's been moved a few times since then and the bot fixed the double redirects).—Ryūlóng (竜龙) 02:48, 14 October 2009 (UTC)

I think it should be "yōni". The word comes from "yōnari" and is considered as a compound word or an auxiliary verb. See [21] and [22]. The "ni" is not a particle. "Yōni" is a conjugated form of the auxiliary verb "Yōda". See [23] and [24]. Oda Mari (talk) 05:09, 14 October 2009 (UTC)
The grammar is actually not relevant: the pronunciation is (single long vowel), so the Hepburn spelling is as well. Jpatokal (talk) 08:44, 14 October 2009 (UTC)
Okay. But I cannot find anything about ように being anything but a phrase and in no way related to a verb "yōnari". And on the grammar front, there appears to be a semantic difference between "様に" and "ように". Most conversions to romaji appear to treat "ように" as two separate words when romanized (resulting in "yō ni" more times than "yōni").—Ryūlóng (竜龙) 19:24, 14 October 2009 (UTC)
That would be because you don't speak Japanese or understand it. 大辞林:
〔補説〕 名詞「よう(様)」に断定の助動詞「だ」が付いたものから。
ように has at least three separate meanings with no clear correlation to whether it's written with 様 or よう, although 大辞林 opines that 様 is favored for expressing similarity and よう for expressing the intent/outcome of an action. Whether you separate and da, ni etc depends on whether you consider the 断定の助動詞 ("concluding auxiliary verb"?) to be a separate word or not. The form "yōnari" is archaic (according to 大辞泉, it hasn't been used since the Muromachi era), but it's the predecessor of today's yōni/na. Jpatokal (talk) 02:41, 15 October 2009 (UTC)
Well, I've rarely seen the phrase parsed in kanji in any way. I usually come across よう and 様 used interchangeably, but I've never come across ように and 様に used interchangably.—Ryūlóng (竜龙) 03:53, 15 October 2009 (UTC)
Fascinating. I think we can safely agree that the macron should be used. I'd suggest separating ni as the more common usage; while historical usage suggests the phrase is one word, I wonder how many people, even native Japanese speakers, would be aware of it. Doceirias (talk) 03:22, 17 October 2009 (UTC)
Well, the Japanese language only uses spacing to denote subtitles and the like. The Japanese Wikipedia only does it to separate a person's surname and given name. The spacing is only done to learn/teach Japanese.—Ryūlóng (竜龙) 03:34, 17 October 2009 (UTC)


I am trying to figure out whether or not the above two kanji in the instance of their usage in the article Ai o Torimodose!! should be read as towa or eien. I am finding that both readings exist online. In the official subtitles for Fist of the North Star (currently on Hulu), they read the two kanji as "Eien". However, in a CD released last year for the Lucky Star soundtrack which covered the two songs they refer to the two kanji as "Towa". The iTunes store has two versions of this song for sale (one by each of the original singers who broke up the original band). There is the version by the singer who goes by Crystal King [25] and a version by Masayuki Tanaka who was the other lead singer of Crystal King in the 80s [26]. Google searches also show that both are used (17,100 hits for Yuria Eien ni, 313,000 for Yuria Towa ni, 81 for "Yuria Towa ni", 10,300 for "Yuria Eien ni"). I can't figure out what it should be.—Ryūlóng (竜龙) 20:42, 19 October 2009 (UTC)

Towa maybe some kind of special reading, for when one is trying to be poetic, or superimposing another meaning or another word onto these characters, but it is by no means a normal reading for these characters. 永 is naga(i) or ei, and 遠 is too(i) or en. Eien is a much more typical, normal, reading for these characters and is the only one that comes up in my dictionary. Then again, in song titles and lyrics and such, you never know. LordAmeth (talk) 21:51, 19 October 2009 (UTC)
I have seen these characters read both as towa and eien, and especially so as in the title of this particular song. However, the song's title does not appear in the verses of the song to make a definitive decision on whether the song is to be read as "Yuria... Towa ni" or "Yuria... Eien ni".—Ryūlóng (竜龙) 21:59, 19 October 2009 (UTC)

Toaru/To Aru

There's an interesting romanization discussion going on at Talk:To Aru Majutsu no Index. All input from the experts here is welcomed. Doceirias (talk) 15:39, 27 October 2009 (UTC)

Moves by User:The Prince of Darkness

The Prince of Darkness (talk · contribs) recently moved several articles on Japanese composers to versions of the title using macrons in the name when there has been extensive evidence in the past that the names of these people in the Latin alphabet have existed for several years. Example pages moved include Yoko Kanno to "Yōko Kanno" and Ryuichi Sakamoto to "Ryūichi Sakamoto". I've undone all but a handful of page moves where I cannot find an official website or a pre-existing English/French/etc. name for the person:

I've also warned the user that if he repeated these moves in the future he may be blocked for it, because this was extremely disruptive for articles on people who have established name spellings in the Latin alphabet.—Ryūlóng (竜龙) 01:45, 29 October 2009 (UTC)

"Extremely disruptive"? Give me a break -- thanks to redirects, the change is almost invisible.
IMHO, the current MOS-J guidelines on when to use macrons are vague and quite possibly unnecessary. In most other languages, the local form with all diacritics is used even for very well known figures (Lech Wałęsa, Václav Klaus etc), and I don't really see a downside to consistently using macrons for Japanese names as well. (Excepting, of course, rare cases like Yasuhiro Nightow where the person has consciously adopted a non-Hepburn romanization.) Jpatokal (talk) 03:07, 29 October 2009 (UTC)
Many people have adopted non-Hepburn romanizations to their names because no one wants to write their name. That's why the most common surname in Japan (佐藤) has so many spellings in the Latin alphabet. By moving several pages from long-standing instances of a non-Hepburn romanized name to the Hepburn romanization without looking for any evidence that the individual's name has a name written in the Latin alphabet is disruptive (Yoko Kanno and Ryuichi Sakamoto both have albums where their names are written in the Latin alphabet and have existed at those titles for several years, but this did not stop User:The Prince of Darkness from moving the pages). These are not diacritics but a writing standard put in place by this project because we are writing about a language that is not written in the Latin alphabet.
The only reason I have brought this up here i beause I am not sure about the four other pages that I did not move because I am not sure if there are standard and long-standing instances of these individual's names having been written in the Latin alphabet (Kohei or Kouhei; Shoji or Shouji; Yuzo, Yuuzo, Yuzou, or Yuuzou; Toru, Tohru, or Tooru), thereby meaning we should leave the pages at the Hepburn name. The standard for this manual of style is this hierarchy: the name officially used in English or other languages that use the Latin alphabet, the most common spelling of the name in similar languages, and then the Hepburn romanization if the other two don't exist.—Ryūlóng (竜龙) 03:40, 29 October 2009 (UTC)
Referring to MOS-J will get you no brownie points, since this is the talk page for MOS-J, I'm telling you I think the current rule is misguided, and we can change it here.
The problem with deciding "official" names is that there are no official references for romanization. I can demonstrate that "judo" doesn't need a macron because any halfway decent English dictionary will include the word, but how do you propose to do that for, say, 佐藤太郎? (The fact that the MOS-J currently tells people to look up names of modern figures in a dictionary just shows that this originates from misguided cut'n'pasting!) People with Japanese names living outside the country need to lock down a macronless version for daily life (hence Sato, Satoh, Satou etc), but in Japan they're 佐藤, which is by default romanized Satō.
I'd thus suggest flipping the rule on its head: use the Hepburn version with macrons unless English-language reliable sources consistently use another version. Jpatokal (talk) 05:19, 29 October 2009 (UTC)
For your example of 佐藤太郎, it would depend on how he (or his talent agency or record label or other organization) has decided to write his name in the English alphabet. It could be "Taro Sato", "Tarou Satou", "Taroh Satoh", "Tarow Satow", "Tarrow Sattow", or any variation thereof. If no such English text exists, then of course Wikipedia should default to "Tarō Satō". For example, we have 小泉純一郎 written almost exclusively as "Junichiro Koizumi" in Japanese sources that have decided to also print the name in English ([27], [28]), and never "Jun'ichirō" or "Jun'ichirou" or variations thereof. Because a majority of the subjects of the pages that I performed the reverts on had such websites and sources that listed their names in the English language, I performed the reverts (such as in these images from Kanno's and Sakamoto's websites: "Kanno Yoko", "Kanno Yoko", "Ryuichi Sakamoto").
I don't think any flip flopping of the guideline needs to be done. What we really don't need in the "Names of modern figures" section is the "trade name" or the "dictionary" things, perhaps to be replaced with something that clearly states that if the subject of the article has an official spelling of his or her (or "its" for organizations) name, the English Wikipedia should use that spelling (Kenichi Matsuyama over "Ken'ichi Matsuyama", Yoko Kanno over "Yōko Kanno", and Ryuichi Sakamoto over "Ryūichi Sakamoto"). This should be followed by the name used most prominently in English language media (Yoko Ono over "Yōko Ono"). And then only if there is no version of those names, we go with the Hepburn romanization (Kenjirō Ishimaru).
The method you are suggesting kinda won't work for a checklist. You are going "Write his name in Hepburn romanization first. Is there an official way to write his name in the English alphabet? No. Is there a common way of writing his name in the English alphabet? No. Stick with Hepburn." It more or less should be "Is there an official way to write his name in the English alphabet? No. Is there a common way of writing his name in the English alphabet? No. Write it with Hepburn romanization." or "Is there an official way to write his name in the English alphabet? Yes. Go no further."—Ryūlóng (竜龙) 06:14, 29 October 2009 (UTC)
"he (or his talent agency or record label or other organization)" are self-published sources, not WP:RS, and I don't see why Wikipedia should pay any attention to their misguided ideas about a technical topic like romanization. Jpatokal (talk) 13:39, 29 October 2009 (UTC)
So you are saying we can't use the subject or those who represent the subject as a reliable source for how to spell the subject's name in the English alphabet? That is one of the most ridiculous things I have heard on this project.—Ryūlóng (竜龙) 00:50, 30 October 2009 (UTC)
Try again. I'm simply pointing out that self-published sources are generally not WP:RS.
As a simple demonstration, this is the official site of the Brother Leader of the First of September Great Revolution of the Socialist People's Libyan Arab Jamahiriya, معمر القذافي. This single site uses the romanizations "Muammar Al Gathafi", "Mummar Al Qaddafi", "Muammar Al-Gathafi" and more... yet Wikipedia chooses to use Muammar al-Gaddafi. Jpatokal (talk) 05:38, 30 October 2009 (UTC)
Obviously, if there is no consistency in the official spellings, we would default to macrons. The exceptions exist for names that ARE consistently romanized in a way that differs from the MOS's preferences. I've never seen the rule as inconsistent or even hard to follow; use macrons unless there's a good reason not to, i.e., a consistently used romanization by which they are widely known. Doceirias (talk) 10:17, 30 October 2009 (UTC)
That's exactly what I'm suggesting, and it's the opposite of the current rule, which states that macrons should not be used except as a last resort! Jpatokal (talk) 02:14, 1 November 2009 (UTC)
That is not what Doceirias is saying. He is saying if there are multiple transliterations/translations of a person's name, we should default to one that uses macrons, which is what the manual of style currently dictates. You are saying that even if we have a consistent non-Hepburn name for a person, we should go with the Hepburn romanization regardless of any reliable sources (or primary sources) we may have to prove it.—Ryūlóng (竜龙) 02:36, 1 November 2009 (UTC)
I find your total mischaracterization of my position borderline offensive. In addition, the MOS says nothing at all about multiple versions. Jpatokal (talk) 05:01, 1 November 2009 (UTC)
You want Hepburn romanization (macrons) to be the first thing we have the page for. However, if a person has a long vowel in their name and if there are sources that show that the subject has a writing of the name in the English language without macrons, we should use the macrons regardless, so we don't have any Taro's or Yuichi's or Sato/Satoh's, but only Tarō's, Yūichi's, and Satō's.—Ryūlóng (竜龙) 05:27, 1 November 2009 (UTC)'re saying we're all on the same page about how this should work, but you think the current wording is misleading? I'm having trouble following the exact nature of the disagreement here. As I understand the current policy (and a proposal to clarify this might be in order) A. If there is a clearly established, consistent romanization available, use that. B. If not, use macrons. C. If there are several competing romanizations available, use macrons. The last clause might need reworking if there are several available readings but one which is demonstrably the most common, but I figure common sense could take care of that. Isn't this essentially how it works now, and how we all agree it should work? Doceirias (talk) 12:45, 1 November 2009 (UTC)
See, the problem is that many non-scholarly sources do not use macrons for Japanese names, not because they're trying to make some profound philosophical statement, but because typesetting them is hard and (as Oda Mari says) because they're unfamiliar with macron usage. According to the current rules, this would mean that we would have to strip these names of their macrons... but we do know what macrons are and how to use them, and if we can render the name more correctly by including the macron, I see no reason not to include it. Jpatokal (talk) 14:37, 1 November 2009 (UTC)
I see. I would not have considered overruling correct macron usage based on academic sources; I was largely talking about writers and artist's names, reflecting the kind of articles I edit. I'm not sure if there's a way to clearly distinguish between the kind of names we want to keep without macrons, and the kinds we would like to macron despite the academic sources disagreeing with us. One other point -- sort of playing devil's advocate -- is that the correct macron usage is found in the nihongo tag, so what does it matter if a less correct, but more common, usage appears in the rest of the article? Doceirias (talk) 14:49, 1 November 2009 (UTC)
It is not a case of "less correct". Subjects of popular culture (writers, musical artists, etc.) may or may not have a specific way to write their name in the English alphabet. If they do, we use them for articles. If not, then we use the Hepburn romanization. If there's a grey area (more than one spelling in the English alphabet and a common one cannot be determined), then use Hepburn. Just because we can render the name with a macron does not mean that for those who may have rendered their name in the English alphabet, and where a macron could have been used but wasn't, does not mean that they simply are unaware that the macron should be used there and we should put it there when referring to them.—Ryūlóng (竜龙) 23:00, 1 November 2009 (UTC)

Okay, so what about people like Kōji Kondō, Ryūji Sasai, and Kenji Itō. BTW, these articles weren't moved by me. The Prince (talk) 11:35, 29 October 2009 (UTC)

I see no official website for Koji Kondo. Unless his name is not written in a standard fashion in the video games he worked on that have made it to the US, then that page title is correct. The page should be at Ryuji Sasai per his official profiles on official websites. Same goes for Kenji Ito.—Ryūlóng (竜龙) 12:06, 29 October 2009 (UTC)
But almost all Japanese people do not use macron when they write romaji. Or they don't know that macron should be used for long vowels. Oda Mari (talk) 13:19, 29 October 2009 (UTC)
We do not because we do not know how. Also, we learn Kunreis[h]iki, and not necessarily Hepburn. That is why Sinzyuku and Huzi are most natural and quite common. And long vowels will often be wrote with a "u" or just ignored because of ignorance, not because of preference or anything official. If Wikipedia is trying please the individual preferences, as it appears, then you are wasting your time. There may be the few special cases to be made, but in general we really have no idea how to romanize Japanese; it is all very 適当. You would be better off just being consistent throughout all articles and mark all long vowels as seems to be norm on academia and professional books. Each publication has the styleguide. Create one and enforce it. The inconsistencies and constant changes make it difficult for your readers. (talk) —Preceding undated comment added 13:39, 29 October 2009 (UTC).
"Sinzyuku" and "Huzi" are not even close to common in Japan. Not anywhere near common. Sure, you see them very occasionally, but "Shinjuku" and "Fuji" are—by far—the most common romanizations. ···日本穣? · 投稿 · Talk to Nihonjoe 02:10, 30 October 2009 (UTC)
I agree - I think it would be good for Wikipedia to have a consistent style that applies across all titles. My preference would be for the Hepburn macrons. Of course, it is also important to recognize common renderings that do not use macrons; however, what if there are multiple spellings that are equally common? (e.g. a person's publicist uses more than one particular romanization, or different media publications may use different renderings) This is where the beauty of redirects kick in. Let's have the article be named in the Hepburn style with macrons, and remember that we can create redirects for commonly recognized spellings so that the user experience will be seamless. Yeah, the article's title might look odd to someone who's never seen a macron before, but it won't interfere with their ability to find the article (#REDIRECT) or read the title, and it provides a surefire way for those who do know the pronunciation to locate the article on Wikipedia. Anyway, that's my two cents - Hepburn macron titles across the board with redirects aplenty. Joren (talk) 21:41, 29 October 2009 (UTC)
If we have a rendering of the name in English, whether it's Kunreishiki or some random spelling that the subject or his or her publicist has chosen (Masamune Shirow, Yasuhiro Nightow, Tite Kubo, Show Aikawa, Yoko Kanno, etc.) we should use what we know exists instead of saying "Everything is Hepburn" when Hepburn is barely used at all in Japan. I've seen instances of what we would force as ū as uu, and even an instance where the subject's preferred spelling of "Yū" as "You". Forcing a system that is only used in an English academic sense is extremely unnecessary. Wikipedia naming policies outside of this manual of style go Common Name and then English Name. The "common name" in this sense, would be the name that the subject of these articles (or their publicist) has chosen for them in the English alphabet. The Hepburn romanization would fall under the "English Name". If we have a common name that appears in all reliable English language content regarding the subjects of these articles then we should use that version of the name over any romanized name that disregards any reliable sources we may have. Unless there are no reliable sources that state the subject's name in the English alphabet or there are multiple versions, then we should default to the Hepburn romanization style, which follows site wide policy.—Ryūlóng (竜龙) 00:34, 30 October 2009 (UTC)
  • Since many translated names are not transliterated but sometimes form new names or alter the romanization to create puns, it would be necessary that names on the English encyclopedia should primarily reflect what the individual prefers to use on their own. A redirect from the direct transliterated name can be created to establish -both-, but the preference per WP:COMMONNAME seems clear. If there is no chosen name by the individual, then the default academic transliteration seems acceptable. Ottava Rima (talk) 01:06, 30 October 2009 (UTC)
    • I agree with Ryūlóng and Ottava Rima. WP:COMMONNAME dictates that all article titles can not be consistent as there are quite a few names which are not consistent or that don't follow revised Hepburn romanization rules. This is all clearly spelled out in WP:MOS-JA. ···日本穣? · 投稿 · Talk to Nihonjoe 02:10, 30 October 2009 (UTC)
      • This is the talk page for MOS-JA, and I'm suggesting that we change it. Jpatokal (talk) 05:01, 1 November 2009 (UTC)
    • Agree. MOS-JA has its general guidelines, but these should not trump common-sense exceptions for "a foolish consistency..." For example, I'm glad to see Edogawa Rampo has been moved from "Rampo Edogawa", which destroyed the pun intended in the pseudonym. Otis Criblecoblis (talk) 17:39, 30 October 2009 (UTC)
      • Nobody is suggesting that we move to 100% scholarly Hepburn and damn the torpedoes. Jpatokal (talk) 05:01, 1 November 2009 (UTC)

Since there seem to be so many thoroughly bashed strawmen about, let me just lay out my suggestion for the exact wording:

The spelling of the name of a modern figure should adhere to the following:

  1. By default, the Hepburn romanization with macrons should be used. (Examples: Shinzō Abe, Asashōryū Akinori)
  2. If the use of an idiosyncratic romanization is widely attested in reliable sources, use that name. (Example: Yasuhiro Nightow, Tadayuki Naitoh)

For avoidance of doubt, use of a Hepburn-romanized name without macrons is not considered idiosyncratic. A Hepburn-romanized name without apostrophes, however, may be considered one (eg. Junichiro Koizumi).

Comments? Jpatokal (talk) 05:01, 1 November 2009 (UTC)

Lack of macrons, which is a system very few people employ anyway, should not be considered not an idiosyncratic usage. If we have sources that explicitly show the same spelling over and over again and they don't use macrons, merely because it should have a macron in the name does not mean the Wikipedia article should have a macron in the page title. Somewhere along the line, 坂本龍一 and 菅野よう子 decided that in the English alphabet they wanted to write their given names as "Ryuichi" and "Yoko" over "Ryuuichi" and "Yohko" or over "Ryuiti" and "Youko", much like (for their pennames) 士郎正宗 went with "Shirow" and 久保帯人 went with "Tite". The lack of knowledge of the Hepburn system in Japan should not be taken as the presence or lack of idiosyncracies.—Ryūlóng (竜龙) 05:27, 1 November 2009 (UTC)
A very simple question: how are Wikipedia's readers better served by not including macrons in names? Jpatokal (talk) 14:29, 1 November 2009 (UTC)
Including macrons does not inform the reader of how the subject (or his/her publicist) has chosen to write his or her name in the English alphabet. Just because a macron should be there does not mean we (as a project) should put it there because we assume the subject didn't know it should be there.—Ryūlóng (竜龙) 23:00, 1 November 2009 (UTC)

One more question: Kō Ōtani has his name spelled as Kow Otani on his official website, but he is more commonly known as Ko Otani without the macrons. What do we go by here? The Prince (talk) 17:42, 6 November 2009 (UTC)

If its more common among RSes then we go with Ko, otherwise we should go with Kow. If its unclear, go with Kow. Better to go with official when its unclear.Jinnai 19:15, 6 November 2009 (UTC)
IMDB: Kô Ôtani (talk) 20:15, 6 November 2009 (UTC)
IMDB automatically chooses the old circumflex notation for any instance of long vowels in Japanese.—Ryūlóng (竜龙) 20:20, 6 November 2009 (UTC)
I thought there'd been some agreement earlier in the thread to go with macrons when the sources conflicted. Doceirias (talk) 20:12, 6 November 2009 (UTC)
If he has "Kow Otani" on his official website and this spelling is featured on his official releases, then use "Kow Otani". If there is no larger preference for "Ko Otani" or "Kow Otani", then you should probably stick with "Kō Ōtani" (or if anything "Kō Otani" as at least "Otani" is common).—Ryūlóng (竜龙) 20:20, 6 November 2009 (UTC)
Why even have a manual of style? There is no style. It is a complete mess inconsistent across the whole encyclopedia, which is sadly the only consistency. There have been many attempts to improve (though not fix) this, such as a above, but never any real progress. Academia uses macrons; end of story. Newspapers remove macrons; end of story. IMDB uses circumflexes; end of story. Wikipedia uses... constantly moving, POV pushes and endless conversations year after year without any stability. You want macrons, use them consistantly; you don't want them, remove them completely. This manual of style is rediculous and serves no purpose.
Back on topic: according to IMDB, "Kô Ôtani, Alternate Names:

Koh Ohtani | Kou Otani". None of his works use Kow Otani. Nor does his passport, as Japan does not allow -w for long vowels. (I have tried.) (talk) 20:48, 6 November 2009 (UTC)

IMDB generally doesn't know shit when it comes to names of Japanese subjects. They just use circumflexes for anything that's a long vowel (this is the case for television programs and personalities). We have Joe Odagiri. They have Jô Odagiri with "Joe Odagiri" as an "alternate name". If we have a name that is common among primary Japanese sources, then we go with that name. And we have a handful of instances where the -w is used for the long O (Shirow and Nightow).—Ryūlóng (竜龙) 20:52, 6 November 2009 (UTC)

@IP address: Shadow of the Colossus, one of his most famous works, credits him as "Kow Otani". The Prince (talk) 21:34, 6 November 2009 (UTC)

Naming order

The recent edit on naming order clarrification reminded me we have an unresolved issue at Wikipedia talk:Manual of Style (anime- and manga-related articles)#Clarification of naming order within this page based on that section. To summarize the problem, the description gives detail on dealing with the real-world. Since most anime and manga do not do so, at times this becomes a dilema. While clearly any historic piece after 1860, modern-day and most future settings would follow this western order and historic peices pre-Meiji would follow traditional Japanese order, the information isn't so clear when dealing with certain fantastical situations.

Specific examples include Inuyasha which time travels between present-day and Sengoku period, Samurai 7 which is set in Sengoku period, but uses a lot of modern technology and terminology, Gin Tama which takes place during the Edo period and is much more modern than Samurai 7, Utawarerumono which is set in future Japan with with a pre-Meiji society more reminiscent of feudal Japan, completely fantasy ones like Record of Lodoss War and timeless pieces like Mushi-shi.

The ultimate descion was look to this MoS for what to do, but its not so clear.Jinnai 04:11, 14 November 2009 (UTC)

When in doubt, use the most commonly used naming order for the character. If there is an official English-language release, then the order used in that would be the most common. ···日本穣? · 投稿 · Talk to Nihonjoe 05:22, 14 November 2009 (UTC)
And when that's unclear?Jinnai 07:40, 14 November 2009 (UTC)
WHich series needs clarification? If you can post it here, people may be able to point you in the right direction. ···日本穣? · 投稿 · Talk to Nihonjoe 07:51, 14 November 2009 (UTC)

Song subtitles

I guess it has been more or less decided that any Japanese text when between two tildes/wavy lines (~, 〜, or ~; there are also instances of the subtitle between hyphens or similar lines) is the subtitle of the initial text. However, when finding articles on Japanese songs, this has been changed to being written as "Title: Subtitle" (I even found an instance of a song's instrumental version denoted as "~Instrumental~" being written as ": Instrumental"). The current policy has an example of the album BEST ~first things~. As an album, this normally would be "Best: First Things".

However, when it comes to songs such as "Mr. Moonlight ~Ai no Big Band~", the standard song subtitle formatting (in English) would be "Mr. Moonlight (Ai no Big Band)" (where I moved the page tonight). My move was immediately reverted by User:MS citing this page and referring to the fact that both "Best: First Things" and "Best (First Things)" are suitable subsitutes for "BEST ~first things~".

I believe the guideline should be modified so it clarifies that when used in song titles, text between or after tildes/wavy lines (~, 〜, or ~) should be parsed as between two parentheses and not after a colon.—Ryūlóng (竜龙) 09:46, 15 November 2009 (UTC)

I've never thought it should use parentheses, and I've always used a colon. Parentheses imply it's an aside and not really part of the title, which isn't the case. A colon keeps it as part of the title. ···日本穣? · 投稿 · Talk to Nihonjoe 18:52, 15 November 2009 (UTC)
What about songs such as (Don't Fear) The Reaper, (I Can't Get No) Satisfaction, etc.? There are not any songs in any Latin alphabet language that utilize a colon as far as I am aware.—Ryūlóng (竜龙) 23:28, 15 November 2009 (UTC)
In those cases, the parts in parentheses are not actually part of the original title, but are included because people often consider them part of the title. This happens on occasion. ···日本穣? · 投稿 · Talk to Nihonjoe 05:25, 16 November 2009 (UTC)
I don't know. These appear to be part of the song titles. They're referred to whenever the song is actually brought up in normal music writing.—Ryūlóng (竜龙) 05:57, 16 November 2009 (UTC)
Those are very special cases, and shouldn't apply to this discussion. In fact, I would suggest removing the parenthesis in both cases, since the words in parenthesis aren't exactly prefixes, just as the "The Reaper" and "Satisfaction" aren't subtitles. I mean, parenthesis aren't even around "I Can't Get No" on the single cover in the infobox...-- 01:09, 16 November 2009 (UTC)
"Don't Fear" and "I Can't Get No" are the subtitles and are how the song titles are parsed in English and how they have been written in English because those are two English songs. However, by simply saying everything between tildes should be changed to behind a colon is unhelpful. For an actual song subtitle, there's Rocket Man (song). Either tildes are undeprecated or changed to parentheses to remove the ambiguity.—Ryūlóng (竜龙) 01:31, 16 November 2009 (UTC)
I feel this proposal is trying to change standard conventions. If we're going to be changing song titles that have tildes, what about titles of books, movies, anime/manga, visual novels, etc? It's been a standard convention to place subtitles after colons in all those cases, so I see no reason to do any different with song titles.-- 03:23, 16 November 2009 (UTC)
The only "standard convention" I am seeking to change is the automatic decision that a song's subtitle has to come after a colon, because song titles rarely ever have colons in them. Forms of visual and written media use colons. Forms of audio media are not treated as such. It is a subtitle, yes, but it is not the kind of subtitle like those used in visual media like television, film, books, etc. It is not like "Star Wars Episode V: The Empire Strikes Back" (a film) or "Bat Out of Hell II: Back Into Hell" (an American album) or "Best: First Things" (a Japanese album) where these items are part of a series, but more like or "Frankenstein, or The Modern Prometheus" (a novel) where it is an alternate title, as in "Rocket Man (I Think It's Going to Be a Long, Long Time)" (a song), however the alternate title is actually part of the song's title. Unless songs are part of a series (such as the movements in Seikima-II's Akuma Nativity), then the subtitle/altenate title should be written between parentheses (because certainly "Song ~Instrumental~" should not be written as "Song: Instrumental").—Ryūlóng (竜龙) 04:05, 16 November 2009 (UTC)
You have to use your brain a little when you encounter those tildes. In most cases, they contain a subtitle for the song and this should be placed after a colon. In the final example you give, it's giving an indication of the version of the song (in this case, an instrumental version). That's all part of the translation process from Japanese to English: understanding and interpreting how the Japanese do something and bringing that into English in a consistent way. ···日本穣? · 投稿 · Talk to Nihonjoe 05:25, 16 November 2009 (UTC)
I do not think the colon is the proper English punctuation character to use when translating a Japanese song title containing tildes or hyphens surrounding what we are considering a subtitle. Yet, there were two instances of "Song: Instrumental" used on the English Wikipedia because of how this manual of style is currently worded. It should not matter if the words are "Instrumental", "Original Karaoke", or "Toki o Koete". If we are going to be writing these in English and we are not going to use tildes or hyphens or whatever to separate the two aspects of the song title rather than using a colon when directly translating.—Ryūlóng (竜龙) 05:50, 16 November 2009 (UTC)
Well, in those cases we can correct the title so it doesn't have the ": Instrumental" in it. Just because someone who doesn't know better gets confused doesn't mean we have to toss out everything. Using parentheses in these cases is incorrect, plain and simple.
Well, why is using a colon correct? Just because we have this one sentence at Tilde#Japanese? A song's subtitle is written between parentheses in all English songs that I've come across. There's this statement (although not official), but it gives examples of songs with subtitles that we can find here on Wikipedia: The Piña Colada Song, Happy Xmas (War Is Over), Orinoco Flow (Sail Away), 867-5309/Jenny. If we are going to treat things as the subtitles of songs, we may as well follow the general practice when it comes to treating these songs as if they were composed by English, French, Spanish, etc., speaking artists and use parentheses instead of the tilde, the dash, or other similar punctuation symbols that are used "decoratively" in Japanese.—Ryūlóng (竜龙) 06:23, 16 November 2009 (UTC)

[leftbounce] Ryūlóng, I don't suppose that I've seen any elpea or CD or single titled "Something: Something Else", but I think this is analogous to the front covers and title pages of books in English: typically (though not always) what's written there is "Something [line break] Something Else", in which the line break is conventionally represented by a colon when the title's written elsewhere. As for lists of songs, on CDs and so forth; well, they're lists, and thus they can indulge in typographic exuberance (often in funky fonts, etc) that would be irritating in our staid encyclopedia. -- Hoary (talk) 05:35, 16 November 2009 (UTC)

But in these cases we do not have line breaks but a separate punctuation item separating one part of the title from the other. The closest analog to the tildes is parentheses.—Ryūlóng (竜龙) 05:50, 16 November 2009 (UTC)
Well, in lists of English titles in English contexts there are two common ways of writing title+subtitle combinations that I can think of. One's to separate them a colon, the other (and far less common) is to separate them with a dash. In those English contexts where one has to be careful about such matters, one normalizes them all one way or the other (usually to a colon rather than a dash), paying no more attention to the choice in the original than one pays to the choice in the original between the "up" and the "down" capitalization style. Meanwhile, in Japanese-language contexts there are three common ways of writing title+subtitle combinations that I can think of. One is simply to plonk a space between the pair, another is to surround the subtitle with horizontal lines that are similar to em-dashes, yet another is to surround it with swung dashes. I'm not aware of any nuance put across by the choice of any one of these rather than the other two. The fact that the way the dashes come in pairs would be imitated by the way in which parentheses too come in pairs may be of very minor interest but doesn't seem to me to be of any significance. -- Hoary (talk) 08:47, 16 November 2009 (UTC)
No song's subtitle has ever been separated from the song's title by a colon in the American or British record industries (as far as I am aware). This is performed by placing the subtitle in parentheses, such as with The Piña Colada Song, Orinoco Flow, etc. Yes, the Japanese record industry varies between using the fullwidth space, the dash, and one of the two wave dashes to separate the two parts of the song's title. Because the English Wikipedia has decided not to copy these, and I have proven that in other instances on the English Wikipedia that it is not correct to separate song title from subtitle with a colon, but rather to place the subtitle between parentheses.—Ryūlóng (竜龙) 11:42, 16 November 2009 (UTC)
Those (more or less) English song titles are interesting, yes. Let me sleep on this one. -- Hoary (talk) 16:26, 16 November 2009 (UTC)

Unaccustomed though I am to agreeing with anyone, it now seems to me that Ryūlóng's right. However this came about, parentheses are now rather often used within, um, anglospheric song titles to mark bits of titles that aren't exactly analogous to the subtitles of books but that do come close. Meanwhile, colons are rarely if ever used in this way. (I don't say they're never used, but I can't think of examples where they are used.) So yes: "Mr. Moonlight (Ai no Big Band)", and so on. -- Hoary (talk) 11:17, 18 November 2009 (UTC)

And in that spirit, this move, of an album by Glay (or "GLAY", as their fans might insist). -- HOARY ~in capitals because I'm so important~ 00:04, 23 November 2009 (UTC)
Ack. That was an album. Not a song. Albums get colons. Like Bat Out of Hell II.—Ryūlóng (竜龙) 00:45, 23 November 2009 (UTC)
"Bat Out: Of Hell"? Oh, an album, right. (Me, I tend to see singles and albums rather interchangeably, as consumer products via which impressionable Japanese yoof can express their love of the popsters' hairstyles.) -- HOARY ~in capitals because I'm so impotent~

in capitals because I'm so impotent

Well, if you say so... ···日本穣? · 投稿 · Talk to Nihonjoe 04:46, 11 December 2009 (UTC)


Today, I saw that Unnecessary stuff (talk · contribs) had changed every single romanization of the two katakana characters ッジ from "jji" to "dji", basing this off of the fact that ッチ is romanized as "tchi" instead of "cchi", as well as off of his knowledge of the IPA. I am almost positive that dji is in no way part of the Revised Hepburn system. I reverted all of the edits after giving him a message that I believed it was wrong.—Ryūlóng (竜龙) 00:32, 11 December 2009 (UTC)

He also removed a number of unorthodox katakana combinations from the katakana article on the grounds that they don't exist. They're certainly unusual, and maybe even unofficial, but I googled a few and they are in use -- part of the struggle to more accurately represent foreign sounds that don't have proper Japanese equivalents. For example, he removed スゥ but here's a video game title that uses the thing. I'm not sure if it's worth just reverting all of his removals, though. Thoughts? Doceirias (talk) 00:51, 11 December 2009 (UTC)
I'm thinking it should be undone.—Ryūlóng (竜龙) 01:39, 11 December 2009 (UTC)
And I've undone it.—Ryūlóng (竜龙) 01:52, 11 December 2009 (UTC)
Thanks for catching and undoing these edits. ···日本穣? · 投稿 · Talk to Nihonjoe 04:44, 11 December 2009 (UTC)
Regarding sokuon, I have a question about Hepburn romanization#Double consonants. It says " Double (or "geminate") consonants are marked by doubling the consonant following the sokuon, っ, except for shssh, chtch, tstts." Are shssh and tstts exception to this rule? ―― Phoenix7777 (talk) 23:30, 11 December 2009 (UTC)
They are not exceptions, because the consonants in question are also read as si and tu. There's also no such combination of consonants available in Latin alphabets or English phonetics to otherwise emulate those sounds in Japanese.—Ryūlóng (竜龙) 00:47, 12 December 2009 (UTC)

The reason why Hepburn romanization doesn't mention anything about っじ is simply because sokuon never comes before voiced consonants in native Japanese, and romanization is only meant to transliterate native Japanese into the Latin alphabet. This is how I concluded っじ = dji:

  1. ち (chi) is voiceless and じ (ji) is voiced; "t" is voiceless and "d" is voiced. So if っち is tchi, then っじ should be dji.
  2. Japanese ch is [t͡ɕ] and Japanese j is [d͡ʑ]. っち becomes tchi is because ch ([t͡ɕ]) is not a single sound; っち is pronounced [tt͡ɕi]. The consonant right before the sokuon is actually t. This explains why っち is tchi in Hepburn. From that, we can deduce that っじ is meant to be pronounced [dd͡ʑi] and should be romanized as dji, not as jji; [d͡ʑ] is not a single sound either. [d͡ʑd͡ʑi] (jji) is extremely hard to pronounce and it's not even pronounced that way.

And I removed all -wu combinations from Katakana article because in standard orthography, -wu is generally replaced with -u (wu itself is a combination of two very similar sounds after all). ja:外来語#一覧表 only has kwu and gwu.

By the way, swi used to be replaced with sui, like in スイッチ (suitchi; "switch"). Nowadays, since around 1990, because スィ is used to write the si sound instead of swi, in order to write swi, katakana combinations like スゥィ or スゥイ are needed. ja:外来語#一覧表 actually shows スゥィ katakana combination. So スゥ itself is not in use, but スゥィ or スゥイ is in use.

And, regarding Katakana article, I've never seen kana combinations such as ホゥ, テァ, テゥ, テォ, etc. I have no idea how to pronounce テァ, テゥ, テォ, and so on. (I guess those th- and dh- lines are derived from wapuro romaji but I haven't got any clue where ホゥ came from.)

--Unnecessary stuff (talk) 09:03, 12 December 2009 (UTC)

Except "dji" does not exist in Hepburn romanization, while the various "wa" forms do.—Ryūlóng (竜龙) 20:59, 12 December 2009 (UTC)
As I said above, the reason why dji doesn't exist is because "sokuon never comes before voiced consonants in native Japanese, and romanization is only meant to transliterate native Japanese into the Latin alphabet." --Unnecessary stuff (talk) 00:50, 13 December 2009 (UTC)
So if the concept does not exist in Japanese, we should not make up a new rule for the usage of the Japanese phonetic alphabet when writing foreign words that feature voiced consonantal morae.—Ryūlóng (竜龙) 02:28, 13 December 2009 (UTC)
Well, actually we already made up rules for romanizing ティ, トゥ, ヴ, etc. So why bother? --Unnecessary stuff (talk) 07:11, 12 January 2010 (UTC)
Those exist in Japanese to begin with. Ti, Tu, and Vu, instead of the wapuro Thi, Twu, and...Vu.—Ryūlóng (竜龙) 08:24, 12 January 2010 (UTC)
Well, but they're not even mentioned in Hepburn romanization at all. So technically, romanizing ティ, トゥ and ヴ as ti, tu, and vu is making up a rule. And since っじ's actual pronunciation is dji rather than jji, as we romanize ティ, トゥ, and ヴ as ti, tu, and vu, we can and we should certainly romanize っじ as dji. --Unnecessary stuff (talk) 13:44, 6 February 2010 (UTC)
They are used as "ti", "tu", and "vu" in modern Japanese. "っじ" has never been "dji". It does approximate the "dge" sound in standard English when it comes to writing "edge" or "hedge" in Japanese, but the yoon is derived from the シ sound, which is part of the "sa" sound family. So if anything, just as "チ" is part of the "ta" sound family, ッジ should be come "zji" or something. Anyway, consensus would be against you if you decided to change any romanization schema without agreement here or elsewhere.—Ryūlóng (竜龙) 22:17, 6 February 2010 (UTC)
Well, even though they are indeed used as ti, tu, and vu, romanizing them as ti, tu and vu would still be making up a rule since it's not mentioned in any Japanese romanizations.
And what do you mean "the yoon is derived from the シ sound"? --Unnecessary stuff (talk) 23:11, 6 February 2010 (UTC)
It is not making up a rule because that is how they are treated in practical usage. And what I meant by the "yoon is derived..." is that "ッジ" is from "シ". Making it "dji" woul not be in line with the fact that "ッシ" is "sshi", however I am seeing that it would be proper considering "ッチ" is "tchi". However, there still needs to be an agreement in some fashion to make it that (also you would have to do more than the word "Edge" and also include "Hedge" such as Sonic the Hedgehog in your edits).—Ryūlóng (竜龙)

The reason that っし is romanized sshi not shshi is just because the consonant "sh" is written with two letters, even though "sh" is technically one single consonant. And し/シ and じ/ジ are totally different characters with different sounds.

You actually can't guess like that from the gojuon system. According to the way you guessed, since ち is ta-gyo i-dan character, should it be pronounced & romanized ti?

The sokuon repeats the following ONE consonant. Japanese ch is [t͡ɕ] (t + Japanese sh). So when repeating, the [t] is repeated because [t] is the consonant that is right after the sokuon. This is why it is tchi not cchi.

And Japanese j is [d͡ʑ]. (Even though it is romanized using one letter, it's not a single sound.) So when repeating, the [d] is repeated, since [d] is the consonant that is right after the sokuon. This is why it should be dji not jji. jji means you're repeating TWO consonants, and it doesn't reflect the correct pronunciation.

And in English, there's no pronunciation difference between "dge" as in "bridge" and "j" as in "jelly."— Preceding unsigned comment added by Unnecessary stuff (talkcontribs)

Still, the っじ phoneme is one that exists in modern Japanese and Hepburn romanization covers how it is romanized and not pronounced. And you keep making non sequiturs to just pick apart at my arguments. ち is in practice "chi", but historically and in the wapuro system it is "ti". It is still pronounced as "chi". The only reason it seems that じ is this [d͡ʑ] sound is because ぢ went out of use, and both are homophonous, which as dakuten becomes "di" in the same vein that ち is "ti". This is why English words like "studio" and "radio" are written with the ジ kana instead of the ヂ kana.
But this is still all a non sequitur. Even after several months later, you should stop beating this dead horse and not change "jji" to "dji" because nothing has certainly come about of this conversation.—Ryūlóng (竜龙) 07:43, 4 March 2010 (UTC)
There's a news program called 速ホゥ! apparently. テァ is used in German romanization, apparently. Stein. ja:サンチェス・バルカイステゥギィ. I'm not getting any real world hits for テォ except for a tank manufacturer that might be fictional. I'm just googling them. Doceirias (talk) 09:50, 12 December 2009 (UTC)
速ホゥ!: They call themselves Sokuhou! in English. No need for Japanese romanization. Besides, it is 速ホウ, not 速ホゥ.
サンチェス・バルカイステゥギィ: Called Sánchez Barcáiztegui. No need for Japanese romanization. (talk) —Preceding undated comment added 09:57, 12 December 2009 (UTC).
The listing I turned up used ホゥ. I'm not giving these as examples where we would need to use them, anyway; I'm showing that they are in use at all. Doceirias (talk) 02:49, 13 December 2009 (UTC)

On ja:速ホゥ! page, it says it is meant to be pronounced as そくほう (sokuhō). (そくほう is written as 速報 in kanji, which means "prompt report" or "quick announcement.") And somewhere below in that article, it says:


And this is my rough translation:

Generally it is written as 速ホゥ!, but on the official website, it is written as 速ホウ!; besides, the "!" was sometimes omitted and sometimes not, so until the end the name wasn't unified.

Obviously, in this case, ホゥ is not used for denoting hu sound. And whenever the Japanese language needs to adopt hu sound, it just uses フ (fu). It's because even though it is romanized as fu in Hepburn, its real pronunciation is between hu and fu.

Even if ホゥ itself is used, because it's not for hu sound, I'm removing it from Katakana article. --Unnecessary stuff (talk) 19:09, 13 December 2009 (UTC)

Again, don't remove things if you haven't vetted them properly. Here's another usage: ペーター ホゥ. If your objections to them is that we aren't sure how to romanize them consistently, then just remove the romanization -- there must be a better way of handling this than to act as if they don't exist. Doceirias (talk) 20:21, 13 December 2009 (UTC)
Seems like Høeg is pronounced [hø]. So then ホゥ is for hø/hö?
Anyway, since we don't know how many katakana combinations are and how they're supposed to be pronounced, I think we only should leave the ones that we are 100% sure about, such as ティ. --Unnecessary stuff (talk) 22:06, 13 December 2009 (UTC)
Why remove any of them when we have them exist (in IME)? I know of an instance where ホァ is used to approximate Mandarin, and several other combinations seem to exist despite not being common.—Ryūlóng (竜龙) 22:08, 13 December 2009 (UTC)
If テァ, テゥ, テォ are supposed to be German, I can imagine it is an attempt to convey tä, tö, and tü. Ä, ö, and ü are rounded versions of front vowels, like i and e. I do not know if this is common. I looked up Goethe and München on the German/Japanese Wikipedias and found ゲーテ and ミュンヘン. imars (talk) 20:50, 12 December 2009 (UTC)

Hm, I didn't think of any other languages besides English so I didn't know テァ and テゥ are actually in use. But how are they supposed to be romanized/pronounced? I'm pretty sure it's not tha and thu. --Unnecessary stuff (talk) 10:05, 12 December 2009 (UTC)

I have no problem with removing romanizations from the articles on katakana or hiragana, if the Hepburn romanization system does not cover certain kana combinations or specify how to romanize them. But please do not remove actual kana digraphs/trigraphs, especially ones that are easily attested in google search results. Just because you have never heard of something doesn't mean that it "doesn't exist". — flamingspinach | (talk) 04:51, 16 December 2009 (UTC)

Hepburn indeed does not suggest romanizations for anything beyond the gojuuon, but instead of removing them entirely, it would be more useful to show the way they are usually pronounced (with a note to say that they are not Hepburn), and leave only some of the really weird/unstandardized ones blank/question marks. And I fully agree with Flamingspinach that our job is to record what's out there, not to try to prescribe whether it "should" be used. Jpatokal (talk) 11:37, 16 December 2009 (UTC)
Only if that record is supported by reliable sources. (talk) —Preceding undated comment added 12:12, 16 December 2009 (UTC).

All right then. Every single katakana combination besides gojuon (+ (han)dakuten), yoon must have reliable sources, which means combinations that are unsourced or don't have reliable sources will be removed. --Unnecessary stuff (talk) 13:44, 6 February 2010 (UTC)

That would be WP:POINTy if you followed through with it. Because this would result in your removal of everything from the katakana table that is in the "Modern digraph additions..." section, despite the fact that there are some that are in standard use, such as "wo", "va", "she", "ti", "tu", "zi", etc.—Ryūlóng (竜龙) 22:17, 6 February 2010 (UTC)
So if those in-use combinations are sourced, then they wouldn't be deleted. --Unnecessary stuff (talk) 23:11, 6 February 2010 (UTC)
Thos combinations are "common knowledge" and therefore don't need to be sourced. You are apparently the only one who doesn't know this. I recommend studying your Japanese a little longer. ···日本穣? · 投稿 · Talk to Nihonjoe 01:59, 7 February 2010 (UTC)
Removing anything from katakana is not necessary, given the fact that you do not have a consensus to do so. Also, is there some reason you do not include the section header in your edit summary, which would certainly make it much easier to reply to your edits?—Ryūlóng (竜龙) 23:29, 6 February 2010 (UTC)
Well, at least ホゥ is apparently not for hu. If unsourced kana combinations are allowed to be included on that page, then people can just make up their own kana combinations. --Unnecessary stuff (talk) 07:05, 4 March 2010 (UTC)
I regret to inform you that people have and will continue to make up their own kana combinations, regardless of what the Wikipedia page says... Jpatokal (talk) 08:33, 4 March 2010 (UTC)
I recommend walking away and giving it a rest for a while. You are very clearly in the minority opinion here on pretty much everything, so beating the desiccated corpse of the horse is not going to help your case any. I strongly recommend dropping the issue for now, going and studying up on it a bit more, and not coming back to this issue for at least several months. ···日本穣? · 投稿 · Talk to Nihonjoe 07:55, 4 March 2010 (UTC)