Aliasing houdzuki_akane to houzuki_akane.

Reason: Proper romanisation

s

shii said:
Isn't Houdzuki more common? What "proper" system are we using here? Writing "hou" for "hō" isn't a part of any official Romanization scheme I know of.

Modified Hepburn (as permitted for personal names by the Japanese Ministry of Foreign Affairs) because people can't type macrons.

Your website is inaccessible for me with a 500 error. 'dzu' isn't a part of any official Romanization scheme I know of, either. Nihon-shiki uses 'du' and most everything else uses 'zu'.

7HS said:
Modified Hepburn (as permitted for personal names by the Japanese Ministry of Foreign Affairs) because people can't type macrons.

Uh... then shouldn't the tag be changed to hoozuki_akane, touhou become toohoo, and so forth?

My site is being DDoSed, I sent an e-mail to Dreamhost.

shii said:
Uh... then shouldn't the tag be changed to hoozuki_akane, touhou become toohoo, and so forth?

Well, wikipedia's source (http://www.pref.kanagawa.jp/osirase/02/2315/hebon.html) says "長音表記とは、氏名に「オオ」、または「オウ」と伸ばす音を含む場合であり、「O」または 「OH」、「OU」、「OO」などの選択が可能です。" So I suppose you could if you wanted to, but there's no particular need.

shii said: Uh... then shouldn't the tag be changed to hoozuki_akane, touhou become toohoo, and so forth?

No, because 'ō' is ambiguous, it can be 'oo' or 'ou', which is exactly why we don't use macrons, in addition to them being difficult to type.

Touhou is, so far as I've ever seen it "spelled" out, とうほう. That's what it's being based on, the actual written syllables. That's our source of disambiguation. I've never seen it written out as とおほう. This as opposed to the Fate/Stay Night character, 遠坂凛（とおさか りん）, Toosaka Rin. The 'ou' and 'oo' are not interchangeable, as for example とうい and とおい are two distinct words.

At any rate the syllable under consideration was the づ, not long vowels, which we set tiles down for awhile ago. According to the Japanese Wikipedia article for Gyakuten Saiban, the character is the following:

宝月 茜（ほうづき あかね）

So 'Hou' and 'ki Akane' are set. As for the づ, it seems like it's been practice here to use the 'zu' romanization, rather than 'du' or 'dzu'.

I can't remember what the specific argument for/against was. I admit to at one point liking unique romanizations for every kana (so not using 'zu' for both づ and ず)　back when I was taking a few semesters of Japanese, just to help me learn more quickly. (Also trying to get back in the habit of using 'jya' for じゃ, not 'ja'.)

But basically, what I want is consistency. 'zu' seems to be pretty consistently used on danbooru.

Japanese people pronounce づ and ず identically, at least since Edo era. They use ほうづき for 宝月 just because 月 is つき.

Romanization of Japanese is really a difficult problem...

s

"houdukiakane " would be the waapuro, wouldn't it? I mean, the "wa-puro ". :P I've seen many Japanese people write じゃ as "zya", しゃ as "sya", じ as "zi", etc. That is technically more consistent with the kana spelling -- さしすせそ = sa si su se so, ざじずぜぞ = za zi zu ze zo, たちつてと = ta ti tu te to, だぢづでど = da di du de do. So じゃ = zi + small ya = zya, I guess.

To English-speaking sensibilities, it looks a little silly, but in the end I guess it is somehow more logical. Once you stop writing in ローマ字 it becomes hard to go back :P In fact, how do you even romanize　ローマ字? ro-mazi ? Hmm...

But in the end, romanization is generally for the purpose of helping people who can't read actual kana or kanji to pronounce Japanese words, and for that purpose, one of the standard romanization systems is probably more utilitarian.

The most confusing point is that Japanese kana has been changed in the long history, and therefore its has not one-to-one correspondence with pronunciation.
So, romanization of Japanese has one of the two different purposes:

1. Express Japanese kana by the Roman letters (i.e. Japanese IME)
2. Express Japanese pronunciation by the Roman letters (i.e. guide for non-Japanese people)

Based on purpose 1, we should use "dzu"(or "du") for づ. But based on purpose 2, I don't agree using those. It depends on the consensus in Danbooru.

Well, since we're using ou and oo to maintain the distinction between おう and おお, even when they're pronounced the same, it would be consistent to do the same with ず and づ. I support "houdzuki".

Then we'd better switch yumizuka_satsuki to yumidzuka_satsuki as well, hadn't we? I dunno, we have very rarely distinguished between づ and ず on danbooru, as far as I've been able to tell.

For people who want to subordinate our system to Hepburn, imagine these future aliases:

kaenbyou_rin to kaembyou_rin

kagamin_bocchi to kagamin_botchi

tenjinmon_kasumi to tenjimmon_kasumi

I'm not saying we have to ignore "standards". The standard I linked above codifies a system, used widely on Danbooru, which is superior in representing in both sounds and morae. "Tenjinmon" is better than "tenjimmon" because we recognize the word "tenjin" within it, and some among us can pick out the kanji　天神. Similarly, "houdzuki" hints at the kanji 月, whereas "houzuki" does not.

ん doesn't become 'm' in revised Hepburn. っち apparently does still become 'tchi', which I was unaware of.

Regardless, I don't see a reason why we should embrace something written on your wiki over anything else. Particularly when the main functional difference is the use of 'dzu' for づ, something which none of the standards do. And it doesn't work in wapuro either, so what's the point?

Well, if we scroll up to the page you linked to above...

「B」「M」「P」の前の撥音については「M」で表記する。
[例] なんば→NAMBA ほんま→HOMMA さんぺい→SAMPEI

"Dzu" is not wapuro but it suggests the correct wapuro Ｄｕ　づ whereas "zu" can be nothing other than　ず in wapuro.

shii said:
In some areas づ is still pronounced "dzu"--Kyoto, for example. It's also useful for etymological purposes.

As is ず. There is absolutely no useful phonetical distinction between ず/づ that would be somehow reflected in writing.

LaC said:
Well, since we're using ou and oo to maintain the distinction between おう and おお, even when they're pronounced the same, it would be consistent to do the same with ず and づ. I support "houdzuki".

No. If anything, we should standardise on "ou" and drop "oo", which has been mentioned previously, and which I've been considering. Stop furthering idiotic romanisation schemes, please.

Bottom line: We're using revised hepburn + some sanity-related modifications like dropping macrons in favour of "ou". Additionally, we let in some deviations from the general "purely phonetic" scheme for the sake of aesthetics and being a bit easier to get right. Currently we have:

• を → wo, not o (but は → wa). IMHO, this one is perfectly safe, since wo occurs in no other contexts, whereas "ha" would be pretty dangerous.
• nm, np, not mm, mp. This is fairly arbitrary, but IMHO looks better this way, and buys us a little bit more predictable spelling for the price of just a tiny bit of pronunciation inaccuracy.
• cch instead of tch. "cch" looks better and is more sensible as doubling of ch, not to mention consistent with ssh. IMHO, the "proper" hepburn fails with this one.
• both ou and oo. This one is actually inconsistent with the policy on zu/dzu, and I've been considering proposing revising this. I'm ambivalent; on one hand, I'm used to the distinction, OTOH, there's no real reason to maintain it, given how おお and おう are totally identical phonetically.

What we don't have:

• "oh" for おう. It's a totally retarded, self-inconsistent and completely unjustified aberration introduced by God-knows-whom (I suspect clueless US fansub typesetters[1] played a big role here) and inexplicably picked up.
• du/dzu vs. zu, as explained above. If you want 1:1 mapping onto kana, use one of the abominations like Nihon-shiki, or better yet, just use kana. And stop pretending that it gives you any information about kanji[2], if you want that, just write the goddamned kanji directly.
• all other sorts of Nihon-shiki idiocies (sya, zya and friends), as well as the even more made-up schemes (like "Tenjho Tenge", I want to burn whoever invented that)

[1] That doesn't mean all typesetters are clueless. But a sizeable portion is, and I've seen all kinds of atrocities committed by them.
[2] Especially given how not a single fucking group could get Azuma Hazuki's name right, with roughly half of them reading it as "Hatsuki".

葉月 said:
No. If anything, we should standardise on "ou" and drop "oo", which has been mentioned previously, and which I've been considering. Stop furthering idiotic romanisation schemes, please.

Now that would be idiotic. If you're going to standardize on one, it would have to be "oo".

This is getting over the top, and I vote we stop tweaking Danbooru's romanization rules, period. What we have been doing works in virtually every situation.

Keep doing what we've been doing, which has included 'zu' for づ and keeping the ou/oo distinction (or whatever it takes to avoid 'oh', though blanket 'oo' policy would bring enormous headaches). I don't care if it's perfectly consistent, because the difference is so minor. Just stop arguing about things that will bring very little gain and much retagging/alias work.

LaC said:
Now that would be idiotic. If you're going to standardize on one, it would have to be "oo".

What? Why? おう is by far the more common one, おお occurs in a couple kanji, most of which only have it in name readings. For practical purposes, it's limited to 大, 多 and not much else.

Because it's a long "o", not an "ou". "ou" is like "si" or "zi": for example, しよう is "siyou" if you follow kana, "shiyoo" if you follow pronunciation.