Help with Chinese Romanization (1147173)

Posted under General

In post #1147173, it's 謝謝你!壞運, or Thank You! Bad Luck in English, as I found out. It's a Taiwanese light novel, apparently. I have no idea what to do with that, though. From browsing Wiki, Xie Xie Ni! Huai Yun or something? Might using the English title be preferred? Either way I'm leaving this to someone who knows better.

Updated by NNescio

Warning: Wall of Text

jjj14 said:
As long as we're romanizing copyrights, post #926051 and post #1074755 are from a Korean series called "일편흑심". post #926051 is currently tagged "ilpyonheukshim", but Google's romanizer gives "ilpyeonheugsim". Which of these is correct?

"ilpyonheukshim" uses the South Korean variant of McCune–Reischauer romanisation. This should be written as "ilpyŏnheukshim" (note the breve over the o), as o denotes another vowel (ㅛ) instead of ㅕ. People tend to drop diacritical marks when typing though, and Danbooru doesn't like those marks either,

McCune–Reischauer is a transcription system, which transcribes the pronunciation of words into a form readable by people familiar with the Latin alphabet. Korean has some sound changes, and the ㄱ is transcribed as "k" here, because that's how it is pronounced. In other places, ㄱ may be transcribed as "g" or "ng".

"ilpyeonheugsim" uses the transliteration variant of Revised Romanization. RR romanises ㅕ as "yeo" instead of "ŏ", and ㅅ as "s" instead of "sh". There are two main variants of RR. The first is transcription, which, like McCune–Reischauer, transcribes sounds, and phonetic changes are reflected in the spelling. ㄱ is usually transcribed as "g", but may be transcribed as "k" or ng" elsewhere. Under this system, 일편흑심 would be transcribed as "ilpyeonheuksim".

The second variant for RR is transliteration, a system that transliterates letters or symbols from one writing system into another. Transliteration doesn't care about sound shifts, and ㄱ is always transliterated as "g" regardless of how it is pronounced, so 일편흑심 is transliterated into "ilpyeonheugsim".

I recommend against using McCune–Reischauer. RR is the currently accepted official standard in South Korea, and McCune–Reischauer has fallen into disuse. While I don't know Danbooru's policy on romanising Korean, we have adapted Pinyin for Chinese and Hepburn for Japanese. The former is the de facto and de jure standard in the PRC, while Hepburn is the de facto official standard in Japan. Similarly, I think RR should be the preferred choice here.

Furthermore,McCune–Reischauer frequently uses diacritic marks and apostrophes. As mentioned, Danbooru doesn't like diacritic marks, and dropping them may cause ambiguity between different vowels. Apostrophes are also used to separate syllables in Danbooru, which may cause confusion when McC-R is adopted as the latter uses them to differentiate between aspirated and unaspirated consonants (similar to Wade-Giles for Chinese). RR, meanwhile, only uses the English alphabet.

I am not that sure whether we should use RR transcription or RR transliteration though. I believe transcription is more common in Korea, and most companies prefer this method to avoid people mispronouncing their names. Transcription also has the advantage of telling you the pronunciation.

Google uses transliteration because it's easier to implement and provides a one-to-one correspondence between Hangeul and the transliterated text. When using RR Transliteration, you don't need to know how a word is pronounced, merely how it is written. This makes it easier on machines, and non-native speakers without a good dictionary (rules exist for sound-shifting but there are some exceptions). And if Google-Translate does indeed use RR Transliteration instead of some modified variant they come up themselves, this allows Danbooru users to use it as an automated tool to transliterate Hangeul easily, even if they can't read Korean.

Danbooru also seems to favour transliteration over transcription; we spell 「先輩」 as "senpai" and not "sempai", 'though this may be a side effect of choosing Hepburn and hence not expressly intended.

In any case, an exception should be made for common Korean names. 李(이) is usually romanised as "Lee", not "I", as would be the case under both RR and McC-R.

TL;DR:

Ilpyonheukshim: Not recommended.
Ilpyeonheuksim: How the word is pronounced.
Ilpyeonheugsim: How the word is written.

1