Well, yes. But I also disagree that our current system is focused on phonetics.
Posted under General
Re-reading the last few entries here I think I misunderstood what Hazuki was saying earlier and what he means by being based on phonetics and not spelling.
In the particular case he cited "Hazuki", we don't have a 1-to-1 correspondance. in the case of both /zu/ (ず) and /du/ (づ) we use "zu" due to pronunciation, likewise both /zi/ (じ) and /di/ (ぢ) we use "ji" due to pronunciation. Because of those cases (and the aforementioned topic and object markers), our system isn't truly 1-to-1. It is however consistent, and that's what I had meant to illustrate.
I stand by my statement of the particular circumstances considering the non-compliance of "Haduki" by some members though.
EDIT: I'm actually being redundant because this is exactly what Zatchii had said earlier. But I think it explains the confusion here.
Updated
Katajanmarja said:
- If I should globally standardize the romanization of Japanese, I would certainly go for a pure transcription (something like Itō Noiji) or for a pure transliteration (something like Itou Noidi), not for a mix-up.
This is one hundred percent correct. Itō Noiji is a phonetic transcription approximating an extension of English orthography. Itou Noidi is a transliteration from kana to Latin. 葉月 or anyone else who disagrees with this is way off the mark, sorry.
葉月 said:
Waapuro schemes have to be unambiguous
Uh... no. Waapuro schemes need to be as inclusive as possible, since by definition they are meant to be input methods, not output methods. This introduces tons of ambiguity. Google renders both "hwaltuku" and "fakku" as ファック, for example, and for good reason.
葉月 said:
You're confusing "transcription" with "using only single characters for sounds". It's perfectly fine to use digraphs; in fact that's exactly what happens when you write "ji" or "shi". Or IPA /tʃ/ for that matter. Our scheme *is* almost pure transcription, with the sole exception of "wo", which is not done because it's infinitely less confusing this way. Oh, and the "oo"/"ou" split, which is just the way things are.
Don't say that a person is confusing X with Y when X has nothing to do with Y, and even if it did, Y has nothing to do with what the person said. And even if you must make such accusations, please do so while making at least 0.5 iotas of sense. "ji" is a digraph, yes. It also represents two sounds. "shi" is a trigraph representing two sounds. Whether you were talking about digraphic representations of consonants or of syllables (I assume the former), something's wrong there. (If I just somehow failed to understand your intricate and hidden point, forgive me.) But you're right, our romanization scheme is mostly a transcription. It's also mostly a transliteration; this is because kana itself is for the most part a transcription of spoken Japanese, unlike English orthography's odd relationship with spoken English.
Now, the reason that Nihonsiki / Kunreisiki are arguably better transcriptions than Hepburn is that the mapping between phonemes and graphemes is as simple as possible (in terms of the representation of the postulated graphemes as graphs, and the relationship between that representation and the structure of the phonemic bestiary of Japanese). Nihonsiki / Kunreisiki are also arguably better transliterations than Hepburn, for much the same reason - because kana is well-attuned to the phonology of Japanese. However, Hepburn is much better-attuned to the phonology of English: the image graphemes it maps Japanese phonemes to are more easily describable in terms of English graphemes (disregarding the bare structure of Latin graphs). So which is preferable is really dependent upon your personal definition of parsimony.
But none of this changes the fact that what you quoted Katajanmarja as saying is unimpeachably correct. Sure, that paragraph doesn't say a whole lot (it just says that Itō Noiji is a transcription and Itou Noidi is a transliteration, and makes some statements about the speaker's beliefs), but it is not wrong.
zatchii said:
I also disagree that our current system is focused on phonetics.
Like I said, kana is pretty close to phonetic as it is, so "focused on phonetics" vs. "focused on spelling" is pretty much a moot point when it comes to romanization of Japanese.
Bastille said:
Interesting. I've never heard anyone say it that way, even amongst my friends who actually know Japanese.
Here's a good reference for the pronunciation of Danbooru.
EB said:
Here's a good reference for the pronunciation of Danbooru.
Audios aids are always so much more effective for this sort of thing, thanks.
Oh yeah -- how and when did Danbooru get its name?!
***
葉月 said:
You're confusing "transcription" with "using only single characters for sounds". It's perfectly fine to use digraphs; in fact that's exactly what happens when you write "ji" or "shi". Or IPA /tʃ/ for that matter. Our scheme *is* almost pure transcription, with the sole exception of "wo", which is not done because it's infinitely less confusing this way. Oh, and the "oo"/"ou" split, which is just the way things are.
I am afraid I do not completely get the digraph thing with ji there, and see no reason to argue about it. Let me just note that I come from a culture where the "one grapheme for one phoneme" ideal has been widely promoted by most linguists.
My main problem is exactly the oo ~ ou split: I find it really hard to understand why anyone would do that without doing di ~ zi as well. In other words, why is the etymology of /ō/ so much more interesting than the etymology of /j/? But around this point I also run out of knowledge of Japanese...
jxh2154 said:
(Also, macrons are worse than genocide. Just sayin'.)
Not too nice towards poor Latvians, who use macrons as a basic part of writing their own language. ;)
Updated
Katajanmarja said:
My main problem is exactly the oo ~ ou split: I find it really hard to understand why anyone would do that without doing di ~ zi as well. In other words, why is the etymology of /ō/ so much more interesting than the etymology of /j/? But around this point I also run out of knowledge of Japanese...
It's not. And in fact I've advocated dropping this distinction in the past, but that proved too much of a change to be acceptable.