Schrobby said:
Hard translated stuff is second rate, but sure better than nothing at all.
If we believed something was better than nothing, we wouldn't be discouraging amateur translations on Danbooru.
Updated
Posted under General
Toks said:
Sharing a writing system means very little as for how easy it is to translate between the two languages while preserving accuracy. The grammars of the two languages would make a much bigger difference.
I don't know Chinese in any detail so correct me if I'm wrong, but from just quickly looking up the basics it seems absolutely nothing like Japanese. Chinese doesn't have inflection, it's SVO, it has prepositions instead of postpositions... if anything it sounds like Chinese and Japanese grammar may be more different than English and Japanese grammar are.
As a native Chinese speaker and a frequent JP->EN translator hereabouts, I can assure you that all your above points are entirely correct.
--
As for the main topic, I'm against second-hand translations. Playing the telephone game can lead to hilarious results, even when both the JP->CN and CN->EN translators are highly proficient. Same goes for any other intermediate language. Some meaning is always lost in translations, as human languages never correspond on a one-to-one basis with one another.
This goes double when idioms and/or figure of speech are involved.
Updated
CodeKyuubi said:
As the saying goes, better something (As long as it isn't blatantly terrible) than nothing.
Schrobby said:
Hard translated stuff is second rate, but sure better than nothing at all.
Danbooru has high standards, both for uploading and translating, so "a bad translation is better than nothing" isn't going to fly. The translation guidelines say this right in the first point, and always have.
Kadoya said:
I'll go 180 and say no to hard translations then. Besides, NNescio made a good point about the translations and it's still a recent doujin.
Now you see, a month has passed, and I've tried all I could to persuade those Chinese who bought this doujinshi and uploaded the Chinese-translated version, to upload the original raw, but it just won't work.
Wouldn't it be best to just go for "The best possible result" rather than limiting ourselves to nothing but the best? If nothing else is available and the usual methods have been attempted, then what else is there to say?
Priority should be given to the highest quality possible, but it should not be considered the only option available if you ask me.
So it seems I have contributed to the problem. My apologies.
But anyway, although the use of grammar maybe different in Chinese and Japanese, many expressions correlate specifically enough in each language that it's possible to get a decent translation out of it. Idioms, figures of speech, puns, and slang notwithstanding of course. Sharing a writing system means sharing a means of expression. And that means a lot to translations too.
In many cases, especially sentences that are kanji-heavy, there are often ways to translate into Chinese using mostly the same characters while keeping most of the meaning.
In fact, I'd argue that most kanji can be translated into Chinese without losing any meaning by a competent translator. It's almost too effective compared to translations between two other languages at least. The ambiguity in Japanese can almost always be kept by using equally expressive and ambiguous words in Chinese. Like Japanese many parts of a sentence in Chinese can be left out, so there is no need for translators to "make stuff up" from the context like JP > EN. Not to mention the ambiguity in Japanese culture will probably never be lost from JP > CN due to cultural similarities.
As for other cases, unless the sentence has idioms, figures of speech, puns, or slang, the result should generally be fine. Context is usually enough to understand something. And for a doujinshi that should be no problem. As an example, even honorifics can be directly translated into Chinese, and you can see it on basically any doujinshi translation nowadays.
As for the case at hand, it seems yohane was never pun, idiom, slang, or figures of speech heavy. So why should he be this time? The Chinese version has no pun, idiom, slang, or figures of speech, so the translation should be competent.
Updated
Even if it were theoretically possible to translate Japanese -> Chinese without losing meaning, there's no guarantee that that actually happened. How do you know that the Jp->Ch translator didn't make a bunch of mistakes? You don't. The only way you could know for sure is if you read the original Japanese and compared it to the translation, but that's not an option here since we don't have access to the original raws.
finalagent said:
limiting ourselves to nothing but the best?
That's the whole point of danbooru. Compromising on quality isn't high quality.
I hate to use the "slippery slope" argument, but this is the exact kind of thing it applies to. You make one exception to the rules then you have to make even more exceptions to legitimize that choice until the rules become meaningless.
So considering that if I were to upload a sample by Yohane himself. Say
http://file.sangensei.blog.shinobi.jp/ata645.jpg
Would the 4-5 pages of sample raws be enough to prove the validity of the Chinese hard-translated? pool #7180 as you can see, has huge sections of the doujin in Chinese. Yet I assume the reason it exists is because it bridges the gap between the sample Japanese raws?
Hollows said:
pool #7180 as you can see, has huge sections of the doujin in Chinese. Yet I assume the reason it exists is because it bridges the gap between the sample Japanese raws?
It's not only hard translated, but also very poorly scanned and not cleaned.
There's literally only one reason it exists on danbooru: because Saladofstones approved it.
And about "bridging the gap". There has been a discussion about discouraging uploading samples of comics since they in turn encourage uploading the whole doujinshi.
Hollows said:
So considering that if I were to upload a sample by Yohane himself. Say
http://file.sangensei.blog.shinobi.jp/ata645.jpg
Would the 4-5 pages of sample raws be enough to prove the validity of the Chinese hard-translated?
That can only prove the validity of those specific pages. You don't know that the translator didn't make a mistake on one of the other pages that Yohane didn't provide a sample of. A translation doesn't have to be 100% correct or 100% incorrect, being partially incorrect is more likely.
So we went from topic #9402 and similar threads to seriously discussing whether hard-translated doujinshi should be allowed?
Seems like we're sliding down the slippery slope.
Fred1515 said:
So we went from topic #9402 and similar threads to seriously discussing whether hard-translated doujinshi should be allowed?
Seems like we're sliding down the slippery slope.
Yeah I have to agree with this, danbooru isn't a site where hard-translated doujin should be uploaded with the hope that a later, original version will be found and uploaded.
There is no guarantee of its accuracy.
Given I have actually approved comics that have turned out to be hard-translated, I'll go through and flag them. I'm going to be more careful in the future.
Its worth noting, however, that its not spelled out that posts should be flagged for being hard translated, but I think there's enough precedent to formally blacklist it. At any rate, it seems like the best option for quality control.
Fred1515 said:
So we went from topic #9402 and similar threads to seriously discussing whether hard-translated doujinshi should be allowed?
Seems like we're sliding down the slippery slope.
just to note, there's a janitor-only function of merging topics (issue #1289), i think it's implemented already (though i can't be sure). if this thread just replicates the earlier thread, it can be safely merged, the last reply there 29 days ago happened to be the same date the OP of this thread.
ghostrigger said:
just to note, there's a janitor-only function of merging topics (issue #1289), i think it's implemented already (though i can't be sure). if this thread just replicates the earlier thread, it can be safely merged, the last reply there 29 days ago happened to be the same date the OP of this thread.
Its related but not enough for a merge.
Both are related to doujin, but the other thread is looking at whether doujins should be approved, how many are being approved, and why they are being approved.
This thread is specifically about hard translations when a raw isn't available.
albert said:
Would it make sense to make another site for manga/doujinshi/comic series with less stringent rules? I agree that a lot of these comics have mediocre art and don't really belong here.
Spinning the stuff off to a separate site has been proposed in the other thread, but how separated or integrated with danbooru would the new site be? Would it impact how moderation is performed? Also where would the line between what is acceptable here and there probably be?
Anyways, I could support the creation of a site dedicated to these kinds of things. In the other thread it was said such stuff could overload those who could translate, but I also think the setup could serve in attracting those who would like to (and can) translate as well.
Specifically on the issue of hard translations, I feel inclined more toward supporting having something versus nothing, but alternatively I also try to avoid approving hard translated multipage comics and don't really make exceptions on not approving them when I've identified them as such. This is so even if I can't find a Japanese RAW to suggest as replacement.
