To All Members:
Please use this thread to discuss any second hand translation issues.
* second hand translation: translations based on English translation of (Japanese) raws
mikkih, TA mod
(This post was moved from TA main board. mikkih)
hi guys .
I'm EN-AR translator, and I've been wondering how is my translation accuracy is since I'm translating from JP-EN-AR, and I came up with 60% accuracy!
anyway, what's your opinion of this? BTW, I only translate from trusted and good reputation teams (Mangastream for Fairy Tail and Naruto, and HAO for Dragonball) .
thanks in advance
Last edited by mikkih; November 24, 2013 at 06:57 PM.
Since the Japanese language and the culture behind it are so different, without a doubt, nuances and connotations of the raws are lost when translated into another language significantly + translation errors can be made easily. Besides, there are too many ways to translate a single Japanese phrase into English imo.
Have you played the game "telephone" when you were a kid? A person wispers a phrase in someone's ear and then the person passes it along to the next. The last person reveals how distorted the original phrase becomes.
JP-ENG translators you mentioned may be great, but that's how I feel about trnasltion in multiple languages in general.
There is one discussion thread in TA, trying to figure out what some English sentences mean. Here is a link if you are interested.[LINK].
Discussion thread on subtle nuances in translation @ TA Classroom: Link
This tells you how difficult it can be to translate JP to ENG.
Last edited by mikkih; December 09, 2013 at 04:18 PM. Reason: Edit/Add
Well, what is the goal of translation? It is to convey the meaning of the original in the target language as accurately as possible, ideally with 100% accuracy. Is it possible to achieve 100% accuracy with double translation? No, not really. Therefore, if you engage in re-translation, from the very start you are accepting that the quality of your work will be sub-par. Due to this, I'm tempted to say that re-translation is unacceptable if you care about the quality of your work. However, there is an additional complicating factor: the fact that we are fan translators here and there are no quality controls of our work. What this means is that there are plenty of J-E fan translators who are so bad that a double translation done by two decent translators would be superior.
In general, if we want to discuss quality of double translations in manga scanlation, we have to look at 3 flavors of them: 1) English scanlators not being able to find Japanese tlers or Japanese raws and translating Japanese manga through an intermediate language; 2) international scanlators who find a translator from Japanese to their native language (as far as I know, a very rare occurrence in the international scanlation scene... I assume only China and Korea would have an abundance of Japanese tlers) and decide to branch out into English scanalation by having someone else translate from their native language to English; 3) the OP's example, international scanlators re-translating from English.
I can't speak for the international scanlation scene in general, since I'm sure it all depends on which language you're talking about, but most likely the key is how wide-spread English is in that particular culture. This means that the larger the country, the less need of English they have, the worse their re-translation quality is going to be -- after all, their native translators would not be very good. In addition to poor translation quality from English, international translators are often plagued by incomprehensible English sources. Like I mentioned previously, there are plenty of bad fan translators. I know that simply because some international groups that I'm friendly with often ask me to help them understand confusing English sentences in English scanlations. What I find in 90% of such cases is that either the J-E translator didn't care enough to translate the original Japanese correctly or their grasp of English grammar is tenuous at best or, most commonly, both. When I ask those international scanlators what they did before they met me and started asking me to check problematic sentences, they invariably reply that they just made something up they thought would fit the context. But challenges don't end there -- what international scanlators often don't know, since they have no way of telling, is that a large chunk of English scanlation is already the result of double translation. When they re-translate that to their native language yet again, they end up already with triple, not double, translation. It's a perfect game of broken telephone. For this reason, in the international scene, anyone who knows English well enough reads English scanlations, not their native language re-translations of them.
As for international groups trying to branch out into English scanlation, the results I've seen were invariably disasterous. No need to discuss their translation quality. For whatever reason, they are never able to find a translator to English who would be able to write in comprehensible English. I'm not even saying that their English is ungrammatical. That could be fixed by a proofreader. I'm saying it is incomprehensible. I've seen such releases from Vietnamese and Indonesian groups, and they weren't pretty. I would speculate that the reason is that those groups are based in their own countries, where even if you find someone able to read English, it would be very difficult to find someone able to write it. Those who can read their language and write well in English live in English-speaking countries, and if they are interested in scanlation, they would likely join the English language scanlation scene.
Now, closer to home, let's talk about double translation to English. Of that, we have two dominant types: J-C-E (C=Chinese) and J-F-E (F=French). They can range in quality from okay to terribad. Based on my (admittedly not very extensive) experience of checking these against the original, the terribad translations come from J-C-E, and J-F-E re-translations are very often quite decent (not without issues, of course). I believe there are two reasons for that:
1) As has been mentioned by mikkih, closeness of languages is important. F and E are rather close, and even though they belong to different language families, the infusion of French vocabulary into English cannot be underestimated, which brought the languages closer after the fact. It is possible to make perfect translations between English and French even for fan translators. J-C-E triplet, on the other hand, is made up of languages none of which have any relation to each other (the fact that Japanese borrowed Chinese writing system is of no consequence really, since Japanese grammar remained utterly unaffected). So from the point of view of pure linguistics, J-F-E translations have a far greater probability of being successful.
2) I believe the second, and more important reason, is the nature of translations from Japanese. The source for F-E translations is professionally translated manga published in France. The source for C-E translations is fan J-C scanlations. As you can imagine, while J-F translations are nearly free of issues, J-C translations can have all the issues inherent in fan projects. That's why groups doing J-C-E translations frequently end up in the position of international scanlators struggling with bad translations to English. Also, in my opinion, the pool of fan F-E translators is superior in quality to the pool of C-E translators. Either because of the general level of language knowledge, or because French is closer to English, I don't know about that.
So to summarize, while in the fan community it is possible for re-translations to be superior to poor first translation, on average, choosing to do double translation is a deliberate downgrade in quality. Moreover, choosing to do any re-translation other than J-F-E is an even further deliberate downgrade in quality. I guess it all comes down to this question: does a certain manga really need to be released so badly that you would agree to reduce the quality of its translation? I'm certainly not saying that J-E translations are perfect, but at least they have the possibility of being perfect in theory. However, choosing to re-translate you remove that possibility before even starting. Again, if you have no chance of getting something right, does it really need to get done? Or are you simply grabbing more projects out of greed or misguided desire for popularity?
Well,... I'm not really a fan of double translations, as the quality varies greatly for sure. But so do Japanese to English translations... saying doing a double translation is just be wrong or not accurate enough, would be stupid.
Once (quite a long time ago) I translated some Japanese lyrics to English and later on I also translate those same lyrics from Japanese to Dutch. When I started to compare the English and Dutch translation, I actually noticed if I had just translated from English to Dutch, my translation would've been essentially the same.
So, I think it really depends on the quality of the "English" source that is used, and your own knowledge of English that decides how accurate your re-translation will be. No matter what, everything something is translated into a different language, something will be lost... (however, sometimes things actually get added... also in professional translations!) Just saying, but just because a translation is done by a large publishing company doesn't directly mean the translation is more trustworthy than a fan translation. I've actually been comparing some official English translations to the original Japanese. And sometimes the English translation really doesn't equal the Japanese (at least, in my opinion. That also happens with fan-translation btw.) Yet, because it somehow fits in the context, no one notices or complains about it.
Anyway, back to the double translations. Just be careful in choosing your English sources and then I guess you'll be able to create a decent translation to your own language.