Carl Bielefeldt on translating Dogen
Found an interesting article by Carl Bielefeldt of Stanford: Circumambulating
the Mountains and Waters. I don’t know Carl, but it seems like he’s one of the pre-eminent Dogen scholars we have in the West. His article is about translating Dogen’s “Mountains and Waters Sutra” in particular, and translating Dogen in general. He has great insight into the whole question of translating loosely versus translating literally. Suddenly my past musings on this issue in the context of computer manuals seem very shallow. He says:
Of course, there are lots of ways to translate, each with its own virtues and vices. When the translator doesn’t understand what the author is talking about, probably the safest approach is to keep as close as possible to the author’s language. Every translator has to cook her text, but the trick in this approach is to try for no more than medium rare, so the reader can still taste some of the raw juices of the original words.
Yes—but there’s a difference here. Eating a picee of meat, raw juices are raw juices and the eater can taste them as they are. But once something is translated, from Japanese to English in this case, the raw juices are gone forever. (In culinary terms, it has at least been “seared”!) It doesn’t really seem to me, for example, that using the word “practice and verification” (for 修証) in your English translation is leaving any “raw juices” for the reader to enjoy. Unless he’s a Buddhist scholar, and can back-translate in his head to the original Japanese—but in that case, why is he reading an English translation in the first place? Carl goes on to say:
The chief virtue here, at least when all goes well, is that the translation will have less of the translator’s own ideas.The chief vice is that the translation will be hard to read,with a foreign feel, full of odd diction and unusual syntax. Sometimes, this minimalist approach may catch more of the author’s style; other times it can distort the style, making what may originally have been smooth and flowing for the native reader into something twisted and clunky. Sometimes, it can make a passage seem more difficult or more exotic than it really is, turning what was fairly easy and idiomatic into something strange and fraught with unintended mystery; but it can also preserve some of the original strangeness and keep open mysteries that are inherent in the text.
It’s nice to objectively list the trade-offs between more and less literal translation like this. But the real question is: how do you weigh those trade-offs, and what side do you come down on? Personally, I think Carl has captured the negatives of literal translation extremely well here and they far outweigh the smaller number of positives. “Hard to read”, “distorted style”, “clunky”, “strange”—I couldn’t have come up with better words to describe the vast majority of Dogen translations I’ve read.
More basically, I think it is a false dichotomy to say that there is a tension between something smooth and something faithful. The implication that a translation might read smoothly just because the translator has “papered over the cracks” is wrong in my opinion. On the contrary, clumsiness often results from a combination of not being faithful (usually due to not understanding the source text), and lack of effort in crafting the target text. If the text is understood, and the effort is taken, the result can naturally become smooth. Carl now goes on:
Every translation is a bunch of trade-offs, every translator is a negotiator between author and audience. But when the negotiations get tough, as they often do with Dogen, I guess I’d rather let the reader wrestle with the difficulties of his medieval Japanese diction and syntax than make her read my own ideas in easy English paraphrase…
But wait. You are going to let the reader “wrestle with the difficulties of the medieval Japanese diction and syntax”? I’m confused then. If you’ve translated something at all, it’s in English now, so you’ve already made some large percentage of the decisions, right or wrong, about how to parse and interpret the Japanese and turn it into English. You can’t avoid responsibility by leaving clumsy structural remnants of the Japanese lurking in the English syntax and vocabulary and let the user sort them out for you.
Let’s say I am arranging a Bach cello suite for the piano. I could simply transcribe each note in the original into the corresponding note on the piano. But why would I bother to do that? I want to make this into a piano piece—with the harmonies appropriate to that instrument. Yes, that requires me to have an understanding, or perhaps even to guess, about what harmonies were in “Bach’s head” when he wrote the piece. But presumably I’m doing the piano arrangement because I am a musician and student of Bach and can make good judgments about such things that will make my piano arrangement pleasing and in some sense causative of analogous emotions to those someone feels when listening to the original cello piece.
Frankly, I think “literal translation” is an oxymoron. It is translating into a language that does not exist.