I found Steiner's take on translation and fidelity to be quite interesting because he doesn't see it as an issue of being faithful to the original text. Rather, he accepts that it's impossible to replicate a text at all and sees it as an issue of equality- whether you can add to the new text as much as you can lose when translating. I actually liked this approach to the fidelity argument the best of the ones we have read so far because instead of trying to choose between two approaches which are both unsatisfactory, this approach accepts the futile efforts of trying to replicate a text and instead focuses on what can be done by the translator. However, Steiner's description of the steps, particularly aggression and incorporation, were quite vivid and seemed rather forceful and violent than what I imagined translation to be, kind of describing it like the forceful removal and assimilation into a new language.
Bellos' article was also quite interesting. I have seen my fair share of comically bad translation, a mix of both human and machine translated. Though I knew that Google Translated used probability to determine which specific words or patterns are generated, I had no idea it was generating translations based off information from the Internet, though it does make sense. I agree with Bellos in that for routine or quite common patterns this works great but for more obscure grammar patterns, less common readings of kanji, or rarely used phrasing, Google Translator fails. Because of this, Google Translate is notorious for weird grammar patterns and I know that some TV shows, when they want a character to sound crazy or speak gibberish-like English, they will put it into Google Translate it and convert it to a bunch of languages before returning it to English.
No comments:
Post a Comment