Bellos spoke about machine translation more specifically about Google Translate. He speaks about how it translates, which was actually kind of interesting to find out, and why it could never replace real translators. At first glance, that is very true. If you have ever used Google Translate you can see that the sentences are a bit off even if you have no understanding of the language you are translating. Furthermore, I do agree that any machine learning will indeed have trouble translating into more creative pieces of work and literature. But that's the thing though. It feels like there is a limited number of ways to write things that eventually at some point machine translation will eventually be able to translate sufficiently. Even then, there could be a development in the future using some different approach that could replace translators. The passage by Steiner itself was a bit of a translation project, but if I understood correctly, he spoke about the approach a translator should take, so basically the typical motion. The first step is putting faith into what you are translating, meaning that there is a purpose to what you are translating, and not just some nonsensical nonsense. Next would be figuring out what the actual meaning in what you are translating actually is and extracting all of it. Then he goes on to speak about creating some sort of balance in the translation, and compares it to that of importing and exporting. It is difficult to balance what you change and how faithful you are to the original. Lastly, he speaks of ways in which to maintain this balance. At the moment, there is not really anything that I disagree with so I think this is a decent approach, although it is possible later on I may see some flaws.
No comments:
Post a Comment