Here we go again. Our attempt to bring "language" to machines. And yet again, we have failed. Indeed, it is one of those experiments where practice, work, and more practice is going to be beneficial, but, as of right now, nothing is happening. Sentences are going awry, and meanings are completely eschewed. The truth is without presenting a context for words, it is almost impossible to translate anything. Even for a human being, a statement not heard in its original context could mean a vastly different thing. And once we start to misunderstand each other, then the importance of language dwindles. For language's primary function is to communicate. It is the best communicate we have at our disposal. But, I must admit, if google or some other company, were to come up with a precise translator, of about 6 or more languages, then I am all for it. Now, I can get a passage from Greek class, scan it unto my computer, and simply click translate. You see, it is a win win situation. I do not have to spend long hours, hours which I could be sleeping or more importantly watch soccer games, translating Greek. Is it bad? No. And it is simple, I see this as the new version of the calculator. Why do math by long hand, when the calculator is there??
http://blogs.zdnet.com/Google/?p=778
http://blog.searchenginewatch.com/blog/071023-115310
Subscribe to:
Post Comments (Atom)
2 comments:
I find it really interesting that though we often learn a new language by translation, speaking a language fluently does not make us good translators. We learn vocabs by understanding their meanings, and yet we can't always reverse the process and state the word when given the meaning. Why do you think this is so?
learning new vocabs by understanding what the word means in our language
Interesting perspective... it would be great to get a little bit more context and history about the topic of machine translation that you are discussing here! How do the past approaches to machine translation reflect past theories of the nature of language and meaning? What are some new ideas for the future in this field?
Post a Comment