2008. május 15., csütörtök

The translator

Jeg skal erfare norsk, men det er ikke grei fordi jeg har ikke ressurs på norsk. Så jeg prøvde finne noe.

Ok, these two sentences were written by me, in Norwegian of course. However I've just found a translation service, so I typed this English text (I wanted to write something like this in my first two Norwegian sentences):
I want to learn Norwegian but it is not easy because I haven't got resources on Norwegian. So I tried to find some.
And then, the translator has come with this result:

Jeg vil lære norsk men det er ikke praktisk fordi jeg ikke har fått ressurser på norsk. Så forsøkte jeg å finne noen.
Well. At least some common words with my version :) (please note: of course I know an automated translation like this is not error free either!). But let's see my original try to be translated into English with that service:
I have to learn Norwegian, but it isn't ok because I do not have any resource in Norwegian. So that I examined Finn something.
Heh. Of course, a computer based translator service would not be able to translate a text exactly, because this process requires to understand the meaning of the text which would mean intelligence, in English. But even this quick try shows something for me (ok, I know I may have some problems with English as well, for example the "have any resource" or "have any resources" question, I should use "have no resource", I guess). First of all, it's quite logical, that my problem is based on the fact, that I choose words randomly when there is more words available in a dictionary (the difference between "vil" and "skal" for example). Second, I may have some grammatical problems, like using "ikke" at the right place.

However it's important to note that the behaviour of a beginner learner of language is quite similar to the behaviour of a computer software try to translate. Why? Because a beginner (like me) only knows some very basic rules on the language, no "real word" experience on using the language, and almost no clue on choosing the right word if there is more for the same meaning according to the dicitionary. A computer software do this, exactly: it uses an internal database as the dicitionary, and it has some rules about the syntax of the language.