Human vs. machine translation
Machine translators for automatic (online) translation, i.e. IT tools for translating websites etc. – are they really in contradiction to the purpose of the “human” translator? ATT INTERWERS presents its opinion concerning machine translation.
What was translated before using a pen and paper, nowadays can in some cases be translated by a machine within just a few seconds, at any time without any work of a human being. The only condition for using such online tools of today is the necessity of creating dictionaries and developing proper translation algorithms which use the vocabulary resources according to predetermined linguistic rules. Currently the creators of online dictionaries, search engines and the scientific institutions, linguists, translators and some translation agencies collaborating with them operate within the so-called field of “localisation”. Their work aims at using the existing dictionaries to write algorithms allowing to… exactly!
At this point we need to make a stop and clarify that the dreams of substituting a real translator with dictionaries and algorithms will never come true! At least when we are talking about a high quality translation which will convey the message of the source text. Of course, the bad quality translation which we can get at an instant is a whole new story. Those are useful in some cases, i.e. you can translate or ”translate” a text in a foreign language anywhere at any time if you have a computer and an internet connection. Why “translate”? Because you would not really call it a translation but just some babble of which you can more or less make out the general idea of the author. Of course it is convenient and easy to do if you only need to understand the key points of a website etc. (which sometimes is not even the case, as the translation is so bad that it’s hard to make out anything of it). However, the benefits of automatic online translation are limited to translating websites using simple (most likely marketing) language.
There is one basic reason from the linguistic point of view: a single language (not to mention two languages) is a very broad communication system at the same time being an open system, i.e. connected to other code systems such as the cultural code where particular terms and phrases only operate in a given culture, time or language etc. These code systems are sometimes closed but in many cases they are open and thus, variable. It is not possible to cover them with just one algorithm because they’re indefinite.