traduteur tradutor translator Machine translation was already created some time ago and allows internet users to instantly translate any page in the desired language. This technology is quite interesting to understand what a page is about but why do we see at the same time an ever increasing number of translators and conference interpreters?

If you have ever tried this kind of tool to understand the main meaning of a page, a report or maybe just to say something funny in a foreign language, then, you have probably already noticed that the results can be very different. With some texts, you can have average results but with others, you will end up laughing for sure! These results depend on several factors as language pairs, complexity and above all abstraction degree.

Translating a text into Portuguese from Spanish will usually give you better results than translating from German. This happens for syntactic and semantic reasons. A text with a clear structure usually allows an almost reasonable translation and it will be possible to understand the general meaning of the text while a text with a complex structure will be much harder to translate for machine translation.

However, the main problem with translation programs is that they don’t understand abstract content. Actually, they don’t understand anything. They just translate separate words from one language to the other. During my studies, I had the opportunity to study linguistics for a while and realized that it is possible to schematize a speech and even establish rules for automated writing through human speech analysis. However, it is not possible (at least for now) for a machine to understand the sense of a text or its ideas.

Therefore, if you write in French «Thomas a fait un peu d'anglais à l'école », machine translation will give you the following result in English: «Thomas made a little English to school». We can all imagine Thomas creating some kind of englishman toy with a bowler hat but what the French text meant is that “Thomas learned a little bit of English at school”. Translation and interpretation require full understanding of the text, of its ideas and intentions and therefore, while machines aren’t capable of thinking, quality machine translation will not be possible.

Translators and interpreters will still be around for many years!