I surf the Net like anybody else, and sometimes I find unexpectedly something I was not looking for. Not a big surprise here, that happens also to everybody. However, in this particular case, an unexpected article in the Zimbabwe Independent actually made me think about the copyright value of machine translations.

Zimbabwe Independent

Now, this newspaper is not something I read usually – when I stumbled on it it was the first time I knew it even existed. But it has an interesting article about “Originality concept in copyright protection” from Richard Pasipanodya that drew my attention. Basically, the author claims that only works that are original can be subject to copyright protection. And that is absolutely true, not only in Zimbabwe, but also in the rest of the world. All laws about copyright -anywhere in the world-  require the concept of originality to claim copyright.

Specially interesting was that Mr. Pasipanodya describes the different types of works subject to copyright protection, and one of them is obviously the translation of an existing work (copyrighted or not). Literally, his article states:

“Translation of works
This entails the verbatim copying of a copyrightable work into another language. Translation of works is most rife with literary works.

The resultant work is said to be possessive of originality where the translation uses much skill and judgement as that of the author of the source, work though of a different type this time around. It does not matter really whether the resultant work is of a mediocre version of the source; work for the merit of the work is irrelevant in adjudging a work being eligible for copyright.

For a translation to be protected as an original work it must not be constituted as a verbatim translation of the source works lock, stock and barrel. Rather, it is the originality embodied in the resultant work such as the addition of subtleties and other creative ingredients into the translated work.

Again, though this is apparently an infringement act, the translator is accorded copyright protection on the same footing as that of the author of the source work, owing to the mental effort they would have applied and invested in it.”

I could not agree more. For a translation to fall under copyright protection, a creative effort must be made by the translator, not just copy word for word. This, curiously, also means that a non-copyrighted work (say, an ancient text) that is translated by somebody would be subject to copyright protection, which is something that happens all the time. Why? Because this translation of the ancient text requires creativity on the side of the translator, and therefore will be protected by copyright laws, even if the original text (which might be even centuries old) is no longer protected by such laws.

And now the big question: If machine translation is performed, would the ensuing translation be protected by copyright laws? The obvious answer is no, because no “creative effort” has been applied. No judge in the world will acknowledge that pasting a text in Google Translator or any other translation software is “creative”. It could be perhaps argued that the “creativity” is on the side of the machine translation (MT) software, but in that case it could be also argued that the copyright holder would be the MT software developer, though I doubt that this could be upheld in court.

So, summarizing, since there is no creativity involved, machine translation results would not fall under copyright protection. Which is bad news for those that want to save the money of paying a human translator.

Be Sociable, Share!