[Legales] Re: [FSFLA-Traductores] http://fsfeurope.org/projects/gplv3/brussels-rms-transcript

Alexandre Oliva lxoliva en fsfla.org
Vie Mayo 4 17:44:41 UTC 2007

Hi, Pesserl,

Thanks for your feedback!

On May  2, 2007, Alexandre Pesserl <alexandre en pesserl.com.br> wrote:

> Translations are also protected as original works without prejudice to
> the copyright in the original work (Berne Convention, Art. 2 (3)) -
> meaning that you need the permission of both author and translator to
> use the resulting work.

Even fully automated translations that don't involve any creative
process whatsoever, which AFAIK is requirement for something to be
subject to copyright?

The issue is not so much what the requirements are for us to publish
translations (we already had a bit of a clue about that), but rather
on whether an automated translation is subject to copyright, such that
we'd need express permission from the copyright holder of the
translation.  And, in case it's fully automated, who could give us
that permission, the juridic person who created the translation
software, or the person who submitted the text for translation?

>> My point is that your work is based on the output of the tool, so it
>> is possible that you're subject to its copyright terms.

> If Babelfish doesn´t state any limitations to the use of the results
> produced by their tool, I can´t see how they could claim any
> copyright at all.

Conversely, if Babelfish doesn't explicitly grant a copyright license
to the result of its copyrightable work, I can't see how anyone else
could claim any rights to publish it.  AFAIK copyright licenses are to
be interpreted restrictively, and, in the absence of a license, you
can't do anything that's not fair use.

> Copyright protection can only fall upon production in the literary,
> scientific and artistic domain. When you use a computer program to
> translate a text, you are actually just using a tool to help you in
> your work, but you - a real person - is the one actually acting out
> the creative process.

Thanks, this matches my understanding of the situation.

However, just like a program preprocessor can copy portions of its own
copyrighted text into its output, an automated translator could claim
to do something similar, and then some uses of its output might
require a license from the author of the automated translator.

But, unlike the case of programs, natural language translations would
probably never include portions that are big enough to be
copyrightable, or to be regarded as creative in themselves, so now it
seems to me that it is indeed safe for us to use the output of
automated translators.

Has anyone come to a different conclusion?

> This discussion becomes more interesting if you throw in the notion of
> artificial inteligence, and if an original work created by an IA
> machine could be subject to copyright, and who would the copyright
> holder be... But then again, I´m no Isaac Asimov and this is hardly
> the appropriate forum :D

And if you take that further, and believe what some scientists who're
stating that free will is an illusion, because our brains are as much
of a deterministic machine as any IA device could be, then one might
come to the conclusion that copyright doesn't really make sense
because creativity is just an illusion as well.

> (b) express permission from the translator,

That's the thing, we don't, if you mean the automated translator, not
the person who takes the automated translator's unusable output and
turns that into something that actually makes sense ;-)

Thanks again,

Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Más información sobre la lista de distribución Legales