This is somewhat hypothetical, but it isn’t completely beyond the pale of what’s currently possible.
Let’s say a programmer wrote a program to take some input text, transform it according to rules and a second random input, and save the result to a file. He looks over the most recent batch of files, and he finds one that’s really quite funny. (Anyone who’s played with dissociated-press on Emacs knows what I’m talking about.) Legally, does he own the text, seeing as how he didn’t, strictly speaking, create it?
Secondly, if he was using other people’s works as the input, can his output files be seen as fair use?
Your entering the murky depths of copyright law and, frankly, we just don’t know because nobody has ever been tried on such a case and, IMHO, the courts would rather keep it that way.
Patent absurdities can be made on both sides of the argument, if I use a random text generator on a royalty free english dictionary and generate the lyrics to a Britney Spears song, does that fall under copyright? We just dont have the legal system in place to deal with such a question.
Something that’s approaching the edge of the question, I suppose - “Millenium hand and shrimp” is a nonsense phrase author Terry Pratchett puts into the mouth of a character, “Foul Ole Ron”. Apparently, Pratchett got this phrase by playing with a text scrambling program and picking out stuff that tickled him. He claimed to have input the menu of a Chinese takeout because it was lying on his desk, and must have also included the lyrics to “Particle Man” by “They Might Be Giants” (Pratchett is an immense fan of the band).
I suspect the phrase I’ve bolded might be important. It seems to me that if you just take the first thing a program randomly comes up with, then you hardly deserve any credit for running the program and claiming the output as your own. But songwriters randomly search lists of rhyming words (either in their minds or in rhyming dictionaries) and then pick the ones that work the best. Or they play notes and chords with some amount of randomness until they find combinations that sound good. So it probably does matter that it’s what he thinks is good, not just what some computer spit up randomly.
Hm. I figured this case would be fairly straightforward, such that I was almost ashamed to have to post it here. But I’m a programmer, not a lawyer, and I figured I had better get others’ opinions on the matter.
I think the element of individual selection is indeed important. However, we can remove that element and still have art. The works of John Cage include such pieces as 4’33" and Radio Music, which are equally random and do not include nearly as much artistic choice on the part of the composer or the musicians. (The first is four minutes, thirty-three seconds of silence, the second is `played’ with one to eight radios, each tuned to a different frequency.) I cannot imagine anyone being successfully prosecuted for violating the copyright of either work, although both have been recorded.
I guess there’s more to copyright than I thought. If the copyright laws can’t handle something like dissociated-press, it’s going to be very interesting when and if someone tries to copyright a self-modifying program and all of the possible results from its execution.
Here’s another scenario, to be answered or ignored or simply discussed on a factual basis: Let’s say someone patents a computer chip that can write its own software as it processes information. Who would own the copyright to that software? Does the patent have any bearing on this matter? (I’m ignoring, for the moment, the contentious issue of software patents.)
I remember reading last year about a computer programme that intelligently generated thousands of tunes that the programmer then copyrighted so he could then sue any unfortunate composer later on . I am not sure what happened there
It’s fairly clear that a writer retains copyright even if she uses aids in creating her work, including a thesaurus, a computerized rhyming dictionary, or a random word chooser. The writer is still making judgements (including picking the input text) to create a brand-new creative work.
It’s a more interesting question whether one could copyright huge amounts of random text/melodies, then wait for the random text to match something another person writes, and sue for copyright violation.
Might be a bigger conflict between intent and the law as written there.
There was a case where a band put a segment of silence in one of their songs, claimed it was a sample from 4’33", and got sued because they didn’t seek permission. I don’t remember the outcome.
It doesn’t matter what the source was; otherwise someone could copyright the dictionary and sue if you use any of the words without permission. As long as the final result is not recognizable as the original, it’s OK. (There could be an argument that this is a derivative work, but that seems to me to be a very shaky case.) It’s the order of the words, not the words themselves, that are copyrightable.
Not that big an issue. The random text must match words the other person writes. It’s unlikely you can get more than a few lines that match.
