The Wall Street Journal has a story (found via boingboing gadgets) on Amazon's new Kindle, which features a thinner design, imput tweaks, better grayscale—and a text-to-speech feature.
It's that last bit that seems to have disturbed the Author's Guild. Paul Aiken, Executive Director of the Authors Guild, was quoted by WSJ:
They don't have the right to read a book out loud…That's an audio right, which is derivative under copyright law.
Which is odd. If I were looking for infringements under every bed, and cast my eye upon the Kindle's text-to-speech, my first thought would have been to attack it for making a public performance of a literary work. After all, reading something aloud certainly counts as a performance—literary works are explicitly mentioned by the Copyright Act as works that must be performed with the authorization of the author or the law.
But then, you run into the problem of the fact that, to be an infringement, such performances have to be public. Watching a recorded broadcast of Battlestar Galactica at home isn't a public performance—no, not even after I've invited a bunch of friends over for a viewing party.
Even if a court decided that the performance was covered under section 106(4); it's almost certainly fall under a limitation or exemption of that right—if not one of the specific section 110 performance limitations (like 110(4)), then likely under fair use.
So ok, if you're desperately searching to find a way in which text-to-speech is infringing, you could turn to derivative works—copyright holders have the exclusive right to make, or let others make, adaptations of their copyrighted works. So while a movie based on a book would not be a reproduction of that book, it could still infringe the author's copyright, as it's derived from that original work. Other examples of derivative works would include an abridgement, language translation, or a satire.
So is reading a book aloud creating a derivative work?
Forget for a moment whether or not this reading is being done by a machine or a human—the law doesn't care so much about the technology as what happens and how—am I creating a derivative work if I pick up a copy of Spook Country and start reading aloud?
Common sense would say no. Reading aloud is something that's one of the things that is done with a text; it's expected that people will, and should, read books aloud. Furthermore, doing so doesn't make something new—the words are the same, the language is the same—there's only one work at issue, and that's the original.
And in this instance, the law and common sense are in perfect accord. Under the Copyright Act, it's not a derivative work because it's not a “work.” For there to be a work at issue, something has to have been fixed in a tangible medium. The noises coming out of my mouth aren't a work; I'm performing an existing work.
It's no different with a book run through a text-to-speech program. The sounds emanating from the speakers are no more permanent or fixed than the sounds I emit. And you're not going to get much traction going after the processing and RAM buffers in the device either—otherwise, every single device capable of playing recorded audio is an infringement machine—even the simplest of CD players today stores buffer copies of whatever it's playing as an anti-skip feature. Text-to-speech is no different. In fact, it's nothing new. Pretty much every Mac comes with a program that can do this.
So let's dispose of this idea of it being an infringement.
Leaving aside the copyright considerations altogether, it's maybe plausible that publishers are concerned that this feature will eat into the demand for audiobooks. To test that theory, here's a little experiment. Grab the digital text of a book from somewhere—there's a great selection of public-domain books at Project Gutenberg and the Internet Archive, for starters—and plug that into your computer's text-to-speech program. Try listening to that on your commute home today. If you think this is a substitute for an audiobook, you have a better tolerance for continued aural assault than I do.
The text-to-speech function could provide a convenience in certain situations, but it's certainly no replacement for a commercial audiobook (which themselves are pale substitutes for a properly digitized text for the visually impaired). In this case, the Author's Guild seems to be looking a gift horse in the mouth. Amazon, a massive buyer of books, and now also a massive licensor of copyrights for digital books, has created a small, but possibly growing revenue stream for authors. Yet the Guild seems ready to start waving around half-baked notions of copyright law in an attempt to extract more money from any additional bit of technology that touches a written work.