Take a look, it’s in a book

This week, Amazon announced the details of its second generation Kindle e-book reader. One of the new features is text-to-speech software that can read aloud the text of a document stored on the Kindle.
The Wall Street Journal reports that the Authors Guild is not happy with the feature, “They don’t have the right to read a book out loud,” said Paul Aiken, executive director of the Authors Guild. “That’s an audio right, which is derivative under copyright law.” New Kindle Audio Feature Causes a Stir
Typically, audio books are derivative works of the original work and are fixed in a recorded medium. The reader adds his or her own interpretation to the text. The work can stand alone as an artistic creation. (See e.g. the accolades that Jim Dale has received for his readings of the Harry Potter books). However, a Kindle 2 owner could foreseeably forego buying the more expensive audiobook to choose to read aloud the e-book.
But is the Kindle text-to-speech reading a derivative work? With the Kindle 2, the computer is generating a reading of the original work dynamically for the portion of that work the Kindle user chooses to have read aloud. Is it ever considered fixed in a tangible medium? If the text-to-speech reading isn’t fixed, then it can not be a derivative work, since a work must be fixed in a tangible medium in order to be copyrightable.
If publishers worry about text-to-speech affecting the market for audiobook rights, perhaps e-book rights (and thus e-books themselves) will become more expensive.
Engadget’s Nilay Patel analyzes, Know Your Rights: Does the Kindle 2’s text-to-speech infringe authors’ copyrights?: “This is actually pretty tough stuff — as far as edge cases go, this one pushes right up against the boundaries of the current law. On one hand, you definitely have the right to read books that you own out loud using whatever tools you want, and on the other, authors definitely have the right to prevent others from selling audio versions of their works. The Kindle’s text-to-speech feature blurs the lines between books and recordings, and that means those two rights are in conflict with each other.”
See also John Siracusa’s take on the past, present and future of the e-book market, The once and future e-book: on reading in the digital age – Ars Technica: “A veteran of a former turning of the e-book wheel looks at the past, present, and future of reading books on things that are not books.”
Tangentially related, Apple’s text-to-speech software, MacInTalk, has a major film credit in its resume. In Pixar’s Wall-E, MacInTalk voiced the character Otto.
Update (2/12). Neil Gaiman weighs in with a Quick argument summary, “When you buy a book, you’re also buying the right to read it aloud, have it read to you by anyone, read it to your children on long car trips, record yourself reading it and send that to your girlfriend etc. This is the same kind of thing, only without the ability to do the voices properly, and no-one’s going to confuse it with an audiobook. And that any authors’ societies or publishers who are thinking of spending money on fighting a fundamentally pointless legal case would be much better off taking that money and advertising and promoting what audio books are and what’s good about them with it.”
Evan Brown, Does the Kindle 2’s text-to-speech feature violate copyright law? “Does Aiken have a legitimate gripe? I say it depends on the technology. And the fact that there could be a difference based merely on a technological setup underscores how digital technology has sent some aspects of copyright fumbling towards absurdity.”