I had rather hoped to stay away from the controversy being generated by the new Kindle 2 Book Reader from Amazon and its “text to speech” feature that will allow the reader to offer a computer-generated audio reading of e-books, but there are copyright issues here too good to ignore. It is hard to make sense of the claims being made in this kerfuffle, but it may be worth the effort in order to clarify what copyright does and does not protect.
In a widely-ridiculed public pronouncement, a spokesman for the Author’s Guild has denounced the audio feature of Kindle as an infringement of copyright, even though the e-books sold by Amazon are, of course, licensed from the publisher. He is quoted as saying the “they,” meaning consumers, “don’t have the right to read a book out loud… That’s an audio right, which is derivative under copyright law.” This led many to trot out a parade of horrible consequences, suggesting that parents might be sued by the Author’s Guild for reading”Goodnight, Moon” to their children. So the President of the Author Guild took to the New York Times Op-Ed page to explain that that was not their intention. Unfortunately, his piece does not really explain what the claim really is. He merely says that the Guild collects separate royalties for audio books and for e-books and that Kindle would “swindle” authors out of that double fee. From a copyright perspective, it is interesting to try and sort out what infringement, if any, is involved in this “swindle.”
One way to look at this, of course, is as a simple contract dispute, and contract provisions are probably the way to settle this. Authors and publishers can simply charge Amazon more for the e-book license to compensate for the potential decline in audio book sales when those e-books are “read” by Kindle. Other e-book platforms would pay a lower price if they do not provide a text-to-speech function, and both sides could monitor to see if audio book sales really do decline. For e-books already licensed to Amazon, the Author’s Guild could try to claim that this feature of Kindle 2 breaches the license terms, and try to demand additional money. The public spat is likely an attempt to force such renegotiation.
But it is more interesting to ask if any copyrights are being infringed. When a parent reads to a child, this is a private performance of a work that does not infringe any of the rights under copyright. It is very important to remember that the performance right in copyright is only an exclusive right to authorize or deny PUBLIC performances, defined as a performance “at a place open to the public or at any place where a substantial number of persons outside of a normal circle of family and its social acquaintances is gathered.” Based on this definition, it is unlikely that Kindle would ever infringe the public performance right under ordinary use.
This distinction of performances is undoubtedly why the Author’s Guild spokesman spoke of a derivative right. Now he was simply wrong to refer to an “audio right” which is “derivative under copyright law.” There is no separate “audio right,” there is only the public performance right discussed above and another exclusive right over the preparation of “derivative works.” So a lot turns on whether an audio reading of a text can be called a derivative work.
It is generally thought that a derivative work must itself be an original work of creative authorship. So a translation of an English text into Hindi involves new creative expression, as does the creation of a film from a novel; these are classic examples of deriviative works, and each involves the reuse of protected expression in combination with new creative authorship. So a translator or a filmmaker must get a license from the original author to create these works, in which there is subsequently two (at least) copyright interests. But an audio reading adds no creative expression, so it is hard to see how it is a derivative work. In this fascinating article, Julian Sanchez analyzes this argument very nicely, and suggests an exception — an abridgment has been held to be a derivative work, and it does not contain original expression that is added to the original. I think there are historical reasons for this, but I will let Sanchez explain the ins and outs of this debate to those who are interested.
What I want to add to this discussion is an additional argument for why an audio reading should not be considered a derivative work. There is a long standing rule of statutory interpretation that instructs court to read laws in wasy which do not make parts of the language used by legislatures irrelevant; we do not want interpretations that make whole portions of a law redundant or unnecessary, since we assume legislatures did not intend those readings. If an audio reading is interpreted as a derivative work, we would have just such a reading, because that interpretation would make the public performance right “mere surplusage.” Why would Congress include a specific right over performance, and limit that exclusive control to public performances, if ALL audio readings were derivative works and therefore subject to the authors control based on a different exclusive right. Audio readings are not derivative works because, unlike abridgments, they are subject to a different right, and we must assume that Congress intended that right (public performance) to circumscribe the control an author should have over readings of his or her work.
POSTSCRIPT — In the interval between writing this post and publishing it, the news has come out that Amazon has agreed to make changes to the Kindle. I am afraid this merely reflects on the chilling effect of a lawsuit threat; it does not change the legal analysis, which suggests that the Author’s Guild won by making a very weak claim, but making it loudly.
Kevin-
If you recur to the article, you’ll see I actually cite a House Report elaborating on the scope of the “derivative work” right, and explicitly asserting that it is meant to cover certain types of performances, though the courts (in this case, basically the Ninth Circuit) have declined to read the statute that way. My own sense is that the work/performance dichotomy will be rendered increasingly unstable as it becomes easier to distribute free-standing “works” that render what amounts to a derivative work without the need for fixation. (I offer the example of a mash-up algorithm.) If I were inclined to make the Authors Guild’s case for them (I’m not particularly, but just putting on that hat) I’d suggest that the relevant principle now is not “embodiment in concrete form” (distinguishing recorded audiobooks from “live” readings, human or mechanical) but rather market displacement. The “embodiment” criterion is a judicial rather than a statutory artifact, and there’s already a long chain of cases incorporating the idea of a “distinct market” in derivative works analysis. So the principle they could push for to distinguish between human reading and the Kindle is that the latter amounts to a market substitute for audiobooks in a way that bedtime stories don’t — especially given that Amazon is distributing both the books and the software that translates them into speech as part of an integrated package. Or so it seems to me, anyway. I’m not actually eager to see copyright take this route, but I do see the force of the argument for it.
Are you arguing that all audio renderings of a work are not derivative works, or only private audio readings? I ask because it seems to me that an audio recording of a book would absolutely contain enough creative expression to constitute a derivative work. The people who voice audio book voices are usually actors, and they make creative choices about how to read each book.
I agree with you that the Authors Guild assertion is absurd. I just wanted to clarify about whether you think a recorded audio book performed by an actor would qualify as a derivative work under copyright law. My experience working at a literary agency was that the right to sell audio recordings of books is usually referred to in shorthand: audio rights. This might be leading to some confusion about what the Authors Guild is actually saying.
I want to argue for two points, and I am grateful for being pushed to be more clear about them.
First, I do not believe that a mechanical audio rendering meets the criteria for a derivative work as a matter of law. I do not need to take a position on the question of whether a human reading can be a derivative, and am happy not to do so.
Second, I think that, as a normative matter, the derivative works right should not be allowed to give a rights holder control over any private performances, live or mechanical. To interpret it so would be to vitiate the Congressional decision to provide an exclusive right only for public performances. In that sense I would argue that the public performance right acts as a kind of boundary on the derivative works right, preventing any private performances from falling into the ambit of derivative works.