28 October 2003

Amazon and Retrograde Motion
There have been a lot of electrons (and even some ink) wasted on facile considerations of the new Amazon "search inside the book[s]" feature. It is a "feature" that is so ill-conceived and bug-ridden that it should never have been released—and it presents major copyright problems for a large class of authors who have largely been silent thus far.

   For those just returning from a short fall break on Alpha Centauri who don't know what I'm talking about, Amazon now allows one to search for text inside a book. What pops up in return is a purported "image" of the page on which that text occurs, along with "forward" and "back" arrows that allow one to go a few pages in each direction. Supposedly, one cannot print these image pages; defeating that "protection" is so trivial that it is actually available by changing a default setting in IE6, if one knows what to look for, and can be done manually in any browser.

   Well, that doesn't sound so bad, does it? Get a few pages out of a book, see if it has the information one needs, then go out and buy the book? I'm afraid not. This feature was implemented without any real thought by the programmers or the legal staff at Amazon, for a very simple reason:

Not all books are single, complete manuscripts.

   It is trivial to retrieve individual stories, essays, and poems in their entirety from collections and anthologies. Particularly as many of these works will be less than the apparent nine-page limit on retrievals embedded in the engine (that is easily circumvented by reinitialization of the search, and I'm not going to describe how—it would take longer to describe it than it would for a reasonably perspicacious third-grader to figure it out), that results in a copyright violation. And the release of an easily-convertible-to-plain-text version of that independent work into the wild. Without the author's authorization or even knowledge. <SARCASM> Except for my clients, of course. </SARCASM>

   As Professor Bainbridge points out, this is also a significant concern for authors of reference works. In his posting, Professor Bainbridge indicates that at least one publisher—the University of California Press—has given at least cursory thought to the issue by excluding "some" reference books and books of poetry. But that's not good enough. Using a 56k dialup connection, it took me less than six minutes to get a "free" copy of a 9,000-word article in investigating one of my clients' collections of academic works. It's an important, indeed seminal, work in that field; and it took less than five minutes thereafter to run the result through OCR software and get a compact, editable version that could easily have been posted on the Internet through any of the various pirate sources. Needless to say, my client was not very pleased.

   On top of all of the above, consider the problems caused for smaller publishers. Based on conversations with people at Amazon in the course of trying to get some of the above resolved, Amazon approached the place from which they get books for their "permission" (realistically, it was probably pretty close to the unstated blackmail of "your competitors have already confessed, so why don't you make it easy on yourself…" used by experienced interrogators) to post the material. There is just one tiny problem: that source, even though it may be a publisher, often is not the publisher with the right to make that decision. A substantial number of smaller publishers, particularly those specializing in reprint collections and anothologies of both fiction and nonfiction, do not act as their own distributors. Instead, they contract out to another publisher to act as the distributor. (If you have any idea what it takes to run a book warehouse, you understand why.) On the initial evidence, Amazon did not contact actual publishers if there was a distribution agreement. Distributors do not have the authority under standard publishing contracts, or under the specific contracts that I have reviewed, to make this decision for either the actual publisher or the author.

   Some of this accounts for the rather tepid response of the Authors' Guild to the situation. This is completely understandable, if a bit short-sighted: the Authors' Guild (yes, I have corrected the punctuation in the organization's name; it's a pet peeve of mine, because one doesn't create a collective organization from a single author) consists almost entirely of authors of book-length manuscripts, even though the eligibility requirements indicate that freelance writers of shorter works are also eligible. Thus, the problem with "extraction" is just not on the Authors' Guild's radar screen, because "extraction" of significant portions of a larger work is not very easy.

   In technical terms, some of my clients and I are really pissed off by the highhandedness with which Amazon's new "feature" has been implemented. Even if the publishers did have the rights in question, Amazon didn't even provide the common courtesy of advance notice to individual authors to allow them to opt out before any damage might have been done. Neither did the publishers. And the arrogant assertions that this cannot result in a copyright violation clearly result from keeping various heads six feet under some very dense beach sand, waiting for the tide to come in.

   Amazon has already been notified with the polite version of The Letter. I expect things will be much less polite by the end of the week if there is no significant and verifiable action. Dammit, this should have been done on an opt-in basis, not an opt-out basis. The publishing contracts and copyright law demand nothing less.