Tuesday, 11 November 2003
Amazon.com cuts back on copyright infringement -- step by step
As of today, my books seem to have been removed (at my publisher's request -- Amazon.com still hasn't made any effort to obtain permission from authors) from Amazon.com's co-called Search Inside the Book program(s). Other books whose publishers have requested their removal seem to be working their way through the same process.
I can't tell if the removal is really complete, or if Amazon.com has retained copies of its "library" of page images and OCR-generated text.
But the 3-stage process by which my books have been removed, and others are being removed, has itself made clear 2 important things about the scheme(s):
First, the fact that it took so long (and wasn't done all at once, once Amazon.com decided to do it) strongly implies that Amazon.com hadn't considered the copyright implications of its plans (as others have also inferred ), hadn't planned to offer even an "opt-out" option, and wasn't technically prepared to do so. In light of this, Amazon.com's claimed legal justifications need to be interpreted as ex post facto attempts to avoid liability, not as statements of their actual thinking or beliefs.
Perhaps more importantly, the step-by-step "removal" process exposed the independence of what are really three separate programs, despite their bundling by Amazon.com into a package it has somewhat misleadingly labeled "Search". By removing them one at a time, Amazon.com has demonstrated that they don't need to be linked and. most important of all, that they need to be independently justified.
- Search to identify books containing specific search text.
- Display of searched-for text in brief (2-4 line) context from the full text of the book.
- Display of full-page images.
In the case of my books, these were removed in inverse order.
First, as I reported earlier , the page image display was suppressed (with an "error") message, while the search results were still shown in context.
Then the search results were removed, but by searching for a text string appearing in my book, and in no other book in the Amazon.com program, I discovered that the book(s) containing that string would still be displayed (as cover thumbnails) in response to the search. This was perhaps the most interesting and promising stage, since it showed that the ability to search for and identify books containing specific text can be offered without needing to display either text excerpts or page images. "Search" and "electronic content delivery" are already, with Amazon.com's current technology, completely separable.
Finally, Amazon.com stopped identifying the books at all in response to searches.
Amazon.com listings for some books not in the "Search Inside the Book" and text and page-display programs now include a "Why can't I search inside this book?" link, which pops up the following message:
Our Search Inside the Book feature includes only books for which we have the publisher's permission to display copyrighted material.
That's a significant admission that the co-called "search" program actually involves the "display" of copyrighted material, of a sort that requires permission from someone other than Amazon.com (i.e that it wouldn't be permitted as "fair use").
Continuing the discussion, Steven Kaye's Thousand-faced Moon blog has had a thoughtful series of articles on this, including lots of good links.
Copyright attorney Neil Isenberg of Gigalaw.com has a column on Cnet's News.com, Steal this book online in which he concludes that Amazon.com's conduct moght be legal, based on the (erroneous, in many cases) assumption that "authors may have granted to their publishers the right to participate in a program such as this." Isenberg also suggests that page-image display, if sufficently limited, might constitute "fair use". But his argument on this latter point is clearly flawed, since he bases it on the impact of digital giveaways on print sales, rather than the impact of e-book and e-excerpt giveaways on potential e-book and e-excerpt sales .Link | Posted by Edward on Tuesday, 11 November 2003, 20:38 ( 8:38 PM)