As I said in my last post, the Google Books settlement has some fairly serious copyright implications. I was going to write up a lengthy discussion of them, but in researching them I discovered that Prof. Lawrence Lessig has an excellent article on the subject up at The New Republic. His argument is too nuanced to capture in a pull-quote, so go read it. It's well-written and not terribly long. I'll wait.
In the settlement, Google really is trying to adjudicate significant changes to copyright law, notably a registry of works and a policy on orphaned works. I've heard a number of authors, in talking about the settlement, say, "It would have been fine if Google asked me before scanning my books, but since Google didn't ask me, they can't have them." The problem is that it's impossible for Google to individually ask every author or rightsholder for every work under copyright for permission to scan their books. Since creative expression is automatically under copyright once it is created, and the United States has no central registry of copyrighted works, every book published after Steamboat Willie is presumptively under copyright to someone. That means that, for Google to ask permission before scanning, they would, for each work, need to go through a lot of trouble and expense to locate the current rightsholder, and if that person can't be found, Google would be barred from scanning the book. It doesn't matter if the book is rare and falling apart — if it's copyrighted, they can't scan it. (And the legal tangle around who got what rights when an author died, and what they did with them, is almost always a serious mess, or there wouldn't be a project up to make sure that authors have wills.)
To make it even worse, if an author has, say, quoted some song lyric in her book — and, as a good- and copyright-observant author, went and obtained the rights to use that lyric from the company which owns it — now, in this proposed world order, Google has two problems — first to track down the author of the book, and second to track down whoever now owns the rights to the song (which may have changed hands since the author did their due diligence) and clear that separately. If the author has quoted multiple songs owned by different companies, Google has to find the rightsholder for each song, no matter how small the quotation. And even if the author allowed Google to scan their book, if even one of the song companies says no, Google is faced with the choice of using the book without the offending quotation or not using the book at all. If it removes the quotation and that happened to be a major plot point? Oops, sorry. Now repeat as necessary if that lyric quotes part of another song, and so on ad infinitum, and combine with ever-increasing copyright terms (pushing a century these days), for a scary picture of what the future could look like.
Basically, if Google had asked permission first, they'd never have scanned anything. Some people would say that would be right — Google shouldn't have scanned anything. I'm too much of an information archivist and a librarian to agree with them.
Lessig is right that, considering the sad state of US copyright law, the settlement is really quite a good patch on it. He's also right that it's another worrying step in the trend towards the world outlined above, where copyright goes recursive and strangle culture abed. Then again, if the settlement falls through or gets amended as the complainants want, it will also be a step towards that world, and perhaps an even bigger one. Everybody is right that the US needs serious copyright reform, preferably spearheaded by Congress, sooner rather than later. Whether that will actually result in a better system — or any system at all — is unclear, given that body's current state of gridlock, but the current system is clearly broken, and Lessig has some good ideas on how to fix it.
 The Open Book Alliance's blog post on it, which brought it to my attention, misses the heart of it, I think. Their goals, in my understanding, are exactly what Prof. Lessig is worried about.