Serraphin: I'm aware that Google would need to keep private copies of the books on their servers. But they already keep a private copy of the web, and nobody seems to think that's wrong. I'm happy to let them copy my (copyrighted, non-CC) websites.

But I'm also noticing some US/UK confusion in this thread. Apparently, the UK has extremely restrictive laws about fair use. For example, UK writers receive royalties from public libraries, a policy which would be laughed out of court on our side of the pond. And under US rules for backups, you presumably could make a private photocopy of a book.

Fundamentally, though, this is an argument about defaults: Should the Google Print's indexing be opt-in, or opt-out?

I've volunteered for the Gutenberg project before, so I have a good idea what it takes to get copyright clearance on older works. If you're dealing with more than a few books, it's basically impossible.

My argument: Claiming that, "Google can only index books with the author's permission," is equivalent to saying, "We shouldn't have a full-text searchable archive of anything besides pre-1920s books and the current front-list." And in the eyes this bibliophile, losing the ability to search obscure fiction from 1954 (say) would be a terrible crime.

Under US law, the solution is simple: Simply decide that indexing books and quoting 30 words is covered under fair use, and call it done. To appease those authors who can't _stand_ the thought that somebody, somewhere might read 30 words without paying, offer a way to opt-out. This approach worked fine on the web.

For me, the argument against building Google Print (at least the "30 words" version) is a time-reversed version of the argument for burning the Library of Alexandria. And if I recall correctly, the Library of Alexandria was built by impounding books at customs, making private copies, and giving them back. :-)

Once again, I agree that Google needs to offer an opt-out, and deal directly with the actual copyright holder.