First of all, I don't think the question of damages is of particular importance right now, either to Google or to the Authors Guild. This is a huge case for everyone who has any interest in copyright law, but it's not particularly riveting in terms of potential damages. Not a whole lot of copyrighted stuff has been scanned so far -- but there is at least enough to give the Authors Guild legal standing.

Secondly, Google has indemnified the University of Michigan for any litigation against them, for the act of providing the books to Google. If Google escapes damages, then the University might not. Look at Section 108 on library and archive copying. The library can make no more than one copy, and only if there is no direct or indirect commercial interest behind that copy.

If a court decides that the University, by virtue of the fact that they own the books and have complete control over what they provide to Google, is the guilty party here, then what is the impact on this case? Actually, this is more than a remote possibility, but for reasons than no one has yet mentioned.

If a judge looks at Section 108, together with the contract that U-M has with Google, it provides a convenient way to decide the case without wading into the very deep waters of "fair use." And yes, U-M is mentioned in the Authors Guild filing. Even if Section 108 isn't specifically mentioned, it is still a part of copyright law, and will probably be considered by some court in this case, sooner or later. There is no case law on Section 108 so far, and that's what has some library law experts nervous. They've been taking liberties with electronic copies for years, but so far not to the extent that it has ended up in court.

This solution might prove attractive to the court. All it has to do is rule that U-M should have known that their contract with Google is unlawful on its face, and therefore Google's copies of copyrighted material were obtained illegally and must be destroyed. End of story. At that point the Authors Guild, if they cared about damages at all, could ask for damages from U-M. Google would, by contract, have to cover these damages. We're back where we started from.

Would this mean that Google has won? No, it means that Google has to find libraries outside the U.S. if they want to grab millions of copyrighted, out-of-print books for scanning purposes. Google can skip Europe, which is already nervous about this. Perhaps Canada? I don't think even Canadian libraries would invite Google in at that point. Google has estimated that it will end up with only 15 percent of the copyrighted material it wants to scan if it has to replace their present opt-out system with an opt-in system.

In D-Lib Magazine, September 2005, there is an article that looks at the five libraries in the Google Library Project and cross-checks their combined holdings with the OCLC's WorldCat bibliographic database of 10.5 million unique books. Only 33 percent of these 10.5 million unique books are held by at least one of the five Google libraries. In other words, Google is a long way from indexing "all of the world's information" even if they win this case. (When U-M brags about 7 million items that they hope Google will scan, U-M's figure is less believable than the WorldCat numbers. WorldCat numbers are based on unique books, and apparently Google is only doing books.)

If Google doesn't win this case, they'll be stuck with public domain material only. Harvard has estimated that about one-quarter to one-third of its collection is not copyrighted, meaning that Google would be cut off from two-thirds of the entire library except for the 15 percent of those copyrighted books that are expected to opt-in. Let me be blunt: If Google has to give up their opt-out system due to a court decision, the entire scanning project is hardly worth the effort. You and I may find value in reading the Great Books from the last century, but that's not going to drive much advertising toward Google's bottom line. No, Google Library Project will probably die if Google cannot use opt-out.

The opt-in is a headache for Google. Google would prefer to deal only with publishers, not individual authors. However, the rights holder for most of the out-of-print, copyrighted material are the authors themselves. There are two reasons for this. One is that many book contracts specify that the rights revert to the author once the book is out of print. The other is that since the Tasini v. NYT case, the only way a publisher can claim electronic rights is if the contract specifically assigned the electronic rights to the publisher. Since Google considers everything since the mid-1920s to be under copyright (except for government publications), how many authors will this involve as opposed to publishers?

What Google should do is stand down on the copyrighted material and settle with the Authors Guild. Then they should join with the Internet Archive and lobby the Library of Congress and various legislators to write into the copyright law some special provisions for orphaned works. This will take years, and a lot of discussions and lots of committees, simply to define "orphaned" in terms that will work in the real world.

But I don't suppose Google will do the reasonable thing. They'll have to get slapped down by a court first, and then they'll kick and scream and pout about how the courts are sabotaging humankind by keeping information away from the people, and they'll feed more grist to the cyber lawyers on the blogging circuit, and then they'll quietly and slowly abandon the Library Project.

I don't expect they'll try to get the law changed, because that would be the socially-responsible approach. The Googleplex is full of geeks, not philosopher-citizens.

Very interesting, thanks. But I disagree on a few points.

Even a minimal amount of statutory damages per book could add up to substantial cost. But an argument for zero damages is significant.

The Googleplex is geek heaven, but they can hire lawyers - or get some to volunteer! (you know the Usual Suspects ...)

I'm not sure you're right about the magnitude of the effect of Tasini - I think a publisher's standard contract is different than the contract in that case.

But you do sketch intruguing scenarios.