A huge problem in working out any reasonable compromises on the Google Print program is that there are millions of works out there that are under copyright as a matter of law, but orphaned as a matter of reality.

Perhaps finer distinctions need to be made than have heretofor been propounded. Perhaps making a digital copy anything published within the past decade without permission is a copyright violation, whereas copying anything earlier is fair use.

Virtually everything formally published after 1923 is still under copyright. And, since the last great reform in 1976, coyright covers almost anything reduced to fixed form (your casual emails are under copyright). Most of these works have long since been forgotten by everyone, including their creators. For older works, the creators are long gone, and formal title is fragmented among numerous and often unknown heirs.

The idea that Google should get permission before including a copyrighted work in its database has a nice ring, but it is a practical impossibility.

Various solutions have been proposed to this dilemma -- re-registration, and databases, etc. But none are real, or on the near horizon. So whatever one thinks of the program -- and we at PFF seem to be splitting in about six different directions -- one should be clear on one point. To insist that Google get permission means that the post-1923 literature cannot be included. And, given the de facto orphan status of so many works, most of it could never be included. It would just sit there moldering.

Now, one can argue, "So what?" If it is not worth the transaction costs, then the value is obviously minimal. Well, maybe, but there are two rejoinders:

(1) Who knows? This argument is circular, because a large part of the problem is that no one knows what is there, or whether it is worthwhile.

(2) This is a long-tail problem. Each individual work may be worth little, but collectively they are worth a lot.

AMENDMENT (10-21-05; 08:30 a.m.): I neglected an important part of Patrick's blog on this issue, where he noted:

The AAP wanted Google to use ISBN numbers -- in use since 1967 -- to identify books under copyright and request permission to scan them. BW reports that AAP would have had a more relaxed perspective for books that were out of print and lacking ISBN numbers.

So the AAP was amenable to the concept of the time-based differential. There may be promise here.