I'm not so much disagreeing with what you have said, as obviously the various courts this has already been through hold a similar position on the matter, howeve

I'm not so much disagreeing with what you have said, as obviously the various courts this has already been through hold a similar position on the matter, however, I would like to emphasise two points: 1) Google continue to possess _copies_ of books that they do not own. Hence my suggesting they purchase the books (i.e. I don't think that partnerships with libraries is enough). To offer the service of Google Books they must retain these copies, i.e. to allow you to read individual pages of the books. 2) Fair use usually entails the production of another copyrighted work that includes as a very small portion the existing copyrighted material. For example, a quote in a book. However, there are some instances such as for criticism or satire that allow a far greater amount of the original material to be used. But my main point is that a new work is created that is now distributable and is complete in and of itself, i.e. it does not rely on the maintenance of any source materials for its continued existence, e.g., a paper, a book, a video, a movie, a poster, a painting, a program, etc. The copyrighted work that Google have created is entirely their own, i.e. the source code related to the Google Books service. The copyrighted material they are using is the data used by their system. I don't think it is right to argue that each time a person accesses Google Books that the transformation from data in the database to webpage output constitutes a fair use as the content output is largely just a subset of whichever book it is. There is no extra criticism, satire, or transformation added, as you can then read the output as you would a normal book, and the entire purpose is to read the book as the author originally wrote it (albeit with some pages missing).