“On one hand, Google has scanned an impressive thirty million volumes, putting it in a league with the world’s larger libraries (the library of Congress has around thirty-seven million books). That is a serious accomplishment. But while the corpus is impressive, most of it remains inaccessible.”