Is this mass copying of a single file for mass distribution? Even if but a few sentences being displayed per query (each being a different user response from the single file copy), I think this is a reasonable question to the model. It appears the digitized copy is sitting (exclusively) on Google servers. I think this has relevance. In effect, Google has acquired the copies, acting as a content aggregator for what is intended by design to be widespread distribution of each single file copy. I think it is reasonable to say that libraries own the rights to make a single copy (they bought the work and paid the royalty). But not sure if this right covers the inherent design of widespread distribution as a result of making a single copy. The technology enables a single copy to be accessed by what can be an infinite number of queries simultaneously. Where are the permissions emulating from for this? Don't know the answer, just asking the question :)