Suppose a friend constructs a big eye. He makes it large. He uses a million photoreceptors and ganglia from frog retinas and some fancy quantum circuitry based on selenium and cesium so that it is not totally unlike the human eye (except for the cesium and selenium). He gives it to me, since I am a scholar of the works of James Joyce (which happen to be still under copyright -- see footnote).
I use this amazing eye to read Joyce's complete oeuvre. But it only takes a moment -- 300 microseconds -- and less than the time the human retina would store the material. What I do next (in just a few more microseconds) is create a meta-concordance. It's not just a word-frequency count but a matrix representing the frequency of co-occurrence of each of the 100,000 distinct words that Joyce used in his life's work. It is a matrix with 10^10 entries, most of which are zero, representing the number of times the i-th word and the j-th word co-occur in a sentence somewhere within the Joycean corpus. It is a scholarly work. It is meant to help analyze Joyce's writing style. It is cool. It is definitive. I publish it, and to further the goals of Joycean scholarship, I put it on the Web. And as a final gesture of good-will I donate it to the public domain. Years pass. The matrix is copied and rebroadcast. Some versions eventually lose the attribution that I created it. Some even lose the fact that it is an analysis of Joyce's work. The matrix is kept intact, but spreads from place to place and venue to jurisdiction and from jurisdiction to medium. It becomes an object of fascination in certain circles. Someone asks the question "what is it?" and they seek to discover an answer. They develop amazing software which decrypts the puzzle -- lo and behold they have recreated the complete works of James Joyce, and amazingly in the public domain. They post the result.
© circa 1989, David Dailey.
Note: At least "Ulysses" was still under copyright when I created this scenario in about 1989 -- the current status of Joyce's work seems a bit vague see http://digital.library.upenn.edu/books/bplist/archive/2000-01-20$1.html.
Back to Dailey on copyright.