Working Paper No. 03-51:
Date Posted: 2003
Abstract (below) | Full text (most recent) on SSRN
This article is a limited defense of copyrights for the contents of factual compilations. The form of protection that I propose, under which the collective factual content of such compilations is protected, differs from an approach that protects individual facts and from the currently accepted approach (as articulated in Feist v. Rural Telephone), under which only selections and arrangements of individual facts are protected.
Although I accept that there are sound economic justifications for refusing to copyright individual facts, my justifications differ from those that have traditionally been offered. The traditional justifications are: 1) that the monopolization costs of protecting individual facts are too great, because facts are too valuable as components for future works to have access to them limited by property rights, and 2) that facts fail the independent creation requirement for copyright protection, because they are not authored by anyone. Both of these justifications fail. Monopolization costs can at most justify limited terms for copyrights in facts. And, far from failing the independent creation requirement, facts (properly understood as representations of reality rather than reality itself) are as much works of authorship as novels are. I argue that transaction and enforcement costs are the real reasons that individual facts are not copyrightable.
Furthermore, some components of factual works - specifically, ground breaking and explanatorily powerful theories like Einstein's theory of relativity - should be copyrightable if our sole concerns were transaction, enforcement and monopolization costs. Instead such theories are not protectable because any work that borrows them is their complement (in the sense that its production makes them more desirable), provided that the work acknowledges the theories' true provenance. This is because it is only through dissemination in other works that such theories can undergo the test of truth.
But nothing about the uncopyrightability of the components of factual works stands in the way of copyrights for the collective factual content conveyed by such works. It might appear that protecting collective content is no different from the Feist approach, in which selections and arrangements of facts alone are protected. After all, collective factual content is created by selecting and arranging individual facts. If individual facts are not protected, then the selections and arrangements, it seems, must be.
But this is a fallacy. Protecting a fictional story is not the same as protecting the methods of selection and arrangement used to generate the story from unprotected elements of character, plot, and setting. Likewise, protecting the collective factual content of a database is not the same as the Feist method. The collective content of databases, I argue, should be protected in the same manner that fictional stories are. Such an approach, far diverging from traditional copyright principles, follows from them.