I'd be interested to know, but somehow I seriously doubt there were librarians taking eigenvalues of citation networks.

Yeah, nothing there or in related articles disputes the assertion that they invented PageRank. Sure, they did it in a community that was developing similar ideas, but they built and developed it.