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