Why doesn't Google just count 'original' as whoever published the corpus first...