The limits of conventional cladistic analysis