Background dielectric permittivity: material constant or fitting parameter?

The concept of background dielectric constant proposed by Tagantsev, together with the distinction between critical and non-critical electric polarizations as a natural extension for the order parameter of ferroelectric phase transitions, is critically discussed. It is argued and exemplified that, even if these quantities can be very useful for obtaining qualitative and semi-quantitative results from phenomenological modeling, they cannot be introduced in a self-consistent manner.