The substance MHPOBC is the oldest and still most important reference antiferroelectric liquid crystal (AFLC). There is still considerable controversy concerning the correct phase designations for this material and, in particular, about the presence or absence of SmC* in its phase sequence. By means of dielectric spectroscopy and polarizing microscopy, we show that whereas the pure compound lacks the SmC* phase, this phase rapidly replaces the SmC*β subphase through the reduced purity resulting from temperature-induced chemical degradation which is hard to avoid under standard experimental conditions. X-ray investigations furthermore show that this change in phase sequence is coupled to a decrease in translational order. This explains the large variations in the reported phase sequence and electro-optic behaviour of MHPOBC, in particular concerning the SmC*β phase which has been said to exhibit ferro-, ferri- as well as antiferroelectric properties. It is likely that the sensitivity of the AFLC phase sequence to sample purity is a general property of AFLC materials. We discuss the importance of optical and chemical purity as well as tilt and spontaneous polarization for the observed phase sequence and propose that one of the key features determining the existence of the different tilted structures is the antagonism between orientational (nematic) and translational (smectic) order. The decreased smectic order (increased layer interdigitation) imposed by chemical impurities promotes the synclinic SmC* phase at the cost of the AFLC phases SmC*α, SmC*β, SmC*γ and SmC*a. We also propose that the SmA* phase in FLC and AFLC materials may actually have a somewhat different character and, depending on its microstructure, some of the tilted phases can be expected to appear or not to appear in the phase sequence. AFLC materials exhibiting a direct SmA* -SmC*a transition are found to be typical ‘de Vries smectics’, with very high orientational disorder in the SmA* phase. Finally, we discuss the fact that SmC*β and SmC*γ have two superposed helical superstructures and explain the observation that the handedness of the large scale helix may very well change sign, while the handedness on the unit cell level is preserved.