The electron configurations (in spite of being a debatable approximation to the wavefunction) predict correctly the number of levels with givenJvalue, and in the case of reasonably valid Russell-Saunders coupling, their bunching together in (S,L) terms. The similarity with the energy levels in condensed matter attenuates from the 4f to the 5f group, and in d-group compounds (like those outside the transition groups) usually only 5 remains a good quantum number. Recent experimental results, such as the Judd-Ofelt treatment of absorption and fluorescence transition probabilities (of importance for lasers), photoelectron and Auger spectra, antiferromagnetic coupling through anion bridges (allowing excitation of two transition-group ions by the same photon) and arguments for definite spectroscopic and conditional oxidation states, are connected with a general analysis of non-totally symmetric states. It is shown that the less frequent occurrence of positive S values in compounds of elements lighter than calcium, distort many opinions about chemical bonding.