This theoretical study focuses on geometries, vibrational spectra, charge distributions, electron affinities, and reaction energies for SO(n)(p-) anions and alkali salts MSO(n)(-), M(1,2)SO(n) in the gas phase (n = 1-3; p = 0-2; M = Li-K). Most of our data for compounds with the S oxidation states 0, 2, and 4 are new in the literature. The bulk of the results are obtained at the B3PW91 level, with CCSD(T)=FC calculations carried out for relative energy calibrations; the 6-311+G(3df) basis set is used throughout. The formation of contact ion pairs is prevalent; they are of type: (i) M(+)(SO(n)(-)) for the π-radicals MSO, MSO(2), MSO(3) of doublet multiplicity; (ii) (M(+))(2)(SO(n)(2-)) for M(2)SO, M(2)SO(2), M(2)SO(3) in their singlet ground states; and (iii) M(ns)(SO(n)(-)) for the radicals MSO(-), MSO(2)(-), MSO(3)(-) in their triplet states. When isolated in matrices, M(2)SO and M(2)SO(2) will facilitate the spectroscopic study of the little known SO(2-) and SO(2)(2-) ions. Divalent M(2)SO(n) salts, due to their large dipole moments, should be highly soluble in polar solvents, first dissociating into MSO(n)(-) + M(+) products. For MSO(3), bidentate coordination OS(O(2)M) is preferred over tridentate S(O(3)M) binding. We confirm that all MSO(2) molecules are planar, at variance with an ESR study assigning to NaSO(2) a nonplanar structure. This study partially support the assignment of an experimental frequency at 918.2 cm(-1) (932 cm(-1), calculated) to the antisymmetric ν(a)(SO) mode of the elusive sulfoxilate ion, SO(2)(2-). A definitive identification, however, would require to record the vibrational spectrum below 800 cm(-1) (apparently not done in the original work) because the missing symmetric ν(s)(SO) mode is here found to lie around 760 cm(-1), exhibiting high intensity in both IR and Raman spectra.