Profiles of the H-alpha line calculated using the full computer simulation method (FCSM) and the standard theory (ST) agree excellently with a measured profile (at an electron density of N_e = 9.0 × 10¹⁶ cm⁻³ and temperature T = 12 600 K) in an argon–hydrogen (minority) arc plasma. Calculated widths (FCSM) and shifts (ST) also agree well with experimental data from flash-tube plasmas and gas-linear pinch plasmas. Some discrepancies are probably caused by experimental flaws. Our analysis of the so-called generalized theory (GT) shows that this theory is burdened with un-physical approximations, and consequently so are all subsequent modifications based on GT. It is shown that the so-called acceleration of the electron by the ion field (AEIF) effect is no new effect, but only a drastically simplified alternative model involving physical and algebraic errors, compared with traditional theory (ST, FCSM, etc) of the ion–electron interactions. A corrected, but still somewhat faulty, AEIF model yields an up to 15% increase of the half width of the H-alpha line (about 10% being contributed by unavoidable approximations and 3–5% by the pure AEIF effect, and worsening the agreement with experimental data) in contrast to the originally reported 25% decrease of the half width. Finally, we find that for plasma parameters Γ≤ 0.25 there is neither theoretical nor experimental evidence for the existence of the new 'warm dense matter effects' claimed in Escarguel et al (2000 Phys. Rev. E 62 2667), Flih et al (2003 J. Phys. B: At. Mol. Opt. Phys. 36 283), Escarguel et al (2000 J. Quant. Spectrosc. Radiat. Transfer 64 353) and Flih and Vitel (2001 Proc. 15th Int. Conf. on Spectral Line Shapes).