Abstract: The ionization mechanism for the low-ionization emission line regions in quiescent galaxies has been puzzling us for decades. Shocks and photoionization models are the two most promising candidates. I will show direct gas temperature measurements from stacked spectra that can help differentiate the two mechanisms. Surprisingly, I found neither models could explain all of the observed line ratios, nor can the combinations of them do. I will also demonstrate how we can measure the gas-phase abundance patterns in them despite the uncertainty about their ionization mechanisms.