Galaxy luminosity evolution: How much is due to model choice?

The cluster and field luminosity functions (LFs) determined on large homogeneous samples (N > 2200 galaxies each) are almost indistinguishable, down to M* + 4 in the r and i filters, suggesting that the effect of the cluster environment on the galaxy properties does not affect the galaxy luminosity function in red bands. The similarity of the red band LFs in different environments suggests that the galaxy mass function is preserved during the galaxy infall in the cluster. By analyzing a large sample of galaxies in clusters, ideal from many points of view (multicolor data, large size, many clusters, metric magnitudes) we found that luminosity evolution is required by the data but its flavour cannot be unambiguously derived if a differential luminosity evolution between bright and faint galaxies is allowed. We show that the LF parameters (slope, characteristic magnitude and their evolution) and errors depend on assumptions in a way seldom recognized in the literature. We also point out logical inconsistencies between hypotheses assumed in deriving literature LF and presented results, suggesting caution in interpreting similar published results.