The presentation exposes a couple of ideas on how models of color categorization in human languages should be evaluated.