It has often been claimed that the distribution of case marking is systematically affected by a universal scale of referential categories. This can be understood as a universal correlation between the odds of overt case marking and scale ranks (a negative correlation for subjects, a positive one for objects), or as an implicational universal proposing that, if a language has a split in case marking, this split fits a universal scale. We tested both claims with various versions of scale definitions against a sample of over 350 case systems worldwide, controlling for confounding factors of genealogical and areal relationships. We find no statistical evidence for a universal correlation that is independent of family membership and has any appreciable predictive power. Formulated as an implicational universal, we find that there are only few areally independent families that show a trend towards fitting scales, and that each family fits different scales. What we do find, by contr ast, is a strong area effect: once genealogical relationships are controlled for, differential argument marking shows a frequency peak in Eurasia and nowhere else. We conclude that the currently available empirical evidence is too weak to reject the null hypothesis that splits in case marking develop through individual diachronic changes – such as innovations of case morphology in nouns but not pronouns (Filimonova, 2005), reanalyses of instrumentals as ergatives on inanimates (Garrett, 1990), contact-induced calquing of definite vs. indefinite contrasts by means of case marking, or other idiosyncracies.