Gender distinction in non-singular numbers is widely believed to imply gender distinction in the Singular, especially that of corresponding genders and in one and the same word inflectingfor these two categories. The bestknown claim to this effect are the universals Nos. 37 and 45 in Greenberg (1963). There arefew universals without exceptions, and the Singular preferencefor genders, though sometimes claimed to be absolute, is after all no exception. However, considering the reputation ofthis universal and similar ones capitalizing on markedness, it is perhaps surprising that instances of genders in non-singular numbers which lack a counterpart in the Singular, äs documented in this paper, are äs numerous äs they are. Offen there are in fact goodfunctional or diachronic reasonsfor genders toprefer non-singular numbers, and one therefore begins to wonder how strong the reasons really arefor the regulär, or at any rate predominant, gender preference to be on the Singular.