Abstract

The traditional notion of linguistic irregularity is unique, without any internal distinction. It covers all the linguistic facts departing from the general patterns of a particular language. Everything falling within this notion is thus considered irregular in the same way. Under the irregularity label, though, traditional grammars exhibit an amount of facts, some of which do not have much in common, and would be better treated as belonging to different sets or to different subsets. Assuming that `proper' irregularity is that related to transgressions of the rules, different kinds of 'improper' irregularity come to light. The aim of this paper is to show this on a morphological basis.