On the interaction of morphological and syntactic ergativity: Lessons from Kurdish

Abstract

It is well known that languages are not necessarily consistent in their alignment, i.e. whether they are ergative or accusative. For example, the morphology may be ergative while the syntax is accusative. Mismatches of this kind led Anderson (1976, 1977) to the conclusion that ergativity is a superficial phenomenon, largely confined to the morphology, with no further consequences for the syntax. His standpoint was sharply criticized by other scholars who insisted that the presence of ergativity was a significant indicator of a language's typological profile. But in more recent work, e.g. Dixon (1994), it is also claimed that the presence of ergativity in some domain of the grammar need not have any further consequences for the grammar as a whole, hence to a certain extent reaffirming Anderson's standpoint and furthermore casting doubt on the validity of alignment as a typological parameter. In this study I present a detailed analysis of ergativity in Kurdish and demonstrate that in Kurdish, ergativity is indeed a relatively superficial phenomenon without further consequences for other levels of grammatical organization. Hence the usefulness of alignment as a typological parameter is questionable. Kurdish data is presented on morphological alignment, several syntactic processes, and voice phenomena. Finally, evidence from the loss of ergativity in Kurdish (Dorleijn, 1996) is discussed which is seen to provide further support for my claims.