Number-based noun classification

Abstract

AbstractNilo-Saharan languages are well-known for their complicated system of nominal number marking, which features a variety of singulative and plural affixes (Dimmendaal 2000). Even though these systems have received some attention in the typological literature, there has been limited theoretical work on their implications for the morphosyntax of number cross-linguistically. The goal of this paper is to fill this gap, by providing an analysis of nominal number morphology in Kipsigis (Nilotic, Kenya), based on data from original fieldwork. First, I show that singulatives in Kipsigis are true allomorphs of singular number, unlike singulatives with a classifier function in languages like Ojibwe (Mathieu 2012). The descriptive term ‘singulative’ is therefore misleading, as it corresponds to two very different types of morphemes. Second, I claim that the tripartite system of number marking of Kipsigis and other Nilo-Saharan languages is due to the classification of nouns into morphosyntactic classes defined by the presence of inherent number features on little n; the interaction of these features with interpretable number features on the functional projection Num (Ritter 1991 a.o.) in the post-syntactic component gives rise to the exponence pattern that we observe. Finally, my analysis corroborates the existence of noun classification based on number, which has only been argued for Kiowa-Tanoan before (Harbour 2007). The existence of three number classes in Kipsigis can only be explained by reference to bivalent number features; number-based noun classification systems thus strongly support the view that number features are bivalent and not privative, which is also argued by Harbour (2007, 2011) for Kiowa.