The current investigation aims to lay the conceptual and basic theoretical foundations of a morphosemantic theory of number. Until now, morphological and semantic theories of number have addressed very different problems and have produced answers of little mutual relevance (see below). So, the notion that the study of Universal Grammar requires a theory that covers both the semantics and the morphology of number constitutes, I believe, a substantial departure from previous work.
To be specific, morphological theories of number have been primarily concerned with inventories of pronouns and agreement of the world's languages (see Corbett 2000, Cysouw 2003 for excellent overview and synthesis). For instance, the morphologist might wonder whether there exist cardinally exact trial, quadral or quintal number, or whether the forms that permit such readings are really paucals, and hence cardinally inexact. More rarely, morphologists concern themselves with the relationship between members of such inventories. For instance, in a system with singular, dual and plural, one can wonder whether each of these is sui generis, or whether, for instance, dual is a type of ‘expanded’ singular or ‘restricted’ plural. Even more rarely, observations that numbers in some languages are not sui generis have led to attempts to specify the features that underlie them. An important example of this work is Noyer (1992), which the current investigation follows both in spirit and in content. So, morphological theories of number examine languages' pronoun inventories and agreement categories, aiming to explain why only certain ones are attested or how members of the inventory are related.
KeywordsWord Order Number Feature Direct Object Syntactic Feature Mass Noun
Unable to display preview. Download preview PDF.