Semantics of the northern khanty salient article: definiteness, salience, and obviation

Abstract

Uralic possessive agreement markers often function as determiners. This paper presents a case study of the Northern Khanty (Kazym dialect) 2sg Possessive that developed into a “salient article”. The Salient Article is definite as it requires informational uniqueness and familiarity, but its distribution is narrower than the distribution of previously described definite determiner types. It is most commonly used with topical Subjects and in noun phrases with demonstratives, but its use is not obligatory across the board in these cases and is not limited to them. Furthermore, the Salient Article is subject to a constraint that is similar to the proximate uniqueness constraint of languages with obviation systems like the Algonquian: there may be at most one noun phrase with a Salient Article per clause (with the exception of noun phrases with demonstratives). I consider and reject two possible syntactic accounts of such distribution and instead propose a tentative semantic analysis that derives all the observed facts: the Salient Article marks the most salient discourse referent in the given context. (I understand salience as a graded property that a referent has to the extent that the referent is being attended to by the addressee following Roberts and Barlew). This study thus supplies another argument for the hypothesis that salience is an important dimension to determiner semantics cross-linguistically.