On the exclusively borderline case-marking in Ainu

This paper argues that the system of marking case relations in Ainu represents a typical example of borderline case-marking since arguments are unmarked and it is difficult to distinguish case markers from adverbs. For instance, dative, instrumental, comitative, and other cases are marked by postpositional adverbs, which can occur without respective NPs and some can even take indexing. It is shown that such an unusual zero anaphora-like behavior of adverbial “case” postpositions can be explained by their relatedness to respective transitive verbs out of which they have not yet fully grammaticalized. I adduce a tentative hypothesis that case-markers in Ainu are relatively young and innovative, while applicatives, which are a functional alternative to case markers, are old and genuine, and that the borderline case-marking may generally correlate with the head-marking language type, which is most broadly attested in the Pacifi c Rim.