More ways to *ABA: the case of Blansitt's generalization


CAHA Pavel

Year of publication 2017
MU Faculty or unit

Faculty of Arts

Description In the talk, I pick up the general topic of whether *ABA patterns lead uniformly to the so-called cummulative decomposition (Bobaljik 2012) or not, and I argue that there are in fact more ways to get a *ABA pattern. As a specific case, I look at the marking of datives, allatives and locatives. Blansitt (1988) has observed that these roles may show various syncretisms, with the exception of dative and locative being syncretic to the exclusion of the allative, an instance of the so-called *ABA generalization. I explore its implementation in terms of feature decomposition, showing first why classical nesting structures (pioneered in Jonathan Bobaljik's work on adjectival suppletion) won't work for this particular case, and proposing an alternative in terms of the so-called overlapping decomposition.
Related projects: