Speaker: Zorica Puskar (ZAS)
Title: Multiple Challenges of Multiple Agreement
Date: Thursday 30 April
Venue: Online (sign up for our mailing list for details)
Time: 15.15 – 16.30 hrs
Abstract
In some languages where finite verbs morphologically index both subject (S) and object (O) agreement, both are argued to be performed by a head (or heads) high in the syntactic structure, mostly T or higher. This type of multiple agreement poses a problem for each of the crucial ingredients of Chomsky’s (2000; 2001) Operation Agree: Matching (multiple Goals must be detected and Matched, without necessarily including their complete phi-set), Minimality (interacting with the lower O argument past the higher S), Locality (the two Goals belong to two different phases, O is in the vP domain, S is at the edge) and Activity (agreement across an active S should be impossible).
In this talk I will present a preliminary survey of languages argued to involve a high object probe (belonging to Uralic, Algonquian, Quechuan, Basque, Guiacuran, Chukotko-Kamchatkan and Sahaptian families). A closer examination of their S and O agreement patterns reveals that O agreement in these languages is not always obligatory, and that it often hinges on certain conditions. Such conditions may include person specification (e.g. only local-person or only third-person objects agree), or specificity/definiteness/topicality (Differential Object Marking). The conditions may be more elaborate, e.g. an affix that typically indexes features of one argument realises features of another argument if that argument has the preferred feature set (omnivorous agreement), or the verb seems to agree with object first and then with the subject (argument from the Mirror Principle where O-agreement marker is closer to the stem than S-agreement marker).
Upon illustrating the range of such preferences, and indicating the challenges they pose for the current theory, I will consider the theoretical avenues for a possible analysis. The idea to be explored is that each of the challenges can be overcome by assuming a combination of (i) multiple Probes on the T-head, (ii) Relativized Probing, and (iii) defining locality derivationally, in terms of agreement paths, as opposed to phases.