Persistent ergativity: Agreement and splits in Tsimshianic

This thesis presents a morphosyntactic analysis of agreement patterns in the Tsimshianic language family (primarily Gitksan, with some focus on Coast Tsimshian). These ergative languages demonstrate several distinct splits in the distribution of agreement, conditioned by differences in clause type,...

Full description

Bibliographic Details
Main Author: Forbes, Clarissa
Other Authors: Bejar, Susana, Linguistics
Format: Thesis
Language:unknown
Published: 2018
Subjects:
Online Access:http://hdl.handle.net/1807/91770
Description
Summary:This thesis presents a morphosyntactic analysis of agreement patterns in the Tsimshianic language family (primarily Gitksan, with some focus on Coast Tsimshian). These ergative languages demonstrate several distinct splits in the distribution of agreement, conditioned by differences in clause type, nominal type, and person. Despite the presence of such agreement splits, the ergative pattern of agreement is largely unaffected: I refer to this property as persistent ergativity, as opposed to split ergativity. A pattern of particular interest is that of agreement switch (Kalin 2014; Kalin and van Urk 2015), or pivoting ergativity (Davis to appear). In this type of split, the morphological paradigm used to mark ergative in one context is used to mark absolutive in another context, with the overall ergative/absolutive alignment of the system remaining unchanged. At the heart of the proposed analysis is the use of feature relativization to generate several different arrays of potential agreement targets. I propose that agreement in Tsimshianic can be modeled with an ergative agreement probe on v which consistently agrees with the transitive subject, plus a higher agreement probe with several options for how it may be relativized, and consequently several potential agreement distributions: absolutive, nominative, or ergative. In particular, I analyze agreement switch as the result of the high probe agreeing directly with the lower ergative probe. This is accomplished by relativizing the higher agreement probe to uninterpretable phi-features, found exclusively on the lower phi-agreement probe. I capitalize on several familiar featural distinctions (including the interpretable/uninterpretable distinction), and apply them in new ways through feature relativization to generate typologically unusual splits without affecting the ergative pattern of v-agreement. The proposed analysis is novel among accounts of Tsimshianic in its approach to the factor conditioning the clause type split, and in its connection between agreement and A'-patterns. It is furthermore comprehensive, in explicitly accounting for agreement and word order patterns across both the Interior and Coast branches of the family. The analysis of the complex agreement patterns presented is of interest to theories of ergativity, as well as theories on the mechanics of features and agreement systems. Ph.D.