Most theories of morphosyntax in use by linguists today make some use of the notion of a feature, a way of encoding the various properties that a word possesses. Using features is a way of formalizing relationships between different forms of one word as well as identifying commonalities between different words. Although the necessity of features in syntactic theory is more or less uncontroversial, there is far less consensus on exactly how features are represented and manipulated in the grammar and how those abstract representations correspond to the realizations of the features that we can see in the words of a given language. My dissertation will address these issues by investigating features in Estonian and Icelandic nominals. The dissertation addresses two domains: nominal phrase internal agreement (or concord), and the structure of partitives and pseudopartitives: constructions like ‘a bag of flour’ or ‘some of the children’. I argue in my dissertation that concord must be distinguished from at least one other form of agreement: subject-verb agreement. I propose a new model for concord within nominals, using partitive structures as a way to shed light on aspects of the phenomenon that simpler nominal structures cannot illuminate. The dissertation thus provides a novel analysis of concord as well as, to my knowledge, the first generative syntactic analyses of partitives in both Estonian and Icelandic.
Mark Norris is a PhD Candidate in Linguistics at UC Santa Cruz.