Agreement and Clause Structure

Abstract

Abstract Part I discussed non configurationality in polysynthetic languages. In the course of the investigation, it became clear that there are normally three distinct elements involved in expressing the argument of, say, a verb: an empty category that is the true argument of the verb; a morpheme in the word containing that verb that makes the argument visible for 0-role assignment; and, optionally, an NP in adjunct position that expresses most of the semantic content associated with the argument. For the most part, part I concentrated on the relationship between the null argument and the adjoined NP. In part II we now turn to the more fundamental relationship: the one that holds between the null argument and the morpheme on the verb. This inquiry yields a better picture of basic clausal relations in the polysynthetic languages and clarifies some of the detailed operations of the Morphological Visibility Condition. It also develops the idea, introduced in chapter I, that much of the phrase structure in polysynthetic languages shows up on the surface as word structure. Recall that there are two types of morphological entities that can satisfy the MVC: agreement elements and incorporated roots. There are also two major categories that one needs to be concerned with: nouns and verbs. Hence, part II naturally divides into four topics: agreement on verbs, agreement on nouns, incorporation of nouns, and incorporation of verbs. I begin with the most basic of these topics-the syntax of agreement relationships within verbal domains.