Coreference Relations in American Sign Language

Abstract

Coreference relations are at the heart of ASL grammar. This statement is of no surprise since it is true of every language, but ASL distinguishes itself in expressing many of its coreference relations overtly by means of a spatial system of co-indexing. This paper reviews characteristics of this overt co-indexing system and goes on to focus upon those factors which condition the absence of pronouns in topic and argument positions. The factors to be considered are spatial agreement markers, the presence of role prominence and object clitics, and the presence of topics (both overt and null). This paper follows from work presented in Shepard-Kegl (1985; Section 6.2 Typological Parameters and Empty Categories) where research by Huang (1984) and McCloskey and Hale (1982) was used as a springboard for distinguishing two types of null pronouns in ASL. The analysis in my dissertation is revised with respect to the behavior of verbs which take object clitics. Previously I argued that these verbs condition pro-drop in object position. However, I argue here that these object clitics are arguments serving as case absorbers which prohibit the presence of any argument in object position. The topicalized NPs that optionally appear in sentences with object clitics are not coindexed with an empty category in object position, but rather bind the object clitic and by virtue of this variable binding relationship inherit the case and thematic characteristics it possesses. I conclude by discussing some of the characteristics of ASL’s overt co-indexing system and how it interacts with Binding Theory, Case Theory and Theta Theory as presented in Chomsky (1981) and its extensions. This is not an exhaustive treatment of the issues. My goal is to bring to light some interesting questions, issues and possible problems raised by a language with overt preferential indices and multiple types of null pronouns.