9411672 The central issues in the study of language involve the relationship between language and nonlinguistic cognition. Accordingly, the aim of the project is to determine the kinds of conceptual and linguistic information that are called on to create grammatical agreement, and to explain their interplay in ongoing speech. Two general agreement relations will be explored. The first is number agreement between subjects and verbs, extending work carried out during the previous grant period. The second is agreement between subjects and pronouns, specifically reflexive pronouns and tag pronouns. Previous research focused on three questions about the production of subject-verb number agreement in English: (1) the nature of the verb's agreement controllers, whether linguistic or pragmatic; (2) the nature of the number information that is used in agreement, whether grammatical or conceptual; and (3) the principal locus of agreement implementation, whether created over grammatical relations or across surface strings. The empirically derived answers to these questions indicate that the verb's agreement controllers can be linguistic rather than pragmatic, that the number used during agreement implementation is grammatical, and that the implementation is driven from a grammatical level, with little influence either from conceptual features of the message or from the phonological features of inflectional morphemes. The new research extends these results in three directions. All three have to do with the cognitive context of agreement. The goal is to better illuminate the transition from conceptual to linguistic features, including the mechanisms that implement the concrete details of agreement. The experiments examine (a) whether the number and gender features that are called on in subject-verb agreement are the same as the number and gender features called on in pronominal agreement; (b) whether the features that are responsible for agreement errors are the s ame features that control canonical agreement; and (c) how speakers create agreement-inflected word forms. In addition to these three main themes, studies are proposed to extend the work crosslinguistically, going beyond the minimal agreement system of English. All of the investigations draw simultaneously from controlled experimental methods for eliciting utterances and from observations of agreement patterns in everyday speech. In these ways, the research systematically enhances our understanding of a cardinal facet of human communication, the production of language. The emphasis on normal production and the role of syntax in it makes the ongoing research program unique in psycholinguistics.
