A cross-linguistic study of syntactic and semantic agreement : polite plural pronouns and other issues
MetadataShow full item record
This dissertation investigates syntactic and semantic agreement in Czech, French, Latvian, Persian, Romanian, Russian, Serbian/Croatian, and Turkish. When more than one target agrees with a controller with a mismatch between its form and meaning, different patterns of agreement are attested: uniform syntactic agreement (e.g. these/*this pants are/*is cute), uniform semantic agreement (e.g. this aircraft is fast/ these aircraft are fast), and mixed agreement (e.g. this/*these committee has/have decided in British English). Diverse agreement patterns are shown to arise through an interaction of three components of agreement: different types of agreement controllers, different types of agreement targets, and the Agreement Marking Principle (Wechsler and Hahm to appear). A distinction between CONCORD, INDEX, and semantic agreement features is adopted (Wechsler and Zlatic 2003). Agreement controllers are specified for different phi-features in CONCORD and INDEX. The types of agreement targets differ in whether they are sensitive to the CONCORD or INDEX features of their controllers. The relation between controllers and targets is governed by the Agreement Marking Principle, which states that an agreement target checks for a feature of its controller, but the target defaults to semantics when its controller lacks the feature. For example, the noun committee in British English is specified for singular in CONCORD but lacks number in INDEX, and thus a CONCORD target (e.g. a demonstrative determiner) agrees in singular syntactically, whereas an INDEX target (e.g. a finite verb) defaults to semantics because it fails to find an INDEX number feature specified for the controller. This research focuses on agreement with polite second person plural pronouns across languages. Such pronouns are plural but can refer politely to a single addressee. All languages discussed exhibit syntactic agreement on finite verb targets when the controller is a polite plural pronoun, while other types of target vary across languages in how they agree. This generalization is explained by two factors: pronouns possess INDEX features and finite verbs are necessarily INDEX targets if they are specified for the person feature, which only belongs to INDEX. A Hybrid Pronoun generalization, which encompasses all types of hybrid pronouns, is supported by evidence from Indo-European and non-Indo-European languages.