LSA.131 | The Syntax of Agreement

Mark Baker
TR 8:15-9:55
location: 32-141

On a casual survey of languages of the world, processes of agreement seem to be among the most idiosyncratic and least universal aspects of grammar. Languages differ greatly in how much agreement they have: some have essentially no agreement (e.g. Chinese), some agree only with the subject (e.g. Turkish), and some have agreement with virtually all grammatical relations (e.g. Mohawk). In some languages agreement is optional, whereas in others it is obligatory. Also, the features that enter into agreement differ widely; some languages show agreement only for person, whereas others have very rich and pervasive gender systems. Nevertheless, underneath all this diversity there are some very striking patterns that point to a strong universal core of syntactic principles that underlie agreement systems, along with a few simple parameters. Particular languages then embellish on this core system in different ways in the morphological component. This class will try to uncover the core system and its major variants by close comparison of such different languages as Kinande (Bantu), Lokaa (Niger-Congo), Mohawk (Iroquoian), and Mapudungun (spoken in Chile ), as well as more familiar languages and languages known from the literature. Specific topics that will be investigated in the course include: (a) developing a unified theory for subject-verb agreement and noun-adjective agreement, (b) characterizing which syntactic nodes can host agreement, and seeing whether these vary from language to language,(c) defining the structural relationship that must hold between the host of agreement and its trigger, and seeing if this varies from language to language, (d) reevaluating the role of agreement in licensing null pronouns, dislocation, and nonconfigurationality, (e) investigating the role agreement plays in licensing adjunct phrases, and (f) considering the special properties of traces with respect to agreement.