Much of my research relies on my original fieldwork on the Bantu language Zulu. I have made several fieldwork trips to KwaZulu-Natal, South Africa, most recently spending April 2012 and summer 2011 in Durban, South Africa. Conducting fieldwork on Zulu is important to my research program in multiple respects. There are many aspects of Zulu grammar that have not been documented; to fully explore the syntactic phenomena I am concerned with, I need access to original data. In addition, as with many Bantu languages, much of the research on Zulu has involved very small numbers of speakers from different regions. As a result, it can be difficult to interpret variations in the data across the literature. My recent work in Durban has given me the opportunity to work with many speakers of a single dialect and to begin to understand the character of that dialect. I hope to be able to continue mapping out the syntactic properties of the Durban dialect, and perhaps to begin to compare it to other dialects, such as the more conservative northern dialect spoken in the region of Maqongqo, where I have researched previously. In addition, much recent research on Zulu has revealed striking differences with the syntactic properties attributed to the language by researchers working 50-100 years ago. Even in my own research, generational differences in the Durban dialect have emerged. Given these apparent syntactic shifts that are underway in Zulu, I am particularly eager to thoroughly document the current syntactic state of the language.
Case and Agreement
Much of my research has focused on the factors that govern nominal distribution and agreement patterns. My dissertation explores these issues primarily in the Bantu language Zulu. In particular, I argue that Zulu has both a system of abstract structural case and a system of morphological case. This conclusion is notable in light of some of Zulu's unusual-looking properties in the domain of nominal distribution and with respect to the long-standing assumption that Bantu languages lack both of these types of case (e.g. Harford Perez 1985). In addition to Zulu, I have also examined nominal distribution patterns in Kinande, another Bantu language. In my dissertation, I argue that the initial "augment" vowel that marks most nouns in Zulu functions as case marker, along with certain prepositional morphemes in the language. I am currently expanding my work on case morphology in Zulu to investigate how similar morphemes in related Bantu languages function with respect to nominal licensing. A large part of my work on case and agreement in Zulu has focused on the existence of optional raising out of agreeing, finite clauses. I am now working to better understand the differences between raising-to-subject and raising to object constructions in Zulu. I am also beginning to compare optional raising and optional agreement constructions in Zulu to similar constructions in other Bantu languages -- and to languages beyond the Bantu family -- to approach the question of what drives argument raising in these constructions where neither agreement nor argument licensing seems to depend on such movement.
In ongoing projects, I have been working with Hadil Karawani and Bronwyn Bjorkman on the cross-linguistic typology of CF marking. In a paper with Hadil Karawani, we compare Zulu and Palestinian Arabic to show that imperfective morphology does not seem to be a necessary ingredient of counterfactual marking in those languages that use past tense as a marker of CFs, contrary to previous generalizations that claimed imperfective aspect appears in addition to the past tense in a subset of languages that use past to mark CFs. In my work with Bronwyn Bjorkman, we propose further amendments to the typology, arguing that all languages that use "fake" temporal morphology to mark CFs require only a single ingredient: either past (English, French, Arabic, Zulu, Russian) or imperfective (Hindi, Persian). In all languages that appear to require both past and imperfective, we demonstrate that the appearance of the second temporal ingredient is an illusion, resulting from morphological underspecificaton in a language's temporal system.
My work on Zulu has also involved the prosody-syntax interface, in particular focusing on the prosody of Zulu post-verbal subject constructions. In these low subject constructions, the subject can easily receive a narrow focus interpretation. As Cheng & Downing (2011) have showed, XPs that receive focus within vP typically require all other XPs to evacuate. They argue that this movement occurs to ensure that prosodic prominence falls on the focus. In low subject constructions, movement of other XPs is syntactically ruled out. In these cases, the grammar allows prosodic prominence to be realized on the focused subject even though there is usually a strict match between prosodic prominence and syntactic structure. I am working on how to model the way in which the requirement for movement over prosodic manipulation is overridden by independent syntactic considerations. I am also interested in the cross-linguistic typology of languages that favor movement to achieve prosody-focus match but where the movement strategy also runs into syntactic limitations.