The Morphosyntax of Bantu Nouns
This project is rooted in the research carried out over the past few years at the Center for Advanced Study in Theoretical Linguistics (CASTL). On the empirical side, this research has emphasized importance of cross-linguistic comparative study. On the theoretical side, a view of morphosyntax has crystalized that is highly decompositional and relies on a specific, non-standard conception of the relationship between morphemes and abstract morphosyntactic structure. The general framework emerging from this has come to be known as “nanosyntax”. Works such as Ramchand (2008) and a growing number of CASTL PhD dissertations bear witness to power of the nanosyntactic world view. Recently, an investigation of the nominal morphosyntax of a small sample of Bantu languages made us realize that this area of study may provide an ideal testing ground for the general theoretical assumptions characteristic of the nanosyntactic framework. Conversely, nanosyntax turned out to provide elegant analyses for many of the otherwise mysterious properties of the morphosyntax of Bantu nouns and agreement markers. This realization led to the project proposal we describe here.
The results of this preliminary investigation of the properties of nominal structures in a small group of Southern Bantu languages are reported in Taraldsen (2009) (to appear in Lingua). In this study, we have developed an analytical technique which we now want to apply to a much larger sample of Bantu languages. The project focuses on the analysis of the noun class prefixes of Bantu and their relationship to different classes of “agreement markers” (concords) affixed onto verbs and adjectives. The analysis of Nguni in Taraldsen (2009) shows that noun class prefixes are decomposable and correspond to fairly rich grammatical structures. It also shows that the different types of concords are built from different subparts of the structure associated with the full noun class prefix. But the theoretical interpretation of these results still needs to be refined, and we also need to know to what extent the results are replicated in other Bantu languages in a way consistent with our preliminary theoretical conclusions. The project has both a descriptive and a theoretical side. We will deliver a descriptively adequate account of the variation in the morphosyntax of nouns and concords within a representative sample of Bantu languages. We will also use the data obtained to provide an assessment of different theoretical accounts of the patterns seen in the morphosyntax of nouns and concords in Bantu. The empirical work will involve linguists working at African universities as well as their graduate students. We also want these to collaborate on the theoretical side, and will try to offer short summer schools for African graduate students to make this possible. Our results will both be a contribution to descriptive Bantu linguistics and a contribution to theoretical linguistics. The project is led by Knut Tarald Taraldsen and Michal Starke, and is funded by Tromsø Forskningsstiftelse. A post-doc position has been associated with the project from the fall of 2010.