A model-theoretic discussion of Bantu actionality

Recent years have seen increased attention to the dimension of actionality (also known as lexical aspect) in Bantu. Bantu verbs are known for their complex lexicalisations of aspectual structures, in which the same verb frequently encodes both a coming-to-be phase and a result state (e.g. ‘get/be angry’). The prevalent framework of actionality in Bantu, developed primarily by Robert Botne and Tiffany Kershner, models these complex lexicalisations as consisting of up to three distinct phases: onset, nucleus, and coda. In this paper, we describe the key tenets of the theory, tracing its development and cross-linguistic applications. We then offer a problematisation of the three-phase model and note some outstanding questions in the domain of Bantu actionality. We suggest that a simpler model might allow for more straightforward analyses and comparative work. We sketch a preliminary proposal for a framework in which the explanatory weight is partially shifted to semantic characteristics of the phases and boundaries. In this model, maximally two phases – a coming to be phase and a resultant phase, along with their boundaries – are lexically encoded.