Seminar within the research area Linguistic Structures, Department of Languages and Literatures. All interested are welcome!
This talk focuses on negation in Ngoreme (JE401) [ngq], a Great Lakes Bantu language spoken by some 55,000 speakers in the linguistically dense Mara region in north-western Tanzania (Aunio et al. 2019). The domain of negation in Ngoreme is in a state of considerable flux; it appears to be caught up in several different and intertwining diachronic cycles of change affecting both standard (main/declarative) as well as non-standard negation.
In contrast to its closest relatives, Ngoreme seems to have moved towards the (re-)instantiation of a more canonical Bantu negative system with a syntactically based and syntagmatically coded division of labor between a “pre-initial” negator te-/ti- for standard negation and a “post-initial” negative prefix -ta- for more “marked” clauses (cf. Güldemann 1999; Nurse 2008:179-201; Devos forthc.). This shift, however, is far from complete, as negative main clauses may still optionally be expressed with the post-initial negator. What is more, the pre-initial negative prefix may be accompanied with a (non-obligatory) post-verbal particle he in a type of discontinuous negation construction more reminiscent of the initial stages of the cross-linguistically renowned Jespersen’s Cycle (see e.g. van der Auwera 2010, Devos & van der Auwera 2013).
Changes are also taking place with regard to non-standard negation: The imperative verb form inflected with post-initial negative morphology has been re-analyzed as expressing an affirmative rather than a negative command. Instead, imperatives – but also several subordinate clause types – are negated through the use of different auxiliaries which vary in functional range.
Finally, there is a separate development where another auxiliary construction marking ‘not yet’ (as well as ‘before’) appears to be neutralizing into a regular negative perfect along the cyclic route sketched by Veselinova & Devos (2021).
The aim of this talk is to account for this intriguing mess and the historical processes behind it, drawing on language-internal as well as both intra- and inter-genealogical comparative data.