Jaspersen cycles (van der Auwera et al. (forthc.)) have been proposed as explanations for the occurrence of double negation (DN) cross-linguistically. Beyer (2009, Volta River Basin languages), Dryer (2009, central Africa), and Idiatov (2018, northern sub-Saharan Africa) discuss contact phenomena as explanation for the proliferation of clause‑final negation (CFN) markers for SVO languages in these regions. Although Kainji is a major branch of Benue-Congo and geographically central to these regions, Kainji languages are not well documented (McGill and Blench 2012), and therefore their diverse data cannot have been taken into account.
Based on three closely related Northwest Kainji languages (U̠t‑Hun/Duka [uth], C’Lela/Dakkakanci [dri], and u̱t‑Maꞌin [gel]), I propose two distinct patterns of negation for Proto‑Northwest Kainji:
Clause-final Negation: SVONeg
Double Negation: SNegVONeg
Further, I propose that the “original” Northwest Kainji construction is CF and that DN is the secondary development. By investigating the detailed negation patterns of these closely related languages, we can see a diachronic explanation for the development of DN.
Note: Originally accepted for 51st ACAL scheduled for 21-24 May 2020; postponed due to Covid-19.