In this article we bring to light one additional factor underlying so-called Jespersen’s Cycle (JC) in Romance which has to date gone unnoticed, namely the varying position of the finite verb within the IP. More specifically, we show that there exists an empirical correlation between the availability of clause-medial/high verb-movement and Stages II–III of JC in which a postverbal negator is licensed. Drawing on novel data, we demonstrate that this correlation holds not only across modern Romance varieties, but also across early varieties. Formally, we explain this link between negation and verb-movement from the (in)active status of the T-domain and the consequent (im)possibility of donating a [Neg] feature to the lower v-VP domain. Although verb-movement in itself is not a sufficient condition to trigger a shift towards Stages II–III negation, we argue that it is a necessary one, a fact which explains the peculiar distribution of negation strategies across the Romània.
- Jespersen’s Cycle