Introduction

Chapter 2 argued for a distinction between two types of ne – ne1 and ne2 – on the basis of patterns of variation and change in diachronic corpus data within a model of change as morphosyntactic competition. These two types of ne have distinct distributions in corpus data and exist in different relationships to not: ne1 competes with not, and ne2 co-occurs with not. This chapter will observe a similar distinction between ne1 and ne2 in contexts of redundant or paratactic negation where ne is used independently of not. Redundant ne provides further evidence of the distinction made between ne1 and ne2 in the preceding chapter. Here, I argue that this evidence allows us to characterise more precisely the distribution – and therefore the syntax – of ne at each stage of the Jespersen Cycle.

The phenomenon of redundant negation, also called expletive negation (van der Wurff, 1999b) or paratactic negation (Jespersen, 1917) is well known and extensively described for European languages, see for example van derWouden (1994), Espinal (2000). Van derWouden (1994, 107) defines paratactic negation as follows:

While the typical contexts for paratactic negation (PN) are well known crosslinguistically, data from two early English corpora – the York Corpus of Old English (Taylor et al., 2002) and the PPCME2 (Kroch and Taylor, 2000) – reveal contexts for PN in Middle English which are not described in the literature. They do not involve the usual triggers for PN: predicates of prohibition, fear or denial, comparatives of inequality or prepositions indicating temporal sequence. Hence they are difficult to accommodate under accounts of PN based on negative polarity item licensing (van der Wouden, 1994); on the syntactic subcategorisation of PN contexts, for example van der Wurff (1999b); or on a characterisation of PN contexts in semantic terms (Yoon, 2011).