Abstract

In seminal work, Zeijlstra has proposed that the sentential negative marker in strict negative concord languages is a meaningless particle (uNeg) that invokes a silent negative operator (iNeg) at the periphery. Negative concord items (NCI) are also supposed to have uNeg. This paper puts forth new arguments to the effect that the Hungarian negative marker NEM has uNeg, but NCIs do not. Their relation to negation is indirect; they need to be exhaustified, which in turn requires an intervening negation to maintain logical coherence (Chierchia 2013). This eliminates the appearance of redundancy in the negative marker co-occurring with NCIs. The paper sets the analysis in the context of the hybrid (strict and non-strict) NC system of Hungarian.