On the seemingly illogical behavior of negation in natural language

Monday, 20 April 2009, Colloquium

Hedde H. Zeijlstra, University of Amsterdam

Abstract

As already observed by Jespersen (1917), negation in natural language seems to deviate from negation in propositional logic (in the Fregean sense): (i) contrary to what might be expected, in many languages (but crucially not all) two negative expressions, able to induce semantic negation by themselves, do not cancel each other out, but rather yield one semantic negation; (ii) contrary to what might be expected, negation seems to be ‘uncomfortable’ in sentence-initial position: in many languages imperatives (with fronted verbs) may not be preceded by negation, and in V-to-C languages sole negative markers may not occupy Spec, CP.

In this talk I argue that these two phenomena are typologically intertwined and therefore call for a unified explanation. I present a theory that is based on the two assumptions that (i) functional categories, including negation, are syntactically flexible, and (ii) operators that encode the illocutionary force of a speech act take scope from the highest clausal head position. I show that this approach makes sense of all previously mentioned seemingly illogical aspects of negation in natural language, and of the typological relations among these phenomena.