Descriptive/metalinguistic dichotomy?: Toward a new taxonomy of negation

This paper proposes, based on Japanese data, that Horn's (1985) descriptive/metalinguistic dichotomy of negation should be replaced with a distinction between OBJECTION and DESCRIPTION, and that the taxonomy with an abstract, logical view of negation should be replaced by a functional, cognitive-pragmatic (relevance-theoretic) view. Horn's descriptive negation (DN) represents a truthfunctional operator, taking a proposition p into a proposition not-p, and "metalinguistic" negation (MN) represents a non-truth-functional operator, objecting to a previous utterance. While it is hard for English to distinguish ON from MN based only on their forms, Japanese external negation is marked by a number of forms, each of which is sensitive to different properties. Based on a close examination of three Japanese negatives, nai (:= not), node wa nai it is not that...) and wake de wa nai it does not follow/mean that...), this paper extracts three properties that contribute to the taxonomy of negation (i.e. objection as a function, attribution in Wilson's (2000) sense, and the conceptuality of the negation target), and concludes that the 'attribution' of the lower representation (and its resulting function 'objection'), but not truth-functionality, is a crucial property for the dichotomy of (at least Japanese) sentence negation. (C) 2013 Elsevier B.V. All rights reserved.