A04

Resolving the Neg-Raising paradox

Project Overview

(1)	a.	I don’t think you’re right.
	b.	I think you’re not right.
(2)	a.	I don’t believe you’re right.
	b.	I believe you’re not right.

Neg-Raising (NR) is the phenomenon, illustrated in (1)–(2), where certain negated predicates (e.g. think, believe, expect) allow readings in which negation seems to take scope in the embedded clause. For example, (1a) can be interpreted roughly as (1b), in addition to the literal interpretation that I do not entertain the thought that you are right. The same holds, mutatis mutandis, for (2a–b).

Research Questions & Theoretical Framework

Other predicates, such as predict or claim, do not show this behavior: the sentences in (3a) and (4a) lack the readings in (3b) and (4b). Predicates that allow such readings are called Neg-Raising Predicates (NRPs); the corresponding interpretations are NR readings.

NR is crucial for understanding how syntactic negation relates to semantic scope and cross-linguistic variation. It has been analyzed from syntactic and semantic–pragmatic perspectives, but neither fully accounts for the data. This project combines earlier work by the PIs (Sailer 2006, 2007; Zeijlstra 2018; Mirrazi & Zeijlstra 2021, to appear) to address the apparently paradoxical behavior of NR.

Methodologies & Data

This research will systematically investigate atypical cases of NR in Dutch, English, and German, and extend the study to other languages. Given the expertise in the project and existing contacts with native speakers, it will include two Iranian languages (Farsi and Kurdish), where preliminary results show interesting contrasts with Germanic languages. In the following periods, the research will broaden the typological coverage and refine the conditions on the use of NR, building on the results of the first funding period.