Negative concord as downward Agree
Amy Rose Deal
December 2021
 

Negative concord (NC) plays a prominent role in debates about Agree directionality. Zeijlstra (2004, 2012) argues as follows. If the NC relation (the relation between sentential negation and negative concord items) is Agree, it is Upward Agree. The NC relation is Agree. Therefore, Upward Agree must exist (modus ponens). Preminger (2013) and Preminger and Polinsky (2015) respond by denying the consequent (modus tollens). They concur that if the NC relation is Agree, it is Upward (so Upward Agree would need to exist). However, they argue that Upward Agree doesn’t exist. Therefore, the NC relation must not be Agree.

In this paper, I argue that the shared premise of both arguments should be rejected. The NC relation can be handled purely with Downward Agree-—a move that is especially natural if we adopt an interaction/satisfaction view of Agree, which de-centers (un)interpretability. I show that treating negative concord as downward Agree is both theoretically viable and typologically productive, allowing variation in NC phenomena to fit neatly into independently known parameters of agreement variation.

Format: [ pdf ]
Reference: lingbuzz/006344
(please use that when you cite this article)
Published in: Draft (comments welcome)
keywords: agree, upward, downward, direction, directionality, negation, concord, italian, french, west flemish, interaction, satisfaction, agree, syntax
Downloaded:208 times

 

[ edit this article | back to article list ]