On an inadequate defense of an ellipsis-based analysis for Right Node Raising
Yusuke Kubota, Bob Levine
February 2018

A recent manuscript by Yatabe and Tam (2017) argues that an analysis of coordination in categorial grammar (CG), of the sort defended in detail by Kubota and Levine (2015) and whose key insights go back to proposals by Steedman (1985) and Dowty (1988) in the 80s, is inadequate on empirical grounds. Of the three pieces of evidence they offer, facts about the so-called 'summative agreement' phenomenon seems at first sight to be the strongest possible candidate for a real problem for the CG approach. In this short paper, we critically assess the argument offered by Yatabe and Tam involving these summative agreement facts, and argue that their refutation of the CG approach critically rests on a particular assumption about morphological number agreement and semantic plurality that can be shown to be too simplistic on independent grounds. Once this assumption about agreement is replaced by a more adequate analysis of the relationship between semantic plurality and morphological number agreement that is in line with the standard assumptions in the contemporary formal semantics literature, summative agreement facts are no longer problematic for the CG approach. Based on this discussion, we conclude that the CG approach still remains to be the simplest and empirically most robust analysis of the coordination facts in English.
Format: [ pdf ]
Reference: lingbuzz/003852
(please use that when you cite this article)
Published in:
keywords: coordination, summative agreement, categorial grammar, hpsg, linearization-based hpsg, right-node raising, semantics, syntax
Downloaded:139 times


[ edit this article | back to article list ]