Diagnosing syntactic structure in ATB and RNR constructions: A reply to Larson (2013, 2014)
Doreen Georgi, Andrew Murphy, Timea Szarvas, Johannes Rothert
October 2024
 

A long-standing puzzle in syntactic research is how (apparent) sharing of an XP in ATB- and RNR-constructions can be derived without construction-specific assumptions. Various types of approaches have been proposed in the literature, e.g., ellipsis and multi-dominance. According to the hybrid (or ecclectic) approach (see Barros & Vicente 2011, Belk et al. 2024) several derivational options may co-exist in a single language. Larson (2012, 2013, 2014) reviews these approaches and concludes that none of them can be correct because they can all explain only a subset of the empirical facts. Larson (2013, 2014) thus proposes a new, non-syntactic account: The filler in ATB and RNR constructions is extracted from the gap site in the linearly closest conjunct; all other gaps do not contain any material in the syntax. The illusion of sharing arises in the interpretation (or parsing) of these structures. In this paper, we test an empirical prediction that Larson’s (2013, 2014) approach makes regarding the postulated absence of syntactic material in gap sites in non-adjacent conjuncts. Based on six diagnostics (clitic doubling, agreement, global case splits, disjoint marking, antipronominal contexts, and case assignment), we show that the prediction is not borne out. The results rather suggest that there is some (potentially silent) syntactic element merged in these positions, as postulated in all previous approaches to sharing constructions.
Format: [ pdf ]
Reference: lingbuzz/008514
(please use that when you cite this article)
Published in: To appear in Proceedings of CLS 60
keywords: atb, rnr, empty elements, syntax
Downloaded:422 times

 

[ edit this article | back to article list ]