Use this URL to cite or link to this record in EThOS:
Title: Accepting preposition-stranding under sluicing cross-linguistically : a noisy-channel approach
Author: Molimpakis, Emilia
ISNI:       0000 0004 8508 0395
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
This thesis investigates the representation and processing of sluicing, a type of ellipsis where an interrogative CP is reduced to its initial wh-element (the remnant), e.g. Mary danced with someone, but I can't remember (with) who. It is debated whether remnants from within a PP (with who) must appear with this P or whether they can appear without it (`Pstranding'). Existing theoretical literature (Merchant, 2001; a.o.) argues that only languages allowing overt CPs to move wh-elements without their embedding P will allow P-stranding remnants (P-Stranding Generalisation/PSG). Anecdotally, many languages appear to defy this pattern, allowing P-stranding remnants despite disallowing P-stranding overtly. None of these examples, however, are supported by adequate experimental evidence, nor o er a cross-linguistically generalisable explanation. This thesis addresses both these issues. Novel large-scale acceptability data show that both Greek and German, previously proposed robust PSG-examples, do indeed defy it. This behaviour is explained by proposing ellipsis is a type of `noisy channel' (Shannon, 1948; Gibson, Bergen & Piantadosi, 2013), through which the parser must estimate the probability of the intended (elided) message. The parser simultaneously considers the prior likelihood of the intended message (a remnant as part of a full PP) as well as the likelihood of this message being corrupted through `noise' (a deleted P). P-stranding is thus considered a form of deletion, given deletion has been shown to be a likely corruption in noisy channels. A series of reading time studies aimed at supporting this noisy channel model in online processing found results overall consistent with this approach, but also discovered previous work on the processing of sluicing was inaccurate in concluding its active prediction by the parser. Collectively, the work argues for a theory of sluicing involving syntactic structure at the e-site together with sluicing being treated as a noisy channel by the parser.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available