Pseudoclefts: A Self-Paced Reading Study Investigating Ellipsis Processing

Abstract

Theoretical linguistic accounts concerning the nature of pseudocleft construc- tions have led to differing perspectives on their underlying mechanisms. Specifically, the coreferential properties of pseudoclefts have led to a theoretical divide between syntactic-based accounts and semantic-based accounts. The theoretical contention surrounding pseudoclefts has led to a lack of empirical research concerning their pro- cessing. This thesis argues that there is strong evidence from the literature to suggest that pseudoclefts, more specifically a sub-type of pseudoclefts known as specificational pseudoclefts, are best viewed through the lens of a syntactic-based ellipsis account. I present three arguments for an ellipsis-based account of specification pseudoclefts: (1) ellipsis-based accounts provide a more parsimonious explanation for their coref- erential properties, (2) Ross (1972) and Schlenker (2003)’s conceptual argument for specificational pseudoclefts as question-answer pairs (QAP) places the burden of proof on any theory that does not posit a QAP analysis, (3) Hirsch (2017) arguements for the existence of VP-ellipsis in pseudoclefts. I then present an experiment that uses a self-paced reading task to investigate the processing of pseudoclefts through the lens of an ellipsis analysis. I hypothesized increased reaction times at the ellipsis sites in specificational pseudoclefts, but not in their counterpart predicational pseudocleft constructions that do not possess ellipsis. There was no significant difference in the reaction times across the conditions. It is unclear if the lack of effect was due to the experimental methodology, the lack of control for the participant’s environment, or the potential lack of ellipsis in the pseudocleft constructions. However, this work provides a foundation for future research to investigate the processing of pseudoclefts and the potential for using pseudocleft paradigms to understand language processing.