Hyperraising and copy raising are structurally different: experimental evidence fromSerbian

Authors

  • Aljosa Milenkovic Harvard University

DOI:

https://doi.org/10.3765/plsa.v10i1.5938

Keywords:

experimental syntax, cross-clausal dependencies, hyperraising, copy raising, Serbian

Abstract

This paper documents a hyperraising-to-subject construction in Serbian, previously argued not to be possible in the language. For a subset of Serbian speakers, the raising verb delovati ‘seem’ optionally allows A-movement from the finite complement clause into the matrix clause, resulting in φ-agreement on the matrix verb. A rating experiment with 835 native Serbian speakers, 519 of whom allow hyperraising, found that hyperraising is structurally distinct from copy raising. The experimental results suggest that hyperraising is an A-movement configuration, highly sensitive to movement constraints, such as islands and embedded A-minimality. By contrast, copy raising was found to be generally insensitive to movement constraints, which indicates that it is a non-movement configuration similar to prolepsis. This finding challenges the unified base-generation account of hyperraising and copy raising, which was proposed to eschew the locality issues inherent to the canonical movement-based analyses of hyperraising.

Downloads

Published

2025-05-21

How to Cite

Milenkovic, Aljosa. 2025. “Hyperraising and Copy Raising Are Structurally Different: Experimental Evidence FromSerbian”. Proceedings of the Linguistic Society of America 10 (1): 5938. https://doi.org/10.3765/plsa.v10i1.5938.