Dependency Locality and Neural Surprisal as Predictors of Processing Difficulty: Evidence from Reading Times

Neil Rathi


Abstract
This paper compares two influential theories of processing difficulty: Gibson (2000)’s Dependency Locality Theory (DLT) and Hale (2001)’s Surprisal Theory. While prior work has aimed to compare DLT and Surprisal Theory (see Demberg and Keller, 2008), they have not yet been compared using more modern and powerful methods for estimating surprisal and DLT integration cost. I compare estimated surprisal values from two models, an RNN and a Transformer neural network, as well as DLT integration cost from a hand-parsed treebank, to reading times from the Dundee Corpus. Our results for integration cost corroborate those of Demberg and Keller (2008), finding that it is a negative predictor of reading times overall and a strong positive predictor for nouns, but contrast with their observations for surprisal, finding strong evidence for lexicalized surprisal as a predictor of reading times. Ultimately, I conclude that a broad-coverage model must integrate both theories in order to most accurately predict processing difficulty.
Anthology ID:
2021.cmcl-1.21
Volume:
Proceedings of the Workshop on Cognitive Modeling and Computational Linguistics
Month:
June
Year:
2021
Address:
Online
Editors:
Emmanuele Chersoni, Nora Hollenstein, Cassandra Jacobs, Yohei Oseki, Laurent Prévot, Enrico Santus
Venue:
CMCL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
171–176
Language:
URL:
https://aclanthology.org/2021.cmcl-1.21
DOI:
10.18653/v1/2021.cmcl-1.21
Bibkey:
Cite (ACL):
Neil Rathi. 2021. Dependency Locality and Neural Surprisal as Predictors of Processing Difficulty: Evidence from Reading Times. In Proceedings of the Workshop on Cognitive Modeling and Computational Linguistics, pages 171–176, Online. Association for Computational Linguistics.
Cite (Informal):
Dependency Locality and Neural Surprisal as Predictors of Processing Difficulty: Evidence from Reading Times (Rathi, CMCL 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.cmcl-1.21.pdf