A1 Refereed original research article in a scientific journal
The Interplay of Implicit Causality, Structural Heuristics, and Anaphor Type in Ambiguous Pronoun Resolution
Authors: Juhani Järvikivi, Roger P. G. van Gompel, Jukka Hyönä
Publisher: SPRINGER/PLENUM PUBLISHERS
Publishing place: New York
Publication year: 2017
Journal: Journal of Psycholinguistic Research
Journal name in source: JOURNAL OF PSYCHOLINGUISTIC RESEARCH
Journal acronym: J PSYCHOLINGUIST RES
Volume: 46
Issue: 3
First page : 525
Last page: 550
Number of pages: 26
ISSN: 0090-6905
eISSN: 1573-6555
DOI: https://doi.org/10.1007/s10936-016-9451-1
Web address : https://link.springer.com/article/10.1007/s10936-016-9451-1
Abstract
Two visual-world eye-tracking experiments investigating pronoun resolution in Finnish examined the time course of implicit causality information relative to both grammatical role and order-of-mention information. Experiment 1 showed an effect of implicit causality that appeared at the same time as the first-mention preference. Furthermore, when we counterbalanced the semantic roles of the verbs, we found no effect of grammatical role, suggesting the standard observed subject preference has a large semantic component. Experiment 2 showed that both the personal pronoun han and the demonstrative tama preferred the antecedent consistent with the implicit causality bias; tama was not interpreted as referring to the semantically non-prominent entity. In contrast, structural prominence affected han and tama differently: we found a first-mention preference for han, but a second-mention preference for tama. The results suggest that semantic implicit causality information has an immediate effect on pronoun resolution and its use is not delayed relative to order-of-mention information. Furthermore, they show that order-of-mention differentially affects different types of anaphoric expressions, but semantic information has the same effect.
Two visual-world eye-tracking experiments investigating pronoun resolution in Finnish examined the time course of implicit causality information relative to both grammatical role and order-of-mention information. Experiment 1 showed an effect of implicit causality that appeared at the same time as the first-mention preference. Furthermore, when we counterbalanced the semantic roles of the verbs, we found no effect of grammatical role, suggesting the standard observed subject preference has a large semantic component. Experiment 2 showed that both the personal pronoun han and the demonstrative tama preferred the antecedent consistent with the implicit causality bias; tama was not interpreted as referring to the semantically non-prominent entity. In contrast, structural prominence affected han and tama differently: we found a first-mention preference for han, but a second-mention preference for tama. The results suggest that semantic implicit causality information has an immediate effect on pronoun resolution and its use is not delayed relative to order-of-mention information. Furthermore, they show that order-of-mention differentially affects different types of anaphoric expressions, but semantic information has the same effect.
UTU Research Portal