Predictability in reference processing: Effects of implicit causality and coherence relations 

 

Project supported by a grant of the Ministry of Research and Innovation, CNCS – UEFISCDI.

Project: PN-III-P4-ID-PCE-2020-0731

Principal investigator: Dr. Sofiana Iulia Lindemann (nee Chiriacescu)

 

Funding period: 36 months, 2022-2024

About the project

 

Reference resolution and related phenomena are highly complex tasks, "as they require the integrated application of syntactic, semantic, pragmatic, discourse and world knowledge. Identifying the factors that contribute to successful referent processing and explaining how theses factors interact within a language and cross-linguistically is an important yet very difficult endeavour. The main aim of the project is to explore the semantic structure of verbs, which constraints next-mention biases (e.g. subject vs. object next-mentions), pronominalization rates and the use of particular sentence continuations in terms of coherence relations. To this end, we experimentally test the impact of these factors in local and global discourse through behavioural measures by means of sentence-continuation studies, questionnaires and corpus investigations. 

 

We investigate reference resolution preferences for different types of referring expressions in Romanian and compare them to those attested for other languages. We intend to gain new insights into these issues, as well as more general questions concerning reference resolution and explore grammatical biases and coherence-driven biases from both the perspective of the speaker and from the perspective of the hearer. On a more general level, the project aims at developing a discourse model that will allow us to describe some key features of local and global discourse processing.

Publish modules to the "offcanvas" position.