Learning Outcomes
Upon successful completion of the course, students will be able to:
Identify and explain the core concepts of formal theories of textual cohesion (SDRT, RST, LTAG).
Apply the criteria of these theories to annotate rhetorical relations in texts.
Evaluate the explanatory adequacy of alternative cohesion analyses using empirical corpus data.
Write R code to import, manage, and statistically analyse corpora annotated with textual relations.
Visualise discourse structures as graphs and compute key network-analysis metrics (centrality, communities) with appropriate R packages.
Design and document a small-scale research project that integrates theoretical cohesion analysis with network metrics, presenting their findings in both oral and written form.
Course Content (Syllabus)
The course addresses central issues in the field of text linguistics, focusing on textual cohesion and coherence. It begins by presenting the main formal discourse semantic theories, e.g., Segmented Discourse Representation Theory (SDRT), Rhetorical Structure Theory (RST) and Lexicalized Tree-Adjoining Grammar (LTAG)—and discusses key theoretical notions such as discourse structure, rhetorical relations and lexical anchors. These theories are then assessed against empirical data from text corpora, with special emphasis on studies that test the explanatory adequacy of the theoretically-driven discourse coherence relations. Finally, the course introduces core concepts of network analysis (graphs, centrality measures, communities) and demonstrates how they can be employed to model and visualise data from corpora annotated with discourse relations.
Keywords
Theories of Discourse Coherence, Coherence, Cohesion, Computational Textual Data Analysis, Data Analysis