Coherence Marking, Comprehension and Persuasion. On the processing and representation of discourse

Abstract

This dissertation focuses on the role of coherence marking in discourse. If we take as an example a 1914 advertisement for Ford, that states: Buy it because it’s a better car. The relation that exists between the first segment [buy it] and the second segment [it’s a better car] is a claim-argument relation, in this case explicitly signalled by the coherence marker because. How does such a coherence marker influence the mental representation that a reader constructs from a text? The six experiments in this dissertation show that coherence marking affects the reader in different ways: coherence marking can influence text comprehension, appraisal and persuasion. Coherence marking improves text comprehension for readers who do not have much prior knowledge about the text topic. For readers who have more prior knowledge about a topic, a more implicit text leads to optimal comprehension. Also, coherence marking positively influences readers’ opinions about the text and text quality. Finally, it becomes apparent that to study effects of coherence marking on persuasion, it is necessary to make a distinction between the marking of objective relations (relations that exist in external reality, such as cause-effect) and subjective relations (relations that are constructed by the speaker or writer, such as the claim-argument relation in the Ford advertisement). Empirical results show that objective marking can have a positive effect on persuasion, whereas subjective marking can cause a so-called forewarning effect: when readers recognize the attempt to influence them, they build resistance to it. The research in this dissertation shows the importance of a subtle text characteristic such as coherence marking. It combines insights from text linguistics, discourse processing, and social psychology, and should therefore be of interest to researchers in any of these domains.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 18 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.