Earlier NLP studies on framing in political discourse have focused heavily on shallow classification of issue framing, while framing effect arising from pragmatic cues remains neglected. We put forward this latter type of framing as “pragmatic framing”. To bridge this gap, we take presupposition-triggering adverbs such as ‘again’ as a study case, and quantitatively investigate how different German newspapers use them to covertly evoke different attitudinal subtexts in their report on the event “European Refugee Crisis” (2014-2018). Our study demonstrates the crucial role of presuppositions in framing, and emphasizes the necessity of more attention on pragmatic framing in the research of automated framing detection.

This paper describes the submission of the team KonTra to the CMCL 2021 Shared Task on eye-tracking prediction. Our system combines the embeddings extracted from a fine-tuned BERT model with surface, linguistic and behavioral features, resulting in an average mean absolute error of 4.22 across all 5 eye-tracking measures. We show that word length and features representing the expectedness of a word are consistently the strongest predictors across all 5 eye-tracking measures.