DiSCourse Seminar with Nicola Brocca
29 November 2024, 12:00 (CET), hybrid
Digital Science Center, Innrain 15, 1st floor, Open Space Area or Big Blue Button
DiSCourse - The Digital Science Seminar Series on:
LLMs and Pragmatics: Insights from Linguistics and Language Education
The advent of Large Language Models (LLMs) is transforming the demands and priorities in language education. While formal aspects of written language can increasingly be delegated to AI, other competencies—such as (meta)pragmatic awareness—are becoming essential. These skills include identifying manipulative intent and critically evaluating LLM outputs. This presentation shares a study investigating implicit persuasive techniques in political discourse, comparing authentic politician speeches with AI-generated imitations created through ChatGPT. The analysis centers on “non-bona fide” presuppositions—subtle linguistic cues that may imply manipulative content. Key questions include how closely ChatGPT’s generated texts mirror real politicians’ linguistic strategies and whether the AI’s discourse functions align with those of the original texts. The presentation will further explore the potential of LLMs in language education to strengthen (meta)pragmatic competencies, supported by expirical explorations in (foreign) language classes.
Nicola Brocca, University of Innsbruck, Department of Subject-Specific Education
Nicola Brocca is a Postdoctoral University Assistant at the University of Innsbruck, specializing in the use of technology to enhance foreign language learning in task-based contexts. His primary research interests encompass learning in telecollaborative environments, the development of digital tasks, and fostering pragmatic awareness in teacher education. He also leads the LadderWeb project, which aims to create an application designed to advance pragmatic annotation.