TEXT Seminar Explores New Insights in Language, Data, and AI

The programme reflected the breadth of TEXT’s activities, spanning editorial practices, speech analysis, artificial intelligence, experimental platforms, and computational approaches to discourse.

Throughout the day, presentations demonstrated how linguistic data and digital methods can generate new knowledge about society, communication, and learning:

• Kenneth Enevoldsen and Yevhen Kostiuk presented analyses based on data from Sara og Monopolet, shedding light on social norms and values in a Danish context.
• Simon Aagaard Enni and Malthe Stavning Erslev examined how established editorial teams shape the AI-driven dissemination of encyclopedic content, using Lex – The Danish National Encyclopedia as a case.
• Agnes B. Nielsen and Mads Rosendahl Thomsen offered a deep dive into the rhetorical and linguistic functions of speeches, highlighting patterns that influence persuasion, tone, and audience engagement.
• Mia Jacobsen and Mads Rosendahl Thomsen introduced an effort to systematise how writing is understood across contexts, contributing to a shared framework for analysing and designing written communication.
• Mina Almasi focused on designing with large language models (LLMs), emphasising the importance of grounding such work in pedagogical principles to support meaningful learning experiences in practice.
• Arthur Hjorth presented Phasetophase, a platform that enables structured chatbot experiments and survey studies, making it easier to move from initial ideas to empirical testing in conversational settings.
• Finally, Márton Kardos, who start as a PhD student in September 2026, introduced his upcoming project on Computational Discourse Analysis.

The seminar highlighted the value of interdisciplinary collaboration and shared knowledge-building within TEXT. We look forward to building on these insights after the summer holiday.