CLASP
The Centre for Linguistic Theory and Studies in Probability

Social XAI: Explaining as a Co-Constructive Process

Abstract

Explainable AI (XAI) works on providing explanations that justify a model’s behaviour or decision. But what is an explanation worth if the user it is meant for cannot understand it? “Social XAI”, a recent interdisciplinary offshoot at the intersection of XAI, dialogue research, and the social sciences (Rohlfing et al. 2021, 2026), shifts the focus to the practice of explaining: the dialogic process through which explanations and their understanding are co-constructed between explainer and explainee. In this talk, I will present empirical and computational work on human explanatory dialogue, drawing on corpora of people explaining board games and scientific topics. Rather than treating understanding as a property of an explanation, I will examine how it is contingently achieved in explanatory interaction. I will look at how it is signalled through multimodal feedback, and at how dialogue partners actively adapt the explanation process across different levels of knowledge asymmetry.

Bio: Hendrik Buschmeier is an Assistant Professor (Juniorprofessor) for Digital Linguistics at the Faculty of Linguistics and Literary Studies at Bielefeld University, Germany. His research group, the Digital Linguistics Lab, works on empirical and computational modelling of dialogue and conversational interaction, whether between humans or between humans and artificial agents. As a principal investigator in the DFG-funded Collaborative Research Centre TRR 318 “Constructing Explainability”, his work focuses on how explanations unfold as dialogic processes and what makes them successful.