Laughter in Conversation: Form, Function and Multimodality

• Applicant: Professorin Dr. Petra Wagner, since 9/2024
• Subject Area: Applied Linguistics, Computational Linguistics
• Term: since 2021
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 461442180

Laughter is one of the most often occurring non-verbal phenomena in spontaneous interactions. Despite its frequency and the interest it has received in the research community, there are still a number of aspects which are not fully understood and which might hold back the improved implementation of laughter in speech technology applications. Continuing the approach established in the LAUGH project, we will examine three different aspects of laughter: its form, its function and its multimodal expression, by considering these aspects not only by themselves, but also through the interactions between them. Each of the three parts of the current project will address existing gaps in our understanding of laughter in conversation. First, we will perform a quantitative investigation of laughter functions employed in a large corpus of dyadic interactions, in order to establish their frequency of occurrence in spontaneous speech. We will also analyze the apparent under-determined nature of laughter, by focusing on the effect that the role of the person in conversation (speaker, listener, observer) might have on the perceived laughter function. In terms of laughter form, although a large body of literature examined its acoustic and temporal realization, little is known about the configuration of the articulators during laughter. We will determine the position of the upper tract articulators (tongue, jaw, lips) for three types of laughter – mirthful, social and volitional – comparing them with positions attained during speech. We aim to further apply these findings, by using articulatory measurements, in conjunction with acoustic information, for the automatic detection of laughter from speech. Next, we will investigate the multimodal expression of laughter, by looking at one type of visual expression often found in relation to laughter, smiles, and in particular how these two phenomena are distributed in spontaneous interactions. For a better understanding of the connection between laughter and smiles, we will also determine whether similarities between them exist in terms of articulatory configurations. Finally, in order to be able to answer these questions, a corpus containing casual conversations and corresponding read speech, together with articulatory measurements for both types of speech will be collected and will be made available for research purposes.

• DFG Programme: Research Grants
• Ehemaliger Antragsteller: Dr. Ioan Bogdan Ludusan, until 9/2024