Speech preparation in conversational turn-taking

• Applicant: Professorin Dr. Christine Mooshammer
• Subject Area: General and Comparative Linguistics, Experimental Linguistics, Typology, Non-European Languages
• Term: since 2026
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 577415247

In everyday conversation, speaker alternation occurs rapidly, with gaps between turns of about 200-300 ms, much faster than reaction times in laboratory naming studies (around 600 ms). To achieve these rapid turn transitions speakers must plan their next utterance during their part-ner's ongoing turn, while also estimating when it will conclude. In this project we will investigate behavioural cues associated with the onset of turn planning and predicting the end of the current speaker’s turn. Until now, the time between the current speaker’s turn end and the start of the next speaker’s turn in conversations, known as the floor transfer offset (FTO), has been meas-ured using the acoustic signal. The same is true for reaction time (RT, measured from trigger onset to response onset) in naming experiments. However, using these acoustic measures to index speech planning can be inaccurate and misleading, as speech preparation begins well be-fore the acoustic onset, with an inhalation (for longer utterances) and the initiation of the articula-tory gestures to produce the initial speech sounds. The aim of this project is to disentangle plan-ning and execution in speech preparation in order to investigate the time course of speech initia-tion. Simultaneous recordings of respiratory activity with Respiratory Inductance Plethysmogra-phy (RIP) and of speech movements with Electromagnetic Articulography (EMA) in a sophisti-cated experimental design will provide more precise timing measures of the behavioural compo-nents of speech preparation, allowing us to better understand how turn-taking works. To this end, we will contrast respiration and speech organ movements in highly controlled naming exper-iments with two spontaneous dialogue scenarios. By separating speech planning from execution, the project aims to provide more accurate estimates of RT and FTO in speech tasks of increas-ing interactivity. The results will thus inform existing psycholinguistic models of turn-taking, and enhance our knowledge of the temporal organisation of speech preparation.

• DFG Programme: Research Grants
• International Connection: Sweden, USA
• Cooperation Partners: Professor Dr. Mattias Heldner; Professorin Dr. Jelena Krivokapic; Dr. Mark Tiede, Ph.D.