Neural mechanisms of communication in conversational interactions

• Applicant: Professor Dr. Tobias Reichenbach
• Subject Area: Otolaryngology, Phoniatrics and Audiology
• Term: since 2026
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 576735417

Oral communication is of central importance to human interactions and to our social and professional lives. Speech comprehension required for oral communication has, however, so far mostly been studied in isolated, controlled laboratory conditions typically executed on single subjects. These investigations have often used speech material that is unrealistic for real-word conversations, did not incorporate the participants' ability to engage in social interactions, and often used unrealistically low signal-to-noise ratios when considering background noise. This study seeks to investigate naturalistic conversations between pairs of participants that are more typical of real-world interactions. We will investigate which conversation factors affect the perceived success of the conversation, as well as underlying neural mechanisms. Regarding neural mechanisms, we will focus on the synchrony in brain activity between the two participants, in particular in the delta and theta frequency bands. These two frequency bands have been found to play crucial roles in speech comprehension, in particular with respect to neural tracking of amplitude fluctuations in speech, which can in turn contribute to synchronizing neural activity between conversation partners. We will quantify both neural synchrony and neural speech tracking and determine how they relate to conversation success. Naturalistic conversations can involve the Lombard effect – an automatic and unintentional change in articulation, such as enhanced voice intensity and raised fundamental frequency, in the presence of background noise. In the first part of the study, we will focus on naturalistic conversations that, through a special experimental design, exclude the Lombard effect. We will thereby (1a) determine the dependency of conversational success on the type of conversation, as well as on the type and intensity of background noise, and (1b) identify neural markers of conversational success. The second part of the study will then include the Lombard effect. We will (2a) quantify the Lombard effect in naturalistic conversations with background noise and its influence on conversation success, as well as (2b) determine the extent to which the markers of conversational success indentified in (1b) can be applied to conversational situations in which the Lombard effect occurs. Future applications of the proposed work lie e.g. in the testing of algorithms for hearing prosthetics. Conversations in background noise can be particularly difficult for people with hearing impairment, even when they wear hearing aids. Algorithms for noise reduction are often tested in simplified laboratory scenarios that do not generalize well to the real world. We hope that the naturalistic conversations studied here and the identified neural markers will aid in the development of future noise reduction techniques that will help hearing-aid users to participate more readily in real-world conversations.

• DFG Programme: Research Grants