During social interactions individuals tend to adjust aspects of their verbal and nonverbal behaviour towards or away those of their interlocutors in communicative. While phonetic convergence (speaker becoming more similar) is a good indicator of cooperation, increased acoustic similarity between interlocutors' acoustics may interfere with their auditory recognizability. In this project we investigate how the trade-off is achieved between sounding cooperative through convergence and being recognized through divergence or maintenance. In game scenario where recognition is at stake we will investigate how the dynamics of convergence or individualization change over time ((More stability/divergence at the beginning of the conversational exchanges? Convergence increase with increased exposure?) and what effect they have on voice recognition. A labelled corpus in Bernese Swiss German is already available.