Action Descriptions in Conflictual Conversations: A Study of Turn-Taking and Perception, Application in Contact Centers

This study investigates how action descriptions and explanatory metaphors impact the perception of conflictual conversations in contact centers. Using a corpus of 30 annotated call center conversations, we analyze the effects of these linguistic strategies on turn-taking dynamics and client reactions. Our approach combines Conversation Analysis (CA) with computational linguistic methods, leveraging Natural Language Processing (NLP) techniques to classify action verbs and evaluate their influence on conversational flow. Findings indicate that action descriptions, which involve the advisor explicitly stating what they are doing or what the client will do next, are associated with more positive, and less conflictual responses. In contrast, explanatory metaphors, particularly those used by the advisor to clarify company processes, require higher cognitive effort and tend to elicit negative reactions. This study provides empirical validation of best practices in contact centers, confirming that structuring agent responses around concrete actions rather than abstract explanations enhances customer engagement, reduces call duration, and improves conflict resolution. These findings contribute to both applied linguistics and customer service training, offering new insights into how linguistic choices shape conversational effectiveness.