Abstract concepts like "freedom" are central to human interaction. Yet, conceptual abstractness poses a unique conversational challenge: lacking direct perceptual referents, abstract word meanings are often indeterminate. While cognitive linguistics documents individual-level difficulty in processing abstract words, and recent models of conversational dynamics identify mechanisms for sustaining conversations, we still lack an integrated account of how conceptual features like abstractness impact real-world dialogue. This project aims to develop a quantitative model of joint meaning-making in conversation. I propose novel two-person communication games to test whether negotiating abstract concepts increases conversational effort and social connection, produces conventional or creative outcomes in verbal (WP1) and material modalities (WP2), and leads speakers to adapt to each other's knowledge (WP3). Across WPs, I will investigate how mentalizing ability modulates joint meaning-making processes. By drawing on insights from experimental cognitive linguistics, social cognition, conversation analysis, and computational modeling, this research will advance theories of concepts and dialogue by revealing how word meanings emerge dynamically through communicative interaction.