Dyadic social interaction is a complex coordination task involving a large number of interconnected variables. Previous research has shown that metastability -- persistence for an extended, but impermanent, period of time in a non-stable state of a system -- can be a useful lens for understanding what makes an interaction successful. However, this framework has thus far only been applied to para-conversational signals like heart rate and prosody -- not to the semantic content of a conversation. Here, we present pink noise analysis of semantic trajectories as a metric for conversational success and apply this technique to a large open conversation dataset. Our results demonstrate that pink noise in a conversation predicts a host of variables representing participants' perception of conversation quality. These results have implications for optimizing a whole host of difficult dyadic conversations -- like those between political partisans -- and human-computer interactions, with applications for improving large language models' adaptability.

Authors:

Kathryn O'Nell: Dartmouth College; Emily S. Finn: Dartmouth College