Expectation Violations as an Effective Alternative to Complex Mentalizing in Novel Communication

Abstract

Effective communication in the absence of a shared language is a fundamental challenge, often addressed through complex cognitive mechanisms such as Theory of Mind, which allows individuals to infer others intentions and beliefs. However, this process is cognitively demanding and may not always be necessary. In this study, we propose that a more parsimonious cognitive mechanism--expectancy violations--can serve as an efficient alternative for communication in novel interactions. We tested this in the Tacit Communication Game, where we simulated Sender behavior using four computational models: the Surprise model based on expectancy violations and three levels of Theory of mind. After human Receivers interacted with these simulated Senders, we assessed the effectiveness of communication by analyzing accuracy and reaction times. Our results revealed that Receivers paired with the Surprise model achieved accuracy rates comparable to those interacting with the most complex Theory of mind model and exhibited more human-like message patterns. Additionally, models associated with higher accuracies also resulted in faster reaction times, indicating a reduced cognitive load. These findings challenge the necessity of complex mentalizing strategies in novel human interactions and suggest that an intuitive mechanism of expectancy violation may be a more plausible cognitive mechanism, while also providing quick responses.