Is Recursive Mindreading Really an Exception to Limitations on Recursive Thinking?

The ability to mindread recursively-for example, by thinking what person 1 thinks person 2 thinks person 3 thinks-is a prime example of recursive thinking in which one process, representation, or idea becomes embedded within a similar one. It has also been suggested that mindreading is an exceptional example, with five recursive steps commonly observed for mindreading, in comparison with just one or two in other domains. However, conceptual analysis of existing recursive mindreading tasks suggests that conclusions about exceptional mindreading are insecure. Revised tasks were devised to provide a more rigorous test of recursive mindreading capacity. Study 1 (N = 76) found significantly worse performance at level-5 recursive mindreading on the revised tasks (17% correct) compared with the original tasks (80% correct), and no effect of moderate financial bonuses for good performance. Study 2 (N = 74) replicated poor performance at level-5 recursive mindreading on the revised tasks (15% correct) in the absence of bonuses, but found better performance (45% correct) when participants were offered large bonuses for accuracy, encouraged to take as much time as needed, and assisted with a strategy for recursive reasoning. These findings suggest that, like recursive thinking in other domains, recursive mindreading is effortful and limited. We discuss how the proposed role for high levels of recursive mindreading in communication, culture, and literature might be reconciled with these limitations.

Automated summary

The ability to mindread recursively, thinking about others' thoughts in a recursive manner, is an example of how one thought or idea becomes embedded within another. It has been suggested that mindreading is an exceptional and complex process, involving multiple recursive steps. However, a conceptual analysis of existing mindreading tasks questions the exceptional nature of mindreading. Revised tasks were created to test recursive mindreading capacity more rigorously. Study 1 found significantly worse performance on the revised tasks compared to the original tasks, suggesting that recursive mindreading is effortful and limited. Study 2 replicated these findings and also found that incentives and assistance improved performance. These findings highlight the limitations of recursive mindreading and raise questions about its role in communication, culture, and literature.