Research on mindreading has been dominated by questions about concepts and structures, but has neglected mindreading performance. We argue that mindreading performance requires generating and selecting mental states that are plausible and appropriate. Mindreading is plausible and appropriate if it can be recognised as such by other people, and so is essentially a joint activity. It is an asynchronous joint activity because, once learned, it can be performed alone. We explain a swath of findings about individual differences in mindreading that are otherwise puzzling, and provide a framework for developing interpretable and psychometrically robust measures to study this variation.
Apperly et al. (Tue,) studied this question.