Abstract
Posłajko?s ?Unreal Beliefs? (2024) provides a lucid and compelling overview of competing ontological views on belief, and his minimal anti-realist proposal is well-argued and persuasive. This article calls into question a less central thesis of the book: that ordinary speakers see beliefs as concrete, real, internal causes of behaviour. I offer a plausible alternative explanation of why everyday discourse appears to treat beliefs as discrete, concrete causes of behaviour. While endorsing Posłajko?s core anti-realist insights, I show that folk attributions of belief arise from a communicative necessity to translate holistic, nonverbal mindreading into discrete linguistic units. This translation does not reflect a genuine ontological commitment; it simply facilitates the exchange of information about one?s own and others? perspectives. Philosophical reconstruction then mistakes these linguistic shortcuts for evidence of folk realism, leading to a theoretical ontology that exceeds the minimal metaphysical stance of everyday speakers. By distinguishing between holistic mindreading, language-driven segmentation, and academic reification, I propose a stance aligned with minimal anti-realism that does not aim to decide the question so much as to move the burden of proof.