Active Inference dissolves the Moore paradox

Moore’s paradox, exemplified by statements such as “It is raining but I do not believe that it is raining”, highlights a tension between belief and assertion that is logically possible yet pragmatically puzzling. While extensively discussed in philosophy and cognitive science, the paradox has lacked a mechanistic explanation grounded in neuroscience or cognitive processing. Recent work suggests that the paradox is pragmatic or perspectival rather than semantic, emphasizing its relation to theory of mind. This paper proposes a mechanistic account using the framework of Active Inference and the Hierarchical Mechanistic Mind (HMM), in which the brain functions as a hierarchically structured predictive engine. From this perspective, Moorean statements generate a conflict in self-model precision weighting, simultaneously encoding high and low confidence in the same belief, which the HMM cannot reconcile without internal instability. This approach reframes the paradox not as a linguistic anomaly but as an illustration of the default workings of the mind. Crucially, it supports a reversal of the commonsense view of perception: experiences arise from high-precision beliefs rather than the other way around. The analysis underscores that belief is constitutive of conscious experience, offering a neurocognitive account that bridges philosophical insight and empirical modelling.