The Predictive Mind and Its Myths: Metaphor, Narrative, and Ritual as Structural Necessities of Scientific Cognition

This article argues that three features commonly regarded as extraneous to scientific rationality — conceptual metaphor, mythological narrative, and institutional ritual — are functionally indispensable components of scientific cognition and practice. Drawing on the predictive processing (PP) framework — specifically hierarchical generative modeling, precision-weighting, and temporally deep prediction — we propose that these elements correspond to distinct strategies for managing prediction error at the cognitive, narrative, and institutional levels. We deploy PP not as a universal metaphysical principle but as a heuristic framework whose empirically constrained commitments generate testable claims about which metaphors prove scientifically productive, why scientific narratives acquire specifically mythological structure (defined by four necessary features: cosmogonic framing, agentive causation, eschatological orientation, and identity-constitutive function), and how institutional rituals relate to precision-weighting in social inference. We distinguish between minimal PP (a modeling toolkit with cortical implementation claims) and maximal PP (a grand unifying picture of mind), anchoring our argument in the former. Two extended case studies — Big Bang cosmology and gene-centered evolutionary biology — demonstrate the co-presence and functional interdependence of all three levels. The article engages critically with scientific realism, structural realism, Feyerabend's epistemological anarchism, and the post-2020 critical literature on the free energy principle. We conclude that acknowledging the mythopoetic dimensions of science does not license relativism but enables a more epistemically humble and self-correcting scientific practice. Keywords: predictive processing; free energy principle; scientific metaphor; myth and science; ritual in science; epistemology; philosophy of science; Feyerabend; Kuhn; cognitive science; precision weighting; structure-mapping; gene-centered view; Big Bang cosmology