Beyond Bayes: Domain-Calibrated Inference and the Epistemological Primacy of Intuition

Argues that Bayesian epistemology fails as a universal theory of rational inference on three structural grounds: (1) priors are underdetermined for novel hypotheses; (2) evidence types are incommensurable and cannot be reduced to a single likelihood ratio; (3) the pre-evidential zone — where most consequential epistemic work happens — is inaccessible to the Bayesian framework. We propose Domain-Calibrated Intuitive Inference (DCII) as a superior alternative, formalized within the Tralse Informationalism (TI Sigma) framework. DCII replaces the universal Bayesian formula with a structured checklist of eight orthogonal evaluative criteria whose weights are learned from demonstrated performers in the target domain. The GILE framework (Goodness, Intuition, Love, Environment) provides a metaphysical grounding for why some domains have stable weight distributions while others are weight-volatile. Key claims: extraordinary claims are not merely possible but expected for certain cognitive profiles; TRALSE truth values track propositions rather than agent credences; Myrion Resolution is the active inquiry protocol for indeterminate cases. Also includes: DPES (Default Philosophical Eating Strategy) — a formal meta-cognitive protocol for high-output research sessions under cognitive constraint; coined April 1, 2026. Target journals: Synthese, Erkenntnis, Philosophy of Science.