The Geometry of Consciousness: Informational Dynamics, Intrinsic Time, and Decision-Theoretic Sufficiency A Unified Dissolution of the Externalist Impasse, with Formal Verification and Cross-Domain Validation

This paper advances a principled dissolution of the externalist impasse in philosophy of mind through five interlocking contributions. First, both physicalist and non-physicalist theories of consciousness are shown to share a common architectural defect: they posit mind-independent primitives — matter, external time, entropy — and invoke those same primitives to close the explanatory gaps they generate. Second, three system-internal functionals — the Information Coupling Functional (ICF), intrinsic time via Wasserstein arc-length, and the Predictive Accuracy Functional (PAF) — dissolve the three central impasses that this circularity produces: compositional emergence, temporal persistence, and representational truth. Third, the Internal Sufficiency Theorem establishes formally that external causal structure is provably redundant: once internal dynamics are fixed, its addition produces no difference in any observable, inferential, or decision-theoretic outcome. Fourth, the Decision Invariance Theorem demonstrates that external anchoring is action-theoretically null — for any admissible policy, expected utility is invariant under the addition or removal of external causal structure — closing the gap between the epistemological, metaphysical, and practical dimensions of the externalist claim simultaneously. Fifth, conceivability arguments against internalism, including the zombie argument and Kripke's necessity objection, are shown to presuppose that physical description exhausts the relevant structural description of consciousness; this presupposition fails under the informational framework, collapsing the arguments into evidence for descriptive incompleteness rather than ontological dualism. All three functionals are numerically validated against analytic benchmarks, and the two central theorems are machine-verified in Lean 4 / Mathlib4. The framework is formally grounded, empirically tractable, computationally implementable, and resistant to the canonical objections that have impeded progress in this domain for four decades.