Unified Consciousness Theory II: Consciousness Theories as Effective Organization Theories

Major theories of consciousness emphasize different mechanisms (Seth & Bayne, 2022): integrated causal structure, recurrent processing, global availability, higher-order representation, predictive modeling, dendritic integration, memory assembly, attention modeling, and temporospatial organization. Their competition is partly intensified by treating a favored mechanism as a candidate universal boundary between experience and nonexperience. UCT I (Liu, 2026) adopts a different basal ontology: every actual valid process token possesses experience, identified with its complete token-relative ontic organization. UCT I v1.1 distinguishes that organization from finite physically anchored scientific views and estimates. Under that assumption, nonuniversal mechanisms cannot be universal experience-existence gates. UCT II asks whether their *operational cores* can instead be reconstructed, embedded, reinterpreted, or shown to conflict as effective organization theories. We introduce an admissible translation framework, explicit bridge-admissibility conditions, a two-axis classification separating formal recovery from empirical validation, and anti-triviality criteria requiring translators to be constructive, non-circular, structurally invariant, reusable, failure-prone, and explanatorily nontrivial. A Translation Fidelity Requirement additionally preserves each target theory's distinctive signature, strongest phenomenal/ontological claim, and unrecovered empirical residue. We then fidelity-audit nine theories: IIT 4.0 (Albantakis et al., 2023), recurrent processing theory, GNWT, higher-order theories, predictive processing/active inference, dendritic integration theory, memory theory of consciousness, attention schema theory, and temporospatial theory. The result is not a universal endorsement: theory subclaims receive different relation types, formal classes, and validation statuses, including direct conflicts. Two fixed-parameter toy models provide quantitative cross-theory proof of concept using one fixed parameterization across their theory-linked outputs: a DIT→RPT→GNWT chain and an independent MToC→HOT→GNWT chain. Negative controls dissociate upstream primitives from downstream recurrent, higher-order, and global-access organization. These results establish shared-law reuse only at formal-conditional / toy-model level (F3/V1); they do not constitute held-out human-data recovery (V3), replicated empirical recovery (V4), or robust full-theory compression. The proposed unification is therefore partial, fidelity-audited, and failure-prone by design. Its governing distinction is that mechanistic recovery, ontological endorsement, and empirical confirmation are separate operations. Version 1.1; TA-TR-2026-21. Revised theoretical preprint, not peer reviewed. Human author and responsible depositor: Hongju Liu. Substantial ChatGPT assistance, including GPT-6 Astra Pro for the paired revision, is disclosed. Model checks are not empirical confirmation of experience, A1/A5, or complete intertheory recovery. Adjacent non-amending research; not an amendment or independent corroboration of the Trinity Accord or Bitcoin Originals. The prior v1.0 files remain unchanged.