This paper addresses the physical status of information at three boundary events: erasure, the onset of consciousness, and biological death. It asks whether information deletion is true annihilation, whether consciousness can be modeled as a thresholded information-condensation phase, and whether death is best understood as immediate informational disappearance or as a boundary transition in information flow. The framework begins from two physical commitments: information is physical, and, under a closed unitary description, the global distinguishability of encoded states is preserved. From these, the paper develops its central principle, "Erasure Is Transfer": when a record disappears from one local (subsystem) description, it is not destroyed globally but is represented in the environment or in system–environment correlations and, by the no-hiding theorem, remains recoverable from the complementary degrees of freedom. Deletion is therefore reframed as boundary movement within a declared record ledger rather than ontological loss. The paper is explicit that this scalar ledger is a bookkeeping representation, not a universal entropy theorem. The paper then applies this architecture to consciousness, modeling it — explicitly as a testable realization hypothesis rather than a derived result — as a phase transition: below a critical information-processing threshold the conscious phase is off, and at or above threshold it turns on. The control parameter is a single dimensionless ratio, x = RL·τ / Iₛys, comparing the Landauer-equivalent ideal-reset throughput available within a coherence interval to the system's information scale; the gate is C = Θ₊ (x − 1). Within this frame, reversible anesthesia and waking consciousness are treated as different regimes around one physical transition boundary rather than merely as graded changes in brain activity. The model deliberately does not classify seizure or overdrive; an upper threshold is retained only as an unestimated schematic extension. The proposed physical realization connects the transition to Fröhlich-type coherence protection, gamma-band dynamics, and microtubule-supported information organization, and it gives chemically diverse anesthetics a common macroscopic endpoint by having distinct molecular routes lower the same coherence interval. Several quantitative and retrospective anchors motivate the framework, each carried with an explicit evidence class rather than as confirmation. The paper reports an exploratory two-anchor cross-species gamma-frequency mapping, an illustrative structural-failure timing module for the dying-brain gamma interval, and a threshold ratio whose value falls near unity only at the upper end of the cited theoretical coherence range — the full range spans roughly an order of magnitude. Crucially, the paper argues against reading these as mutually reinforcing: the contributing quantities are parallel dependency paths that share assumptions and therefore cannot be counted as independent convergences. A dependency audit separates algebraically independent inputs, dependent checks, and the cited source range so that no item is double-counted as evidence. The simulation and retrodiction layer is presented as retrospective consistency, not final prospective proof. Its role is to show that one threshold-and-transfer architecture can organize several otherwise disconnected observations — the common endpoint of chemically diverse anesthetics, discontinuous rather than graded loss of reportable consciousness, death-associated gamma bursts, coherence-protection requirements, and cross-species neural scaling — while labeling each as retrospective, internal, prospective, or exploratory. Sensitivity analysis identifies whole-brain information capacity and in-vivo coherence protection as the quantities that dominate the threshold and that future experiments must measure directly; the paper specifies its falsification thresholds, including that a measured coherence interval below 10⁻⁷ s would reject the proposed microtubule realization at the nominal inputs. The conclusion is that information physics should treat "erasure" as a transfer problem and consciousness as a thresholded-condensation problem, so that anesthesia, waking consciousness, and biological death are not unrelated phenomena but different crossings or deformations of the same information-boundary structure. Paper 8 thereby sets out the axiomatic foundation for the later Information Physics Series — existence gate, information continuity, phase transition, and the typed loop linking mass–energy, Landauer-equivalent capacity, the consciousness threshold, and structured physical records — while keeping formal proof and prospective experimental test as explicitly future work. Keywords: information physics, erasure is transfer, consciousness, information conservation, unitarity, no-hiding theorem, Landauer principle, Fröhlich condensation, anesthesia, gamma oscillations, biological death, phase transition, coherence protection, information continuity, dying brain, quantum thermodynamics, falsifiability, evidence classification
Taekyung Lee (Tue,) studied this question.