Gaps and Contradictions

Status: Living catalogue — refereed, cumulative, open to revision. Last updated: 2026-06-08

This page registers the principal gaps, tensions, and apparent contradictions among the accepted frameworks of physics. Its central methodological commitment — see EPISTEMICS.md — is to distinguish four very different things that are routinely conflated in informal discourse:

`kind`	Meaning	Status of the underlying theories
logical-inconsistency	A contradiction derivable within a single, fixed, consistent axiom system (rare; almost never what is actually meant).	At least one framework, as stated, is internally broken.
domain-of-validity mismatch	Two frameworks are each internally consistent but rest on incompatible idealizations; the clash appears only where their domains are forced to overlap.	Both frameworks fine in their regimes.
unsolved-but-consistent problem	A well-posed question with no accepted answer, but no contradiction.	Frameworks consistent; we lack a derivation/mechanism.
conceptual tension	A clash of interpretation or principle, with no operational/empirical contradiction.	Empirically fine; the dispute is about meaning or mechanism.

Every entry carries a REFEREE VERDICT distilling an adversarial review of the original claim: a refined statement, the corrections applied, a recommended epistemic tag, and a severity. Where the referee judged the framing materially wrong (e.g., a kind mislabel), the refined statement governs and supersedes the headline. No entry below was rejected outright (keep=false); the single entry whose internal logic was found unsound (GC-17) is retained but corrected, and its erroneous sub-claim is quarantined in Corrected / not actually contradictions.

Cross-references: OPEN_PROBLEMS.md (the forward-looking research agenda), THEORY_MAP.md, UNIFICATION_LANDSCAPE.md, ASSUMPTIONS_LEDGER.md, CONSTANTS_AND_SCALES.md.

Index

ID	Title	`kind`	Tag	Severity
GC-1	Background dependence vs. background independence (QFT/GR)	domain-mismatch + OPEN foundational	ESTABLISHED/OPEN	major
GC-2	Perturbative non-renormalizability of quantized gravity	established theorem / open UV	ESTABLISHED/OPEN	minor
GC-3	Cosmological-constant problem	domain-mismatch / fine-tuning	OPEN	minor
GC-4	Black-hole information paradox	domain-of-validity tension	ESTABLISHED/OPEN	minor
GC-5	AMPS firewall puzzle	conceptual tension (rigorous no-go)	INFERENCE/CONTESTED	minor
GC-6	Measurement problem	structural incompleteness	ESTABLISHED/OPEN	minor
GC-7	Thermodynamic arrow vs. time-symmetric microlaws	established structure / open origin	ESTABLISHED/OPEN	minor
GC-8	Problem of time (Wheeler–DeWitt)	conceptual tension	OPEN/CONTESTED	minor
GC-9	Haag's theorem vs. the interaction picture	rigor gap / domain-mismatch	ESTABLISHED	minor
GC-10	Triviality / Landau poles vs. "renormalizable = fundamental"	established refutation / open SM-UV	ESTABLISHED/INFERENCE	minor
GC-11	GR singularities (geodesic incompleteness)	domain-mismatch	ESTABLISHED	minor
GC-12	Quantum nonlocality vs. relativistic locality	conceptual tension (no operational clash)	ESTABLISHED/CONTESTED	minor
GC-13	Holographic area-law vs. volume-extensive QFT Hilbert space	conceptual tension on ESTABLISHED parts	OPEN	minor
GC-14	Stone–von Neumann uniqueness vs. inequivalent representations	domain-mismatch	ESTABLISHED	minor
GC-15	Semiclassical gravity vs. quantum superposition of mass	domain-mismatch / breakdown	INFERENCE/CONTESTED	minor
GC-16	ER=EPR / "spacetime from entanglement" generality	conceptual tension / evidential gap	SPECULATIVE on INFERENCE base	minor
GC-17	Ensemble statistical mechanics vs. long-range gravitating systems	domain-mismatch	ESTABLISHED	minor (one excised sub-claim)

GC-1 — Background dependence vs. background independence: QFT and GR

Frameworks: QFT, GR, QM, Particle Physics / Standard Model. Kind: domain-of-validity boundary + OPEN foundational problem (explicitly not a logical inconsistency).

Precise statement. Standard Minkowski QFT presupposes a fixed, non-dynamical metric $\eta_{\mu\nu}$ . That metric anchors (i) the global Poincaré group, whose unitary irreducible representations classify particles via the Casimirs $P^2 = m^2$ and $W^2 = -m^2 s(s+1)$ (mostly-minus signature, $c=1$ ; the sign is convention-dependent) ESTABLISHED; (ii) global microcausality $[\phi(x),\phi(y)]=0$ for spacelike separation ESTABLISHED; and (iii) a Poincaré-invariant vacuum satisfying the spectrum condition ESTABLISHED. GR's defining feature is the opposite: diffeomorphism invariance / background independence, where the metric is dynamical, there is no prior geometry, and physical states are diffeomorphism-equivalence classes (Leibniz equivalence). ESTABLISHED

Technical root. In the ADM formulation the GR Hamiltonian is a sum of constraints, $H \approx 0$ , generating diffeomorphisms; there is no fixed causal structure to anchor microcausality, and "particle" loses observer-independent meaning even semiclassically (Unruh effect; no preferred vacuum in curved spacetime). The two frameworks treat the metric as fixed input vs. dynamical output. ESTABLISHED

Why it matters. This contrast is the deepest structural reason the naive union of the two frameworks fails; it underlies the problem of time (GC-8), the non-renormalizability of quantized gravity (GC-2), and the difficulty of even formulating "QFT on a superposition of spacetimes." See UNIFICATION_LANDSCAPE.md, OPEN_PROBLEMS.md.

Current status. Both frameworks are confirmed to extreme precision in their domains (electron $g\!-\!2$ ; binary-pulsar decay; gravitational-wave detections). They coexist successfully as (a) QFT on fixed curved backgrounds (Hawking, Unruh; algebraic/Haag–Kastler nets on curved spacetime) and (b) perturbative graviton EFT, $g = \eta + h$ , valid below the Planck scale. The clash bites only when the metric must be both quantum and dynamical and non-perturbative.

REFEREE VERDICT — isSound: false (overstated as written); severity: major; keep. Corrections. The original "mutually exclusive kinematic foundations / cannot combine in principle" is false as an in-principle claim and is downgraded. QFT does not in general require Poincaré invariance, a preferred vacuum, or a flat metric: QFT on fixed curved backgrounds is rigorous and already makes "particle" observer-dependent, replacing the global spectrum condition with the microlocal/Hadamard condition. Microcausality needs a definite Lorentzian causal structure, not a flat or non-dynamical one. The hole-argument paraphrase ("determinism FORCES identifying $\phi^*g$ with $g$ ") overstates a philosophically contested inference: Leibniz equivalence is the conventional resolution chosen to retain determinism, not a logical compulsion. Wigner's classification is a flat-space crutch, not a universal QFT requirement. Refined statement. Standard Minkowski QFT and GR rest on opposite treatments of geometry. This opposition is the deepest structural source of the difficulty in quantizing gravity, but it is a non-perturbative/UV and conceptual incompatibility, not an in-principle kinematic mutual-exclusion. Perturbative graviton EFT consistently places a dynamical, superposable metric perturbation inside QFT below $M_{\rm Pl}$ , failing by non-renormalizability (Goroff–Sagnotti), not by inconsistency. The genuine, unsolved obstruction: a superposed/fluctuating causal structure deprives microcausality, the global $S$ -matrix, and the global particle/vacuum concepts of their usual definitions, and no background-independent, UV-complete quantum theory of the metric with well-defined local diffeomorphism-invariant observables (the problem of time) has been constructed. Recommended tag. ESTABLISHED for the structural background-(in)dependence contrast and the non-renormalizability of perturbative gravity; OPEN for a background-independent UV-complete quantum theory of geometry and the resolution of superposed causal structure.

GC-2 — Perturbative non-renormalizability of quantized gravity

Frameworks: GR, QFT, Particle Physics / SM. Kind: ESTABLISHED theorem (perturbative non-renormalizability) + OPEN UV completion.

Precise statement. Treating the metric as a quantum field, $g = \eta + h$ , and quantizing $h$ yields a perturbatively non-renormalizable theory: graviton-loop divergences require infinitely many independent counterterms, so the perturbative expansion loses predictivity in the deep UV. ESTABLISHED

Technical root. Newton's constant $G$ has mass dimension $-2$ in 4D natural units, so the loop expansion is organized in $E^2/M_{\rm Pl}^2$ (higher-derivative graviton operators are irrelevant about the Gaussian fixed point). Concretely: pure gravity is one-loop finite on-shell ('t Hooft–Veltman 1974, the counterterm vanishing on the vacuum equations / removable by field redefinition); gravity + matter diverges already at one loop; pure gravity diverges at two loops with a nonvanishing Weyl-cubed ( $C^3$ ) counterterm (Goroff–Sagnotti 1985, confirmed by van de Ven). ESTABLISHED Contrast the renormalizable Standard Model (dimensionless or positive-dimension couplings).

Why it matters. By QFT's own internal criteria this sets an upper bound on the validity of the perturbative GR+QFT description at or below $E_{\rm Pl}\sim 1.2\times10^{19}$ GeV (see CONSTANTS_AND_SCALES.md), and reframes GR as a predictive low-energy EFT (Donoghue): the leading Einstein–Hilbert term plus $M_{\rm Pl}$ -suppressed $R^2$ , $R_{\mu\nu}R^{\mu\nu}$ corrections, with unambiguous leading quantum corrections (e.g., to the Newtonian potential).

Current status. The perturbative non-renormalizability and the EFT status are proven mathematics. The UV completion is OPEN (string theory, asymptotic safety, LQG; see UNIFICATION_LANDSCAPE.md). This is not a logical inconsistency: an EFT with a cutoff is fully consistent below it.

REFEREE VERDICT — isSound: true; severity: minor; keep. Corrections. The original kind: unsolved-problem is split: the perturbative non-renormalizability is ESTABLISHED, the resolution is OPEN. The inference "GR cannot be a fundamental QFT" is weakened to "cannot be a perturbatively renormalizable QFT in the Dyson sense": a nontrivial UV fixed point (Weinberg asymptotic safety SPECULATIVE/OPEN) would render gravity non-perturbatively renormalizable and predictive at all scales; higher-derivative gravity (Stelle 1977) is perturbatively renormalizable but contains a unitarity-violating ghost. Predictivity is not lost everywhere — only the perturbative expansion / finite parameter set fails in the deep UV; low-energy quantum-gravity predictions are unambiguous. Recommended tag. ESTABLISHED (theorem + EFT status); OPEN (UV completion / "fundamental QFT" question).

GC-3 — The cosmological-constant problem: QFT vacuum energy vs. observed $\Lambda$

Frameworks: QFT, GR, Cosmology, Particle Physics / SM. Kind: domain-of-validity mismatch / extreme fine-tuning (not a logical inconsistency, not a falsified prediction).

Precise statement. A Lorentz-invariant QFT vacuum carries energy density $\rho_{\rm vac}$ with $T_{\mu\nu} = -\rho_{\rm vac}\, g_{\mu\nu}$ (equation of state $w=-1$ ); by Lovelock's theorem (1971–72) the term $\Lambda g_{\mu\nu}$ is the unique additional divergence-free metric piece beyond $G_{\mu\nu}$ in 4D, so a constant vacuum energy is indistinguishable from $\Lambda$ and gravitates. Estimates of $\rho_{\rm vac}$ vastly exceed the observed dark-energy density $\rho_\Lambda^{\rm obs}\sim 10^{-47}\ \text{GeV}^4 \approx (2.3\ \text{meV})^4$ . ESTABLISHED tension

Technical root (robust vs. heuristic).

Heuristic (regulator-dependent): a zero-point sum cut at the Planck scale gives $\rho_{\rm vac}\sim M_{\rm cut}^4 \sim 10^{76}\ \text{GeV}^4$ , exceeding observation by $\sim\!10^{123}$ ( $\sim$ 120 orders); an electroweak cutoff still leaves $\sim$ 50–60 orders. The quartic estimate is regulator-dependent (dimensional regularization yields an $m^4\ln$ structure, not $M_{\rm cut}^4$ ).
Robust (regulator-independent): finite physical condensates already overshoot — the QCD chiral/gluon condensate $\sim \Lambda_{\rm QCD}^4 \sim 10^{-3}\ \text{GeV}^4$ (~44 orders too large) and the electroweak Higgs potential $\sim (100\ \text{GeV})^4$ (~59 orders). ESTABLISHED
Unbroken SUSY gives exactly zero vacuum energy, but SUSY breaking at $\gtrsim 1$ TeV reinstates $\rho \sim M_{\rm SUSY}^4 \sim 10^{12}\ \text{GeV}^4$ (~59 orders too large) — SUSY does not rescue it.

Why it matters. It may signal the naive QFT vacuum-energy calculation is the wrong object, that gravity "degravitates" vacuum energy, or that selection (anthropic/landscape) or sequestering is required. Weinberg's anthropic bound predicted a small nonzero $\Lambda$ before detection, but relies on the contested multiverse + measure. See OPEN_PROBLEMS.md, ASSUMPTIONS_LEDGER.md.

Current status. Cosmic acceleration is ESTABLISHED (SNe Ia, CMB+BAO). That dark energy is exactly a constant $\Lambda$ ( $w=-1$ ) is the simplest fit but a CONTESTED assumption; some recent BAO analyses have been read as hinting at dynamical dark energy — not established. All proposed solutions are SPECULATIVE.

REFEREE VERDICT — isSound: true; severity: minor; keep. Corrections. No technical errors; the canonical numbers check out. Sharpen the driver: vacuum energy gravitates as $\Lambda$ because a Lorentz-invariant vacuum has $w=-1$ (the equivalence-principle phrasing is a heuristic). In strict logic there is no failed prediction and no inconsistency: in EFT the cosmological constant is a renormalized free parameter (a bare- $\Lambda$ counterterm absorbs the divergence, exactly as for the electron mass). What "fails" is naturalness — the renormalized value must cancel radiative/condensate contributions to ~100+ decimal places with no known dynamical mechanism. "Largest mismatch in physics" is a defensible standard characterization (Weinberg 1989), but a rhetorical superlative whose magnitude depends on the unjustified cutoff choice. Recommended tag. OPEN — severe fine-tuning / domain-mismatch at the QFT–GR interface; the original domain-mismatch kind is correct.

GC-4 — Black-hole information paradox: unitarity vs. semiclassical thermal evaporation

Frameworks: QM, QFT, GR, Thermodynamics, Statistical Mechanics, Information Theory. Kind: domain-of-validity / inter-framework tension (not a logical inconsistency).

Precise statement. Hawking's leading-order semiclassical calculation (Bogoliubov mode-mixing between in/out vacua across the horizon) yields an outgoing radiation state that is mixed — a thermal/graybody density matrix at Hawking temperature $T_H = \frac{\hbar c^3}{8\pi G M k_B}.$ Naively iterating to complete evaporation maps a pure collapsing state to a mixed final state, in apparent conflict with the unitarity of global QM/QFT evolution. ESTABLISHED tension

Technical root. Unitarity requires the radiation's fine-grained von Neumann entropy to follow the Page curve (rise, then return to $\sim 0$ ), whereas the naive semiclassical entropy rises monotonically. The Bekenstein–Hawking entropy $S_{\rm BH} = \frac{k_B c^3 A}{4 G \hbar}$ should count microstates that the semiclassical computation does not exhibit. [Microstate counting ESTABLISHED for specific supersymmetric/BPS black holes via Strominger–Vafa; general case OPEN.]

Why it matters. It is the cleanest place where QM, GR, QFT, and thermodynamics collide simultaneously; resolving it constrains quantum gravity and forces the question of whether semiclassical QFT is being used outside its domain near/inside the horizon. See GC-5, GC-13, OPEN_PROBLEMS.md.

Current status. OPEN/CONTESTED as to mechanism. Unitarity is now widely favored: AdS/CFT is manifestly unitary, and post-2019 quantum-extremal-surface (QES) / island and replica-wormhole computations reproduce a unitary Page curve in controlled (largely AdS, lower-dimensional, large- $c$ ) settings INFERENCE. No agreed first-principles bulk mechanism exists for a realistic, asymptotically-flat, dynamically-formed 4D black hole.

REFEREE VERDICT — isSound: true; severity: minor; keep. Corrections. The original kind: logical-inconsistency is wrong and corrected to a domain-of-validity / inter-framework tension: each framework is internally consistent; the conflict arises only when semiclassical gravity is pushed past its validity (precisely what the island/QES results indicate by recovering the Page curve). "Exactly thermal to all orders" is overstated — the spectrum is a graybody (Planckian × frequency/spin-dependent transmission factors), and the mixedness is a leading-order, finite-order statement, not exact/robust. Unitarity violation is not demonstrated; the leading calculation, naively iterated, is what is incompatible with unitarity. "Where information is" (interior, horizon, entanglement wedge, soft hair) is interpretation-dependent. Recommended tag. ESTABLISHED as a structural tension between semiclassical gravity and unitarity; OPEN/CONTESTED as to resolution. Not a logical inconsistency.

GC-5 — The AMPS firewall puzzle

Frameworks: GR, QM, Information Theory, QFT. Kind: conceptual tension built on a rigorous no-go argument.

Precise statement. AMPS (2012) showed that for an old (post-Page-time) evaporating black hole, four individually well-motivated assumptions cannot all hold: (1) unitarity of evaporation/ $S$ -matrix; (2) validity of low-energy EFT outside the stretched horizon; (3) "no drama" at the horizon for an infalling observer (the near-horizon state is the local Hadamard vacuum, from the equivalence principle plus the short-distance vacuum entanglement structure of QFT); and (4) the black hole is a conventional quantum system with $e^{A/4}$ states (so the Page argument applies). INFERENCE — rigorous given premises

Technical root. The obstruction is entanglement monogamy, made rigorous via strong subadditivity of von Neumann entropy applied to $\{$ early radiation $E$ , exterior late mode $b$ , interior partner $a\}$ : $S(b,a) + S(b,E) \ \ge\ S(a) + S(E).$ A smooth horizon requires $b$ nearly maximally entangled with $a$ (small $S(b,a)$ ); unitarity past the Page time requires $b$ nearly maximally entangled with $E$ (small $S(b,E)$ ). SSA forbids both. Severing the $a$ – $b$ entanglement to restore unitarity excites the would-be vacuum, depositing high-energy quanta at the horizon — a firewall — violating (3).

Why it matters. It shows the information paradox threatens the equivalence principle itself at the horizon of an old black hole, where curvature is low and GR "should" be reliable. Proposed escapes — black-hole complementarity, ER=EPR (GC-16), state-dependence — each pay a conceptual price.

Current status. OPEN/CONTESTED: no consensus on whether old black holes have firewalls. ER=EPR (Maldacena–Susskind) is a SPECULATIVE proposed resolution with no first-principles derivation for generic entanglement. The island/QES program is read by many to suggest no firewall, but the interpretation is disputed.

REFEREE VERDICT — isSound: true; severity: minor; keep. Corrections. The rigorous engine is strong subadditivity; "monogamy of maximal entanglement" is the heuristic special case (the physical modes are only nearly maximally entangled, so the contradiction is quantitative). Postulate (3) is precisely "no drama," following from EP plus the local-vacuum (Hadamard) condition — calling it "the equivalence principle" alone is loose. The canonical framing includes the distinct fourth postulate (dimension $e^{A/4}$ ), which the original folds into "unitarity." AMPS proves only joint incompatibility; the firewall is one resolution, and which postulate fails is genuinely contested (complementarity, ER=EPR, state-dependence, islands/replica wormholes all argue the smooth horizon can survive). Recommended tag. INFERENCE for the AMPS no-go itself (a rigorous, widely accepted derivation of joint incompatibility given its premises); CONTESTED/OPEN for whether a firewall actually forms.

GC-6 — The measurement problem: unitary linear evolution vs. definite single outcomes

Frameworks: QM, QFT, Classical Mechanics. Kind: structural incompleteness / domain-demarcation gap (not a bare logical inconsistency).

Precise statement. Unitary Schrödinger evolution (axiom A5; linear) maps a von Neumann premeasurement $\sum_i c_i\,|s_i\rangle|\text{ready}\rangle \to \sum_i c_i\,|s_i\rangle|\text{pointer}_i\rangle$ — an entangled macroscopic superposition, never a single branch. The Born/projection rule (A6) instead replaces the state by one eigenprojection with probability $\mathrm{Tr}(\rho E)$ — non-unitary, non-linear, stochastic. If A5 is taken as a universal dynamical law covering apparatus + observer, it conflicts with the observed single definite outcome (Schrödinger cat, Wigner's friend). ESTABLISHED tension

Technical root. Linearity of $U(t)=e^{-iHt/\hbar}$ forbids a single branch. Standard QM avoids formal contradiction only by leaving "measurement" a primitive, undefined notion, with no rule fixing when each law applies (Bell's "shifty split"). Decoherence (GKSL/Lindblad dynamics) explains rapid, robust suppression of interference in a pointer basis and the emergence of an effective classical regime, but yields an improper mixture — the global state stays pure-entangled — so it does not by itself select an outcome. Gleason's theorem (dim $\ge 3$ ) fixes the form $\mathrm{Tr}(\rho E)$ of the probability rule but not why one outcome occurs.

Why it matters. It is the central foundational problem of QM, bounds the QM→classical correspondence, and is where the only currently-distinguishable alternatives (objective-collapse models) make testable deviations. See OPEN_PROBLEMS.md, HYPOTHESES.md.

Current status. OPEN/CONTESTED as to resolution; ESTABLISHED as a structural tension. Interpretations pay different prices: Everett (keep unitarity; pay with branching + an unsolved account of Born weights), de Broglie–Bohm (definite positions; pay with explicit nonlocality/contextuality and, relativistically, a preferred foliation), GRW/CSL objective collapse (modify dynamics; pay with new stochasticity and Lorentz-covariance tension — but empirically testable and being constrained by matter-wave/optomechanics/spontaneous-radiation experiments). All except objective collapse are empirically equivalent to standard QM so far.

REFEREE VERDICT — isSound: true; severity: minor; keep. Corrections. The original kind: logical-inconsistency is corrected to structural incompleteness / domain-demarcation gap: A5 and A6 have disjoint stipulated domains and are not formally contradictory; the contradiction appears only under the added premise that A5 is universal — precisely the contested premise. The "single experienced outcome" imports an extra determinacy premise that Everettian readings deny. The concrete technical claims (premeasurement entanglement; non-unitarity of projection; decoherence → improper mixture; Gleason fixing form not occurrence) are all correct. Optional sharpeners (Frauchiger–Renner 2018; extended-Wigner's-friend no-go theorems) tighten the tension but are themselves interpretation-dependent — do not cite as settled. Recommended tag. ESTABLISHED (the structural tension/incompleteness is real); OPEN/CONTESTED (its resolution). Not a settled logical inconsistency.

GC-7 — Thermodynamic arrow of time vs. time-symmetric microscopic laws

Frameworks: Thermodynamics, Statistical Mechanics, Classical Mechanics, QM, Cosmology. Kind: ESTABLISHED structural result + OPEN cosmological origin (a consistent-but-incomplete problem, not a contradiction).

Precise statement. The microscopic laws (Hamiltonian/Schrödinger) are time-reversal invariant up to the tiny CP/T-violation of the weak interaction, which is irrelevant to the thermodynamic arrow: a perfectly T-invariant world would still exhibit it given a low-entropy past. Yet isolated macroscopic systems show entropy non-decrease. Monotonic entropy increase cannot follow from time-symmetric dynamics alone — Loschmidt's reversibility objection and Zermelo's Poincaré-recurrence objection both rule out any purely dynamical derivation. ESTABLISHED

Technical root. Liouville's theorem (classical) and unitarity (quantum) conserve the fine-grained Gibbs/von Neumann entropy $S = -k_B\,\mathrm{Tr}(\rho\ln\rho)$ exactly for a closed system. Irreversibility appears only for a coarse-grained entropy (or, for open subsystems, via environmental entanglement/decoherence — a second route the naive "coarse-graining only" story omits), and in Boltzmann's $H$ -theorem only after inserting the molecular-chaos assumption (Stosszahlansatz), applied asymmetrically to pre- vs. post-collision correlations — the precise point where time-asymmetry enters (rigorous only in the Boltzmann–Grad limit for short times, per Lanford 1975). The resolution requires a time-asymmetric input: a special low-entropy boundary condition (the Past Hypothesis) plus a coarse-graining/typicality choice.

The fluctuation theorems refine the second law into exact statistical statements — Crooks: $P_F(+W)/P_R(-W) = e^{\beta(W-\Delta F)}$ ; Jarzynski: $\langle e^{-\beta W}\rangle = e^{-\beta\Delta F}$ — making transient entropy decreases exponentially rare but nonzero. They presuppose an equilibrium initial state (itself a time-asymmetric input) and so illustrate, rather than independently derive, the arrow.

Why it matters. The second law is not a fundamental microlaw but a consequence of dynamics + a cosmological boundary condition; this exports the deepest part of the puzzle — why the early universe had such low, especially gravitational, entropy — to cosmology and quantum gravity. See GC-17 (gravitational non-extensivity), Cosmology, OPEN_PROBLEMS.md.

Current status. That a time-asymmetric input is necessary is ESTABLISHED. Why the early universe had extraordinarily low gravitational entropy (near-zero Weyl curvature; Penrose's Weyl-curvature hypothesis is one SPECULATIVE proposal) is genuinely OPEN. A complete first-principles derivation of thermalization for realistic Hamiltonians is also OPEN (the eigenstate thermalization hypothesis is strong but unproven and fails for integrable / many-body-localized systems).

REFEREE VERDICT — isSound: true; severity: minor; keep. Corrections. Sharpen the CP/T remark: weak-interaction T-violation is essentially irrelevant, not merely insufficient. Fine-grained entropy conservation is exact for the total isolated system; open-subsystem entanglement (decoherence) is a distinct second route to observed irreversibility. The Stosszahlansatz is inserted as molecular chaos and applied asymmetrically; Lanford's theorem justifies the Boltzmann equation only for short times in the Boltzmann–Grad limit. The original unsolved-problem label applies only to the cosmological residue (origin of the Past Hypothesis), not to the established structural result. Recommended tag. Mixed: ESTABLISHED (necessity of a time-asymmetric input; Loschmidt/Zermelo; fine-grained-entropy conservation; Stosszahlansatz as the locus of asymmetry; exactness of fluctuation theorems); OPEN (origin/justification of the low-entropy Past Hypothesis).

GC-8 — The problem of time in canonical quantum gravity vs. QM's external time

Frameworks: GR, QM, QFT. Kind: conceptual / structural tension (not a logical inconsistency, not an in-principle impossibility).

Precise statement. QM presupposes an external classical time parameter $t$ : $i\hbar\,\partial_t|\psi\rangle = H|\psi\rangle$ , and by Pauli's theorem a lower-bounded $H$ admits no self-adjoint operator canonically conjugate to it ( $[T,H]=i\hbar$ ), so $t$ is a fixed background c-number. GR, being background-independent, has no preferred external time; its canonical quantization formally yields the Wheeler–DeWitt equation $\hat H_\perp|\Psi\rangle = 0$ — a constraint with no $\partial_t$ term and hence no Schrödinger-type evolution (the "frozen formalism"). ESTABLISHED facts

Technical root. The total ADM Hamiltonian is purely a sum of constraints, $H = N\,\mathcal H_\perp + N^i\,\mathcal H_i, \qquad \mathcal H_\perp \approx 0,\quad \mathcal H_i \approx 0\ \text{(weakly)},$ with lapse $N$ and shift $N^i$ as Lagrange multipliers. On the standard reading, coordinate-time reparametrization is part of the diffeomorphism gauge group, so "evolution" in coordinate time is pure gauge — though whether $\mathcal H_\perp$ generates genuine gauge or genuine many-fingered-time dynamics is itself contested (the Kuchař critique). Dirac observables must commute with all constraints, making them diffeomorphism-invariant, generically nonlocal, and hard to construct.

Why it matters. It is a direct obstruction to writing a Schrödinger equation for the universe and to defining unitary evolution and probabilities in quantum gravity — one of the sharpest conceptual barriers any canonical program must overcome. Connected to GC-1 (background independence). See OPEN_PROBLEMS.md.

Current status. OPEN/CONTESTED. Proposed (unconfirmed) resolutions: relational / Page–Wootters conditional-probability time (time as correlation with an internal clock subsystem), Rovelli's partial observables / evolving constants, and an emergent semiclassical WKB time from a Born–Oppenheimer expansion. Within classical GR the absence of preferred time is a feature; it becomes a problem only upon quantization. Cosmology's preferred cosmic time relies on FLRW symmetry and does not exist in an inhomogeneous universe.

REFEREE VERDICT — isSound: true; severity: minor; keep. Corrections. "Canonical quantization yields WDW" → "formally yields": the WDW operator is heuristic (operator-ordering ambiguities, regularization/UV ill-definedness; no rigorous self-adjoint definition). Flag that the pure-gauge reading of $\mathcal H_\perp$ is contested (Kuchař) and that this contest is constitutive of the difficulty. Pauli's theorem is used correctly. The tension is a deep unsolved-but-consistent structural problem, not a demonstrated impossibility nor a logical contradiction. Canonical references: Dirac's constrained-Hamiltonian analysis, DeWitt (1967), and the Isham and Kuchař reviews of the problem of time. Recommended tag. OPEN/CONTESTED structural problem (conceptual tension). Underlying technical facts (ADM constraint structure, Pauli's theorem, the formal WDW equation) are ESTABLISHED; the gauge-vs-dynamics status of $\mathcal H_\perp$ and the resolution are CONTESTED.

GC-9 — Haag's theorem vs. the working interaction picture of perturbative QFT

Frameworks: QFT, QM, Mathematics for Physics. Kind: rigor gap / domain-of-validity mismatch — a no-go for a formal construction (not a logical inconsistency of QFT or of renormalized perturbation theory).

Precise statement. The textbook interaction picture presupposes a unitary $U(t)$ , on one Poincaré-covariant Hilbert space with a unique vacuum, intertwining free and interacting fields at fixed time. Haag's theorem (Hall–Wightman 1957, after Haag 1955; see Streater–Wightman, PCT, Spin and Statistics, and All That) proves no such $U$ exists for a genuinely interacting field: were it to exist, the interacting Wightman functions would equal the free ones (so the "interacting" theory would be free). ESTABLISHED

Technical root. Stone–von Neumann uniqueness holds only for finitely many canonical pairs (in the Weyl form, with regularity); QFT has infinitely many degrees of freedom, so it fails, yielding uncountably many unitarily inequivalent representations of the CCR. The free and interacting theories generically inhabit different ones — the interacting theory does not live in the free Fock space. (See GC-14.) The textbook derivation of the Dyson series via a global interaction-picture unitary is therefore ill-defined as literally stated.

Resolution / repair. Rigorous content is recovered with no global $U$ : (a) IR/UV regularization (finite volume breaks the load-bearing translation-invariance hypothesis); (b) LSZ / adiabatic switching defines asymptotic in/out fields in their own representations; (c) Epstein–Glaser causal perturbation theory builds the perturbative $S$ -matrix directly from microcausality with no interaction picture and no Haag obstruction; (d) the invariant content lives in Haag–Kastler nets (generically type III $_1$ factors) and Doplicher–Haag–Roberts superselection theory. Dyson's observation that the QED series is asymptotic (not convergent) is a logically separate point.

Why it matters. Textbook QFT, despite its empirical triumphs (electron $g\!-\!2$ ), is not the rigorous theory it appears to be; this motivates algebraic QFT and the constructive program, and clarifies that spontaneous symmetry breaking and distinct phases live precisely in inequivalent representations.

Current status. ESTABLISHED as a theorem and as a foundational subtlety — not a falsification: it does not impugn the empirical correctness of renormalized perturbation theory, recovered as an asymptotic expansion controlled order-by-order (BPHZ). 4D constructive existence remains OPEN.

REFEREE VERDICT — isSound: true; severity: minor; keep. Corrections. The original kind: logical-inconsistency is downgraded to a rigor-gap / domain-mismatch: Haag's theorem is a no-go for one formal scaffolding, not a contradiction within a fixed consistent axiom system. "Mathematically inconsistent" → "the textbook derivation is not well-defined / rests on a nonexistent object"; the output is an asymptotic expansion of quantities definable by other means (Epstein–Glaser, Wightman/OS). Stone–von Neumann uniqueness is conditional (requires the bounded Weyl form and regularity); it can fail even in finite dof for non-regular representations or on non-simply-connected configuration spaces (Aharonov–Bohm $\theta$ -vacua). The sharp dividing line is "Weyl form + regularity + finiteness," not merely "finite vs. infinite dof." Recommended tag. ESTABLISHED (as a rigor gap / structural no-go and its resolution). The "logical-inconsistency" label is itself an overclaim, replaced by "rigor-gap / domain-of-validity mismatch."

GC-10 — Triviality / Landau poles refute "renormalizable = fundamental"

Frameworks: QFT, Particle Physics / SM, Mathematics for Physics. Kind: ESTABLISHED refutation + OPEN (Planck/gravity-dominated) SM UV completion.

Precise statement. Perturbative renormalizability does not guarantee a nontrivial UV-complete continuum theory. $\phi^4$ in 4D is proven trivial (Aizenman 1981 for $d>4$ ; Aizenman–Duminil-Copin, Annals of Math. 2021, for the marginal $d=4$ case, for reflection-positive lattice $\phi^4$ / Ising-class models): the continuum limit is the free (Gaussian) theory. QED is strongly believed trivial (perturbative Landau pole + lattice evidence), though its pole lies far above $M_{\rm Pl}$ and is academic. [ESTABLISHED for the refutation and $\phi^4$ ; INFERENCE for QED]

Technical root. The $\beta$ -function $\beta(g)=\mu\,dg/d\mu$ is positive for QED and $\phi^4$ , driving the coupling toward a Landau pole. Contrast asymptotic freedom in QCD: $\beta_{\rm QCD} = -\frac{g^3}{16\pi^2}\Big(11 - \tfrac{2}{3} n_f\Big) < 0 \quad (n_f < 33/2),$ so the coupling vanishes logarithmically in the UV — QCD is UV-safe where QED/ $\phi^4$ are not. Caveat: the Landau pole is a one-loop perturbative diagnostic, distinct from the rigorous triviality theorem (the ADC proof does not proceed via the pole).

Why it matters. It overturns renormalizability-as-fundamental-axiom (Wilson's RG already reinterpreted renormalizability as an emergent low-energy property) and contributes to the view that the SM is an EFT. It connects to the still-OPEN constructive question of whether any realistic interacting 4D QFT exists in the continuum. See GC-9, OPEN_PROBLEMS.md.

Current status. $\phi^4_4$ triviality is ESTABLISHED (theorem). QED triviality is INFERENCE. Asymptotic safety as a possible nontrivial UV fixed point is SPECULATIVE/OPEN. This is a consistent statement about domain of validity, not a logical contradiction.

REFEREE VERDICT — isSound: true; severity: minor; keep. Corrections. Re-categorize: the headline content is ESTABLISHED, not unsolved-problem; only the SM-specific UV completion is open, and that is gravity-dominated at $M_{\rm Pl}$ , not driven by triviality. Two overclaims fixed: (1) the SM-as-EFT conclusion rests primarily on empirical incompleteness (gravity, neutrino masses, dark matter, baryon asymmetry), with triviality a contributing motivation; (2) the SM Higgs sector's dominant UV concern is quartic metastability — for the measured top mass the Higgs quartic $\lambda$ runs slightly negative near $\sim 10^{10}$ – $10^{11}$ GeV (electroweak-vacuum metastability) — not a quartic Landau pole; the $U(1)_Y$ Landau pole is academic, far above $M_{\rm Pl}$ . State the ADC hypotheses (lattice, reflection positivity, Ising universality class) rather than presenting triviality as unconditional. Recommended tag. ESTABLISHED (the refutation "renormalizable $\ne$ UV-complete" and $\phi^4_4$ triviality); INFERENCE (QED triviality; SM-as-EFT inheritance); the SM's own UV completion is OPEN but Planck/gravity-dominated. Not unsolved-problem.

GC-11 — Singularities: GR predicts its own breakdown (geodesic incompleteness)

Frameworks: GR, Cosmology, QFT, QM. Kind: domain-of-validity limit / breakdown (not an internal logical inconsistency).

Precise statement. The Penrose–Hawking singularity theorems prove, via the Raychaudhuri focusing equation plus an energy condition and an initial focusing condition, that classical solutions are geodesically incomplete:

Penrose (1965): null energy condition (NEC) + a trapped surface + global hyperbolicity (non-compact Cauchy surface) $\Rightarrow$ null geodesic incompleteness;
Hawking / Hawking–Penrose (1970): strong energy condition (SEC) (or the generic condition) + cosmological expansion or a trapped set $\Rightarrow$ timelike/null incompleteness. ESTABLISHED

Incompleteness means the equations cease to define evolution. Curvature invariants (e.g., the Kretschmann scalar $R_{\mu\nu\rho\sigma}R^{\mu\nu\rho\sigma}$ ) diverge in the symmetric exact cases (Schwarzschild $r\to 0$ ; FLRW $a\to 0$ ), though geodesic incompleteness does not by itself entail curvature blow-up in general (cf. conical/quasiregular singularities).

Technical root & the energy-condition caveat. The energy conditions are hypotheses on matter that quantum fields violate pointwise (Casimir effect, squeezed states, near-horizon Hawking fluxes for the NEC; positive $\Lambda$ and slow-roll inflation, with $\rho + 3p = -2\rho < 0$ , decisively for the SEC). Crucially:

SEC violation by $\Lambda>0$ and inflation is a genuine, decisive loophole that evades the SEC-based cosmological theorems — this is why inflationary cosmology does not need an initial singularity from those theorems. The eternal-inflation case is instead closed by the Borde–Guth–Vilenkin theorem (a kinematic average-Hubble-rate condition, not an energy condition): even eternal inflation is past-incomplete.
Pointwise NEC violation does not generally defeat the collapse theorem: quantum fields obey averaged / quantum energy inequalities (Ford–Roman QEIs; ANEC), and the singularity theorems have been re-derived from these weaker averaged conditions (Tipler–Borde–Roman; Fewster–Galloway; Fewster–Kontou) and largely survive.

Why it matters. Singularities mark precisely where GR must hand off to quantum gravity (Big Bang, black-hole interiors) at curvatures $\sim \ell_P^{-2}$ ; whether they are resolved (bounces, replacement) is OPEN and currently untestable. See GC-1, GC-7, CONSTANTS_AND_SCALES.md.

Current status. ESTABLISHED that GR is geodesically incomplete (the theorems are rigorous). The pointwise energy conditions are known too strong but are robustified by averaged/quantum conditions. What replaces the singularity is OPEN. This is a domain-mismatch/breakdown, not a contradiction within classical GR.

REFEREE VERDICT — isSound: true; severity: minor; keep. Corrections. Penrose's theorem is conditional on the existence of a trapped surface plus global hyperbolicity; that trapped surfaces form in realistic collapse is a separate result (Christodoulou). Hawking radiation's NEC violation is a late-time near-horizon evaporation effect, not relevant to the collapse-stage premises — do not list it alongside Casimir/squeezed states as a threat to the theorem's hypotheses. Distinguish: SEC violation is a decisive loophole for the cosmological theorems; pointwise NEC violation is largely repaired by ANEC/QEIs for the collapse theorem. Overstated "premises in tension with quantum matter" → "pointwise violated, but the theorems are substantially robust under the physically appropriate averaged/quantum energy conditions." Recommended tag. ESTABLISHED (theorem content + domain-of-validity reading); the energy-condition sub-claim is ESTABLISHED as to pointwise violation but INFERENCE/partially-resolved given the averaged-energy / QEI robustification program.

GC-12 — Quantum nonlocality (Bell violation) vs. relativistic locality

Frameworks: QM, Special Relativity (see GR, Classical Mechanics), QFT. Kind: conceptual tension with no operational/empirical contradiction.

Precise statement. Loophole-free Bell/CHSH experiments (Hensen, Giustina, Shalm et al., 2015) decisively violate the local-hidden-variable bound $\text{CHSH:}\quad S = |E(a,b) - E(a,b') + E(a',b) + E(a',b')| \le 2,$ with QM correlations bounded above by Tsirelson's bound $S \le 2\sqrt2$ (Cirel'son 1980); real experiments record values strictly between $2$ and $2\sqrt2 \approx 2.828$ (the ideal QM ceiling is not attained). ESTABLISHED Entanglement correlations and (on collapse-type readings) the projection postulate appear to act instantaneously across spacelike separation.

Technical root. Bell's theorem (local causality, Bell 1976) shows the empirical correlations cannot arise from any theory simultaneously satisfying local causality, measurement-independence (free choice / no superdeterminism), and no retrocausality, holding the QM predictions fixed; experiment forces abandoning at least one, and which is interpretation-dependent. ("Realism" is not a separate premise — it is part of local causality; noncontextuality is the distinct Kochen–Specker assumption and should not be conflated with the Bell premises.) The clash is not operational: QM respects parameter independence — the no-signaling theorem, $\rho_A = \mathrm{Tr}_B(\rho_{AB})$ is independent of $B$ 's setting choice — so no usable information propagates superluminally. In the Jarrett/Shimony decomposition of local causality, QM violates only outcome independence, which carries no signal. At the QFT level, microcausality $[\phi(x),\phi(y)]=0$ for spacelike separation enforces no-signaling. ESTABLISHED

Why it matters. It pins down exactly which classical intuitions must be abandoned and shows the clash with relativity is conceptual (about hidden mechanism), not operational. Making collapse manifestly Lorentz-covariant is genuinely hard and is where objective-collapse models (GRW/CSL) face their stiffest challenge; relativistic QFT measurement also confronts frame-dependent state-update, Sorkin's "impossible measurements," and Reeh–Schlieder vacuum nonlocality. See GC-6, GC-15.

Current status. ESTABLISHED that Bell violation is real and that no-signaling holds — so no contradiction with special relativity's operational content. The tension is interpretational. Measurement-independence/free choice is load-bearing; denying it (superdeterminism/retrocausality) is a logically OPEN but widely-regarded-as-conspiratorial escape, actively CONTESTED by a minority. A fully covariant, agreed account of relativistic state-update is OPEN.

REFEREE VERDICT — isSound: true; severity: minor; keep. Corrections. State the Bell premises precisely: $\{$ local causality, measurement-independence, no-retrocausality $\}$ with predictions held fixed. "Realism" is subsumed in local causality (CHSH = "local realism"); noncontextuality is the distinct Kochen–Specker premise and must not be listed inside the Bell trichotomy. Tsirelson's $2\sqrt2$ is the ideal QM ceiling, not an experimental value. The Jarrett/Shimony split (parameter independence respected = no-signaling; outcome independence violated = no signal) sharpens the SR reconciliation. Collapse is interpretation-dependent (absent in Everett and de Broglie–Bohm; stochastic/Lorentz-challenged in GRW/CSL). Recommended tag. ESTABLISHED (a conceptual tension with no operational contradiction); CONTESTED/interpretation-dependent (which premise of local causality fails; the physical status of collapse).

GC-13 — Holographic area-law degrees of freedom vs. volume-extensive QFT Hilbert space

Frameworks: Information Theory, QFT, GR, Statistical Mechanics, Thermodynamics. Kind: conceptual tension built on ESTABLISHED components; resolution OPEN. Domain-of-validity mismatch, not a logical inconsistency.

Precise statement. Black-hole and covariant entropy bounds — $S_{\rm BH} = k_B A/(4\ell_P^2)$ ; the Bousso bound $S \le A/4$ in Planck units — cap the entropy of a gravitating region by its boundary area, implying a finite, area-extensive count of physical gravitational degrees of freedom. Local QFT, by contrast, assigns independent oscillators to every point, giving a volume-extensive Hilbert-space dimension. The conceptual tension is over "how many degrees of freedom live in a region." ESTABLISHED components

Technical root. Ground-state entanglement entropy of a region obeys an area law but with a UV-divergent, cutoff-dependent coefficient, $S \sim \frac{\text{Area}}{\epsilon^{D-2}} \quad (D = \text{spacetime dimension; e.g. } S\sim \text{Area}/\epsilon^2 \text{ in 4D}),$ reflecting infinitely many short-distance modes; only mutual information, entropy differences, and universal subleading coefficients (central charge in 2D; log / $a$ -anomaly terms) are physical. Rigorously, local algebras in QFT are type III $_1$ von Neumann factors (Reeh–Schlieder; cyclic-separating vacuum): they admit no trace and no tensor factorization of the Hilbert space into local subsystems, so there is no well-defined local density matrix or number operator. The resolution of the apparent over-counting is gravitational: a region whose field-theoretic entropy would exceed $A/4$ has enough energy to have already collapsed to a black hole (the Susskind spherical-entropy-bound argument) — the QFT count is being applied outside its gravitational domain of validity.

Why it matters. It suggests continuum QFT badly over-counts degrees of freedom and is effective, not fundamental (connecting to GC-3 and the species/Bekenstein bounds), and underlies the holographic principle and "spacetime from entanglement" programs (GC-16).

Current status. Area-law BH entropy is ESTABLISHED-theory, near-universally accepted across independent derivations (Hawking radiation, Euclidean action, string microstate counting for extremal/BPS black holes — Strominger–Vafa; loop area spectra), but lacks a general background-independent proof INFERENCE. The type-III structure is ESTABLISHED (algebraic QFT). Recent crossed-product / large- $N$ constructions (Witten; Chandrasekaran–Longo–Penington–Witten) yield type II algebras with a well-defined trace and finite renormalized (generalized) entropy — a INFERENCE partial bridge that does not yet derive the $A/4$ microstate count. How to define entanglement entropy in full quantum gravity, and whether spacetime is fundamentally discrete, is OPEN.

REFEREE VERDICT — isSound: true; severity: minor; keep. Corrections. Fix the dimensional slip: the leading divergence is $S \sim \text{Area}/\epsilon^{D-2}$ in $D$ spacetime dimensions (not $\epsilon^{d-1}$ ). The Hilbert space is volume-extensive, but the ground-state entanglement entropy obeys an area law (a volume law holds only for generic excited/random states) — the original verbally conflated these. The relationship is a domain-of-validity mismatch (QFT without gravity vs. with gravity), not a contradiction: once gravitational backreaction is included, the over-counting is removed. The crossed-product type-II bridge addresses factorization/trace and matches generalized entropy but does not derive the finite count — "partial bridge" is fair; more would overclaim. Recommended tag. OPEN conceptual tension built on ESTABLISHED components (S $_{\rm BH}$ ; the area-law UV divergence and which pieces are universal; type III $_1$ algebras + Reeh–Schlieder; the Susskind resolution; the Bousso bound); the crossed-product bridge is INFERENCE; the full reconciliation is OPEN.

GC-14 — Stone–von Neumann uniqueness vs. inequivalent representations (infinite dof)

Frameworks: QM, QFT, Statistical Mechanics, Mathematics for Physics. Kind: domain-of-validity boundary (not an inconsistency).

Precise statement. ESTABLISHED The Stone–von Neumann theorem (von Neumann 1931): for a finite number of canonical pairs, every irreducible representation of the Weyl (exponentiated) form of the CCR that is weakly/strongly continuous (regular) is unitarily equivalent to the Schrödinger representation. Both the Weyl-form and the continuity hypotheses are essential. ESTABLISHED For infinitely many degrees of freedom (QFT, the thermodynamic limit) the theorem fails: there are uncountably many unitarily inequivalent irreducible representations of the CCR.

Technical root. The deep reason is not "compactness" but the nonexistence of an infinite-dimensional translation-invariant (Lebesgue) measure: in finite dimensions the canonical Gaussian construction is essentially unique up to unitary equivalence, while in infinite dimensions Gaussian measures with different covariances are mutually singular (disjoint), yielding inequivalent representations.

Why it matters. This is the precise mathematical seam between the "one Hilbert space" intuition of textbook QM and the algebraic reality of QFT and many-body physics. It underlies — as related but logically distinct phenomena, each needing extra hypotheses — inequivalent vacua and spontaneous symmetry breaking; superselection sectors (charge, particle number, distinct thermodynamic phases); Haag's theorem (GC-9, which additionally requires Poincaré covariance and vacuum uniqueness); and the necessity of the algebraic C*/von Neumann formulation. In statistical mechanics, distinct phases live in inequivalent GNS representations and genuine non-analyticities (phase transitions) appear only in the thermodynamic limit (illustrated for ferromagnetic lattice models by the Lee–Yang circle theorem on partition-function zeros).

Current status. ESTABLISHED on both sides: Stone–von Neumann is a theorem (finite dof); inequivalent representations are a rigorous, physically essential feature of infinite-dof systems. A clean domain mismatch, fully understood mathematically (algebraic QFT / quantum statistical mechanics), not an open problem.

REFEREE VERDICT — isSound: true; severity: minor; keep. Corrections. State both load-bearing hypotheses (Weyl form and regularity/continuity) — without continuity, inequivalent representations exist even in finite dof. The mechanism is the missing infinite-dimensional Lebesgue measure / mutual singularity of Gaussian measures, not "compactness" (a garbled attribution). SSB, superselection, and Haag's theorem are logically distinct, not uniform direct corollaries (superselection can arise without infinite dof; Haag needs Poincaré covariance + vacuum uniqueness). SvN gives uniqueness of the irreducible rep up to unitary equivalence; "Fock vacuum" is anachronistic for the bare SvN (Schrödinger-representation) setting. Lee–Yang specifically concerns partition-function zeros of ferromagnetic Ising/lattice-gas models. SvN is a theorem about the chosen representation, not a foundational axiom QM "rests on." References: von Neumann 1931; Reed–Simon; Haag, Local Quantum Physics; Bratteli–Robinson. Recommended tag. ESTABLISHED.

GC-15 — Semiclassical gravity $G_{\mu\nu} = 8\pi G\,\langle T_{\mu\nu}\rangle$ vs. quantum superposition of mass

Frameworks: GR, QM, QFT. Kind: domain-of-validity breakdown of an approximation that becomes inconsistent if promoted to fundamental.

Precise statement. Semiclassical gravity (Møller 1962, Rosenfeld 1963), $G_{\mu\nu} = 8\pi G\,\langle\psi|T_{\mu\nu}|\psi\rangle$ , couples a classical metric to quantum matter via the expectation value of the stress-energy — a source quadratic (hence nonlinear) in the state. For a macroscopic superposition $(|\text{mass at }A\rangle + |\text{mass at }B\rangle)/\sqrt2$ , the equation produces a single mean-field geometry sourced by both locations at once — not a superposition of two geometries. INFERENCE that this fails for superpositions

Technical root. The mean-field source is nonlinear in $|\psi\rangle$ (the map $|\psi\rangle\mapsto|\psi\rangle\langle\psi|$ is nonlinear). The mean-field prediction is empirically disfavored (Page–Geilker 1981), and — combined with stochastic state reduction — the nonlinearity is pathological: the general theorem that deterministic nonlinear modifications of quantum dynamics permit superluminal signaling (Gisin 1990; Polchinski 1991) applies, and the non-relativistic Schrödinger–Newton limit conflicts with Born-rule statistics under collapse (Bahrami–Großardt–Donadi–Bassi 2014).

Why it matters. It is a concrete internal reason semiclassical gravity is at best an approximation, motivating either quantizing the metric or modifying QM via gravitational collapse. It drives gravitationally-induced-collapse proposals (Diósi–Penrose) and tabletop gravity-mediated-entanglement experiments aiming to test whether gravity must be quantized. See GC-1, GC-12, OPEN_PROBLEMS.md.

Current status. INFERENCE/OPEN. The near-universal expectation is that the metric must be quantized (or QM modified by gravitational collapse), but this is not experimentally settled. Diósi–Penrose collapse is being constrained (not excluded) by underground spontaneous-radiation and interferometry experiments. Proposed gravity-mediated-entanglement tests would, if entanglement were observed, argue gravity is non-classical — but they have not been performed at the required sensitivity. Semiclassical gravity remains an excellent approximation in its domain (Hawking, Unruh; the leading term of a $1/N$ expansion).

REFEREE VERDICT — isSound: true; severity: minor; keep. Corrections. The signaling/Born-rule pathology does not follow from the nonlinearity of $\langle T\rangle$ alone: unitary-only semiclassical gravity is nonlinear but does not by itself signal; the inconsistency arises specifically when the mean-field source is combined with stochastic state reduction (then Gisin/Polchinski applies). The "single mean-field geometry" is not self-evidently inconsistent in isolation — it is empirically disfavored (Page–Geilker) and pathological only with collapse. "A classical metric cannot consistently couple to quantum sources" is too strong as a universal claim: it is the mean-field (expectation-value) coupling that fails; fundamentally stochastic classical–quantum models (Diósi–Penrose; Oppenheim-type CQ dynamics with intrinsic decoherence/diffusion) evade the deterministic-nonlinearity no-go and remain CONTESTED-but-open. Do not lean on Eppley–Hannay (1977), whose gedanken argument is criticized as relying on unphysical idealizations. Recommended tag. INFERENCE for the breakdown (mean-field failure for superpositions + nonlinearity-implies-signaling are well-established); the universal no-go ("no classical metric can couple to quantum matter") is downgraded to CONTESTED/SPECULATIVE.

GC-16 — ER=EPR / "spacetime from entanglement" vs. its unestablished generality beyond AdS

Frameworks: Information Theory, GR, QFT, Cosmology. Kind: conceptual tension / evidential gap (domain-of-validity / generality-of-evidence limitation, not an internal logical inconsistency).

Precise statement. The Ryu–Takayanagi formula $S(A) = \text{Area}(\gamma_A)/4G_N$ , entanglement-wedge reconstruction, the "first law of entanglement $\Rightarrow$ linearized Einstein equations" result, and the ER=EPR conjecture jointly motivate the program that spacetime geometry emerges from the entanglement structure of an underlying quantum theory. The tension: all of this is rigorously formulated and derived essentially only within AdS/CFT (asymptotically anti-de Sitter, large central charge, strong 't Hooft coupling — the bulk Einstein-gravity regime). INFERENCE within AdS

Technical root. RT/HRT and entanglement-wedge reconstruction are derived (Lewkowycz–Maldacena 2013 gravitational replica trick; FLM/JLMS modular-Hamiltonian relation; QES via Engelhardt–Wall; islands via Penington / Almheiri–Engelhardt–Marolf–Maxfield / Penington–Shenker–Stanford–Yang 2019–20) — but every derivation presupposes the holographic dictionary plus the Einstein-gravity (large- $c$ , large- $\lambda$ ) regime, with a homology constraint and extremal-surface prescriptions. Crucially: (i) the gravity-from-entanglement derivations recover only the linearized Einstein equations (Lashkari–Van Raamsdonk; Faulkner–Guica–Hartman–Myers–Van Raamsdonk 2014; second order only under restrictive assumptions); the full nonlinear, background-independent reconstruction is not achieved; (ii) ER=EPR has no first-principles derivation for generic, non-holographic entanglement; (iii) de Sitter holography (dS/CFT, relevant to our positive- $\Lambda$ universe) is far less constrained and speculative, and flat-space / finite- $N$ holography lack established RT analogs; (iv) an algebraic-QFT obstruction (local algebras are type III $_1$ factors lacking factorization and a finite entanglement entropy — see GC-13) sits beneath the heuristic "entanglement-as-fabric" discourse.

Why it matters. It is the most ambitious proposed reconciliation of GR and quantum information — potentially reframing the GR/QFT clash (GC-1) by making geometry emergent — but its applicability to real cosmology (where the deepest puzzles, including $\Lambda$ and the arrow of time, live) is exactly where it is least grounded. See UNIFICATION_LANDSCAPE.md, OPEN_PROBLEMS.md.

Current status. RT within AdS/CFT is INFERENCE (well-supported, extensively checked, proven in some lower-dim/BPS cases; AdS/CFT itself lacks a general nonperturbative proof). "Spacetime from entanglement" as a universal principle, ER=EPR for generic entanglement, and de Sitter holography are SPECULATIVE/OPEN. Whether this is profound universal insight or an AdS-specific artifact is arguably the central open question of the program.

REFEREE VERDICT — isSound: true; severity: minor; keep. Corrections. No genuine technical errors; the claim is deliberately anti-overclaiming. Minor: within AdS/CFT these results are derived, not merely "tested" (the stated domain restriction is unaffected). The status hierarchy is endorsed: RT/entanglement-wedge in AdS INFERENCE; linearized-Einstein-from-first-law INFERENCE, linearized only; full nonlinear background-independent gravity-from-entanglement SPECULATIVE/OPEN; ER=EPR for generic entanglement SPECULATIVE; dS/CFT and flat/finite- $N$ analogs SPECULATIVE/OPEN. The label is more precisely a generality-of-evidence limitation than a logical tension — there is no contradiction, only an unwarranted extrapolation outside AdS. Recommended tag. SPECULATIVE for the program's generality (AdS-restricted RT and entanglement-wedge results at INFERENCE); the meta-claim that this generality is currently unestablished beyond AdS is ESTABLISHED (contested only by optimists).

GC-17 — Standard ensemble statistical mechanics vs. long-range gravitating systems

Frameworks: Statistical Mechanics, Thermodynamics, GR, Cosmology. Kind: domain-of-validity mismatch (not a logical inconsistency).

Note. This entry's original framing was the only one the referee found internally unsound (isSound: false). It is retained and corrected; one erroneous sub-claim (black-hole area-entropy as a consequence of Newtonian non-additivity) is excised to Corrected / not actually contradictions.

Precise statement. Equilibrium statistical mechanics and thermodynamics presuppose interactions decaying faster than $r^{-d}$ in $d$ dimensions (short-range), so that energy and entropy are extensive/additive, the standard fixed-density thermodynamic limit exists, and the microcanonical and canonical ensembles are equivalent. Self-gravitating systems (star clusters, galaxies, the cosmological matter distribution) violate this: the potential $\sim -G m^2/r$ is long-range and unscreened, making energy non-additive. The standard ensemble machinery breaks down — as a domain mismatch, not a logical inconsistency. ESTABLISHED

Technical root & genuine consequences (classical self-gravitating systems).

(a) Ensemble inequivalence: the microcanonical and canonical ensembles are inequivalent.
(b) Negative microcanonical specific heat — impossible in the canonical ensemble, where $C_V = \mathrm{Var}(E)/(k_B T^2) \ge 0$ identically; the negative microcanonical $C_V$ is precisely the diagnostic signature of (a), not a logically independent fact.
(c) The gravothermal catastrophe / Antonov instability (Antonov 1962; Lynden-Bell–Wood 1968), with long-lived metastable local equilibria below threshold (the Antonov radius) but no global entropy maximum for the idealized confined point-mass model.
(d) Failure of the Euler and Gibbs–Duhem relations, which rest on first-order homogeneity (extensivity).

The "entropy unbounded above" feature is a property of the unregularized point-mass model and stems from the short-distance ( $r\to 0$ , UV) singularity of $-Gm^2/r$ — distinct from the long-range (IR) non-additivity that is the headline subject. Short-distance cutoffs (quantum degeneracy pressure, hard cores; Padmanabhan) restore boundedness and metastable equilibria. A nontrivial mean-field thermodynamic limit does exist under Kac rescaling (coupling $\sim 1/N$ ), though non-additivity and ensemble inequivalence persist.

Why it matters. The foundations of thermodynamics/statistical mechanics do not straightforwardly apply to the dominant long-range force at large scales, complicating any global entropy budget for the universe (relevant to the arrow-of-time / Past-Hypothesis problem, GC-7).

Current status. ESTABLISHED as a domain breakdown: non-additivity, negative specific heat, and ensemble inequivalence of gravitating systems are well understood and not in dispute (Antonov 1962; Lynden-Bell–Wood 1968; Thirring 1970; Padmanabhan 1990; Ruelle/Fisher stability theory). Treating the universe as a global thermodynamic system with well-defined temperature/entropy for the gravitational sector is INFERENCE/CONTESTED and not rigorously founded.

REFEREE VERDICT — isSound: false (one sub-claim miscategorized); severity: minor; keep (with excision). Corrections. The headline domain-mismatch claim is sound and well-established. Excised category error: black-hole area-entropy $S_{\rm BH}\sim A$ was listed as a consequence of Newtonian self-gravity non-additivity on par with (a)–(d). It is not — Bekenstein–Hawking area scaling is a relativistic/semiclassical horizon/holographic result (Bekenstein 1973; Hawking 1975), a different domain that does not follow from the Euler-relation failure; ordinary self-gravitating gas entropy does not scale as area (see quarantine below). "Entropy unbounded / no equilibrium" is a property of the idealized unregularized point-mass model (a UV effect), not of the long-range non-additivity. The standard fixed-density thermodynamic limit fails, but a Kac-rescaled mean-field limit exists. Negative $C_V$ is the signature of ensemble inequivalence, not an independent consequence. Recommended tag. ESTABLISHED (the core domain-mismatch claim and consequences (a)–(d) for classical self-gravitating systems).

Corrected / not actually contradictions

This section quarantines specific sub-claims the referee judged miscategorized. The parent entries are retained; only the flagged sub-claim is moved here with an explanation.

CN-1 — Black-hole area-entropy is not a consequence of Newtonian self-gravity non-additivity (from GC-17)

Excised claim. "Self-gravitating systems exhibit … area- rather than volume-scaling entropy for black holes ( $S_{\rm BH}\sim A$ )" — listed in the original GC-17 as a consequence of the non-additivity of classical Newtonian self-gravitating systems, alongside negative specific heat and ensemble inequivalence.

Why it is not a valid consequence. Bekenstein–Hawking entropy, $S_{\rm BH} = \frac{k_B c^3 A}{4 G \hbar},$ is a result of relativistic / semiclassical gravity — event horizons, the holographic bound, Hawking 1975 / Bekenstein 1973 — a different domain from the classical statistical mechanics of self-gravitating point masses. It does not follow from the failure of the Euler relation that drives the gravothermal catastrophe (Antonov 1962; Lynden-Bell–Wood 1968). Moreover, ordinary self-gravitating gas / star-cluster entropy does not scale as area — it has its own non-extensive (and, for unsoftened point masses, unbounded) behavior. The two phenomena share only the heuristic slogan "gravity defeats naive volume-extensivity."

Disposition. Black-hole area scaling is genuine, established physics and a real instance of gravity defeating volume-extensivity — but as a separate, relativistic phenomenon. It belongs with GC-13 (holographic area-law vs. volume-extensive QFT) and GC-4, not as a corollary of the Newtonian argument in GC-17. No contradiction is lost by removing it from GC-17; one was merely misattributed.

References

See BIBLIOGRAPHY.md for full citations. Landmark sources invoked above (real, standard) include: 't Hooft–Veltman (1974) and Goroff–Sagnotti (1985) on graviton-loop divergences; Stelle (1977) on higher-derivative gravity; Donoghue on GR as an EFT; Weinberg (1989) review of the cosmological-constant problem; Lovelock (1971–72); Hawking (1975) and Bekenstein (1973) on black-hole radiation/entropy; Page on the Page curve; Strominger–Vafa (1996) microstate counting; Almheiri–Marolf–Polchinski–Sully (AMPS, 2012); Bell (1964, 1976), CHSH (1969), Cirel'son/Tsirelson (1980), Hensen/Giustina/Shalm et al. (2015); von Neumann (1932) measurement; Gleason (1957); GRW/CSL objective-collapse literature; Boltzmann's $H$ -theorem, Loschmidt, Zermelo, Lanford (1975), Crooks and Jarzynski fluctuation theorems; Dirac constrained-Hamiltonian theory, DeWitt (1967), Isham and Kuchař problem-of-time reviews; Page–Wootters, Rovelli; Haag (1955), Hall–Wightman (1957), Streater–Wightman, Epstein–Glaser, Doplicher–Haag–Roberts; Aizenman (1981) and Aizenman–Duminil-Copin (2021) on $\phi^4$ triviality; Penrose (1965), Hawking–Penrose (1970), Borde–Guth–Vilenkin, Ford–Roman quantum energy inequalities, Fewster–Galloway, Fewster–Kontou, Christodoulou; Møller (1962), Rosenfeld (1963), Page–Geilker (1981), Gisin (1990), Polchinski (1991), Bahrami et al. (2014), Diósi–Penrose; Ryu–Takayanagi, Lewkowycz–Maldacena (2013), Faulkner et al. (2014), Maldacena–Susskind (ER=EPR), Penington / Almheiri–Engelhardt–Marolf–Maxfield / Penington–Shenker–Stanford–Yang (2019–20), Witten and Chandrasekaran–Longo–Penington–Witten on crossed-product algebras; Antonov (1962), Lynden-Bell–Wood (1968), Thirring (1970), Padmanabhan (1990), Ruelle/Fisher stability theory; Haag, Local Quantum Physics; Bratteli–Robinson; Streater–Wightman, PCT, Spin and Statistics, and All That.