Skip to content

Structural Admissibility ​

We built Structural Admissibility to answer one question and we ask it before we ask whether a claim is right: has this piece of work earned a hearing as the kind of thing it says it is? We catalogue the kinds first and we let the audit follow from that catalogue, so ontology comes before evaluation and the test never shifts to fit the work.

We apply it in three moves and we keep all three in view at once: (1) we name the bucket the work claims (Mathematical exploration / Effective model / Theory proposal / Physical theory), (2) we split the work into components-in-regimes since most works live in different buckets in different regimes, (3) we test each component through three lenses, primitives at the right level (Leg A), operating at the claimed level (Leg B1), correctly positioned relative to established physics (Leg B2), with closure in one of three modes (reaches ground / defers / dissolves). The decomposition is the verdict. A pass means we have a candidate worth testing further while a fail means the work is not yet a candidate in the bucket it claims.


Admissibility asks whether a work is what it claims to be ​

We place admissibility before correctness and we ask whether a work has the structure of a candidate before we ask whether it is empirically adequate or predictive or confirmed. A pass can still be wrong while a fail is not wrong yet, since it is not yet a candidate.

We do not ask "is this true?" and we ask "is this what it says it is?" A paper presents itself as Physical-theory unification or as effective phenomenology or as Mathematical exploration, and we check whether its actual structure backs that kind-claim. We catch work that lives in one bucket while it claims another because the higher bucket pays better.

We apply admissibility to claims that point at reality. Physics and empirically anchored sciences and descriptions of actual systems carry an ontological anchor since they say they describe the actual world, while mathematics carries none since a mathematical construction stands on its own without describing anything and a physics claim cannot. We catch mathematical structure dressed as physics that never earned its anchor, and we call out the category error when a reviewer judges a physics claim "as mathematics" while it is deployed as physics.

We ignore sociological markers and we look at structure. Credentials and venue and citation count and mathematical density test acceptance, not structure. A dense and peer-reviewed and much-cited document can fail while an informal and eccentric-looking document can pass. We catch one specific failure here: a theory that gets qualitatively right behaviour across several phenomena by positing enough primitives to fit them, while it never derives those primitives and never touches the deeper machinery the claimed problems turn on.


Four buckets, components-in-regimes, three closure modes ​

Four buckets of work that touches reality ​

We sort work that touches reality into four buckets. Each bucket claims a different thing and succeeds on different terms, while pure mathematics sits underneath as the foundation Physical theory draws on, with no claim to describe reality on its own, and outside the audit as a candidate.

Mathematical exploration. We develop mathematical structures from physics questions and we accept judgement as mathematics. We claim only the conditional: if this structure describes reality then these are the mathematical consequences. Examples include topos-theoretic approaches to QM, category-theoretic reformulations of gauge theory, and algebraic QFT in its purely formal aspects. We succeed through mathematical richness tied to physical motivation, and we leave empirical anchoring aside until we make a physical claim.

Effective model. We fit observed phenomena with measured parameters and we work inside a stated regime and we make no claim to derive our primitives from a deeper layer. We claim only the bounded result: inside this regime, with these parameters, we describe these phenomena. Examples include MOND, ΛCDM, nuclear-physics shell models, the Fermi theory of weak interactions before the electroweak unification, most of the foundational scaffolding of QM, and most of the content of the SM.

Theory proposal. We put forward mathematical structures as candidate descriptions of reality and we leave the empirical demonstration in progress or open. We claim without qualification but without anchoring yet: these structures describe reality and their predictions follow in principle. Examples include most of string theory in its current state, loop quantum gravity, and causal dynamical triangulation.

Physical theory. We make positive claims about reality across our full claimed regime and we derive from named anchors and we recover established physics at the boundaries. We claim the full structure: our primitives are derived and our regime is bounded by physics rather than convenience and the established results come back out. GR vacuum field equations belong here. The Dirac equation for free fermions in flat space belongs here. The spin-statistics theorem belongs here. The bucket holds less than institutional talk suggests.

The four buckets do not rank worth. A great Mathematical exploration is great and a great Effective model is great and a great Physical theory is great, since each kind defines its own success and none sits above the others in value.

We need the lower-anchor buckets to reach the higher ones. GR needed Riemannian geometry already in place as pure mathematics. QM needed Hilbert space theory and operator algebra. Gauge theory needed fiber bundles. The mathematicians who built those tools were doing mathematics, not physics, and physics drew on their work when physics needed it. We do not exclude lower-anchor work but we ask it to name its bucket out loud.

We catch misrepresentation of bucket: work that claims a higher bucket than it supports because higher buckets bring more institutional reward. A bad paper is not one that sits in a low bucket but one that fails the criteria of the bucket it claims.

Components-in-regimes are the unit, not theories ​

A theory is not one kind of thing. A nominal theory usually holds components that belong in different buckets in different regimes. We run on components-in-regimes, not on theories as wholes.

We see it cleanest in the Dirac equation. In flat Minkowski spacetime for free fermions, it is Physical theory since it follows from relativistic covariance plus first-order time evolution, and it predicted spin-½ and antimatter and g = 2 before observation, with full retrodiction of Schrödinger-Pauli in the non-relativistic limit. In the SM matter sector, the same equation carries Effective-model parameters since Yukawa couplings fit mass spectra and mixing matrices tabulate. In condensed matter, the same mathematical form appears as a derived low-energy effective theory of band structure, with no claim about fundamental physics. Three buckets for one equation, with regime and role deciding each time.

The ontology has to respect that fact: theories are not atoms, while components-in-regimes are, so we evaluate the component-in-regime rather than the theory.

What QM and the SM are, under this split, is Effective models with Physical-theory components embedded. QM builds its scaffolding (Born rule, commutators, measurement postulate) from operational definitions. It embeds derived components (the Dirac equation in flat space, spin-statistics, identical-particle statistics) that are Physical theory. The SM fits its content (gauge group choice, generations, Yukawa structure). It embeds derived components (gauge invariance forcing minimal coupling, anomaly cancellation, the structural form of the Higgs mechanism) that are Physical theory. Both frameworks reach as far as they do because derived nuclei do the structural work while operational scaffolding makes the frameworks usable.

This split does not demote them. Effective models that work this well are extraordinary achievements. The split names what kind of achievement they are. The bucket-collapse view calls them Physical theory, and that sets a bar no successor can clear (it has to beat a bar nobody met) while it hides the work that would count as progress (finding the Physical-theory layer beneath the Effective models).

Closure comes in three modes ​

We mean by closure the relation between a claim and the ground it stands on. A chain closes when it reaches ground rather than trailing off into posits. Closure comes in three modes.

Reaches ground. The chain ends at named anchors: empirical, structural, or established theory in its domain of validity. The primitives are anchored, and the derivation runs through to a layer it does not need to justify further.

Honest deferral. The chain hands a piece explicitly to a layer the work does not yet reach, and it names the handoff at that layer. The deferral does the work, it does not hide a gap. A theory that says "we recover the SM at tree level, the precision residue at loop level is open" has deferred the precision-recovery piece by name. A theory that says "we recover the SM" without naming the residue has buried it.

Dissolution of the presupposed question. The chain shows that the question assumed structure that does not exist past a certain regime. The question stops applying. This is the strongest mode. The theory does not promise future work on the question but shows the question cannot survive its own framework. The worked example is Wave Relativity's treatment of the QM-in-curved-spacetime question sub-Planck. Wave Relativity derives that matter wave and metric are aspect-projections of a single field Ψ, with both projections failing together at the Planck scale because they project the same underlying structure. The question "how do we reconcile QM and curved spacetime at the Planck scale" assumes two separate things that need reconciling. Wave Relativity derives that this assumption fails past a specific regime.

Failures of closure look different from each mode. A chain that trails into undeclared posits does not close. A chain that buries the handoff inside a layer it claims to have closed does not defer by name. A chain that asserts dissolution without deriving where the assumed structure fails does not dissolve.

The three modes say what closure is, and the legs test whether it happened.


Three legs test whether a bucket claim is supported ​

We take the bucket a work claims and we split the work into components-in-regimes and we state what closure would have to look like for each component to hold its bucket, and then we check whether that closure is there.

A claimed bucket needs three structural conditions to hold, and the three legs give three views on whether they hold, with each view catching a different way a bucket claim can come apart.

Leg A: Are the primitives at the right level? ​

Each bucket sets a relation between the work's primitives and a deeper layer. Pure mathematics takes axioms as primitives, so it needs no deeper layer. Effective models define primitives operationally inside a stated regime, with the deeper layer deferred by name. Physical-theory primitives derive from a substrate the work commits to, so the deeper layer is the substrate, and the derivation runs through it.

We use Leg A to press the move from "asserted" to "derived" that the claimed bucket requires. Work that claims Physical theory has to show its primitives deriving from a named substrate. Work that claims Theory proposal has to show mathematical structure in its primitives that is consistent and motivated. Work that claims Effective model has to name its primitives as operationally defined at the layer where they live free.

Probe. Write the component's primitives as functionals of the named substrate. If the work can't, ask it to name the substrate. If it says the substrate is documented elsewhere, ask for it on the table. Off-table substrates are where the load-bearing posits hide.

Fails when undefined densities, operators, gates, thresholds, or exponents are introduced as if derived but aren't. Self-naming as a derivation does not make it one. A theory that calls itself "derived from first principles" while introducing five undefined primitives is not derived, it is asserted, and the Physical-theory claim is unsupported. The same work as a Theory proposal or Effective model would be admissible at the lower bucket.

Leg B1: Does the work operate at the level its claimed bucket implies? ​

Each bucket sets a level of description where the work operates. Pure mathematics works at the level of formal structure. Effective models work at the level of measured phenomena. Physical theory works at the level of derived dynamics in the substrate the work commits to.

We use Leg B1 to check whether the work has earned its place at the level its bucket requires. Work that takes on quantum-mechanical problems has to carry quantum machinery: operator algebra, probability or flux structure, measurement-theoretic content. Work that takes on curved-spacetime problems has to carry curved-spacetime dynamics. Work that takes on QM-plus-GR problems has to carry both, in the regime where they interact.

Probe. Name the level at which the problem is formulated. Does the work do its work at that level, or borrow the vocabulary from one level while computing at another?

Fails when the deeper-level vocabulary is invoked rhetorically while the actual computation happens at a shallower level. The vocabulary is borrowed, the structure isn't there. A classical-dynamics framework that talks about Hawking radiation in QFT vocabulary is not addressing Hawking radiation but relabeling it.

B1 cannot be evaded by self-categorisation. Work that takes on intrinsically QM-plus-GR phenomena (Big Bang dynamics, Hubble-tension and expansion-history claims, Hawking radiation, inflation) inherits B1 from what it explains, not from how it labels itself. "Effective model," "phenomenological cosmology," and "we restrict to boundaries where standard description fails" framings do not lift B1 when the phenomena need QM-plus-GR machinery to address them. What the work claims to explain fixes the level where it has to operate, not what bucket it claims to sit in.

Leg B2: Is the work correctly positioned relative to established physics? ​

Each bucket sets a relation to established physics. Pure mathematics may connect to established physics or not, and the connection does not bear on its admissibility as mathematics. Effective models recover the established physics they touch, at the boundaries of their named regime. Physical theory either (a) derives established physics from a deeper substrate, with recovery as consequence, or (b) defers the recovery by name to a layer the work does not yet reach.

We use Leg B2 to check whether the work has earned its place in the stack of theories that its bucket requires. Work that claims to lie beneath the Standard Model cannot take the Standard Model as primitive in its derivation. Work that claims to underlie QM cannot feed on QM's results as fitted inputs.

Probe. In the limit where established theory is empirically anchored, how does the framework reduce to it?

Fails when established physics is treated as primitive while the work claims to derive what lies beneath it. "Ordinary matter" cannot be a primitive in a theory that claims to derive what matter is. "We label the SM as 'resolved'" is relabeling, not derivation. The SM's specific content (gauge group, chirality, fermion spectrum, mixing matrices, mass spectrum) is what would have to be recovered.

The three legs are three lenses on one question ​

The legs are not three separate audits. They give three views on one question: does the work's actual structure support its bucket claim? Each view catches a different failure.

A theory can fail because it asserts primitives where it claimed derivations (A). It can fail because it works at a shallower level than its bucket requires (B1). It can fail because it takes established physics as a primitive of the derivation rather than a consequence (B2). These fail in structurally different ways, and a theory can fail any one on its own. A theory that derives its primitives beautifully (A closes) can still work at the wrong level (B1 fails) or fail to recover established physics (B2 fails).

The legs conjoin at the component level. Failure on one leg fails that component at the claimed bucket. The component then either belongs in a lower bucket (the common case), or belongs in no bucket that fits what the work actually does (the worst case). The verdict is the decomposition, not a single pass or fail for the whole work but a map of which components close at what bucket.

We get our diagnostic value from seeing which component fails which leg. Averaging across legs or across components destroys it. Component-level splitting lets us see what works and what does not, and that opens either incremental repair (one component is at fault) or wholesale recategorisation (no component closes at the claimed bucket).


Instantiation for quantum gravity ​

QG claims live where matter quantization and curved spacetime and possibly trans-Planckian regimes meet: the cosmological-constant problem, the Big Bang singularity, the Hubble tension, the BH information paradox, inflation, structure formation. The established machinery there is QFT-on-curved-spacetime plus GR plus the Standard Model.

QFT-on-CS does not solve QG, since it computes quantum fields on a classical background metric, and the regime where geometry must also go quantum is exactly where back-reaction lives, which is exactly why QG stays open. So QFT-on-CS is the calculational framework we have because QG is unsolved, and treating it as part of the solution mistakes the category.

The three legs instantiate as:

A (QG). Primitives have to derive from the substrate the candidate commits to, and "Information," "substrate," "field" have to do specific structural work (what they are, what equations they satisfy, what they couple to), not stand in for missing derivation.

B1 (QG). The candidate has to do QM in curved spacetime, with a named regime of validity: a quantum object (field, state, operator algebra), a curved-spacetime evolution, and a probability or flux structure. The cannot-be-evaded clause applies in full force here: Big Bang dynamics, Hubble tension, Hawking radiation, and inflation are intrinsically QM-plus-GR, and any work addressing them inherits B1 from the problem, not from its self-labeling.

B2 (QG). The candidate has to recover the Standard Model (gauge group, chirality, fermion spectrum, generations, electroweak masses) in the regime where the SM is empirically established, or defer this recovery by name to a layer it does not yet reach.

QG respecified as the enumerated task of deriving (a)-(d) ​

The taxonomic split of established physics gives a sharper statement of what QG would need to be than the standard framing.

QG is not "a theory that unifies QM and GR." That framing treats QM and GR as Physical-theory objects waiting to be unified. But under the split, most of QM is effective at its foundations (Born rule, commutators, measurement are operationally defined), and the cosmological-scale and high-curvature regimes of GR are exactly where it breaks down. The unification framing has run for eighty years with no closed result, which tells us the framing itself blocks the way.

The component-level specification runs instead: QG is the project of identifying the Physical-theory layer that produces, as derived consequences from a deeper substrate,

(a) the currently-derived components of QM: the Dirac equation, spin-statistics, gauge invariance forcing minimal coupling, anomaly cancellation (b) the currently-derived components of GR: the vacuum field equations from equivalence principle and self-consistency (c) the currently-effective components (Born rule, commutator structure, measurement, the cosmological constant, ΛCDM parameters) with named regimes of validity (d) the currently-open structural items (back-reaction, the SM gauge group choice, generation count, strong CP problem, hierarchy problem) either as derived consequences or as deferrals by name at the deeper layer's natural open boundary

This specification tests component by component. We evaluate a candidate QG framework on whether its primitives produce each item, leg by leg, regime by regime.

The bucket mismatch we catch in failed QG candidates almost always repeats one pattern: the candidate claims Physical-theory status but takes parts of (a)-(d) as primitives. A candidate that posits geometry rather than deriving it (LQG) fails to recover the GR backbone. A candidate that posits matter content rather than deriving it (most cosmology proposals) fails to recover the SM components. A candidate that works only in restricted regimes (AdS/CFT, string theory in perturbative formulations) does not touch the Physical-theory question that QG-in-our-universe asks.


Worked component decompositions ​

We show how the ontology runs on major theories. Placement at the theory level throws information away, while the component-level view keeps it.

Dirac equation ​

Regime / roleCategory
Flat-space free fermionPhysics proper. Derived from relativistic covariance + first-order time evolution. Predicts spin-½, antimatter, g = 2 before observation.
Perturbative QED tree levelPhysics proper. Minimal coupling from gauge invariance is derived.
Perturbative QED loop levelEffective model with derived structure. Renormalization is operationally defined, and Wilsonian justification places this in EFT territory.
Curved-spacetime via tetrad/spin-connectionPhysics proper within QFT-on-CS regime. Same regime-bound as QFT-on-CS.
SM matter sectorPhysics-proper construction on effective-model substrate. Dirac structure closes, gauge group is fitted.
With Yukawa couplingsEffective model. Yukawa matrices fitted.
Quantum gravity regimesHonest deferral. Equation requires a metric.
Condensed-matter analogsDerived effective model. Same mathematical form from band structure.

We use the free-fermion-in-Minkowski case as leverage on any work that claims to derive QM from a deeper layer. The deeper layer has to give back what we already derived.

General Relativity ​

Regime / roleCategory
Vacuum field equationsPhysics proper. Derived from equivalence principle and self-consistency.
With stress-energy sourcePhysics-proper geometric structure, inherits source-theory status.
With cosmological constantPhysics-proper structure with effective-model parameter. Λ is fitted.
FLRW + ΛCDM applicationsEffective model. Cosmological principle as auxiliary assumption.
Near singularities, Planck scaleHonest deferral. Framework names where it stops working.
Strong-field astrophysicsPhysics proper within regime.

Quantum Mechanics ​

Regime / roleCategory
Schrödinger as non-rel limit of DiracInherits Dirac's Physical-theory status.
Born ruleOperational/effective. No derivation of why squared amplitudes give probabilities.
Commutation relationsOperational/effective. Postulated by analogy with classical Poisson brackets.
Measurement postulateEffective with structural admission of incompleteness. The measurement problem is the open item.
Specific HamiltoniansDerived from classical limits + quantization rules. Quantization rules themselves operational.
Path integralMathematical physics exploration deployed as physics.
Spin-statistics theoremPhysics proper. Derived from Lorentz invariance + locality.

QM runs in layers. The bucket-collapse view treats it as one block. But the split shows derived strata set inside operational scaffolding. The derived strata carry the predictive load, while the scaffolding makes the framework usable.

Standard Model ​

Regime / roleCategory
Gauge invariance forcing minimal couplingPhysics proper. Derived from local symmetry.
Anomaly cancellation constraining hyperchargesPhysics proper. Structural requirement.
Spin-statisticsPhysics proper. From Lorentz invariance + locality.
Gauge group SU(3)×SU(2)×U(1)Effective. Empirically motivated.
Three generationsEffective. Posited.
Yukawa structure / mass spectrumEffective. Tabulated.
Mixing matrices (CKM, PMNS)Effective. Measured.
Higgs mechanism (structural form)Physics-proper construction.
Higgs sector specificsEffective. Potential form posited. Hierarchy problem is the structural admission.
Quark confinementEmpirically present, derivation incomplete. Lattice QCD numerically, closed analytical derivation is Millennium Prize.

The SM works as Effective overall, with Physical-theory components embedded. Its empirical reach comes from the combination.

QFT on Curved Spacetime ​

Regime / roleCategory
Fields on fixed classical backgroundsPhysics proper within regime. Matter quantization derived, metric is classical input.
Hawking radiationPhysics proper within regime. Bogoliubov coefficients on Schwarzschild background.
Unruh effectPhysics proper within regime.
Cosmological particle productionPhysics proper + inherits cosmological inputs.
Back-reactionOpen / hand-off to QG. Exactly why QG remains open.

We count it as calculationally Physical theory inside its named regime, and it explicitly leaves gravity unquantized.

ΛCDM Cosmology ​

Regime / roleCategory
GR backboneInherits GR status.
Cosmological principleAuxiliary assumption.
Cold dark matterEffective parameter. No particle identification.
Cosmological constantEffective parameter. Fitted value.
InflationEffective with multiple model families.
BBN, CMB acoustic peaksPhysics proper within regime. Derived from framework + parameters.
Hubble tension, σ_8 tensionOpen empirical residues. Not yet structural failures.

String Theory ​

Regime / roleCategory
Perturbative amplitudes on flat backgroundsTheory proposal / exploration. Closes as mathematical structure.
AdS/CFT as mathematical statementTheory proposal. Closes within named regime.
As Physical-theory QG candidateCategory mismatch. Empirical anchoring not closed.
Landscape / compactificationStructural open item. ~10^500 vacua, no selection principle.
SM recoveryOpen. No compactification yields SM with derivation.

We read this as the cleanest case of institutional misrepresentation. It passes as Theory proposal and fails at the Physical-theory level claimed.

Loop Quantum Gravity ​

Regime / roleCategory
Spin-network formalismTheory proposal. Mathematically well-defined.
Quantization of geometryPosited, not derived. Fails A as Physical-theory QG.
Semiclassical limitUnresolved after four decades. Fails B1 as Physical-theory QG.
Matter contentNot addressed structurally. Fails B2 as Physical-theory QG.

It fails all three legs as Physical-theory QG. It passes as Theory proposal with interesting mathematical results.


What admissibility does not test ​

Not a consensus test. A theory in line with consensus still fails when it fails admissibility. LQG built institutional infrastructure and fails as Physical-theory QG. Where it sits sociologically runs independent of whether it holds structurally.

Not a complexity test. Mathematical density does not close a derivation. A 600-page document of dense formal moves can still posit its load-bearing content, while a short clean derivation can close.

Not a finality test. A theory can pass with open items, provided it names them at the layer where they stand open and does not bury them inside chains it claims to have closed.

Not a falsification test. Falsification asks whether the theory predicts something checkable. We ask whether the theory has the build to predict at all rather than fit.

Not a mathematical-consistency test. A theory can hold together as mathematics and still fail as physics. We run on the physics claim (including how it deploys the mathematics as a description of reality) and not on the internal consistency of the mathematical structure.

Not a verdict-producer. Admissibility tells you whether work belongs in the candidate bucket for whatever bucket it claims. It does not tell you whether the work is right or important or worth your time.


Worked diagnostics ​

Diagnostic 1: Misrepresentation of bucket (a cosmology paper) ​

A 2026 paper, One Mechanism, proposes that dark matter, dark energy, inflation, and the BH information paradox are four regimes of a single signed coupling exchanging structured information ρ and unresolved potential u under a conservation law E = ρ + u + χ. It claims Physical-theory status for the unified mechanism.

Category claim: Physical-theory QG.

Component decomposition: the paper's framework has no component-level structure to decompose, since every primitive (the tri-partition, signed coupling S(ρ), critical density ρ_c, coupling strength λ_k, resolution gate C(t), kernel scaling exponent 25/12) is posited at the same level. The "underlying Paradox Engine framework" is referenced but held off-table.

  • Leg A: five posited primitives sitting on an undeclared substrate. Paper's own Open Problems names ρ_c, S functional form, kernel scaling, gap calibration, and propagation invariant's numerical value as undefined or conditional. The bucket claim of Physical theory is unsupported on Leg A, since the primitives are asserted, not derived.
  • Leg B1: dynamics classical throughout. Hawking radiation invoked rhetorically via QFT-on-CS language. No quantum object, no operator algebra, no probability structure. The work operates at a shallower level than its claimed bucket requires.
  • Leg B2: "ordinary matter" primitive throughout. No particle spectrum, no gauge group, no chirality, no generations. Established physics treated as primitive while the work claims to derive what lies beneath it.

Three-leg fail at the claimed bucket. As Theory proposal or Effective model, the same content would be admissible at the lower bucket. The failure sits in the bucket claim, not the underlying work. That is where admissibility pays: the work has not earned the kind it claims to be.

Diagnostic 2: Mathematics passing, physics failing (AdS/CFT) ​

Conjecture: Type IIB string theory on AdS₅ × S⁵ is dual to 𝒩 = 4 super-Yang-Mills on the conformal boundary in the large-N limit.

As mathematics (Theory proposal at the Mathematical-exploration level): a well-posed conjecture about two structures. Admissible.

As physics (the bucket claimed in QG discourse, holography arguments, BH information arguments):

  • Leg A: substrate (AdS₅ × S⁵, large-N 𝒩 = 4 SYM) is not the substrate of physical reality. Bucket claim of Physical theory unsupported.
  • Leg B1: regime is AdS-asymptotic, large-N, and neither describes our universe. The work operates in a regime its physics-claim doesn't actually inhabit.
  • Leg B2: boundary theory is 𝒩 = 4 SYM, not the SM. Recovery of established physics in our universe is not addressed.

Three-leg fail as physics. The fault lies in the deployment: citing AdS results as if they settled facts about QG in our universe turns the work into a physics claim, and that move carries no anchor. "AdS/CFT teaches general principles about QG even if our universe isn't AdS" is the deployment we catch. The principles stay mathematical until someone shows the anchoring to reality.

Diagnostic 3: Physics-proper claim with full decomposition (Wave Relativity) ​

Wave Relativity (Tan 2026) claims Physical-theory status: Ψ is the substrate, the master equation derives QM and GR as projections in their regimes, the SM is recovered.

Bucket claim: Physical theory. The work claims to be the deeper layer beneath QM, GR, and the SM.

Component decomposition: layered structure where Ψ on spacetime is the master object and existing theories are recovered as projections in named regimes.

  • Leg A: empirical anchors named (Aharonov-Bohm, Newton, Davisson-Germer + Stern-Gerlach + Pauli). Self-consistency stated as structural demands on any equation for a physically-real potential. Master equation derived through these. Open items catalogued at the layer they're open. Primitives anchored, bucket claim supported on Leg A.
  • Leg B1: Dirac on curved spacetime in Ch. 13. Schrödinger-Newton as weak-field non-relativistic closure in Ch. 19. Metric functional g_μν = G_μν[Ψ] in Section 20.8 with named regime λ ≫ ℓ_P. Probability as conserved Dirac flux in Sections 21.6-21.7. Sub-Planck closes via dissolution (Section 21.12): both projections (matter wave as WKB scalar amplitude of Ψ, metric as Einstein-functional of Ψ) fail simultaneously and for related reasons, leaving Ψ as single invariant object. The QM-in-curved-spacetime question doesn't get deferred past ℓ_P. It dissolves, because the two things the question asks to reconcile are derived as aspect-projections of one field. Work operates at the level the claim requires, B1 closes via dissolution.
  • Leg B2: SM gauge group derived in Ch. 23 as substrate-phase inheritance plus non-abelian local-basis closure. Fermion spectrum and generations from linearized master equation in Chs. 25-26. Electroweak masses through composite chiral background in Chs. 28-29. Tree-level recovery in Ch. 30 with precision residue named as next-iteration work. Established physics recovered at tree level, precision residue deferred by name.

Admissibility holds at the claimed bucket. Open items remain, and they sit named at the layer where they stand open, not buried. The contrast with the cosmology paper sits at the level of structure, not conclusion. Both address overlapping problem-space, and only one carries the architecture to be a candidate solution at the bucket it claims.


Applying the audit ​

Admissibility runs at the level the work claims, not the level it actually performs. We grade a paper that claims QM-plus-GR problems on B1 and B2 even when it never enters those problems. It fails by claiming the territory without walking it. A paper that claims only a phenomenological description gets the legs at the level it claims instead, not at the QM-plus-GR level.

Bucket claims propagate down to components. Work that claims Physical theory claims each of its load-bearing components as Physical theory in its regime. When some components run effective (a fitted parameter, an empirical input), we ask the work to name them as such. Components that parade as derived while running fitted give the most common bucket fraud inside otherwise careful work.

The diagnostic value is in seeing which component fails which leg. Averaging destroys it. A theory that fails Leg A on three components but closes Leg B1 and B2 on those same components fails differently from a theory that fails all three legs straight across. The split lets us see what works and what does not, and that opens either incremental repair (one component is at fault) or wholesale recategorization (no component closes at the claimed bucket).

Passing is not endorsement. Admissibility opens the door to evaluation, it is not the prize. Work that passes moves on to the next filters: empirical adequacy, predictive power, parsimony, explanatory reach. Work that fails does not yet carry the build to face those filters. It has to earn its bucket or step down to one it can hold.

Admissibility is portable. The buckets and the three legs travel across domains. Only how B1 and B2 instantiate changes when the domain changes. For chemistry, B1 asks whether the claim works at the molecular or electronic level, and B2 asks whether it gives back established equilibrium thermodynamics or quantum chemistry. For economics, B1 asks about the level of agent behaviour or aggregate dynamics, and B2 asks about established empirical regularities. We keep the audit and we re-instantiate only the domain part.

Competent application requires reading. Each leg asks something that takes domain expertise. Leg A asks whether primitives are anchored, and that means reading the anchoring work. Leg B1 asks whether the regime is covered, and that means understanding the regime. Leg B2 asks whether established physics comes back out, and that means knowing what recovery would look like. A reviewer who runs this seriously has to do the work. Running it as a rubber stamp leaves us worse than the current system: with the look of structural rigour and none of the substance.


Framework-level open items ​

Component-level scoring. Our decomposition tables run categorical. A finer instrument might score components on a continuous scale: how many steps from named anchors, how robust the regime call. The categorical version is easier to use, while the finer version might track truth more closely.

Cross-component dependencies. Components lean on each other. The SM's effective parameters lean on the gauge invariance derivation holding. We do not yet track dependency structure explicitly. For most uses this does not matter. For judging fundamental claims it might.

Historical drift. Components have moved bucket over time. Quantization of geometry passed as Physical-theory-once-discovered in early LQG. We now read it more accurately as a posit. Tracking how components migrate between buckets over time would make us historically aware.


Ontology before audit, buckets before legs ​

We catalogue what kinds of theoretical work exist (four buckets) and how theoretical structure splits up (components-in-regimes) and what it means for a derivation chain to close (three modes). Then we apply it: three legs that test whether a work's bucket claim holds up in its actual structure. The legs do not run three separate audits but give three views on one question (does the bucket claim hold?), with each view catching a different failure.

The verdict is the decomposition. A pass means the work stands as a candidate in the bucket it claims. A fail means it does not yet stand as a candidate at that bucket, though it may belong in a lower bucket if presented there. The QG project, under admissibility, turns into the enumerated task of producing items (a)-(d) from a deeper Physical-theory layer, with candidates judged component by component on whether their primitives close at their claimed substrate.

We put admissibility before correctness and ontology before audit and buckets before legs. That order shapes the audit: first what exists, then what splits, then what closure is, then how to test.

Markdown source available in the wave-relativity repository. Run by Daniel Tan Fook Hao.