From constraints to scale
The Shape can hand back patterns — ratios, angles, indices, exact fractions, and structural maps. But a pattern does not know whether it is measured in GeV, meters, seconds, or Planck units. Scale supplies the ruler. The framework therefore treats dimensionless quantities as the safest derived outputs, and charges every absolute magnitude to a measured anchor, a paid normalization, a named scheme object, or an open bridge. A number is only a prediction if it is unit-invariant, or if its ruler, scheme, window, and bridge are all named before comparison.
What Scale does
Scale is not the claim that “the theory predicts every number.” It is the discipline that decides which numbers are meaningful, which are dimensionless outputs, which are measured anchors, which are paid normalizations, and which are ill-posed predicates. It has two jobs.
Meaning — is this even a valid observable?
Before asking what a number equals, Scale asks whether it is a valid observable at this dimension, in this scheme, and in this energy window. Some expressions that evaluate to a GeV-style magnitude are not observables at all. A local coefficient, a scheme-dependent quantity, or a magnitude in a dimension with no canonical finite predicate does not become physical just because it has units attached.
Source — what supplies the magnitude?
If the number is a valid observable, Scale asks where its magnitude comes from. Is it a dimensionless ratio the Shape hands back directly? A measured anchor? A paid normalization? A consistency coefficient? Or a magnitude that is still open pending a named bridge? Every absolute magnitude must be tagged with its source before it counts as a prediction.
The unit-gauge theorem
The requirement for a ruler is not a convenience — it follows from unit-gauge invariance.
- Physics is invariant under a rescaling of units. Whether we measure in GeV, joules, meters, or seconds cannot change the physics.
- A naked dimensionful number is not invariant under that rescaling — it changes when the units change.
- Only dimensionless ratios, or dimensionful values stated relative to a named anchor, are invariant and therefore meaningful.
- Therefore at least one absolute scale anchor is required before any magnitude claim can exist.
The existence of an absolute anchor is theorem-grade inside the framework: without it, dimensionful claims are simply not well-defined. But the value of that anchor is a separate matter — it is measured, not derived. Needing a ruler does not derive the length of the ruler.
The anchor convention
Declared anchor: the ordinary Planck mass
The framework declares its absolute anchor to be the ordinary Planck mass, MPl = 1.2209 × 1019 GeV — not the reduced Planck mass. Its status is PAID as a measured anchor: the value is measured, not derived. The electroweak scale, vEW ≈ 246 GeV, is used as a second measured ruler where the electroweak/Higgs sector is specified. Both are rulers the framework openly declares and pays for; neither is a from-nothing output.
The layer-aware application
Scale does not act on a bare metric. It acts across the full three-layer Shape — the geometric Stage, the finite Rulebook, and the physical Actors. Each layer produces things that may carry magnitude, and Scale assigns, charges, or fixes them accordingly.
Stage — assigns geometric magnitudes
The Stage is the metric carrier: M4 × K₆=SU(3)/T² × S² × S¹Y/ℤ₂. Scale assigns magnitudes to its geometric outputs — radii, volumes, the compactification scale, thresholds, and curvature magnitudes. Dimensionless ratios of these are the safest outputs; their absolute sizes are set only against a declared anchor.
Rulebook — charges paid normalizations
The Rulebook is the finite, zero-dimensional admissibility layer: flavor chamber, chamber modulus, generation basis, sector projectors, parity and anomaly conditions, and the freeze-before-compare firewall. Scale charges the normalizations and windows that live here — chamber angles, sector scales, tolerance windows — and records each as paid, so no fitted number can masquerade as a free output.
Actors — fixes readout schemes
The Actors are the zero-dimensional physical content: matter, gauge, Higgs and proton bundles, connections, and the Dirac/index and graviton/ghost operators. Scale fixes their operator schemes, RG windows, mass readouts, seesaw scales, and threshold bridges, and names the comparison scale at which each observable is cashed.
Claim categories
Every candidate magnitude is tagged with exactly one of six Scale categories. The tag records how a number earns its status — it is not a quality grade.
| Category | What it means |
|---|---|
| SCALE-FREE-DERIVED | A dimensionless ratio, index, or exact rational invariant handed back by the Shape. Unit-invariant, exact-checkable, and the strongest column. |
| SCALE-ANCHORED | A dimensionful value stated relative to the declared measured anchor. Meaningful once the anchor is named; inherits the anchor’s measured status. |
| SCALE-PAID | A normalization, window, or scheme quantity that is charged as measured input rather than claimed as an output. |
| SCALE-CONSISTENCY-COEFFICIENT | An internal coefficient that checks self-consistency but is not, by itself, a physical observable. |
| SCALE-DISSOLVED | A candidate magnitude that turns out to be an ill-posed predicate — no canonical observable exists at this dimension/scheme, so it is dissolved rather than predicted. |
| SCALE-OPEN | A magnitude that could become a prediction if a target-blind bridge were supplied, but which is left honestly open until it is. |
Worked examples
Worked example
Why a ruler is required
Unit-gauge invariance kills naked GeV predictions. A dimensionful number has no invariant meaning until it is stated relative to a ruler. The framework therefore proves that at least one absolute scale is required, then charges the Planck mass as a measured anchor. Caveat: the need for a ruler is derived; the ruler’s value is measured.
Full worked example — why a ruler is required
Question
Can a theory derive a dimensionful number from no measured scale at all?
Naive answer
Yes — a beautiful formula might directly output a mass in GeV.
Why the naive answer fails
A naked dimensionful number changes when units change. Physics cannot depend on whether we use GeV, joules, meters, or seconds. Only dimensionless ratios are invariant, unless a dimensionful value is stated relative to a named anchor.
Constraint set
- Unit-gauge invariance — physics is invariant under unit rescaling.
- Dimensionless-first — ratios and pure numbers are the safe derived outputs.
- Anchor requirement — at least one absolute scale is required before any magnitude claim exists.
- Measured-value separation — needing a ruler does not derive the ruler’s value.
Candidate class
- A. No absolute scale anchor.
- B. One measured absolute anchor.
- C. A hidden scale smuggled through a formula.
Layer check
Scale is load-bearing here, acting across all three layers. The Stage supplies geometric formulas and ratios but no units by itself. The Rulebook must state whether a number is derived, paid, measured, consistency-only, dissolved, or open. The Actors produce observables — masses, couplings, rates, thresholds — but Scale decides how their magnitudes are cashed.
Root split
Shape supplies the structures and dimensionless readouts. Granularity prevents hidden precision or smuggled real-valued knobs. Scale supplies the anchor requirement and tags the measured value.
Elimination ledger
| Candidate | Constraint result | Verdict |
|---|---|---|
| No scale anchor | dimensionful claims are not unit-invariant | eliminated |
| Hidden scale in formula | unpaid measured information | forbidden |
| Declared anchor | unit-invariant ratios-to-anchor possible | survivor |
Survivor
At least one absolute scale anchor is required. In this framework the declared anchor is the ordinary Planck mass, MPl = 1.2209 × 1019 GeV. The value is measured, not derived.
Endpoint
Existence of a scale anchor: FORCED within the declared grammar (theorem-grade). Value of the anchor: PAID as a measured anchor.
Caveat
Do not merge “a ruler must exist” with “this exact ruler value was derived.” Only the first is a theorem; the second is a measurement.
Worked example
The hierarchy is a ratio, not a third ruler
Once the Planck scale and electroweak scale are both declared measured rulers, their hierarchy is not a third number to fit — it is their ratio. A deeper target-blind bridge would be an upgrade, but the framework does not pretend it already has one. Caveat: this is honest scale accounting, not a from-nothing derivation of vEW.
Full worked example — the hierarchy as two measured rulers
Question
Why is the electroweak scale so much smaller than the Planck scale?
Naive answer
The theory must derive the ratio from nothing, or the hierarchy remains a fatal failure.
Why the naive answer fails
Once the framework openly uses two measured rulers — the Planck scale and the electroweak scale — their ratio is not a third independent input. It is arithmetic. A theory may still seek a deeper bridge, but it may not pretend the ratio is both measured and separately predicted.
Constraint set
- Absolute-scale anchor rule — dimensionful values need measured rulers.
- No double-counting — the ratio of two paid rulers is not a third input.
- No laundering — measured rulers cannot be rebranded as predictions.
- Upgrade-path discipline — a target-blind bridge would strengthen the theory but is not silently assumed.
Candidate class
- A. Derive vEW and MPl from nothing.
- B. Measure both and count the ratio as arithmetic.
- C. Measure one and derive the other by a target-blind bridge.
Layer check
Scale is load-bearing. The Stage supplies geometric structures that can produce ratios and compactification maps. The Rulebook tracks which values were declared before comparison and prevents double-counting. The Actors generate mass observables only after the electroweak/Higgs sector is specified. Scale does the decisive work: MPl and vEW are measured anchors, and their hierarchy is their ratio.
Root split
Shape supports the structural relation and the Higgs/electroweak readout. Granularity prevents hidden continuous tuning or extra precision in the rulers. Scale carries the load: two measured anchors, one ratio.
Elimination ledger
| Candidate | Constraint result | Verdict |
|---|---|---|
| Derive both rulers from nothing | violates the anchor floor | rejected |
| Count the hierarchy as a third input | double-counts measured rulers | rejected |
| Treat hierarchy as ratio of two measured anchors | honest endpoint | survivor |
| Target-blind bridge from one ruler to the other | possible upgrade path | open unless supplied |
Survivor
H = vEW / MPl ≈ 2 × 10−17. The ratio is a consequence of two measured rulers, not a separate predicted magnitude.
Endpoint
PAID as measured anchors for the two rulers; the “third mystery” demand is DISSOLVED given scale discipline. A target-blind bridge remains an OPEN upgrade path.
Caveat
This does not derive the electroweak scale. It prevents the ratio from being double-counted or overclaimed.
Worked example
a₆ and the wrong kind of number
The geometry can produce exact scale-free heat-kernel checks, and those are meaningful audit objects. But a naked dimensionful a₆ magnitude in odd dimension is not a canonical physical observable. Scale keeps the rational invariant and dissolves the GeV-style overclaim. Caveat: exact does not mean observable; a local coefficient is not automatically a measured magnitude.
Full worked example — the a₆ coefficient and predicate validity
Question
When the geometry produces heat-kernel / a₆ coefficients, which part is a real audit result and which part is an ill-posed dimensionful magnitude?
Naive answer
Any formula that evaluates to a GeV-style magnitude is a prediction.
Why the naive answer fails
A local coefficient or a scheme-dependent dimensionful expression is not automatically a physical observable. In odd dimension, the dimensionful a₆ magnitude has no canonical finite local log/anomaly predicate. The scale-free rational checks can be meaningful while the GeV-style number is consistency-only or dissolved.
Constraint set
- Predicate validity — is the number a valid observable at this dimension?
- Local-vs-integrated distinction — a local coefficient is not automatically a measured record.
- Scheme/window accounting — dimensionful coefficients need a named scheme object.
- Object identity — the operator, ghost, connection, and bundle must match.
Candidate class
- A. Exact scale-free rational invariant.
- B. Local dimensionful coefficient.
- C. Integrated observable with full scheme / window / readout.
Layer check
The Actor layer is decisive. The Stage supplies the K₆ / S² / folded-circle geometry and curvature objects. The Rulebook supplies convention manifests, admissibility, and exact-check requirements. The Actors supply the physical graviton+ghost operator — not a Bochner ghost or a canonical proxy. A numerical match on the wrong operator closes nothing; object identity is enforced by exact fingerprints.
Root split
Shape supplies the correct geometry/operator object when all layers are included. Granularity prevents object swaps and requires two-route / exact checks. Scale decides that the scale-free rational is meaningful while the dimensionful odd-D magnitude is not a gate-closing observable.
Elimination ledger
| Candidate | Constraint result | Verdict |
|---|---|---|
| Scale-free rational invariant | unit-invariant, exact-checkable | bankable audit result |
| Dimensionful odd-D a₆ magnitude | no canonical finite predicate | dissolved / consistency-only |
| Observable magnitude | requires full integration, scheme, window, readout | open unless supplied |
Survivor
Scale-free rational invariants and exact identities are retained. The naked dimensionful magnitude is not promoted.
Endpoint
SCALE-FREE-DERIVED for the rational checks; SCALE-DISSOLVED / AUDIT-ONLY for the ill-posed dimensionful magnitude.
Caveat
Do not call the dimensionful a₆ magnitude a prediction. It is not a physical observable without the missing predicate/readout chain.
Honest endpoint
Scale is where the framework refuses to let a formula masquerade as a measurement. The dimensionless column is the strong one; the dimensionful column is honestly small and clearly labelled.