Master / Fundamental Anchors

The deep roots are too abstract to close gates directly. They become useful through master anchors: finite records, no-unpaid-label accounting, measured scale, frozen branch identity, given-$E$ discipline, and status-limited cost ledgers.

What this page is for

The seven deep roots — the Record Interface, Invariance, Granularity, Scale, Shape, Causal Order, and Nonseparability — are the abstract bedrock of the program. But you cannot hand a deep root to a gate page and ask it to close a gate. A gate test needs something concrete: a measured number, an exact ledger entry, a frozen object to test against, a charged spectrum, a reproducible hash. The master anchors are exactly that concrete layer. They are the observable, auditable, measured, charged, or bridge quantities that translate the deep roots into something a gate evaluation can actually use.

This page sits in the middle of a strict, one-directional hierarchy:

Deep roots → master anchors → gate-specific anchors

Read it this way: a deep root is a principle about how nature must be described; a master anchor is the operational handle that principle gives us; a gate-specific anchor is the particular measured value, charged label, or open residual a single gate page then consumes. Nothing flows the other way. A gate anchor never validates a root; it inherits its status from the master anchor above it, and that status is always one of the allowed labels below — never a silent promotion to "derived."

The discipline that governs every line on this page is the program spine: selection ≠ derivation · given-$E$ ≠ derivation-of-$E$ · frozen/reproducible ≠ proven-unique. Master anchors make gate anchors admissible, traceable, measured, charged, or open. They do not automatically generate gate anchors, and they never upgrade a DERIVED-GIVEN-E result into a derived-E one.

The master anchors

Each master anchor below names the deep root it descends from, its honest status (using only the allowed status labels), and how gate pages use it. This is the central table of the page; gate pages cite into it.

A note on reading the table: the Status column is binding. Where it says MEASURED, the value is an accepted experimental input, not a shape-derived output. Where it says GIVEN-E / CHARGED, the object is bundled into Shape as a primitive unless and until a separate uniqueness theorem removes the "given." Where it says AUDIT ONLY, the artifact certifies object identity and freeze integrity — it certifies nothing physical. Where it says the metric is still OPEN, the comparison rule is a bridge under pressure, not a proven law.

The bridge logic — from roots to anchors

The middle three rows of the hierarchy are not decreed; they are bridged. Below are the load-bearing calculations and the honest status of each bridge.

Granularity → no unpaid labels

The Granularity root, in shorthand, is the assertion that there is a finite operational cell — distinctions below it are not physical records:

$$\exists\,\epsilon_{\rm cell}>0.$$

From this the no-unpaid-labels accounting rule follows. If a theory asserts an exact physical distinction, that distinction must be specified, and specification against a finite cell costs bits:

$$\text{exact physical distinction}\;\Rightarrow\;\text{nonzero specification burden}.$$

Applied to the actual exact labels a candidate theory uses, this becomes the master accounting rule that gate pages enforce. Every gauge label, charge table, bundle $E$, coset quotient, Wilson/mixing angle, or selector rule is, for each such label, either generated by declared invariant structure, charged as primitive input, measured, or left open:

$$ \begin{aligned} &\text{gauge label, charge table, bundle }E,\ \text{quotient, Wilson angle, selector rule}\\ &\qquad\Rightarrow\;\text{GENERATED, CHARGED, MEASURED, or OPEN.} \end{aligned} $$

There is no fourth option and in particular no free exact label. This is the master anchor "no unpaid exact labels." Its status is an audit theorem / accounting rule — it is a statable, defended theorem about what counts as reproducing a target, not a claim that any particular label is generated.

Scale → $M_{\rm Pl}$

The Scale root fixes that there is an absolute physical scale at which the program's geometry and quantum-gravitational regime live. The master anchor it yields is the Planck mass, related to the other scale constants by

$$M_{\rm Pl}\sim\sqrt{\frac{\hbar c}{G}}.$$

The essential honesty here: $M_{\rm Pl}$ is a MEASURED / accepted value anchor. It is read in from experiment (via $\hbar$, $c$, $G$) and used to normalize scale in gate tests. It is not shape-derived — the frozen 13D geometry does not output $M_{\rm Pl}$; it consumes it. Gate pages that normalize against $M_{\rm Pl}$ are using a measured input, not claiming a derivation of scale.

Shape → 13D branch

The Shape root, made concrete, is the single frozen branch the whole program tests. The branch is

$$M_4\times K_6\times S^2\times S^1_Y/\mathbb Z_2,\qquad K_6=SU(3)/T^2.$$

Its dimension count is fixed by the factor dimensions:

$$\dim SU(3)=8,\qquad \dim T^2=2,\qquad \dim K_6=6,$$

$$D=4+6+2+1=13.$$

The master anchor here is the frozen 13D branch, with status DECLARED ROOT / AUDIT: it is the object under test, selected and frozen, not derived or proven absolutely unique. Within the internal-isometry grammar the branch is the strongest known survivor (the weak carrier $S^2$, the chirality-forcing $\mathbb Z_2$ fold, and the abelian-isotropy color coset $K_6$ eliminate cheaper shelves; the $CP^2=SU(3)/U(2)$ route was built end-to-end and breaks at the gauge gate). That in-grammar minimality is partial and hardening, not closed. Architecture-neutral minimality — "no simpler competitor across all grammars" — is OPEN and likely unprovable in full, because it is a universal negative. On the requirement board — 33 resolved at +0 · 0 anchored at +1 · 0 open, with 0 of 33 physics-closed stated plainly on it — the corresponding gate, DeepRoot — Shape, the 13D selector, stands RESOLVED at +0 (DERIVED-GIVEN-anchor): its one economy axiom is charged openly, and the universal-negative demand dissolves rather than standing as a debt. The uniqueness caveat above is the residual shown beside that closure, never a hedge rolled into it.

Nonseparability → local/global discipline

The Nonseparability root says the world does not factorize into independent local pieces. Written for two subsystems, exact independence would be the product state

$$\rho_{AB}=\rho_A\otimes\rho_B,$$

but the generic physical case is correlated / nonseparable:

$$\rho_{AB}\neq\rho_A\otimes\rho_B.$$

The master anchor this yields is global consistency / no local-isolated closure, an audit principle. Its consequence for gate evaluation is a discipline rule: closing a single local leg of a gate is not the same as closing the whole gate. A local gate result must be embedded back into the global quantum/consistency residuals before the gate as a whole can be called closed. Local closure ≠ whole-gate closure — gate pages must carry the global residual explicitly.

Note carefully what this does not license: nonseparability is a statement about correlations in the joint state, not about signaling. It does not send a usable superluminal signal, and no gate may invoke it as if it did.

Secondary-anchor clarifications

Several constants are routinely treated as "naked numbers." On this program they are not roots, and they are not free fitted dials. Each is typed honestly:

• $c$ is not a naked number. It is the operational expression of local causal-cone invariance (Causal Order + Invariance). Gate pages use it as an invariant defining frame/causal admissibility, not as a tunable scalar.
• $\hbar$ is not just a fitted number. It is the action-to-phase / cost-floor residue (Granularity + quantum phase): the bridge that converts action into quantum phase and marks the floor below which classical record-distinctions dissolve. Its status is a residue / scale bridge. (Full accounting: the granularity deep-root gate — RESOLVED at +0, CERTIFIED-IRREDUCIBLE — including the two counter-models proving the cost floor cannot be had for free.)
• $k_B$ is not the root of record cost. It is an entropy–temperature / state-counting bridge — a secondary bridge that lets thermodynamic and information-theoretic record-cost language be stated in common units. It is explicitly demoted to a bridge, not a root.
• $G/M_{\rm Pl}$ is meaningful only through scale relations. Neither $G$ nor $M_{\rm Pl}$ carries physical content as a naked, unit-dependent value; they enter only through dimensionless comparisons and scale relations (e.g. via $M_{\rm Pl}\sim\sqrt{\hbar c/G}$), never as raw numbers a gate "explains."

Anti-overclaim box

This box is binding on every gate page that cites a master anchor. It records the four claims this layer specifically forbids:

• Master anchors do not automatically generate gate anchors. They make gate anchors admissible — they do not manufacture them.
• Master anchors make gate anchors admissible, traceable, measured, charged, or open — and nothing stronger by default.
• derived-given-E is not derived-E. A result that follows given the observed spectrum $E$ is not a derivation of $E$. The "given-$E$" qualifier may never be dropped.
• MDL pressure is not absolute MDL proof. Full-generator / description-length cost is a pressure used for competitor comparison; the metric that would make it binding is still OPEN, and it is not proven against every competing metric.

Allowed status labels

Every claim on every gate page that cites this layer must carry exactly one of these labels — no others, and no unlabeled claims:

• DECLARED ROOT — a frozen, selected anchor (e.g. the 13D branch as object under test).
• GENERATED — produced by declared invariant structure.
• DERIVED-GIVEN-E — follows given the frozen shape and the observed spectrum $E$.
• MEASURED — an accepted experimental value anchor (e.g. $M_{\rm Pl}$).
• CHARGED — paid as primitive input (e.g. $E_{\rm SM}$ / the matter bundle).
• AUDIT ONLY — certifies object identity / freeze integrity, not physics (e.g. hashes, manifests).
• OPEN — a named, finite, unclosed target (e.g. bundle-uniqueness, the metric bridge).
• BLOCKED — a leg whose closure is gated on another unclosed residual.
• ANTI-CLAIM — an explicitly forbidden overclaim recorded so it cannot be smuggled back in.

Global forbidden overclaims

These are ANTI-CLAIMs. They must be absent from this page and from any gate page citing it:

• Do not claim all deep roots are theorem-grade. (Some roots — granularity, the chosen metric, scale — are declared anchors, not theorems.)
• Do not claim the 13D shape is absolutely unique. (Architecture-neutral minimality is OPEN, a universal negative.)
• Do not claim $E$ is derived unless a separate bundle-uniqueness proof is supplied. ($E$ is GIVEN-E / CHARGED; "$E$ is forced" is currently refuted, since anomaly-freedom is a filter with infinitely many solutions.)
• Do not claim local gate closure equals whole-gate closure.
• Do not claim hashes validate physics; hashes validate frozen-object identity / audit integrity only.
• Do not claim anomaly cancellation selects the Standard Model. (It is a filter — $A(E_{\rm SM})=0$ means $E_{\rm SM}$ passes; it does not single $E_{\rm SM}$ out.)
• Do not claim MDL / full-generator cost is absolutely proven against every competing metric.
• Do not claim measurement / nonseparability sends a usable superluminal signal.

Admissible / traceable vs derived — the distinction that holds the page together

The single most important thing this layer does is keep two ideas apart that are easy to blur:

• A gate anchor is admissible / traceable when a master anchor makes it usable in a gate test — it can be measured, charged, frozen, or marked open, and its provenance traces cleanly up the hierarchy to a deep root. This is what master anchors provide.
• A gate anchor is derived only when declared invariant structure actually generates it (status GENERATED), or generates it given the spectrum (status DERIVED-GIVEN-E) — and even then the "given-$E$" qualifier stays attached.

Admissible is not derived. Traceable is not derived. A measured input is not a derivation; a charged primitive is not a derivation; a frozen object under test is not a derivation of itself. The hierarchy deep roots → master anchors → gate anchors is a hierarchy of admissibility and accounting, and only the explicitly GENERATED / DERIVED-GIVEN-E rows cross the line into derivation — under their stated qualifiers.

Cross-links

• The Anchor Hierarchy — overview — where master anchors sit between deep roots and gate anchors.
• The Seven Deep Roots — the abstract roots this page makes operational.
• Master / Fundamental Anchors — this page.
• The Logical Endpoint Proof — how far the chain honestly reaches and where it bottoms out on given-$E$.
• Internal methods (transparency) — the working documents behind these pages (the per-gate ledger template and the page-completion protocol), kept public so the build discipline itself can be inspected.
• Blind Spots & Implicit Assumptions — the false-closure and over-promotion traps this layer guards against.
• The Anchors — overview — the master anchor A0 and the full bridge stack at a glance.

Completion report

Tests passed. - P1 — all required master anchors present (the full eleven-row master-anchor table is included verbatim-in-substance). - P2 — the three required calculations present: granularity → no unpaid labels ($\exists\,\epsilon_{\rm cell}>0$ and the specification-burden chain), scale → Planck ($M_{\rm Pl}\sim\sqrt{\hbar c/G}$), shape → 13D ($M_4\times K_6\times S^2\times S^1_Y/\mathbb Z_2$, $\dim$ count, $D=13$). - P3 — nonseparability local/global discipline present ($\rho_{AB}=\rho_A\otimes\rho_B$ vs $\rho_{AB}\neq\rho_A\otimes\rho_B$; local closure embedded in global residuals). - P4 — $k_B$ demoted to a bridge, not a root (secondary-anchor clarifications + table status). - P5 — $E$ labeled GIVEN-E / CHARGED unless separately derived (table + bridge logic + anti-overclaim box). - P6 — hashes stated as AUDIT ONLY (table + anti-overclaim box). - P7 — "admissible / traceable" explicitly distinguished from "derived" (dedicated section + anti-overclaim box). - P8 — completion report present (this block). - U1 — status honesty: every major claim carries an allowed status label. - U2 — no floating anchors: every named anchor is typed (root, master anchor, gate anchor, measured input, audit artifact, open residual, or anti-claim). - U3 — no root inflation: the page explicitly states not all roots are theorem-grade. - U4 — no E-smuggling: $E$ is given / charged / bundled into Shape unless separately derived. - U5 — no hash-overclaim: hashes are audit-integrity only. - U6 — local/global distinction: local closure is explicitly not whole-gate closure. - U7 — forbidden claims absent: no absolute uniqueness, no anomaly-selector claim, no $\rho$/hierarchy overclaim, no FTL nonseparability claim. - U8 — completion report present with the four required headings.

Tests failed. None.

Open items. - The full-generator / MDL metric bridge remains OPEN: "shortest generator" is metric-relative until the selecting theorem (granularity ⇒ MDL is the right simplicity order) is proven. - Bundle-uniqueness remains OPEN: removing the "given-$E$" qualifier requires a separate theorem; until then $E$ stays GIVEN-E / CHARGED. - Architecture-neutral minimality (Level 3) remains OPEN and likely unprovable in full (universal-negative wall); in-grammar minimality (Level 2) is partial and hardening.

Assumptions made. - Sibling-page filenames follow the slugs supplied in the section’s build specification (anchor-hierarchy-overview.html, deep-roots.html, logical-endpoint-proof.html, gate-traceability-template.html, blind-spots.html, ai-agent-completion-protocol.html); cross-links target those, plus the existing ./index.html. - The deep-root names (Record Interface, Invariance, Granularity, Scale, Shape, Causal Order, Nonseparability) match the "seven deep roots" used by the sibling pages and the source-of-truth. - Status labels and forbidden overclaims are used exactly as enumerated in the section specification; no additional labels were introduced.