Layer 4 — Carrier-forcing and the given-E wall — rendered package. Rendered from layer-4-carrier-forcing.md; frozen technical content unchanged by rendering.

Layer 4 — Carrier-forcing and the given-E wall

Inside the "forces are geometry" game, the three internal carriers are forced — and the one cheaper rival, built end-to-end, breaks. What stays open is exactly the matter content $E$.

This is the layer where the program does its strongest concrete work, and where it meets its hardest honest limit. It splits cleanly into two halves. Layer 4A is the carrier-forcing result: within the rule "gauge forces are the isometries of compact internal factors," the weak, hypercharge, and color carriers of the 13D shape are each pinned down by a real structural theorem — and the most natural cheaper alternative, the eleven-dimensional $\mathbb{CP}^2$ variant, was constructed all the way through and breaks for a clean, non-circular reason. Layer 4B is the given-E wall: the entire stack bottoms on $E$, the observed Standard-Model chiral spectrum (in particular, exactly three generations), which no known principle forces. 4A is the result; 4B is the central open question. We state both as sharply as we can.

The question this layer answers

Suppose you accept the previous layers: that "simplest" should be measured by description length (Layer 1), that no exact Standard-Model label may be used for free (Layer 2), and that comparisons only make sense inside a declared search grammar (Layer 3). Then within Grammar A — geometry where gauge forces are the isometries of compact internal factors — one question dominates:

Given the rule "forces = isometries," and given the observed matter content $E$, are the specific carriers of the 13D shape forced, or merely chosen?

The honest answer has two parts. The carriers $S^2$ (weak), $S^1_Y/\mathbb{Z}_2$ (hypercharge), and $K_6=SU(3)/T^2$ (color) are forced inside the grammar by structural theorems — two of them architecture-neutral, the third now backed by an end-to-end elimination of its cheapest rival. That is 4A. But the index that delivers three families is a rigid integer given a chosen bundle on $E$, not a selection of $E$ itself. That is 4B, and it is where the program is honest about what it has not done.

The full 13D shape is

$$\mathcal{M}_4 \times K_6 \times S^2 \times S^1_Y/\mathbb{Z}_2, \qquad K_6 = SU(3)/T^2, \qquad \dim = 4+6+2+1 = 13.$$

Layer 4A asks whether the internal block $K_6 \times S^2 \times S^1_Y/\mathbb{Z}_2$ is forced once you fix the grammar and demand clean, separable Standard-Model gauge factors.

4A — Carrier-forcing inside Grammar A

The genuine content of the program is not "we recover the Standard-Model gauge group." Recovering $SU(3)\times SU(2)\times U(1)$ is a tie across every serious framework — string constructions, noncommutative geometry, and a plain 4D effective theory can all host it. The real content is carrier-forcedness: within "forces = isometries," can you show that no cheaper carrier does the job for each factor? For two of the three carriers the answer is a clean general theorem; for the third it is a constructed elimination of the only serious cheaper rival.

The weak carrier $S^2$ — theorem F1

The weak force is $SU(2)_L$, which is non-abelian. The cheaper directions one might try — a circle $S^1$, a torus $T^k$, any flat or abelian carrier — are all closed at once by a general fact:

F1. No abelian or toroidal carrier, of any dimension, has non-abelian $SU(2)$ among its isometries.

The isometry group of a flat torus is abelian (translations plus a discrete point group); it cannot contain $SU(2)$. So the cheaper-by-dimension direction does not merely lose to $S^2$ — it cannot host the weak force at all. This is an architecture-neutral result: it closes whole shelves of candidate carriers, not one named competitor. The minimal carrier whose isometry group is $SU(2)$ is the round two-sphere $S^2$, with isometry group $SO(3)$ (Lie algebra $\mathfrak{su}(2)$). The weak rung is forced, and forced cheaply.

The hypercharge carrier $S^1_Y/\mathbb{Z}_2$ — theorem F2

Hypercharge $U(1)_Y$ is abelian, so a single circle $S^1$ supplies it by isometry. The subtlety is chirality. A bare odd-dimensional closed factor keeps both handednesses of every fermion:

F2. A bare odd closed circle keeps both chiralities → a mirror fermion for every Standard-Model fermion → excluded by experiment.

A full mirror sector at or below the electroweak scale is ruled out by the measured invisible width of the $Z$ boson at LEP (the $Z$ couples to a fixed, counted number of light neutrino species; mirror generations would show up). So a bare $S^1$ is not admissible. The fix is the orbifold fold: $S^1_Y/\mathbb{Z}_2$. The $\mathbb{Z}_2$ identification projects out one handedness — it is needed for chirality, not decorative. The hypercharge rung is therefore forced to be the folded circle, and F2, like F1, is architecture-neutral: it closes all cheaper bare-circle carriers, not one rival.

These two — F1 (weak) and F2 (hypercharge) — are the strongly-forced rungs: two of the three internal dimensional choices are pinned by general theorems that close whole shelves. That is already a stronger and more defensible statement than "the carriers are merely category-relative choices."

The color carrier $K_6 = SU(3)/T^2$ — abelian-isotropy uniqueness

Color is $SU(3)_c$, the hard rung, because there is more than one compact $SU(3)$-homogeneous space and the cheapest is lower-dimensional than $K_6$. Classify the carriers $M = SU(3)/H$ by $\dim M = 8 - \dim H$:

Carrier Form $\dim$ Chiral index?
$\mathbb{CP}^2$ $SU(3)/U(2)$ 4 yes — Spin$_c$ Dirac index
$S^5$ $SU(3)/SU(2)$ 5 no (closed odd-dim → no net chiral index)
Wu manifold $SU(3)/SO(3)$ 5 no (closed odd-dim)
$K_6$ $SU(3)/T^2$ 6 yes — Borel–Weil–Bott index

The two five-dimensional candidates ($S^5$, the Wu manifold) are killed by odd-dimensionality: a closed odd-dimensional manifold has no net chiral Dirac index, so it cannot carry chiral generations. That leaves a genuine, cheaper rival: $\mathbb{CP}^2 = SU(3)/U(2)$, four-dimensional against $K_6$'s six — the minimal-dimension $SU(3)$ carrier, nothing competes below it. Under a dimension-first metric $\mathbb{CP}^2$ wins this rung outright, and an eleven-dimensional total shape $\mathcal{M}_4 \times \mathbb{CP}^2 \times S^2 \times S^1_Y/\mathbb{Z}_2$ ($4+4+2+1 = 11$) would be two dimensions cheaper. So the color rung is the one that has to be earned, not asserted.

It was earned by building the $\mathbb{CP}^2$ variant end-to-end, as a faithful parallel of the 13D method, and testing honestly whether it reproduces the full Standard-Model target — built neither to succeed nor to fail. The result is that $\mathbb{CP}^2$ breaks, and for a reason deeper and cleaner than the program's original objection. The principle that emerged:

Abelian-isotropy uniqueness. Among the $SU(3)$ cosets $SU(3)/R$ used as the color carrier, the maximal torus $T^2$ is the unique isotropy that is purely abelian — its centralizer is $C_{SU(3)}(T^2) = T^2$ (the Cartan subalgebra only). It injects no spurious non-abelian factor and keeps color, weak, and hypercharge in independent commuting factors. Any larger isotropy ($U(2)$, $SU(2)$) is non-abelian and collides with the Standard Model's own $SU(2)\times U(1)$. Hence $K_6 = SU(3)/T^2$ is the unique clean $SU(3)$ color carrier.

The mechanism is the centralizer rule of coset-space dimensional reduction: when you reduce on $G/R$, the surviving 4D gauge group is the centralizer $H = C_G(R)$ of the isotropy $R$ in $G$, and the isotropy is an active determinant of the gauge group and matter — not a passive label. This single corrected premise is what produces the break rather than rigging it.

For $K_6 = SU(3)/T^2$ the isotropy is the abelian maximal torus. The Lie algebra splits as $\mathfrak{su}(3) = \mathfrak{t}^2 \oplus$ (six root spaces), each root space a one-dimensional $U(1)\times U(1)$ character; the centralizer is the Cartan alone. The isotropy injects no non-abelian content, and color stays a clean separate factor.

For $\mathbb{CP}^2 = SU(3)/U(2)$ the isotropy $U(2) = (SU(2)\times U(1))/\mathbb{Z}_2$ is non-abelian and a maximal-rank subgroup of $SU(3)$. The adjoint of $SU(3)$ branches under $U(2)$ as

$$\mathbf{8} \to \mathbf{3}_0 \oplus \mathbf{1}_0 \oplus \mathbf{2}_{+3} \oplus \bar{\mathbf{2}}_{-3},$$

and the four tangent dimensions of $\mathbb{CP}^2$ are the fundamental $\mathbf{2}$ of an $SU(2)$ sitting inside color $SU(3)$. By the centralizer rule this isotropy is gauge-active, which forces a fatal fork when you try to build the full theory:

Sharpened by a direct centralizer computation: $C_{SU(3)}(U(2)) = U(1)$ — a single $U(1)$, not $SU(2)\times U(1)$ (by Schur's lemma, $U(2)$ acts irreducibly on the top two-block, so the centralizer is the residual diagonal $U(1)$). So $\mathbb{CP}^2$ does not even hand you $SU(2)_L$ as a surviving gauge factor from the coset.

Four sectors of the variant — the gauge group, the families/chirality index, the center-kernel and anomalies, and proton/Higgs/normalization — were built independently and converged on the same break at the same place (the gauge group), through the same mechanism (a non-abelian $U(2)$ isotropy inside color $SU(3)$). Convergence of independent builds on one mechanism is the signature of a genuine structural break, not a coincidental failure. The honest counterpoint was kept: several results do cohere on $\mathbb{CP}^2$ — the electric charges $Q = T_3 + Y$ for all charge eigenstates, the $\mathbb{Z}_6$ center-kernel, all six per-family anomaly coefficients, and one full Standard-Model generation as Dirac zero modes. But these are all properties of $E$ (true on either carrier because $E$ is input on both sides), not discriminators. The discriminator is factor-separability, and there $\mathbb{CP}^2$ breaks while $K_6$ holds.

A note on the family count, to be precise rather than to over-claim. The program's original objection to $\mathbb{CP}^2$ — "its family count is a tunable real" — is weaker than stated. On $\mathbb{CP}^2$ (which is Spin$_c$, with $\sigma(\mathbb{CP}^2)=1$) the Dirac index is

$$\mathrm{ind} = \frac{c_1(\mathcal{L})^2 - \sigma}{8} = \frac{(2r+1)^2 - 1}{8} = \frac{r(r+1)}{2}, \qquad r=2 \Rightarrow \mathrm{ind} = 3,$$

a discrete topological integer, not a continuously tunable modulus. But $K_6$ also yields three only after a bundle choice: the Borel–Weil–Bott index on $SU(3)/T^2$ is

$$\mathrm{ind}(L_{a,b}) = \pm\frac{(a+1)(b+1)(a+b+2)}{2},$$

which gives $|\mathrm{ind}| = 3$ at weight $(a,b)=(1,0)$, and the frozen branch carries $\chi(K_6, E) = -3$. So the family-count comparison is symmetric — both carriers reach three only by selecting a bundle — and is therefore not the decider. The clean, asymmetric, non-circular decider is the factor-separability break above. Keeping that distinction straight is exactly the discipline these pages are about: we let the gauge-structure argument carry the weight, and concede that the counting argument, on its own, is symmetric.

One more honesty flag, because over-claiming here would be tempting. The eleven-dimensional $\mathbb{CP}^2 \times S^2 \times S^1/\mathbb{Z}_2$ ansatz is a legitimate Kaluza–Klein / Freund–Rubin supergravity construction in the unique 11D dimension — but it is not M-theory on a $G_2$-holonomy manifold. A Riemannian product has reducible product holonomy, whereas $G_2$ acts irreducibly on $\mathbb{R}^7$; $\mathbb{CP}^2$ is positively curved (Fano) while $G_2$ holonomy demands Ricci-flatness; and Kaluza–Klein gauge symmetry comes from continuous isometries while non-abelian $G_2$ gauge comes from codimension-four singularities. Matching dimension counts $(4+7)$ is necessary, never sufficient. We say this is a dimension coincidence, not a mesh — that is the honest read.

What 4A establishes

Putting the three rungs together:

$$G_{\text{isometry}} + T + C_{\text{MDL}} \;\Rightarrow\; K_6 \times S^2 \times S^1_Y/\mathbb{Z}_2 \quad \textbf{inside the grammar.}$$

Two rungs (weak, hypercharge) are forced by architecture-neutral theorems that close whole shelves. The third (color) is forced by abelian-isotropy uniqueness, now backed by an end-to-end build showing the only cheaper $SU(3)$ carrier breaks at the gauge group for a real structural reason. This is the strongest current result in the program — and it is the cleanest concrete elimination it contains. Selection $\neq$ derivation: what is shown is that within "forces = isometries," and given $E$, the internal carriers are forced; it is not a proof that this grammar, or this shape, is the only one nature could use.

4B — The given-E wall

Now the hard limit, stated as plainly as the result above. The entire carrier-forcing stack bottoms on $E$ — the Standard-Model chiral spectrum, and in particular exactly three generations. Read 4A back carefully and the dependence is explicit: F1 and F2 assume we are matching the observed non-abelian-weak, chiral, mirror-free content; the color rung's three families come from the index $\chi(K_6, E) = -3$, which is a rigid integer given a chosen bundle on $E$ — a property of the selected bundle, not a selection of $E$.

This is the difference the discipline lines protect. Given-E $\neq$ derivation-of-E: the framework reads three families off the geometry once a bundle (selected by the Standard-Model charges) is fixed, but it does not derive the charge content, the gauge representations, or the number "three" from a principle that operates before $E$ is known. The integer is rigid; the bundle that makes it rigid is chosen to fit $E$.

To make this concrete: "3" requires a bundle choice on both carriers — weight $(1,0)$ on $K_6$, a particular Spin$_c$ twist on $\mathbb{CP}^2$. The number is forced given the bundle; the bundle is selected by the matter charges. So the generation count traces back to $E$, exactly as the spectrum does. The honest standing is "carrier-forced geometry, given selected $E$," not "the shape forces $E$."

And the stronger claim — "$E$ is forced" — is currently refuted, for a reason worth stating because it shows the discipline is real and not rhetorical. Local anomaly freedom is sometimes advertised as selecting the Standard Model. It does not. Anomaly cancellation is a filter, not a selector: $A(E_{\text{SM}}) = 0$ tells you the observed spectrum passes, but the equation $A(E) = 0$ has infinitely many solutions. To upgrade filter to selector one would have to prove a singleton theorem,

$$\ker A \cap \mathcal{C}_{\text{admissible}} = \{E_{\text{SM}}\},$$

i.e. that within the admissible class the observed spectrum is the only anomaly-free one. That is not proven — and on the contrary, anomaly-free families form an infinite set. So $E$ remains a primitive input. (This is treated in full in the hypercharge-anomaly dossier; the short version is that "anomaly cancellation selects the Standard Model" is false.)

The one route that could move the wall — the bundle-uniqueness fork

There is exactly one route that could remove the "given-E" qualifier on the family count, and it is a named, finite target, not a wish. It is a bundle-uniqueness theorem:

Enumerate the admissible homogeneous bundles (weights, lifts) on $K_6$; apply the index machinery; and show that a minimality / admissibility criterion uniquely selects a weight of index magnitude 3 — without invoking "three generations" as input.

Equivalently in the comparative form that was tested directly: does $E$ (the Standard-Model charges) canonically force the $K_6$ Spin$_c$ bundle, via the $T^2$ isotropy embedding of those charges, while leaving genuine residual freedom on $\mathbb{CP}^2$'s $U(2)$ isotropy bundle? The fork:

The honest prior is that this is hard and may not hold. Equivariant bundles over $G/H$ are governed by the representation theory of $H$, so the question reduces to comparing $\mathrm{Rep}(T^2)$ against $\mathrm{Rep}(U(2))$. The abelian $T^2$ carries only characters and cannot host an irreducible weak doublet without extra pairing data, whereas $U(2)$ carries weak-isospin and hypercharge directly. By that measure $U(2)$ is, if anything, the more natural host — so no architecture-neutral rule obviously privileges $T^2$ over $U(2)$ at the bundle level. (This is precisely why the decisive elimination of $\mathbb{CP}^2$ was relocated upstream of the bundle, to the gauge-group factor-separability break of 4A, which does not depend on how "3" is obtained.) A clean result either way — closing the theorem, or showing it cannot close — is decisive and valuable; a refutation here is as informative as a proof.

So 4B is not a hole we hide. It is the sharpest open question the program has, with the one move that could change it named precisely.

Status — reached vs open

Reached (4A — the strong result). Inside the internal-isometry grammar, given $E$, the internal carriers of the 13D shape are forced:

This is the cleanest concrete elimination in the program. It is an in-grammar result.

Open (4B — the central question). The whole stack bottoms on $E$. $\chi(K_6, E) = -3$ is a rigid integer given a chosen bundle, not a selection of $E$. "$E$ is forced" is currently refuted (anomaly freedom is a filter with infinitely many solutions, not a selector). The one named target that could move the wall is the bundle-uniqueness theorem; the honest prior is that it is hard and may not close, and the realistic ceiling may remain "a rigid integer given a selected bundle." The honest top line for Layer 4: forced within the declared grammar and metric — not absolute, and not E-derived.

The three discipline lines hold throughout:

The honesty here is the methodology: 4A is a real win, stated confidently and bounded exactly; 4B is a real wall, stated openly with the single theorem that would change it named in full.

Cross-links

Sibling anchor pages

Gate dossiers