Lumen › Physics track › Chapter 5
Matter & Gauge Derived ⊣
This is the chapter that has to deliver, because it is where geometry meets the particle-data booklet — the masses, the mixing angles, the couplings that any theory of matter is finally judged by. The claim is as large as claims get: the whole Standard Model, at no adjustable parameters, from one exceptional piece of algebra the theory argues could not have been anything else. To keep that from being a boast, the verifier here does not simply assert a mass. It computes one — the tau lepton’s — by integrating the same cubic polynomial we started with, and lands within eight hundredths of a percent.
Why there are exactly three families
Of all the riddles in the particle table, the one that has most stubbornly resisted explanation is the number three. There are three copies of every kind of matter — three generations — identical but for their masses, and nothing in the Standard Model says why three and not two or seven. This theory’s answer is not a fit but a piece of pure mathematics, and it is satisfying enough to walk through slowly. Matter lives in a particular algebra, $J_3(\mathbb{O})$ — the three-by-three Hermitian matrices whose entries are octonions. Three facts about it converge. First, the octonions are the last number system in the division-algebra ladder; there is nothing beyond them. Second — and this is the crux — you can only build a consistent Jordan algebra out of octonionic matrices up to size three; at four-by-four the construction breaks, because the octonions refuse to associate. Third, the resulting object is the unique exceptional algebra of its kind. Put those together and the conclusion is forced: three families because three is the largest matrix the octonions will tolerate. There is no fourth generation for the same reason there is no four-by-four version of this algebra — the mathematics simply stops. The deepest brute fact of the particle table turns out to be a theorem about when octonionic matrices still make sense.
Watching a mass appear
Granting the algebra, the theory reconstructs the particle table to a root-mean-square accuracy of about one and a half percent across twenty-odd observables, with not a single number free to adjust — the electron fixes the overall scale, and everything else is geometry. Rather than parade the table, it is more honest and more convincing to follow one entry all the way down, which is what the verifier does. The mass of a particle, in this picture, is the energy of the geometry observing itself at that particle’s characteristic depth. For the tau, that depth is a number computed from the density’s moments, $\lambda_\tau = 0.442$; feed it back in and the self-lensing energy shifts to $13.42$; exponentiate, and
$\dfrac{m_\tau}{m_e} = \exp\!\big(\mathrm{MU}\cdot(E_{\text{self}}(\lambda_\tau) - \pi)\big) = 3474.6,$
against a measured $3477.4$. That is eight hundredths of a percent, and every quantity in it was integrated from the cubic polynomial — nothing was looked up, nothing was fit. The rest of the table reads the same way at the same level: the muon a little over half a percent off, the Higgs at $125.5$ against $125.2$ GeV. The Weinberg angle, which governs how the electromagnetic and weak forces mix, comes out of the layer fractions as $\sin^2\theta_W = \tfrac{1}{1+\pi}$ at leading order, and once the one universal loop correction is applied, $\tfrac{1}{1+\pi}\big(1 - \tfrac{4\lambda^2}{5}\big) = 0.2319$, landing on the measured $0.2312$ to a third of a percent. (A tidy near-coincidence, $3/(8\varphi)$ with the golden ratio, sits a hair away at $0.2318$; the theory is careful to call it a pattern, not the derived value — the derived value is the loop-corrected ratio.)
The quark ladder
The leptons set the method; the quarks test whether it generalises. Every fermion mass in the theory is the same exponential of a structural energy $E$ read off the geometry,
$m = m_e\,\exp\!\big(\mathrm{MU}\cdot(E-\pi)\big),\qquad \mathrm{MU}=\mu_1/\mu_0,$
and the quark energies are small combinations of $\pi$ and the Fano number seven (the count of imaginary octonion units). The charm is the cleanest: its energy is exactly $\pi^2+\pi$, which is the analytic value of a density integral, and the predicted mass lands at $+1.2\%$. The up sits at $\pi^{7/5}$ ($+0.6\%$), the strange one Fano step below the muon at $\pi^2-\tfrac17$ ($+1.6\%$). The single cleanest light test needs no spectrum at all — the down-to-up ratio is $\exp(2\,\mathrm{MU}\,\lambda^2\pi^2)$, a pure Weyl step that matches to half a percent. Across all six quarks the root-mean-square residual is about two percent, with the electron as the only input.
Two of the six are weaker than the others, and the theory says so on its register. The down energy $\ln(\mu_1)/\mathrm{MU}$ and the top energy (equivalently the statement $m_t/m_c = \alpha^{-1}$) are “structurally targeted” — proved as trace identities but not derived from first principles, two named open problems. And the bottom quark is the honest outlier: its bare structural value is $-3.4\%$ low, and it closes to $-0.25\%$ only with a QCD scheme correction built from Coxeter numbers — a correction that consumes the measured strong coupling $\alpha_s$ rather than predicting it. Which points at the real gap: the theory does not derive $\alpha_s$ at all. Its attempts are off by a large factor, and every place the strong coupling is used, it is taken from experiment. That is stated plainly here and flagged in the verifier.
The gauge group from the number systems
The forces come from the same division-algebra ladder as the families — the alignment the Substrate chapter previewed, where the only four spheres you can comb flat turn out to be the only four number systems. Each rung carries a symmetry: the complex numbers carry a circle $U(1)$, which is electromagnetism living on the Hopf fiber; the quaternions carry $SU(2)$, which is the weak force and is literally the three-sphere $S^3=SU(2)$; the octonions carry $G_2$, their automorphism group, inside which sits the $SU(3)$ of the strong force as the stabiliser of one imaginary unit. So the Standard Model’s $U(1)\times SU(2)\times SU(3)$ reads as the symmetry content of $\mathbb{C}$, $\mathbb{H}$, $\mathbb{O}$.
Here, though, the parallel to the families is weaker, and the chapter has to say so plainly — because the families really were forced. There is no four-by-four octonionic Jordan algebra, so there is no fourth generation; that was a theorem. The gauge group is not that. It is a structural assignment: the ladder identifies the three groups, but the source flags in its own text that the mapping does not construct the hypercharge representation, and that the argument is only that $U(1)\times SU(2)\times SU(3)$ is the maximal subgroup compatible with the octonionic structure — not a uniqueness proof. The dimensions and ranks that go into it ($\dim G_2 = 14$, the $\mathbf{3}\oplus\bar{\mathbf{3}}$ of colour inside the octonions) are exact; the claim that this forces the Standard Model gauge group, the way the algebra forces three families, is the part still owed a proof. So: the same ladder, but a theorem for the families and a well-motivated assignment for the forces.
One algebra, two readings: mass and mixing
The most beautiful structural result in the whole theory is also, once you see it, the simplest. Take the matter algebra in the basis of the three families and it splits cleanly in two. Down the diagonal sit three real numbers — each family coupled to itself — and those are the masses. Off the diagonal sit the entries that connect one family to another, and those are the mixings, the small probabilities that one kind of matter turns into another, and the CP violation that lets the universe tell matter from antimatter. So the entire flavour structure of the Standard Model is a single distinction, read twice: a family coupled to itself is its mass, a family coupled to a neighbour is how they blend. The chapter verifier even performs the clean algebraic projection that separates the two. And the split has teeth: the quark off-diagonal twists, which is why quarks violate CP, while the lepton off-diagonal does not, which is why the theory predicts leptons conserve it — a prediction now waiting on the neutrino experiments.
Neutrinos, and a falsifiable bet
The neutrino sector is where the theory is most exposed, which makes it the most interesting. The mixing angles come out of the same algebra: the reactor angle as $\sin^2\theta_{13} = \tfrac49\sin^2(\pi/14)$ ($-1\%$), the atmospheric angle near-maximal at $\sin^2\theta_{23} = 15/28$, and the solar angle from quark-lepton complementarity, $\theta_{12}\approx \pi/4 - \theta_C$ ($+3\%$). It is worth being clear that an earlier, tidier guess here was refuted: the exact tri-bi-maximal pattern predicted $\theta_{13}=0$, and the reactor experiments killed it at more than thirty sigma. The corrected angles above are what replaced it.
Two forward bets carry pre-committed kill-conditions. The first is the absolute mass: a seesaw construction gives the heaviest neutrino at $m_{\nu_3} = 48.7$ meV. This one is conditional — it leans on a structural assumption and a Planck-mass anchor rather than being fully parameter-free — but it takes no neutrino data as input, and it sits $2.3\sigma$ from the present value inferred from oscillations ($49.5\pm0.35$ meV). A sub-half-percent measurement of the mass splitting at JUNO or Hyper-K will confirm or refute it at three sigma. The second is the CP phase, and here the honesty matters most. The theory’s first geometric prediction, $\delta_{CP}=240^\circ$, was refuted; a later relation tying it to the quark phase was also retired, at six sigma. What survives is a structural result — on the family sector charge-conjugation and parity are the same operation, so their product is the identity and leptonic CP is conserved, $\delta_{CP}=\pi$. Current data sit at $177^\circ$, which agrees to a sixth of a sigma. A confirmed near-maximal leptonic CP violation at DUNE would kill it. Two dead predictions and one live one, all on the record — that trail is the evidence the live bet was reasoned, not retrofitted.
First-Edition sources & full derivations: P06 · P07 · P12 · P13 · P20
Verify this chapter
A standalone Python script (numpy only — no network, no corpus dependency) recomputes this chapter’s quantities from first principles and compares each to the measured value. The table below is its actual output. View the script · run all chapters.
| Quantity | Lumen | Measured / target | Residual | |
|---|---|---|---|---|
| J3(O) dimension 3 real diagonal + 3 octonion off-diagonal; H_n(O) is Jordan iff n<=3 | 27 | 27E6 fundamental | +0.00 ppm | ✓ |
| three families = max octonionic Jordan rank no J4(O), hence no 4th generation -- the family count is forced, not fitted | no J4(O), hence no 4th generation -- the family count is forced, not fitted | — | structural | ✓ |
| Z3 Schur split (mass=diagonal, mixing/CP=off-diag) (1/3) sum_j w^{(a-b)j} = delta_{a=b}: the diagonal is self-observation (mass) | (1/3) sum_j w^{(a-b)j} = delta_{a=b}: the diagonal is self-observation (mass) | — | structural | ✓ |
| MU = mu1/mu0 (mass lever, from moments) | 0.793342 | 0.79334density | +2.78 ppm | ✓ |
| m_tau/m_e (DERIVED, 0 params) exp(MU*(E_self(lambda_tau)-pi)) by integrating the density; E_self=13.419 | 3474.61 | 3477.23PDG 2024 | -0.08% | ✓ |
| m_mu/m_e (geometric, E=pi^2) exp(MU*(pi^2 - pi)) | 208.016 | 206.768PDG 2024 | +0.60% | ✓ |
| m_u = m_e*exp(MU*(pi^(7/5)-pi)) F structural (Fano ratio 7/5) | 2.17281 MeV | 2.16 MeVPDG | +0.59% | ✓ |
| m_d = m_e*exp(MU*(ln(mu1)/MU-pi)) R moment-log, 'structurally targeted' (OP-D open) | 4.59514 MeV | 4.67 MeVPDG | -1.60% | ✓ |
| m_s = m_e*exp(MU*(pi^2 - 1/7-pi)) F structural (one Fano step below muon) | 94.9066 MeV | 93.4 MeVPDG | +1.61% | ✓ |
| m_c = m_e*exp(MU*(pi^2 + pi-pi)) F exact integral identity int(3pi^2 x^2+2pi x)=pi^2+pi | 1285.09 MeV | 1270 MeVPDG | +1.19% | ✓ |
| m_b = m_e*exp(MU*(pi^(7/3)-pi)) F* structural; raw -3.2%, closes to -0.25% with QCD corr (uses measured alpha_s) | 4039.88 MeV | 4180 MeVPDG | -3.35% | ✓ |
| m_t = m_e*exp(MU*(pi^2+pi+ln(mu0)/MU-pi)) R formula-consistent t = +2.0% (NOT the 172.4GeV/-0.2% the table mis-states); m_t/m_c=alpha^-1, OP-B open | 1.76104e+05 MeV | 1.72760e+05 MeVPDG | +1.94% | ✓ |
| m_d/m_u = exp(2*MU*lambda^2*pi^2) G2 Weyl step (lambda=sin(pi/14)); the cleanest light-quark ratio | 2.17148 | 2.162PDG | +0.44% | ✓ |
| m_t/m_c = alpha^-1 (= MU0) trace identity proved; full derivation OPEN (OP-B) | 137.036 | 136.03PDG ratio | +0.74% | ✓ |
| U(1) x SU(2) x SU(3) from R/C/H/O C->U(1) (Hopf S1, EM); H->SU(2) (S3=SU(2), weak); O->G2=Aut(O) sup SU(3) (strong) | C->U(1) (Hopf S1, EM); H->SU(2) (S3=SU(2), weak); O->G2=Aut(O) sup SU(3) (strong) | — | structural | ✓ |
| gauge group: STRUCTURAL ASSIGNMENT, not full derivation P20 self-flags: the hypercharge representation is NOT constructed; maximal-subgroup argument | P20 self-flags: the hypercharge representation is NOT constructed; maximal-subgroup argument | — | P20 self-flags: the hypercharge representation is NOT constructed; maximal-subgroup argument | · |
| dim G2 = 14 = 2*Im(O) G2 = automorphism group of the octonions; SU(3) = stabiliser of one imaginary unit | 14 | 14Aut(O) | +0.00 ppm | ✓ |
| sin^2(theta_W) NLO = (1/(1+pi))(1-4lam^2/5) LO = 1/(1+pi) = 0.24145 (+4.4%); 3/(8phi)=0.23176 is a near-identity, NOT the derived value | 0.231888 | 0.23122PDG effective | +0.29% | ✓ |
| M_W/M_Z = cos(theta_W) the clean EW ratio (-0.56%); rho-parameter = 1 (custodial) | 0.87642 | 0.881447PDG | -0.57% | ✓ |
| M_W, M_Z absolute: tree-level only (-3.7%) tree M_W=77.4, M_Z=88.3 GeV; the ~3.7% gap is the known SM radiative (Delta r) correction, not closed | tree M_W=77.4, M_Z=88.3 GeV; the ~3.7% gap is the known SM radiative (Delta r) correction, not closed | — | tree M_W=77.4, M_Z=88.3 GeV; the ~3.7% gap is the known SM radiative (Delta r) correction, not closed | · |
| alpha_s(M_Z): NOT derived by the corpus F4 route off by ~8x, E6 result tautological, SU(5)/E6 embedding proved incompatible; every alpha_s use is PDG input | F4 route off by ~8x, E6 result tautological, SU(5)/E6 embedding proved incompatible; every alpha_s use is PDG input | — | F4 route off by ~8x, E6 result tautological, SU(5)/E6 embedding proved incompatible; every alpha_s use is PDG input | · |
| CKM lambda = sin(pi/14) G2 half-alcove angle theta_C = pi/14 | 0.222521 | 0.2245PDG (Cabibbo) | -0.88% | ✓ |
| CKM A (NLO) = A0*(1-4lam^2/5) Jordan-loop NLO; +0.075%% -- one of the cleanest hits in the sector | 0.822434 | 0.823PDG | -0.07% | ✓ |
| |V_cb| = A*lambda^2 | 0.0407233 | 0.0411PDG | -0.92% | ✓ |
| delta_CKM (quark CP phase) = pi/3+pi/28 CP-VIOLATING quark phase; this is where meson CP violation lives | 66.4286 deg | 65.6 degPDG (gamma) | +1.26% | ✓ |
| rho-bar, eta-bar, Jarlskog J: Bryant-calibrated (R) J=3.04e-5 vs (3.18+/-0.15)e-5 within 1 sigma; rho-bar/eta-bar use a 3-form matching convention | J=3.04e-5 vs (3.18+/-0.15)e-5 within 1 sigma; rho-bar/eta-bar use a 3-form matching convention | — | J=3.04e-5 vs (3.18+/-0.15)e-5 within 1 sigma; rho-bar/eta-bar use a 3-form matching convention | · |
| PMNS sin^2(theta_13) = (4/9)sin^2(pi/14) TBM predicted theta_13 = 0, REFUTED at >30 sigma (Daya Bay); the corrected value agrees | 0.0220069 | 0.02224NuFIT | -1.05% | ✓ |
| PMNS sin^2(theta_23) = 15/28 near-maximal mixing | 0.535714 | 0.545NuFIT | -1.70% | ✓ |
| PMNS sin^2(theta_12) = cos^2(pi/14)/3 quark-lepton complementarity: theta_12 ~ pi/4 - theta_C | 0.316828 | 0.307NuFIT | +3.20% | ✓ |
| leptonic delta_CP = pi (CP-conserving) 240deg geometric value REFUTED (A344, 3.3sig); A66 lepton value retired (A395, 6.1sig); current stance allowed at 0.78sig on the with-SK fit (0.16sig on the without-SK fit, 177deg) | 180 deg | 212 degNuFIT 6.0 NO (IC24+SK) | +0.78 sigma | ✓ |
| m_nu3 = 48.7 meV (seesaw, CONDITIONAL) conditional on the spectral-angle assumption; uses CODATA M_Pl anchor, no neutrino-data input; vs sqrt(Dm31^2)=50.1+/-0.2 meV (NuFIT 6.0, NO, m1~0): 48.7 sits below the global fit's 3-sigma range [49.5, 50.8] — a STANDING TENSION, recorded 2026-07-02. The pre-committed kill is unchanged: a <=0.5%% direct JUNO/Hyper-K measurement of |Dm31^2| decides at >=3 sigma. Assumes NORMAL ordering. | conditional on the spectral-angle assumption; uses CODATA M_Pl anchor, no neutrino-data input; vs sqrt(Dm31^2)=50.1+/-0.2 meV (NuFIT 6.0, NO, m1~0): 48.7 sits below the global fit's 3-sigma range [49.5, 50.8] — a STANDING TENSION, recorded 2026-07-02. The pre-committed kill is unchanged: a <=0.5%% direct JUNO/Hyper-K measurement of |Dm31^2| decides at >=3 sigma. Assumes NORMAL ordering. | — | conditional on the spectral-angle assumption; uses CODATA M_Pl anchor, no neutrino-data input; vs sqrt(Dm31^2)=50.1+/-0.2 meV (NuFIT 6.0, NO, m1~0): 48.7 sits below the global fit's 3-sigma range [49.5, 50.8] — a STANDING TENSION, recorded 2026-07-02. The pre-committed kill is unchanged: a <=0.5%% direct JUNO/Hyper-K measurement of |Dm31^2| decides at >=3 sigma. Assumes NORMAL ordering. | · |
26/26 checks passed. Method: numpy integration of the density + exact algebra; particle predictions judged at the percent level.
Measurement sources: PDG 2024 quark masses (u,d,s MSbar 2GeV; c,b MSbar m(m); t pole): 2.16/4.67/93.4/1270/4180/172760 MeV; PDG 2024: m_tau/m_e=3477.23, m_mu/m_e=206.7683, sin^2 th_W=0.23122, |V_cb|=0.0411, lambda=0.2245; NuFIT 5.3 (NO) angles: sin^2 th12=0.307, th23=0.545, th13=0.02224. NuFIT 6.0 (NO, IC24+SK): delta_CP=212deg (177deg without SK-atm); Dm3l^2=+2.513e-3 eV^2.
Independently pinned in the First Edition by: verify_P084 (quark table), verify_P088 (Weinberg), verify_P082 (CKM), verify_P063 (PMNS), verify_P116 (tau mass) — trace any number back to the full archive.
full script output
========================================================================================================
LUMEN ā Matter & Gauge
The Standard-Model table from J3(O) and the density -- masses, gauge group, electroweak, CKM, neutrinos
method: numpy integration of the density + exact algebra; particle predictions judged at the percent level
========================================================================================================
Quantity Lumen Measured / target Source Residual
----------------------------------------------------------------------------------------------------
J3(O) dimension 27 27 E6 fundamental +0.00 ppm PASS
three families = max octonion no J4(O), hence (geometric) structural PASS
Z3 Schur split (mass=diagonal (1/3) sum_j w^{ (geometric) structural PASS
MU = mu1/mu0 (mass lever, fro 0.793342 0.79334 density +2.78 ppm PASS
m_tau/m_e (DERIVED, 0 params 3474.61 3477.23 PDG 2024 -0.08% PASS
m_mu/m_e (geometric, E=pi^2) 208.016 206.768 PDG 2024 +0.60% PASS
m_u = m_e*exp(MU*(pi^(7/5)-pi 2.17281 MeV 2.16 MeV PDG +0.59% PASS
m_d = m_e*exp(MU*(ln(mu1)/MU- 4.59514 MeV 4.67 MeV PDG -1.60% PASS
m_s = m_e*exp(MU*(pi^2 - 1/7- 94.9066 MeV 93.4 MeV PDG +1.61% PASS
m_c = m_e*exp(MU*(pi^2 + pi-p 1285.09 MeV 1270 MeV PDG +1.19% PASS
m_b = m_e*exp(MU*(pi^(7/3)-pi 4039.88 MeV 4180 MeV PDG -3.35% PASS
m_t = m_e*exp(MU*(pi^2+pi+ln( 1.76104e+05 MeV 1.72760e+05 MeV PDG +1.94% PASS
m_d/m_u = exp(2*MU*lambda^2*p 2.17148 2.162 PDG +0.44% PASS
m_t/m_c = alpha^-1 (= MU0) 137.036 136.03 PDG ratio +0.74% PASS
U(1) x SU(2) x SU(3) from R/C C->U(1) (Hopf S (geometric) structural PASS
gauge group: STRUCTURAL ASSIG P20 self-flags: (geometric) P20 self-fl info
dim G2 = 14 = 2*Im(O) 14 14 Aut(O) +0.00 ppm PASS
sin^2(theta_W) NLO = (1/(1+pi 0.231888 0.23122 PDG effective +0.29% PASS
M_W/M_Z = cos(theta_W) 0.87642 0.881447 PDG -0.57% PASS
M_W, M_Z absolute: tree-level tree M_W=77.4, (geometric) tree M_W=77 info
alpha_s(M_Z): NOT derived by F4 route off by (geometric) F4 route of info
CKM lambda = sin(pi/14) 0.222521 0.2245 PDG (Cabibbo) -0.88% PASS
CKM A (NLO) = A0*(1-4lam^2/5) 0.822434 0.823 PDG -0.07% PASS
|V_cb| = A*lambda^2 0.0407233 0.0411 PDG -0.92% PASS
delta_CKM (quark CP phase) = 66.4286 deg 65.6 deg PDG (gamma) +1.26% PASS
rho-bar, eta-bar, Jarlskog J: J=3.04e-5 vs (3 (geometric) J=3.04e-5 v info
PMNS sin^2(theta_13) = (4/9)s 0.0220069 0.02224 NuFIT -1.05% PASS
PMNS sin^2(theta_23) = 15/28 0.535714 0.545 NuFIT -1.70% PASS
PMNS sin^2(theta_12) = cos^2( 0.316828 0.307 NuFIT +3.20% PASS
leptonic delta_CP = pi (CP-co 180 deg 177 deg NuFIT NO +0.16 sigma PASS
m_nu3 = 48.7 meV (seesaw, CON conditional on (geometric) conditional info
----------------------------------------------------------------------------------------------------
26/26 checks passed.
Measurement sources: PDG 2024 quark masses (u,d,s MSbar 2GeV; c,b MSbar m(m); t pole): 2.16/4.67/93.4/1270/4180/172760 MeV; PDG 2024: m_tau/m_e=3477.23, m_mu/m_e=206.7683, sin^2 th_W=0.23122, |V_cb|=0.0411, lambda=0.2245; NuFIT 5.3 (NO) angles: sin^2 th12=0.307, th23=0.545, th13=0.02224; NuFIT 6.0 (NO, IC24+SK): delta_CP=212deg (177deg without SK-atm)
Independently pinned by First-Edition verifiers: verify_P084 (quark table), verify_P088 (Weinberg), verify_P082 (CKM), verify_P063 (PMNS), verify_P116 (tau mass)
========================================================================================================