The Weinberg Angle and Fine-Structure Constant
The Standard Model treats these as measured inputs — numbers that cannot be predicted. GTE derives both with zero free parameters — from two related but algebraically distinct GF(7) cellular-automaton rules (the 19-bit polynomial for α; a separate, 76-bit interaction rule for sin²θ_W).
The Standard Model's Embarrassing Numbers
The Standard Model is the most successful physical theory ever written. But it contains roughly 25 numbers that it cannot predict — you measure them and insert them by hand. Among the most fundamental are:
α ≈ 1/137.036
The electromagnetic coupling strength. It controls how strongly charged particles interact via photons. One of the most precisely measured constants in all of physics.
sin²θ_W ≈ 0.231
The mixing angle between the weak and electromagnetic forces. It determines the mass ratio of the W and Z bosons — and every SM calculation involving neutral currents.
α_s(M_Z) ≈ 0.118
The strength of the strong nuclear force at the Z boson mass scale. Also a free parameter — also derived in GTE.
None of these can be explained by the Standard Model. You measure them in experiments and insert them. GTE derives all three from the orbit structure of the GTE arithmetic substrate — though, as the note below explains, α and sin²θ_W come from two different, non-interchangeable rules within that substrate. This lesson covers α and sin²θ_W.
The Weinberg Angle: sin²θ_W = 3/13
A note on which rule is used here. This section uses a rule called f_MDL, not the 19-bit polynomial p used below for the fine-structure-constant identity. The two are proved to be algebraically and dynamically distinct: f_MDL is the unique minimum-description-length map (76 bits to specify, vs. 19 bits for p) that agrees with Wolfram's binary Rule 110 on purely binary inputs and reproduces the observed three-generation Standard Model orbit; a companion paper's Schwartz–Zippel zero-counting argument shows f_MDL is not a low-degree polynomial at all (only 14 nonzero outputs among 343 possible inputs, vs. 300 for p). The two rules agree only on the 8 purely binary inputs they share as a common Rule 110 sublayer.
What is the Weinberg angle?
Electromagnetism and the weak nuclear force are actually two aspects of a single electroweak force. At high energies they were unified; at low energies they look completely different because the symmetry broke.
The breaking is geometric. There are two underlying gauge fields, $W^3_\mu$ (SU(2)_L) and $B_\mu$ (U(1)_Y). At low energies these mix at a specific angle $\theta_W$:
Electroweak mixing
photon (A_μ) = cos θ_W · B_μ + sin θ_W · W³_μ
Z boson (Z_μ) = −sin θ_W · B_μ + cos θ_W · W³_μ
The angle θ_W ≈ 28.7° determines how much of each underlying field becomes the photon and how much becomes the Z. The ratio sin²θ_W = g′²/(g² + g′²).
The GTE derivation
In the GTE framework, sin²θ_W is the ratio of the generation count to the Higgs-sector orbit count:
Where do 3 and 13 come from?
Comparison to experiment
The tree-level result (pure arithmetic) deviates by −0.20% from PDG. With two-loop electroweak threshold corrections — all inputs from GTE, no PDG parameters borrowed — the prediction reaches +0.030% from the PDG central value.
Lean cert: weinberg_sin_sq_from_ngen_ch (CatAL, zero sorry)
The Fine-Structure Constant: α⁻¹ = 2⁷ + 3² = 137
What α controls
The fine-structure constant α ≈ 1/137 sets the overall strength of electromagnetism. It appears in every QED calculation: the binding energy of atoms, the magnetic moment of the electron, the width of spectral lines.
Richard Feynman called it "one of the greatest damn mysteries of physics." Its value is measured to extraordinary precision but never explained.
The GTE identity
In GTE, the electromagnetic coupling comes from counting how many ways virtual loops can arrange themselves under the winding constraints of the polynomial:
3² = 9: lepton-sector loop contribution from N_gen² = 3²
Where each term comes from
The GTE polynomial operates over GF(7) — clock arithmetic with 7 positions. This gives rise to ℤ₇ with 7 winding sectors.
In QED, the photon emission probability is determined by summing over all virtual configurations of the winding sectors. Each of the 7 sectors is either "active" or "inactive" in a given loop. The total number of binary configurations:
2^|ℤ₇| = 2^7 = 128 configurations
The more ways a loop can be arranged, the smaller each individual channel's
probability → the weaker the coupling. The 128 is the combinatorial weight
that sets the electromagnetic coupling strength. Lean cert: alpha_em_inverse_structural_identity (CatAL).
The Standard Model has exactly three generations of fermions. In GTE, N_gen = 3 is derived — not assumed — from the orbit depth of the generation sequence under the f_MDL map.
When computing quantum loop contributions to the electromagnetic coupling, you sum over all fermion species running around the loop. For leptons, this sum acquires a double factor of N_gen:
N_gen × N_gen = 3 × 3 = 9 = 3²
One factor from the generation index of each lepton in the loop; one from each vertex. The net lepton-sector contribution is proportional to N_gen² = 9.
With two free integers one could always search for a representation of 137. But these integers are not free:
- |ℤ₇| = 7 is forced by MDL minimality of the polynomial — the only prime giving the five-role winding class property
- N_gen = 3 is forced by four independent orbit-arithmetic mechanisms unrelated to α
Both numbers were determined by independent physical reasoning before the fine-structure constant was considered. Their combination giving α⁻¹ is a genuine prediction, not a fit.
Lean cert: alpha_em_inverse_structural_identity (CatAL, zero sorry,
AlphaEMStructuralIdentity.lean)
Comparison to measurement
The GTE prediction gives the integer part exactly. The 0.036 fractional correction comes from higher-order loop contributions — the same QED radiative corrections that apply to any tree-level calculation.
A second route to the same identity
The first-generation orbit value b₁ = 73 satisfies b₁ = 2φ(7) + Ngen², where φ(7) = |GF(7)*| = 6 is the order of the multiplicative group of GF(7). Since φ(7) = 2 × Ngen (a structural identity relating the GF(7) group order to the generation count), the two known routes to α⁻¹ = 137 — the direct route 27 + Ngen² and the orbit route via b₁ — are algebraically equivalent. The two expressions count the same arithmetic structure from different angles.
Bonus: The Strong Coupling α_s(M_Z)
The strong coupling is also derived in GTE — from the β-function coefficient
$b_0 = |\mathbb{Z}_7| = 7$ (Lean cert: b0_eq_z7_order, CatAL):
Why These Are Predictions, Not Fits
The distinction between a prediction and a fit is crucial for scientific credibility:
You choose parameters after seeing the experimental value. The agreement is guaranteed. The fit has zero predictive power.
Example: SM — measure sin²θ_W, insert it into the theory.
The numbers N_gen = 3 and |ℤ₇| = 7 are derived from independent physical requirements (orbit structure, MDL selection) that have nothing to do with α or sin²θ_W.
The theory is then confronted with measurement — and agrees.
Both α⁻¹ = 137 and sin²θ_W = 3/13 are falsifiable predictions. An experiment that measures sin²θ_W = 0.250 or α⁻¹ = 150 would refute GTE in this sector. The values agree with PDG to within fractions of a percent — with zero free parameters.
See Also
weinberg_sin_sq_from_ngen_ch— sin²θ_W = N_gen/c_Hew_c_staircase— c_H = 13 from the EW boson staircasealpha_em_inverse_structural_identity— α⁻¹ = 2⁷ + 3²b0_eq_z7_order— β-function coefficient from ℤ₇