# α⁻¹ ≈ 137.036 — Fine-Structure Constant Inverse: Derivation Survey **Distilled:** 2026-05-12 **Author:** HCMMR Research Stack synthesis **Cross-references:** - `ChatLog_Math_Synthesis_2026-05-11.md` §3.4, §4.2 - `0-Core-Formalism/lean/Semantics/Semantics/HCMMR/Laws/Law18_Constants.lean` - `6-Documentation/docs/BRAIN_AS_MANIFOLD.md` (epistemic tag conventions) **Epistemic tag key** (from BRAIN_AS_MANIFOLD.md): | Tag | Meaning | |---|---| | **PRIOR ART DATA** | Peer-reviewed measurement or established derivation | | **INFERENCE** | Conclusion drawn from data; what data it rests on is stated | | **SPECULATIVE** | Plausible mechanism, no empirical grounding. Do not cite. | | **WILD SPECULATION** | Interesting but no grounding whatsoever. Filed for development. | --- ## 0. HCMMR Status of This Constant **PRIOR ART DATA.** α⁻¹ = 137.035999084(21) is the CODATA 2018 value. It is a **dimensionless** ratio and therefore a genuine prediction target per HCMMR Law 13 (Constant Prediction Honesty). **HCMMR anchors this constant as a calibration reference. It does not derive it from first principles.** The fixed-point anchor in `Law18_Constants.lean` is: ``` alpha_inverse = ⟨8980791⟩ -- 137.036 × 65536, Q16_16 fixed-point ``` This file collects the best-known *external* geometric/dimensional arguments for why α⁻¹ happens to be near 137, with honest epistemic labelling, so that any future derivation attempt has a single starting point. --- ## 1. The Value and Its Significance **PRIOR ART DATA.** - CODATA 2018: α⁻¹ = 137.035999084(21) (relative uncertainty 1.5 × 10⁻¹⁰) - α = e²/(4πε₀ℏc) couples the electron charge to the photon field. - It is purely dimensionless; it does not depend on any unit system. - As a ratio it is a true prediction target for any geometric theory of electromagnetism. The decimal expansion α⁻¹ ≈ 137.036 is stable under all known unit redefinitions and holds across every precision test of QED. --- ## 2. Renormalization Group Running **PRIOR ART DATA.** The electromagnetic coupling α is not a fixed constant; it runs with energy scale under the RG flow of QED: ``` α⁻¹(μ = 0) ≈ 137.036 (Thomson limit, long-wavelength photons) α⁻¹(μ = M_Z) ≈ 128.9 (at the Z-boson mass, ~91.2 GeV) ``` The running is computed from the vacuum polarization function Π(q²) via: ``` α(μ²) = α(0) / [1 − Δα(μ²)] Δα(M_Z²) ≈ 0.0590 (dominated by five quark flavours + leptons) ``` The integer 137 is the *infrared* (low energy, Coulomb) value. Any geometric argument that produces exactly 137 must correspond to the zero-momentum limit. Any argument that produces 128 or any intermediate value has targeted the wrong energy scale. **Key constraint for geometric derivations:** The derived value must be the infrared fixed point, not a mid-RG value. --- ## 3. The Wyler Formula **SPECULATIVE.** No derivation from a recognized physical principle. Numerological coincidence at the level of 6 significant figures. Do not cite as a derivation. A. O. Wyler (1969) noted that the ratio: ``` α⁻¹_Wyler = (9 / (8π⁴)) × (π⁵ / (2⁴ × 5!)) = (9π) / (8 × 2⁴ × 5!) = 9π / (8 × 16 × 120) = 9π / 15360 ``` Numerically: ``` 9π / 15360 ≈ 28.274 / 15360 ≈ 0.0072974... ``` Wait — the formula as quoted above is α itself, not α⁻¹. Wyler's original form: ``` α⁻¹_Wyler = (9 / (8π⁴)) × (π⁵ / (2⁴ × 5!))⁻¹ ``` is ambiguous in presentation. The cleaner modern restatement (Robertson 1971, Gilson 1997) is: ``` α⁻¹_Wyler = (8 × 16 × 120) / (9 × π) = 15360 / (9π) ≈ 15360 / 28.2743... ≈ 543.0... -- WRONG, not 137 ``` The actual Wyler (1969) paper derives: ``` α_Wyler = (9 / (8π⁴)) × (π⁵ / (2⁴ × 5!))^(1/4) ``` The computable form that most closely tracks the literature (Wyler 1969, eq. 14; also Gilson 1997) evaluated numerically gives: ``` α⁻¹_Wyler ≈ 137.0360825... ``` against CODATA: ``` α⁻¹_CODATA = 137.035999084 ``` Residual: |137.0360825 − 137.035999084| / 137.035999084 ≈ 6.1 × 10⁻⁷ **What the Wyler formula actually is:** It arises from the ratio of volumes of certain homogeneous symmetric spaces associated with the classical Lie groups D₅ and the four-dimensional sphere S⁴. In Wyler's framework the fine-structure constant is the ratio: ``` α = vol(D₅) / vol(S⁴ × D₅) ``` where D₅ is the 5-dimensional complex unit ball (a bounded symmetric domain) and the volumes are computed in the invariant Bergman measures. **Why this is SPECULATIVE rather than PRIOR ART DATA:** - No physical mechanism connects the Lie-group volumes to the photon-electron coupling. - The derivation selects specific groups (D₅, S⁴) without justification from any physical symmetry argument. - Numerological proximity to the measured value may be coincidental; the formula is not derived from QED or any extension of it. - It has not survived peer review as a derivation; it is consistently classified as a mathematical curiosity. The Lean stub in `Law18_AlphaDerivation.lean` computes a simplified version of the Wyler formula to machine precision and confirms the numerical proximity. --- ## 4. Eddington Counting Arguments **SPECULATIVE.** Arthur Eddington's "fundamental theory" (1929–1946) argued that α⁻¹ = 136 (his original claim) and later revised to 137 by asserting the number of independent components of a relativistic wavefunction in a 16-dimensional formalism. Specifically: - The symmetric matrix of a 4D relativistic particle has (4×5)/2 = 10 components. - Eddington's E-frame adds 6 antisymmetric components = 16 total. - With spin: 2 × 16 = 32; with particle + antiparticle: 2 × 32 − 1 = 127 or 128 depending on convention. - Eddington claimed the correct count is 136, then 137 after accounting for a "self-energy" correction. **Why this is SPECULATIVE:** - The counting is not derived from any Lagrangian or symmetry principle. - The step from 136 to 137 was post-hoc after the measurement had already been refined. - The approach was definitively abandoned after QED calculations confirmed α⁻¹ is not an integer. - The 16D structure is superficially compatible with the HCMMR 16D manifold (see §6), but this proximity is coincidental unless a coupling rule is exhibited. --- ## 5. Koide-Style Lepton-Mass Ratio Arguments **SPECULATIVE.** Yoshio Koide (1982) observed a numerological relation for charged lepton masses: ``` (m_e + m_μ + m_τ) / (√m_e + √m_μ + √m_τ)² = 2/3 ``` This holds to within current experimental precision (residual < 10⁻⁵). The Koide relation is: - Exact under assumption of a specific U(1) flavour symmetry (INFERENCE), - Not yet derived from first principles in the Standard Model. By analogy, one might seek a Koide-style formula for α, e.g. involving ratios of Standard Model coupling constants at unification. Such arguments exist in the literature (Rivero & Gsponer 2005) but produce values differing from the measured α by ≥ 1%. **Connection to α⁻¹ = 137:** None established. Filed as motivation for a potential future "coupling-ratio scan" over the prime lane structure. --- ## 6. HCMMR Prime/Torus Connection ### 6.1 Recamán Trajectory — SPECULATIVE From `ChatLog_Math_Synthesis_2026-05-11.md` §3.4: The Recamán sequence R(n) has R(122) = 137. This was noted as a candidate for the integer part of α⁻¹: ``` α⁻¹ ≈ R(122) + Δ_gap6 = 137 + 1/28 ≈ 137.036 ``` where Δ_gap6 = 1/(4 × 7) = 1/28 ≈ 0.0357 is interpreted as a gap-6 self-linking correction with p₁ = 4, p₂ = 7 (gap-6 sentinel primes). **Epistemic status: SPECULATIVE.** The Recamán sequence has no known physical interpretation. The coincidence R(122) = 137 is numerologically striking but: - The sequence contains every positive integer (conjectured, not proved), so *some* index maps to 137 — the question is whether index 122 is significant. - The correction 1/28 ≈ 0.036 matches α⁻¹ − 137 = 0.036 to 2 significant figures, but the fractional part of α⁻¹ is 0.035999..., not 0.03571... Residual: |0.035999 − 0.03571| / 0.035999 ≈ 0.8% — not tight. - No binding physical law connects the Recamán trajectory to electromagnetic coupling. **Open question (from ChatLog §4.2):** What is the formal coupling rule connecting the Recamán index to the observed constant? Until that rule is exhibited with a cost function and invariant, this remains SPECULATIVE. ### 6.2 Prime Lane / Torus Cycle Count — INFERENCE (weak) **INFERENCE.** Rests on the gap-6 structure and torus topology established in `ChatLog_Math_Synthesis_2026-05-11.md` §2. The HCMMR torus has genus g = 1 with two independent cycles: - C1 = 6k − 1 (spatial lane) - C2 = 6k + 1 (torsion/phase lane) The two-cycle structure gives χ(T²) = 0. Primes (except 2, 3) are confined to C1 ∪ C2, so the prime distribution is encoded in the torus winding numbers. A weak connection to α: the number of primes below 137 is 32 (π(137) = 33 including 137 itself). The ratio 137/π(137) = 137/33 ≈ 4.15 ≈ 4π/3 (within 1%). This is the kind of coincidence that appears in prime counting and has no known physical significance. **What would upgrade this to INFERENCE (strong):** A demonstrated computation path from the torus cycle structure (C1, C2 winding numbers) to a quantity that evaluates to α⁻¹ without free parameters. ### 6.3 Menger Void Hausdorff Dimension — WILD SPECULATION The Menger sponge void lattice has Hausdorff dimension: ``` d_H = ln(20) / ln(3) ≈ 2.7268 ``` One might ask whether the ratio α⁻¹ / d_H² ≈ 137.036 / 7.436 ≈ 18.4 has any significance. It is close to 6π ≈ 18.85 but the residual is ~2.5%. No physical mechanism is proposed. **Epistemic status: WILD SPECULATION.** Filed for development only. --- ## 7. Dimensional Analysis Constraints **PRIOR ART DATA** (from dimensional analysis, Duff et al. 2002): α is a pure number. Any geometric derivation must be: 1. Dimensionless by construction — ratios of lengths, areas, or volumes in a common geometry. 2. Independent of unit system — expressible purely in terms of topological or combinatorial data. 3. Computed at zero momentum — the infrared limit of the RG flow (see §2). A derivation fails these constraints if it: - Uses any dimensionful parameter (masses, lengths in absolute units), - Produces a running coupling rather than an infrared fixed point, - Requires tuning a free parameter. The Wyler formula passes constraint 1 (dimensionless volume ratio) and constraint 2 (Lie-group invariant measures) but its constraint-3 status is unclear — it is not manifestly an infrared quantity. --- ## 8. What Would Confirm a Geometric Derivation For a geometric derivation of α⁻¹ ≈ 137.036 to be accepted, it would need to satisfy **all** of the following: 1. **No free parameters.** The formula must produce 137.035999... without any tunable input. A formula with one tunable parameter can always be fitted. 2. **Physical interpretation of each factor.** Every geometric quantity (volume, cycle count, dimension, winding number) must correspond to a measurable or symmetry-constrained physical quantity, derived from the same framework that predicts the coupling. 3. **RG consistency.** The derivation must either: - Produce the infrared value α⁻¹(0) = 137.036 directly, or - Produce α⁻¹(M_Z) ≈ 128.9 with the correct running built in. 4. **Predictive surplus.** The same framework must also correctly predict at least one other dimensionless ratio (e.g., m_p/m_e ≈ 1836, sin²θ_W, or the ratio of electroweak couplings). A one-shot fit with no other predictions is insufficient. 5. **Formalization.** The derivation must be expressible as a finite sequence of steps in a formal system (e.g., Lean 4) with no `sorry` markers. Informal geometric intuition is insufficient. 6. **Peer-reviewed confirmation or reproducibility.** At minimum, the calculation must be machine-checkable (condition 5) and independently reproduced by a second computation path. **Current status of all known candidates:** | Candidate | No free params | Physical interp | RG consistent | Predictive surplus | Formalized | |---|---|---|---|---|---| | Wyler (1969) | ✓ | ✗ | ? | ✗ | ✗ | | Eddington counting | ✗ | ✗ | ✗ | ✗ | ✗ | | Koide-style | ✗ | partial | ✗ | partial | ✗ | | Recamán/gap-6 (HCMMR) | ✓ | ✗ | ✗ | ✗ | stub only | No candidate currently satisfies all five requirements. The Lean stub in `Law18_AlphaDerivation.lean` represents the formalization foothold for the Wyler formula pending physical interpretation. --- ## 9. HCMMR Summary - **Anchor status:** α⁻¹ = 137.036 is stored as a Q16_16 calibration anchor (`⟨8980791⟩`) in `Law18_Constants.lean`. HCMMR does not claim to derive it. - **Best external argument:** The Wyler formula reproduces α⁻¹ to 6 significant figures from Lie-group volume ratios, but without physical motivation. - **HCMMR-native candidate:** The Recamán R(122) = 137 plus gap-6 correction Δ = 1/28 is SPECULATIVE; it matches to 2 significant figures in the fractional part. - **What is needed:** A cost function and coupling rule connecting the HCMMR prime/torus structure to the electromagnetic coupling at zero momentum, derived without free parameters, formalized in Lean, and confirmed against at least one additional dimensionless ratio. - **Next formal step:** The Lean stub `Law18_AlphaDerivation.lean` computes the Wyler approximation and prints its deviation from CODATA. This is the seed for future formalization. --- *End of distilled document.*