Randomized trial investigates the fine-structure constant in quantum electrodynamics, suggesting a significant measurement outcome.
A construction containing no physical input and no measured value outputs two numbers from one quadratic: its physical root 1/alpha_S = 137.035 999 074 948 275 854, and the harmonic sum of that quadratic's two roots, 1/alpha_D = 137.035 999 167 273 924 798 (equivalently its no-closure limit), separated by exactly alpha_S L^4 = 9.2326 x 10^-8 in 1/alpha (92.33 ppb). The electron g-2 determination — the most precisely measured quantity in physics — gives 1/alpha = 137.035 999 166(15): agreement with alpha_D to eleven significant figures, at 0.085 sigma. Both statements are checkable and neither is conditional on anything argued here. Every structural choice in the construction — kernel, domain, threshold, exponent, branch, and each term of the assembly — is pinned by a stated theorem or textbook physical fact that never references the output, and every named deformation — slot, exponent, branch, physics — is priced at the end; none lands near. The construction's signature is convergence: the same constant is characterized three independent ways, the same closure carries five equivalent faces, and the same fixed-point object is met again, unmodified, inside the quantum harmonic oscillator, the heat trace of a shifted line Laplacian, and the Bargmann–Fock space (Part IV) — one foundation, many walks. The closure condition is shown to be a level set of the inversion invariant u + u^-1, from which it follows that the predicted methodological split between recoil and g-2 extractions of the fine-structure constant equals, exactly, the inverse coupling of the construction's second fixed point; the sorting of the two extraction classes onto the two roots is itself derived, conditional on the paper's one identification and on nothing else. Three routes reach the same constant — a geometric construction carrying no physics, a physical argument carrying no geometry, and an inversion of the measured coupling; the first two have disjoint inputs, and the third is independent of the defining equation it recovers rather than of the other two (scope in §12.1). One identification remains open — that the construction's coordinate is quantum electrodynamics' vacuum polarization; it is not small, this paper does not claim it, and it is provably not resolvable by perturbation theory, hence a measurement. The abandoned identifications and an erratum against the author's own published proof are recorded in Part VIII; numbered falsification thresholds with hard numerical cut-lines are stated in §1, including the next-generation recoil measurement that decides. The verification script printed in Appendix F is included in this record as verify_alpha.py, together with its certified output. Developed with AI assistance (Anthropic's Claude) for high-precision verification, literature search, and drafting; all original content and structural decisions are the author's, who takes responsibility for all claims.
No takes yet. Share an insight, caveat, or question.
C. De Meyer (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: