Aspden-Hearld: Quality Engineering (Step 4b)

QE run 2026-10-03. Inputs read: _text/manifest.json, all 9 digests in digest/, facts.json, claims.md, physicist.md, mathematician.md, contrarian.md, arbiter.md, primer.md, example.py. Spot checks against _text/Unified-Physics-8f6841.txt, _text/Aspden-s-A-Problem-in-Plasma-Science-dea62b.txt and one page image (PDF p.184). No deep/ folder exists: this project was reviewed before digests and deep reads were introduced.

Page convention used below: "PDF p.N" is the page in the PDF and the === page N === marker in _text. The printed book page number is PDF page minus 7 (PDF p.184 = printed p.177, confirmed from the page image). The arbiter and reviewers mostly cite printed pages (e.g. "p.151" = PDF p.158; "p.119" = PDF p.126-127).

Verdict: MINOR

The plasma-essay claims (1-3) and the cosmic-background claim (6) were reviewed carefully and their verdicts are well supported. The weak point is the book. About half of Physics Unified was never read in depth, and the arbiter and primer give "no verdict" on those chapters because of a "source gap". In fact the full OCR text exists in _text/ and is good quality (checked against a page image). The unread chapters hold many of the book's most specific numerical claims: the deuteron binding energy, neutron beta-decay energy, pion and muon masses, muon lifetime, proton magnetic moment, the Mercury-perihelion and retardation derivation, the Jupiter/Sun and Moon/Earth mass ratios and Earth's magnetic moment. None of these was checked, and claims.md and the primer are silent about them. That is a coverage defect in the project's main document.

The headline verdicts are unlikely to change. The new material belongs to claims 4-6 and is most likely post-hoc fitting of the kind already judged for alpha and m_p/m_e. So adding these verdicts through a "QE revision" section is enough, and Steps 2-4 do not need to be redone. If the Chapter 1-2 deep read shows that the perihelion derivation is sound and independent of general relativity, claim 4 would need a new sub-verdict. Even then the overall picture would hold.

1. Coverage

• Both manifest documents were digested and nothing was skipped. The folder has no junk files.
• facts.json lists both documents. Its how_read and text_available fields for the book are inaccurate. It says "how_read": "page images" and "text_available": "none", but the book was read mainly from Tesseract OCR, and that OCR now exists as full text in _text/ (the manifest has 212 OCR pages and 424,689 characters). The correct values are "partial (OCR)" and "OCR text, sampled; page images for pp.1-12 and eqs 157-168".
• facts.json gives the essay's year as "web page printed 6/11/2019". The essay is signed 30 October 2005 [read: essay, line 181 of the extracted text]. The arbiter caught this, but facts.json was never corrected.

2. Shortcuts

3. Missed material (from the digests; none of it appears in facts.json, claims.md, the reviews, the arbiter or the primer unless noted)

Precision numerical claims in Chapters 7-8. These are the book's strongest-looking evidence. Each needs a verdict. - Deuteron binding energy calculated as 2.22464 MeV, "exactly" the measured 2.22464(4) MeV [read: PDF p.135-136, OCR line 6035]. This is not mentioned anywhere. - Neutron beta-decay electron kinetic energy calculated as 1.531173 electron-mass units, against 1.53116(8) measured [read: PDF p.140, OCR line 6276]. Not mentioned. - Charged pion mass 273.1262 m_e against 273.1266(39), and muon mass 206.7688 m_e against 206.76859(29) [digest p.142-144]. Only the mathematician mentions these, in one bullet. The arbiter and primer do not. - Muon g-factor 1.001165884 against 1.001165895(27) [digest p.144-145]. Not mentioned. - Proton magnetic moment 2.792846 nuclear magnetons against 2.7928456 measured, "one part in two million" [read: PDF p.149, OCR line 6705]. Not mentioned. - Muon lifetime 2.1973 microseconds against 2.197134(77) [read: PDF p.153, OCR line 6860], together with a non-relativistic explanation of the lifetime increase of fast muons (PDF p.155). Not mentioned. - Neutral pion 134.959 MeV against 134.9645, and quark masses 330.6/336.9 MeV matched to MacGregor's fit [digest p.160-162]. Not mentioned. - Psi decay thresholds of 3551, 3503, 3455, 3415 and 3340 MeV "exactly" matched [digest p.127-128]. The contrarian and arbiter discuss only the 3.095 and 3.684 GeV states.

The book's own revised alpha (affects the arbiter's reasoning). Appendix III revises eq. 157 for lattice motion at 390 km/s and gives alpha^-1 = 137.035953 [read: PDF p.211, OCR line 9692], not 137.035915. The arbiter and two reviewers say "the formula was already under strain c.1980" because 137.035915 lies 3.2 sigma from Williams-Olsen's 137.035963(15). Against the book's final value, 137.035953, the gap is only 0.7 sigma [calc: (137.035963 - 137.035953)/0.000015]. The modern comparison still fails: CODATA 137.035999177 minus 137.035953 is 4.6e-5, about 0.34 ppm and roughly 2,100 standard uncertainties. The verdict "Tested, not supported" stands. The "under strain in 1980" remark and the 3,800-sigma figure should name which of Aspden's values they use. The primer should mention the 390 km/s correction, because it shows how the formula was adjusted after the fact.

Gravitation and relativity (Chapters 1-3). - Aspden derives a retardation factor 1 + 3(v/c)^2 and claims to recover Einstein's perihelion formula without general relativity, citing Gerber (1898) [digest p.25-36]. The book's back cover makes this the book's selling point [read: PDF p.212]. It is not reviewed anywhere. The contrarian notes "Chapter 2 ... not read". - Trouton-Noble is cited as evidence for Maxwell's law over Lorentz's for isolated charges [digest p.14-15]. This is directly relevant to claim 1 and is not mentioned. - Aspden proposes tests: a speed-dependent Rydberg constant, R proportional to 2 - (1 - v^2/c^2)^(1/2), checked against Mandelberg-Witten and Snyder-Hall data (0.502 +/- 0.003) [digest p.69-72]; a repeat of the Silvertooth experiment with ruby lasers [digest p.65-68]; and a re-reading of Michelson-Gale-Pearson and the Townes maser test [digest p.61]. These are "something the reader could test" items that the contrarian would want. None is mentioned. - The 1977 U-2 CMB anisotropy result (390 +/- 60 km/s, "Aether drift detected at last", Nature 3 Nov 1977) is used as evidence for the C-frame [digest p.54]. The contrarian mentions it once. The arbiter's claim 3 discussion covers the CMB dipole only in general terms. - k = 2 from the Sucksmith-Bates (1923) gyromagnetic anomaly [digest p.46]. Not mentioned.

Cosmic chapters (Chapter 9). - Planet/Sun mass ratio from eq. 213: Jupiter 1/764 against 1/745 observed (k = 2), and Moon/Earth 1/83 against 1/81 (k = 1). The factor k changes between the two cases [digest p.165-166]. Only the contrarian mentions this, in one line. - Earth's magnetic moment predicted as 7.86e25 against 8.06e25 cgs measured [digest p.176]. Not mentioned. - Geomagnetic reversals at space-domain boundary crossings: a 700,000-year period, fossil extinctions, a reversal 12,000 years ago, and Steiner's galactic periodicities [digest p.177-181]. These are named in claims.md, but no verdict was given. - Ball lightning as rotating vacuum (8e9 J/m^3 against 2-5e9 J/m^3 observed), Tunguska, tornadoes, Zinsser's 1975 anomalous-torque work, and earthquake lights [digest p.195-201]. Only facts.json mentions ball lightning. - Aspden's own publication history: a 1960 first statement with lattice-particle mass 3.7e-29 g, and the 1966 second edition with the 2.587 GeV graviton [digest p.199]. This matters for which numbers were predictions and which were fits.

Essay. Nothing important is missing. All citations and the Correa link are carried through.

4. Accuracy spot checks

No load-bearing number was found to be wrong. The one substantive accuracy issue is the omitted Appendix III value (section 3).

5. Process

• Source tags. Tags are present and used honestly in every file. Tag counts: physicist 10 [web] and 22 [memory]; mathematician 12 and 17; contrarian 9 and 21; arbiter 18 and 18; primer 16 and 13. The arbiter lists the medium-confidence facts it could not check (Lorentz-invariance bounds, beam-ion consistency, coronal-hole ion temperatures, the fate of the 2.26 GeV particle), and that is acceptable. One gap: the physicist's "Tested, not supported" for claim 1 rests on [memory, medium] that was not web-checked. The arbiter correctly downgraded it rather than relying on it. The fate of the 2.260 GeV state is a medium-confidence [memory] fact behind a verdict and should be web-checked during remediation.
• Web checks. These were done within budget for the key verdicts: CODATA alpha, m_p/m_e and G; Borexino; the Sethian abstract; CMB T(z); Gaia escape speed; MICROSCOPE.
• Verdict scale. It is applied consistently, with split verdicts clearly labelled. Two places need care:
• Claim 4 G mixes "Internally inconsistent" (as a parameter-free claim) with "Tested, not supported" (numerically). That is defensible but should be stated as an explicit split, the way the other claims are.
• "No verdict" for unread chapters is not on the scale. It is honest, but it must be replaced after the deep reads.
• The "source gap" note is repeated in claims.md, the reviews, the arbiter and the primer, as Step 1 requires. It is now overstated. The book's text is available as OCR of checked quality. The real remaining gap is the cited external papers and patents.
• Primer. It follows the Step 4 order, sections 1-9 plus answers. It suits the reader profile: symbols are defined and acronyms expanded (OCR on first use), prerequisites have refreshers and textbooks, and the worked numerical example is backed by code. It says nothing on the particle-mass and perihelion material (see section 3).
• example.py. It runs cleanly with .venv/bin/python. It reproduces the fly-by cancellation (net about 1e-20 eV against a ±7.3e-5 eV swing), the second-order directional term, T = 2.557 K, the annual swing of ±0.040 K, the Galactic 267-453 K, alpha for 1841, 1843 and 1845, and G = 6.6650e-8.
• One small inconsistency: example.py prints "g that reproduces CODATA G exactly = 5062.12", while the arbiter gives 5062.16. Align them; this is a rounding or input difference in e/m_e.

6. Remediation actions

Deep reads (all on slug Unified-Physics-8f6841, from _text/ with page images for equations; PDF page numbers):

1. pp.130-162, Opus. Chapters 7-8, particles of matter. For each claimed match (deuteron 2.22464 MeV, neutron decay 1.531173, pion 273.1262, muon 206.7688, muon g 1.001165884, proton moment 2.792846, muon lifetime 2.1973 us, pi0 134.959 MeV, quark masses 330.6/336.9), list: - the inputs, and which parameters were chosen or fitted (k, form factor 0.8, 2-alpha factor, model choice A-D); - the measured value Aspden compared against and its date; - the modern value.

Also say whether any number was published before its measurement. Check the key equations (166, 175, 181, 190, 198-199, 202) against page images. 2. pp.7-36, Opus. Chapters 1-2. Is the derivation of eq. 14 consistent, given that the Newton III assumption is used and then dropped? What does the Trouton-Noble argument claim? Does the retardation factor 1 + 3(v/c)^2 really reproduce the GR perihelion advance (43"/century for Mercury) without circular inputs? How does it relate to Gerber 1898 and its known problems? 3. pp.94-118 plus p.211, Opus. Chapters 5-6 and Appendix III. The Planck law, Schroedinger equation and g-factor derivations: what is assumed and what is derived? Compare the g-factor series g = 1 + alpha/2pi - 0.32735(alpha/2pi)^2 with QED (Schwinger term and the -0.328478965 coefficient). Check the lattice sum r/d = 0.302874 and the 1843 selection. Check the Appendix III revision to 137.035953. 4. pp.119-129 plus p.158, Sonnet. The graviton (5062.75 to 5063) and its decay-pattern table; the order in which g and G were fixed; the psi decay-threshold table; the 2.260 GeV and 3.772 GeV claims. What does PDF p.158 (printed p.151) say about the 1843 route putting the early graviton derivation "in doubt"? 5. pp.163-181, Sonnet. Chapter 9 cosmic material: - the eq. 213 mass ratios (Jupiter 1/764, Moon 1/83) and the k = 2 versus k = 1 switch; - eq. 221 and the derived densities rho_0 of 257, 248 and 288 g/cc; - Earth's magnetic moment of 7.86e25 (eqs 227-229) and the choice of boundary height (100 km versus 200 km on PDF p.194); - the 700,000-year reversal prediction against the geomagnetic polarity timescale. 6. pp.54-80, Sonnet. Chapter 3 aether experiments: - Aspden's treatment of the U-2 anisotropy, Michelson-Gale-Pearson, Townes et al. 1958, Silvertooth, Champeney-Isaak-Khan, Kennedy-Thorndike, Ives-Stilwell, and the Fresnel-drag derivation (eq. for u(1 - 1/n^2) = v); - his proposed Rydberg-constant speed test against the Snyder-Hall 0.502 +/- 0.003 result.

Which of these give a measurable prediction that differs from special relativity?

File changes after the deep reads (MINOR path): - facts.json. - Essay year: 2005, signed 30 October 2005; the printout is from 2019. - Book: how_read should say "OCR text (Tesseract, all 212 pp.), sampled in Steps 2-3, plus deep reads of pp. ...", and text_available should be "partial (OCR; equations need page images)". - Book key_claims: add the deuteron, neutron decay, proton moment, muon lifetime, perihelion and geomagnetic-moment claims, and the Appendix III alpha of 137.035953. - claims.md. Rewrite the "Source gap" note so it says the book's OCR text is available and of checked quality, and that the gap is now the external papers, patents and the 1996 report. Extend claim 4 to add the retardation/perihelion derivation. Extend claim 5 to add the particle-mass and magnetic-moment matches. Extend claim 6 to add the geomagnetic moment and planetary mass ratios. - arbiter.md. Add a "QE revision" section that: - gives verdicts in place of every "no verdict" (Planck, Schroedinger and g-factor derivations; geomagnetism and domains; field reversals); - adds sub-verdicts for the particle-mass matches (most likely "Historically superseded but instructive" or "Tested, not supported", depending on modern values and the fitted-parameter count) and for the perihelion derivation; - corrects the "under strain c.1980" remark using the Appendix III value of 137.035953 (0.7 sigma from Williams-Olsen), and states the modern gap for both 137.035915 and 137.035953; - makes the claim 4 G verdict an explicit split; - web-checks the fate of the 2.26 GeV state. - primer.md. - Rewrite the source-gap paragraph. - Add to the evidence table (section 6) rows for the particle-mass matches, the perihelion derivation and the geomagnetic moment. - In section 7, discuss "many precise matches with several chosen parameters" as the general lesson, using one particle example with its fitted inputs counted. - Mention the Appendix III 390 km/s correction to alpha. - Optionally add a check-yourself question on counting free parameters. - example.py. Make the "g for exact CODATA G" value (5062.12) agree with the arbiter's 5062.16, or explain the difference. Optionally add the Appendix III alpha and the count of parameters in one particle-mass formula.

Recheck

Recheck run 2026-10-03 against current facts.json, claims.md, arbiter.md (with its "QE revision" section), primer.md, example.py and deep/.

Defects

Remediation actions

Remaining open issues: D1, D6 (primer part), D7 (primer part), D9, D10, D11, D12. The headline verdicts are unchanged by the work so far.

Final: OPEN (7 issues)

Recheck 2

Recheck 2 run 2026-10-03 against current facts.json, claims.md, arbiter.md (QE revision, Round 2), primer.md, example.py and deep/Unified-Physics-8f6841.md.

Remediation actions R1, R3, R4, R5, R6 and R10 are done through the deep reads and the primer rewrite. R11 is done by explanation. R12 remains, the same item as D10.

Final: OPEN (1 issues)