Evans-Myron: Science Contrarian review (Step 2)

Reviewer role: Science Contrarian. Standard of judgement: experimental evidence first; separate "tested and failed" from "never properly tested" from "dismissed without testing"; build the strongest honest case; apply equal rigour to the researcher and to the critics.

Source gap (repeated from Step 1, narrowed)

The folder holds only two books: Evans et al., Criticisms of the Einstein Field Equation (CEFE, 2010, 548 pp.) and Felker, The Evans Equations of Unified Field Theory (2005, a popularisation, not by Evans). There are no primary AIAS papers (e.g. papers 63, 93, 94, 122), no experimental data, no device descriptions, and no peer-reviewed journal papers. All statements about what critics said or whether anything was replicated rest on one secondary web source (Wikipedia) plus what I could check myself. Where a conclusion rests only on [memory], it is flagged.

What I did myself for this review: - Read in full: CEFE preface (Eqs. 1-22), Ch.1, Ch.2 intro and Sec. 2.6 (Eqs. 2.148-2.167; the page was rendered and read as an image [read: page image p.47 of the PDF]), Ch.4 intro and method (Eqs. 4.1-4.13), Ch.3 intro and Secs. 3.1-3.2 start. Grepped Felker for Croca, the electrogravitic section and the "difficult to test" statements. - Ran SymPy (in .venv) on two of Evans's equations (details below).

Corrections to Step 1

1. Step 1 left the antisymmetric-connection argument as a [memory] suspicion ("seems to conflate..."). It is now checked [read]: preface Eqs. (5)-(6) define torsion T^λ_{μν} = Γ^λ_{μν} − Γ^λ_{νμ}, then assert that "the connection therefore has the same symmetry as the commutator", Γ^λ_{μν} = −Γ^λ_{νμ}. That is a non sequitur (see claim 1). The [memory] low item that Hehl/Obukhov showed the dual identity fails is not needed: the failure is demonstrable from Evans's own equations (below).
2. Step 1 and claims.md describe Ch.4 as a "computer algebra test" that "shows" violation of the dual identity. That framing hides what is computed. Eq. (4.8) defines R^{κμ}{}{μν} = R^{κ0}{}{0ν} + R^{κ1}{}{1ν} + ..., which is just a contraction of the Riemann tensor, i.e. the Ricci tensor with an index raised (up to sign). I verified this with SymPy on a spatially flat FLRW metric (a(t)): Evans's contraction equals exactly minus the Ricci tensor component for every component. So "R^{κμ}{}{μν} ≠ 0" means only "Ricci ≠ 0", which the Einstein equation already says holds wherever there is matter. The test is circular, not an independent discriminator. claims.md claim 2 should say this.
3. Step 1 omitted that Ch.4 also reports that "several well known line elements failed" the Ricci cyclic identity (4.12) and metric compatibility (4.13) [read, CEFE Sec. 4.2]. For any metric computed with the Christoffel connection both are identities, so a failure indicates a coding/input problem or a non-metric line element, not a problem with general relativity. Step 1 also repeats "the code reproduced known results" as if neutral; it is Evans's own statement.
4. Step 1 sources mention paper 93 / "Physica B" abuse as sociological evidence. That is Evans's unsupported statement [read, preface]; no referee reports are in the folder. The [web] record (Wikipedia, below) says the main publications were in Foundations of Physics Letters 2003-2005 and that the editor 't Hooft's 2008 editorial ended support; it does not mention Physica B. Treat the Physica B story as unverified.
5. Minor: facts.json says Sec. 2.6 attributes the Bedini motor to "papers 63 and 94". Per the text, paper 63 is the Coulomb-law resonance derivation and paper 94 is the Bedini/patented-device application [read, CEFE p.38].
6. claims.md cites "Felker Ch.6-8", "Ch.7-12", "Ch.14" for chapters I did not verify; I confirmed only that the electrogravitic material (TOC pp. 254, 290) and Croca references (Felker text lines about the uncertainty chapter) exist.

Calculations used below

A. Evans's spin-connection equation (CEFE Eq. 2.156, from ω_r = −1/r): φ'' + φ'/r − φ/r² = −ρ/ε0. SymPy dsolve of the homogeneous equation gives φ = C1/r + C2 r. Write r = e^s: the equation becomes φ_ss − φ = −r²ρ/ε0, which has exponentially growing/decaying (hyperbolic) homogeneous solutions and no oscillation in real r. Eq. (2.158), κr = exp(iκR), makes R = −i ln(κr)/κ complex for real r; the oscillator form (2.159) appears only in that complex variable. A "resonance in R" is a rotation of the coordinate into the complex plane, not a resonance of any physical field at real positions.

B. Evans's general case (CEFE Eqs. 2.160-2.162 [read: page image p.47]). Substituting ω_r = ω0² r − 4β ln r − 4/r (Eq. 2.161) into the printed Eq. (2.160) with SymPy does not give φ'' + 2βφ' + ω0²φ = −ρ/ε0 (Eq. 2.162): the φ' coefficient is 2/r + ω_r, not the constant 2β, and the φ coefficient is not constant. Also Eq. (2.165) uses κ0² where (2.161) uses ω0², and (2.164)-(2.165) do not match (2.161). The ln r also takes the logarithm of a dimensional quantity. This is a derivation error in the printed text, unless a typo hides a different intended formula; I cannot see paper 90.

C. Evans's "dual identity" test (point 2 above), FLRW check: R^{00}{}_{0 0}-contraction = −3ä/a; Ricci^0_0 = +3ä/a, and likewise for the spatial components.

D. Antisymmetric connection test (reasoning, no code): Γ is not a tensor; in flat space in plane polar coordinates the Christoffel symbols include Γ^r_{θθ} = −r ≠ 0. Evans (preface) says Γ^λ_{μν} must vanish for μ = ν. Γ^r_{θθ} = −r contradicts that, though the space is flat and torsion-free. [memory] high: standard textbook result (Γ^r_{θθ} = −r, Γ^θ_{rθ} = 1/r).

---

Claim 1: The Einstein field equation is wrong because it assumes zero torsion and so violates the dual Bianchi ("Cartan-Evans") identity; the connection must be antisymmetric

a. Mathematics: Internally inconsistent in its key step. T^λ_{μν} = Γ^λ_{μν} − Γ^λ_{νμ} is antisymmetric by construction, whatever Γ is. A general Γ splits into a symmetric part and an antisymmetric part and T sees only the antisymmetric part; nothing forces Γ itself to be antisymmetric (Eq. 6 does not follow from Eq. 5) [read: preface Eqs. 5-7]. Imposing it breaks flat-space polar coordinates (Calc D). The "dual" machinery (Eqs. 8-16) then defines a Hodge-dual connection, calls Γ̃ a connection in Eq. (11), then in Eq. (15)-(16) replaces a three-term cyclic sum by a single contracted term "which is the same as" it. A sum of three terms vanishing does not mean one term vanishes. For the torsion-free case the cyclic sum is the ordinary first Bianchi identity, which holds exactly for the Levi-Civita connection. Eq. (16) as used in Ch.4 (4.1) with T = 0 reads "Ricci = 0", which is not an identity of geometry. Mainstream differential geometry texts contain no such dual identity [memory, medium; consistent with Wikipedia's summary of Hehl/Obukhov below].

b. Experiments: Nothing experimental is claimed for this claim; it is geometric. Torsion-free GR passes all precision tests (below, claim 2). Experiments do not rule out small torsion; they bound it.

c. Tested: Only by critique, not experiment. [web https://en.wikipedia.org/wiki/Einstein%E2%80%93Cartan%E2%80%93Evans_theory] lists Trovon de Carvalho & Rodrigues (2001), Rodrigues & Souza (2005), Bruhn (2008), Hehl & Obukhov (2008) and 't Hooft's 2008 editorial as critiques, and says the theory has no documented replication. I did not read those papers; I rely on Wikipedia's summary.

d. If true: GR's field equation and its solutions (Schwarzschild, FLRW) would be geometrically invalid; but GR's agreement with data (claim 2) would then be a puzzling accident. The identity would have to be an actual theorem of Riemann-Cartan geometry, and it is not.

Strongest honest case for the claim: the contrarian respects one true thing here. Torsion is a legitimate generalisation (Cartan 1922; Einstein-Cartan-Sciama-Kibble theory, [memory] high), and torsion-free GR is an assumption, not an experimental result for all scales. Evans is right that the symmetric connection is a choice. The honest criticism of him is that he then asserts that choice is an error and that the resulting equations contradict an identity which they do not.

Verdict: Split. Strong form (EFE is geometrically wrong; connection must be antisymmetric; Ch.4 shows violation): Internally inconsistent and in the Ch.4 test Refuted by calculation (the test reduces to "Ricci ≠ 0 with matter", which is the Einstein equation itself; Calc C). Weak form (torsion may exist and GR is the zero-torsion limit): Plausible, untested at the precision relevant to a testable effect; this is the standard Einstein-Cartan idea, not specifically Evans's.

Claim 2: Computer algebra on ~62 metrics shows every matter solution violates the identity, so GR, black holes, gravitational waves, the Big Bang and dark matter are invalid; Schwarzschild was misderived (Crothers)

a. Mathematics: Two separate pieces. - Eckardt/Evans Ch.4: The check is circular (Calc C). The code may well be correct at computing Riemann tensors ("reproduced known results" is their own report, not seen by me; 62 tables not read), but what is computed is the Ricci tensor, so the "violation" in matter solutions is just the field equation's source term. The paper also self-contradicts: it says vacuum solutions pass "fortuitously" (Sec. 4.1) while Crothers (Ch.3) says vacuum solutions are meaningless; Evans uses Schwarzschild as test case and Crothers says it is nonsense. - Crothers Ch.3: The historically accurate point, which I confirmed in a primary source, is that Schwarzschild's own 1916 form used R = (r³ + α³)^{1/3}, with the singularity at r = 0 corresponding to R = α, so R never goes below α. Crothers' claim that "α was undetermined and not 2m" is not what the paper says: Schwarzschild writes that the solution "contains only the constant α that depends on the value of the mass at the origin" [web https://arxiv.org/abs/physics/9905030, English translation of "On the Gravitational Field of a Mass Point"]. That R = α is the point where g_tt vanishes, and the Hilbert form uses R as areal radius with α = 2GM/c² [memory, high; the Newtonian limit fixes it]; it is a coordinate feature, and extension of the manifold through it is the standard Kruskal-Szekeres construction, which Crothers rejects but which does not conflict with the algebra. The statement that "Ric = 0 contains no matter, so it is meaningless" is like saying Laplace's equation outside a charge is meaningless; the mass enters through boundary matching to an interior solution (e.g. Oppenheimer-Volkoff [memory, high]). I did not read all of Ch.3 and did not check its Gaussian-curvature argument in full; I cannot say whether every step is wrong, only that the headline "no black holes, no gravitational waves" does not follow from what I read.

b. Experiments: GR's predictions are tested: [web https://en.wikipedia.org/wiki/Tests_of_general_relativity] solar light bending to ~0.03% by VLBI, Cassini time-delay, binary-pulsar periastron and orbital decay, Event Horizon Telescope images of M87 (2019) and Sgr A* (2022). Gravitational waves were directly detected on 14 Sept 2015, announced 11 Feb 2016, as a binary black hole merger [web https://en.wikipedia.org/wiki/First_observation_of_gravitational_waves], and indirect evidence came from the Hulse-Taylor pulsar, Nobel Prize 1993 (same source). Dark matter is a separate question (see d).

c. Tested? GR was tested by many independent groups (LIGO/Virgo, VLBI teams, pulsar timers, EHT). Evans/Crothers themselves ran no experimental test; their "tests" are algebraic. The black-hole/gravitational-wave denial is therefore in the class "tested, not supported", but not because a test was designed against it; mainstream measurements contradict it. Whether the rejection of Crothers' arguments was by reasoned rebuttal or by non-engagement, I cannot say from the folder: Wikipedia does not discuss Crothers. [memory, low]: that Crothers' papers were circulated mostly outside mainstream journals.

d. If true: LIGO's detection of merger waveforms, EHT shadows and pulsar decay would all need an alternative explanation; GR's field equation would need replacement. Note: dark matter is mainly inferred from galaxy rotation, lensing and CMB data [memory, high]; it does not follow from the Einstein equation alone, and "dark matter is mathematically impossible" is not an argument Evans makes in any passage I read. It is a stretch in the claim list itself; and the problem of dark matter is genuinely open and is the place where a contrarian might legitimately push on mainstream.

Strongest honest case: GR has been tested in strong fields only since 2015; the weak-field results do not prove horizon structure; alternatives (modified gravity) are live research. Crothers' historical observation about Schwarzschild's original coordinates is accurate and under-told in textbooks. None of that gets to "impossible".

Verdict: Split. Ch.4 "test": Refuted by calculation (circular; Calc C). Crothers' "black holes/gravitational waves are mathematically impossible": Tested, not supported (data from LIGO, EHT, pulsars point the other way) with the underlying algebraic argument being Internally inconsistent with Schwarzschild's own statement that α depends on the mass. Dark matter and Big Bang being invalidated by this: Not testable as stated in the connection claimed.

Claim 3: Electromagnetism is the torsion of spacetime (A = A(0) q, F = A(0) T), giving a generally covariant unified field theory; Maxwell-Heaviside is the zero-spin-connection limit; gauge theory is incomplete

a. Mathematics: The ansatz is dimensionally consistent (A(0) is a voltage-times-time-per-length scale, a free constant) and is a postulate, not a derivation [read, CEFE Sec. 2.1, 2.7]. The text says the spin connection ω "must ultimately be determined experimentally" [read, p.40]; with ω a free function the theory can mimic many outcomes. Per Wikipedia, Bruhn (2008) found the O(3) law not Lorentz invariant and Hehl/Obukhov identified fundamental flaws [web, not read directly]. I cannot say from the folder whether the field equations are consistent with each other.

b. Experiments: Standard Maxwell electrodynamics is confirmed to enormous precision (Coulomb's law exponent, photon mass bounds). ECE says it recovers all of these off resonance [read, p.40]; this is asserted, not demonstrated here. The Faraday-disk argument in Sec. 2.7 ("no explanation in the standard model") is wrong as a statement of the standard model: homopolar induction is covered by the Lorentz force on the moving conductor [memory, high; Griffiths ch. 7]. I note that this is a point where Evans overstates the failure of the mainstream, as part of the rigour I owe the researcher.

c. Tested: ECE's unique predictions (B(3) spin field is an Inverse Faraday effect idea from 1992 [read]; gravitational effect on the Sagnac effect; polarisation of light grazing a white dwarf) were not tested by anyone I can find; Wikipedia records no documented replication of ECE-specific effects [web]. Never properly tested, and for the most part no one ran the experiments; the reasons given by critics were mathematical, not that the experiments failed.

d. If true: electromagnetism would not be a U(1) gauge field; QED's success would need re-derivation within a torsion framework.

Verdict: Not testable as stated (free spin connection, postulated ansatz; no unique, falsifiable prediction in the text I read). The critics' stronger result, that the field equations are internally inconsistent, is plausible but I did not verify it; it rests on [web] summary only.

Claim 4: A nonzero spin connection makes the Coulomb/Poisson equation resonant; this explains Tesla and lets circuits (Bedini motor) draw usable power from spacetime torsion; electrogravitic relation E = φ(0) g / c²

a. Mathematics: Refuted on the text's own example (Calc A): with ω_r = −1/r the real-r equation has solutions C1/r + C2 r and exponentials in ln r; no oscillation. The oscillator appears only after the complex substitution (2.158), which maps real r to complex R. The general case (2.160-2.162) does not follow from the printed equations (Calc B). E = −(∇ + ω)φ also needs ∇×E = 0 for statics, which restricts ω to being parallel to ∇φ in a way not discussed; the radial choice works only for spherical symmetry. "Spin connection resonance amplifies power" in the text conflates a resonance in a mathematical variable with energy gain; no energy source or balance is calculated. The paper states "No violation of the laws of conservation of energy and momentum occurs" [read, p.41] but gives no energy budget. Under accepted laws, a resonant LC circuit stores and exchanges energy; it does not create it (a circuit Q factor amplifies voltage, not average power) [memory, high].

b. Experiments: Tesla's resonant transformers are explained by ordinary LC resonance, with no need for ω [memory, high]. Over-unity from magnetic motors has never been demonstrated in an independently verified, controlled test [memory, medium; no source checked for Bedini; Wikipedia ECE page lists the "energy from spacetime" circuits UFT364/382/383 but "no documentation of successful replication" (web)].

c. Tested: I found no controlled, neutral test of the Bedini motor or the SCR circuits; Wikipedia records none [web]. Evans's own papers 63 and 94 (not in folder) are the sole sources; none was read. So the experimental claim is never properly tested (and not dismissed on the basis of a test either; it was dismissed on mathematical grounds). A contrarian would say: if a device is claimed to give excess power, a simple input/output measurement with calorimetry or a calibrated power meter would settle it, and nobody credible has run it. The burden for that test lies on the claimant, who supplies none in these books.

d. If true: energy conservation would have to be restated (spacetime torsion as an energy reservoir) and electrostatics (Poisson's equation, Gauss's law) would need an extra free field.

Electrogravitic E = φ(0) g/c² [read, Felker]: dimensional check: φ(0) has units of volts; g/c² has units 1/m; E has V/m, so it is dimensionally consistent. Felker admits "the effects will be difficult to test since they are quite small" [read, Felker lines about electrogravitic section]; no test is cited.

Verdict: Split. Resonance-from-the-math: Refuted by calculation (Calc A, B). Free-energy claim: Never properly tested, and on energy-conservation grounds Refuted by calculation in any form that claims net output without an input. Electrogravitic: Not testable as stated at the magnitude Felker admits.

Claim 5: The Heisenberg uncertainty principle is refuted and replaced by the causal Evans wave equation (□ + kT)q = 0 giving Dirac, Klein-Gordon, Schroedinger; the photon has a B(3) longitudinal field and a small mass; electroweak masses follow without a Higgs

a. Mathematics: Not demonstrated in what I read. The "Lemma" □q = Rq with R = −kT is a postulate; deriving three different wave equations (spin-1/2, spin-0, non-relativistic) from one equation for a tetrad needs specific limits and the text gives them in summary [read, Felker TOC and sampled chapters]. Uncertainty follows from the Fourier properties of wavefunctions (a mathematical theorem, Δx Δp ≥ ħ/2 for the standard deviations; [memory] high), so refuting it as mathematics requires showing the Fourier theorem false; it is not that. What can be disputed is the interpretation (measurement disturbance vs. intrinsic spread), which experiments tested (Ozawa-type error-disturbance relations [memory, medium]), and which is a different thing from the standard-deviation relation.

b. Experiments: Heisenberg-type bounds are confirmed in countless measurements. Felker quotes Evans as saying it was "refuted experimentally by the Croca group in Lisbon" [read, Felker]; I found no source for the Croca results other than Felker's references to Croca's book and 1996 microscope paper. I did not verify or replicate them; I note they are not mainstream-confirmed. [memory, low]: that Croca's claims have not been reproduced by independent groups. Photon mass: The experimental upper bound is about 1e-18 eV [web http://pdg.lbl.gov/2026/listings/rpp2026-list-photon.pdf, summarised from a search result; the PDG listing was not opened]. A "very small nonzero mass" is therefore not excluded by experiment; the issue is that ECE gives no number or prediction to test. Higgs: the Higgs boson was discovered at the LHC in 2012 [memory, high; not rechecked].

c. Tested: Uncertainty relations: tested by many groups, replicated. B(3) spin field: Evans infers it from the inverse Faraday effect, which is real (circularly polarised light magnetising a medium [memory, high]) but is explained in mainstream theory without a longitudinal photon field. Never tested as a discriminator.

d. If true: quantum mechanics' probabilistic basis, QED's treatment of the photon (massless, two transverse polarisations) and electroweak symmetry breaking would all be replaced.

Strongest honest case: interpretations of quantum mechanics are open; deterministic, causal alternatives exist (de Broglie-Bohm, [memory] high), and a tiny photon mass is a legitimate experimental question. But Evans offers no equation that yields numbers for either.

Verdict: Split. Uncertainty principle refuted: Tested, not supported. Unification wave equation reducing to Dirac/KG/Schroedinger: Never properly tested and not derived in the text I read; as stated, Not testable as stated. Photon mass nonzero: Plausible, untested (bound exists, no ECE-specific value). Electroweak masses without Higgs: Never properly tested (and in tension with the 2012 Higgs discovery).

Claim 6: Academic physics has ignored or abused these results through "belief retention"; wide readership of AIAS papers shows acceptance

a. Mathematics: Not applicable. It is a sociological claim.

b. Evidence: "Fully accepted" is supported only by website hit counts [read, CEFE Sec. 2.1]; readership is not replication. The preface says paper 93 was met with "crude personal abuse" at Physica B [read]; no referee reports are present. The [web] Wikipedia summary says ECE appeared in Foundations of Physics Letters 2003-2005 and that, in 2008, its editor 't Hooft wrote that further discussion would cease "unless very good arguments are presented", noting ECE work remained "limited to personal web pages and absent from the standard electronic archives" [web https://en.wikipedia.org/wiki/Einstein%E2%80%93Cartan%E2%80%93Evans_theory]. Specific written critiques exist (Bruhn; Hehl and Obukhov; Rodrigues and collaborators) [web].

c. How it was rejected: On the evidence I could access, the mathematical claims were not ignored: they were publicly criticised with specific objections, and I have independently confirmed two of them (Eq. 6 non sequitur; Ch.4 test circular). Funding or publication blocking: Evans says abuse and ostracism; I have no source beyond the preface; Wikipedia documents an editorial closing of the discussion in a journal that had earlier published Evans. The experimental claims (Bedini, Tesla resonance) were dismissed without a test, but were also not backed by a clearly specified device and measurement that a neutral lab could run. This is not equal to a well-posed claim that was suppressed.

d. If true: the sociology would matter less than the physics; even if mainstream behaviour was harsh, the errors above stand on their own.

Researcher's own conduct (same rigour as critics): the books present "The British and Commonwealth Civil List Scientist" as authority, call ECE "fully accepted" and "well accepted" while quoting its critics' dismissal as abuse, call Einstein's symmetric connection "inexplicable" without engaging the standard answer, and in Ch.4 name coded outputs "charge density" and "current density" without any physical derivation [read, Sec. 4.2]. Appeals to readership figures are not evidence.

Verdict: Not testable as stated as physics. As history, the record I can see is "criticised on specific mathematical grounds", not "dismissed without testing", for the geometry claims; for the energy-device claims it is dismissed without testing on theoretical grounds, but with no test offered by the claimants either.

---

Source summary

• [read]: CEFE preface, Ch.1, Sec. 2.1, Sec. 2.6 (incl. page image p.47), Ch.3 intro and 3.1-3.2 start, Ch.4 intro and Sec. 4.2; Felker greps (Croca, electrogravitic, "difficult to test").
• [calc]: SymPy checks A, B, C above (run locally).
• [web]: https://en.wikipedia.org/wiki/Einstein%E2%80%93Cartan%E2%80%93Evans_theory (critics, journal history, no replication); https://en.wikipedia.org/wiki/First_observation_of_gravitational_waves (GW150914; Hulse-Taylor); https://en.wikipedia.org/wiki/Tests_of_general_relativity (light bending, binary pulsars, EHT); https://arxiv.org/abs/physics/9905030 (Schwarzschild 1916 translation, "constant α depends on the value of the mass at the origin"); http://pdg.lbl.gov/2026/listings/rpp2026-list-photon.pdf (photon mass bound, via search snippet; the Wikipedia Photon page fetch did not give a number).
• [memory] high: Γ^r_{θθ} = −r in polar coordinates; Levi-Civita connection and first Bianchi identity; Fourier origin of the uncertainty relation; Tesla coil operation as LC resonance; Lorentz-force account of the Faraday disk; Einstein-Cartan theory exists. [memory] medium (not checked): Oppenheimer-Volkoff matching, error-disturbance relations, Higgs discovery year 2012 (stated in Step 1 too), no verified Bedini replication. [memory] low (not checked): that Croca's results were not reproduced; Crothers' publication venues.
• Not checked: all AIAS primary papers; the 62 Ch.4 tables; Crothers Ch.3 beyond 3.2 start; Felker beyond sampled passages; Hehl/Obukhov and Bruhn papers themselves.

Verdict table