RankingsPrimerArbiterPhysicistMathematicianContrarianQEClaimsSources

Gardi-Lori: Science Contrarian review

Reviewer role: Science Contrarian (judges by experiment; separates "tested and failed" from "never tested" from "dismissed without testing"; applies the same rigour to the researcher as to the critics). Inputs: facts.json, claims.md (Step 1), and the source PDFs in source folder: Gardi-Lori.

Source tags: [read] from the document; [read: page image p.N] from a rendered page; [web] with URL; [memory] with confidence. Where a check was attempted and failed, it says so (the WebSearch budget was exhausted; I could use WebFetch for arXiv abstracts and NIST only).

Source gap (repeated from Step 1, with what I added)

  • The numbered equations of the Yee-Gardi papers (Electric Universe, Mass and Charge, Geometry of Particles) are images and are not in the text layer. I read the text layer of the Spacetime v2.3 paper in full around the key sections and rendered pages 8-9 (fine-structure ratio) and 28-29 (G derivation) of the PDF [read: page image]. The Geometry of Particles paper was only sampled, so my judgement of the "magic number" and decay claims rests on Step 1's summary plus the abstract-level text, not on the equations.
  • The companion papers (Mole and Charge; Fine Structure Constant and Pi; Distinti; Yee's wave constants; Gardi 2016 "Calibrating the universe") are not in the folder and were not read. No conclusion below treats them as read.
  • Conclusions that rest on [memory] alone are marked.

Corrections to Step 1

  1. Claim 3 note is wrong for two of the items. claims.md says "each checked case reduces to a definition or Planck-unit identity." That is true for c, h, G, Coulomb's constant (and the Bohr/Avogadro items are not checked by me, see below). It is not true for: - Fine-structure constant. Spacetime v2.3 eq. 2.2.4 gives alpha = 1/(4 pi^3 + pi^2 + pi) = 0.00729734 [read: page image p.8 of the PDF, label page 8]. This is not a definition: it is a numerological closed form. I computed it: 4 pi^3 + pi^2 + pi = 137.036304 vs measured 1/alpha = 137.035999, a relative error of 2.2e-6. The CODATA 2022 relative uncertainty is 1.6e-10 [web: https://physics.nist.gov/cgi-bin/cuu/Value?alph], so the formula misses by roughly 14,000 standard uncertainties. The paper's "match" is to six digits only. Step 1 missed that this is a falsified numerical coincidence, not a tautology. - Anomalous magnetic moment (OM particle, eq. 8). Step 1 correctly noted it is alpha/(2 pi) = 0.0011614; it is not a tautology but the one-loop Schwinger term. I confirm: e^2/Q_p^2 with Q_p = sqrt(h c / k_e) equals alpha/(2 pi) = 0.00116141 [read: OM particle p.4; calculation]. The measured value is 0.00115965 [memory: high], so the formula is off by about 0.15% (1.8e-6 absolute) while experiment and QED agree at the 1e-10 level or better [memory: high]. The paper claims to "calculate" the quantity; it does not match.
  2. Claim 5 wording. claims.md says "all future LIGO events should show the same gamma-ray follow-up." The paper says: "if all future 'gravitational waves' are followed by a gamma ray signal, then the idea ... should be revisited" [read: Medium paper p.4-5]. That is a revisit criterion, not a stated prediction. claims.md also omits that the Fermi signal was weak: see claim 5 below.
  3. Checked and confirmed: the NaCl arithmetic (sqrt(1.3052e-18/9.538e-26) = 3699 m/s), the e+e- number (1.637421e-13 J / 1.82188e-30 kg gives exactly c^2 when the inputs are 2 m_e c^2 and 2 m_e), 376.73/(2 pi) = 59.96 ohm, m c^2/r_e = 29.05 N (claims.md/facts say 29.1 N elsewhere; both consistent).

Not verified by me: the Bohr-radius, Bohr-magneton and Avogadro "derivations" (equations are images; I read only the surrounding text). Step 1's reduction of these to identities is plausible but unchecked.

Claim 1. Light and all forces propagate in a physical medium (e+e- lattice, later Planck-scale granules)

a. Mathematics. The central calculation is v = sqrt(A/(B+C)) with A the "binding energy" of an e+e- pair and B+C the pair mass [read: Medium paper p.2-3]. A = 1.637421e-13 J is 2 m_e c^2, the pair's total rest energy [read; calculation: 2 x 9.1094e-31 x c^2 = 1.63742e-13 J], and B+C = 2 m_e. So v = sqrt(2 m_e c^2 / 2 m_e) = c for any value of m_e. This is E = mc^2 rearranged, not a derivation. It is not a binding energy (annihilation energy is the rest energy released, not the energy holding a bound state). The paper itself admits the formula "applies to all particle-antiparticle pairs" and gives c for proton-antiproton or quark pairs: a formula that returns c for every input explains nothing about which medium exists. The rewrite c = sqrt(mu0^-1 / eps0) is a legitimate algebra step; assigning mu0^-1 the role of stiffness and eps0 that of density is an analogy, not a result. Also the NaCl number (3700 m/s) uses mass per ion pair rather than mass per volume: dimensionally it is energy per mass, which has units of m^2/s^2 so the check "works", but a real longitudinal sound speed in NaCl is higher [memory: medium], so the NaCl calibration is itself loose.

b. Experiments. Michelson-Morley (1887) and its descendants show no direction-dependence of c to a very small level. Modern optical-resonator tests reach about 1e-17 relative sensitivity [memory: medium; web check failed]. The paper's reply is that MM is "circular" because light detects the medium of light. That argument is wrong as stated: MM does not need the medium to be detected as an object, it tests whether the round-trip speed of light depends on the apparatus's orientation or velocity, which any medium whose frame differs from the lab would affect. The acoustic analogy offered also fails: an air-based acoustic MM with apparatus at rest in air gives a null because the air is co-moving, not because "air does not exist"; with a wind it does give a signal [memory: high; standard acoustics]. What the contrarian can legitimately say is that Lorentz's ether theory (a preferred frame with undetectable effects) is empirically equivalent to special relativity for these tests [memory: high], so "MM does not exclude every aether" is true for that narrow class, but it is not an argument for Gardi's medium, whose specific e+e- composition would carry extreme density and be affected by pair production and by gamma-gamma scattering limits, and which has no stated dynamics.

c. Tests. No test of the e+e- lattice itself exists. The MM-type tests are neutral and repeated (many labs, many decades) [memory: high]; they did not test this specific model.

d. If true. Special relativity would become an effective description of a preferred frame; QED vacuum would need a literal lattice. The vacuum polarisation and Lorentz-violation bounds from astrophysics would have to be reworked.

Verdict. Split. - Specific e+e- (and any-pair) lattice derivation of c: Internally inconsistent (it returns c for any particle-antiparticle input because it is E = mc^2 restated; the "binding energy" is rest energy). - "Light needs a medium" as a general philosophical claim: Historically superseded but instructive (a Lorentz-ether version is empirically equivalent to relativity, so this form is also close to Not testable as stated). - The "MM is circular" argument: refuted on the logic above, no experiment needed.

Claim 2. Charge is wave amplitude in metres; mass is standing-wave energy; Coulomb, electron mass and gravity follow from the same waves

a. Mathematics. Per the Mass-and-Charge paper as summarised by Step 1 and the Spacetime paper: m_e = mu0 e^2 / (4 pi r_e) and k = mu0 c^2 / (4 pi). Both are exact standard identities: the classical electron radius is defined by m_e c^2 = e^2 / (4 pi eps0 r_e), and 1/(4 pi eps0) = mu0 c^2/(4 pi) [memory: high; standard]. So they hold by definition. Renaming coulomb to metre is a change of unit assignment, not a prediction. In the G derivation I read eqs. 5.8.1-5.8.7 [read: page image p.28]. Eq. 5.8.1 begins F_g = mu0 c^2 (e^2 / 4 pi r^2)(alpha l_P^2 / r_e^2): that extra factor alpha l_P^2 / r_e^2 is inserted without derivation (the "shading effect" is described in prose only). After substituting e = 4 pi m_e r_e / (mu0 e) and the Planck definitions, the paper arrives at F_g = (l_P c^2 / m_P)(m_e^2 / r^2) (eq. 5.8.7). But l_P c^2 / m_P is identically G: with l_P = sqrt(hbar G / c^3) and m_P = sqrt(hbar c / G), l_P c^2 / m_P = G [calculation]. So "G derived from five constants" is Planck units read backwards; Planck length and mass are defined from G. Fair point for the researcher: the paper's own Planck units use h, not hbar, in the OM paper; the identity still holds with a consistent definition.

b. Experiments. Charge quantisation and the electron's identity are measured (Millikan; Thomson e/m; Tolman-Stewart for metallic conduction [memory: high]). Treating charge as a length is not contradicted by these, since a model could redefine units. A fair contrarian point: charge as an independent dimension is a convention; in Gaussian/CGS units charge has mechanical dimensions (g^1/2 cm^3/2 s^-1) [memory: high], and since the 2019 SI redefinition e is a fixed number. So "charge is not an independent dimension" is not heretical per se. What is not shown is any quantity the wave picture predicts that differs from Coulomb's law.

c. Tests. None. No measurement is predicted that distinguishes this model from the standard one.

d. If true. Gauge symmetry and the independent status of the electromagnetic coupling would be consequences of lattice geometry; QED's agreement to 1e-10 would need to be reproduced, which the papers do not attempt.

Verdict. Split. Charge-as-length as a unit relabelling: Not testable as stated. The "derived" Coulomb law, electron mass and G: Internally inconsistent as derivations (they are circular: they assume the definitions they claim to produce, plus an inserted factor in eq. 5.8.1). The physical content (waves in a lattice produce these) is Never properly tested because no quantitative test is stated.

Claim 3. Derivations of constants (c, k_e, alpha, Bohr radius, G, anomalous moment) show the model reproduces measured values

a. Mathematics. Check by item: - c, h, G, k_e from Planck units (OM particle eqs. 9-13): identities; the OM paper itself states "to calculate a constant, all you have to do is replace the units with Planck natural units" [read: p.5]. Planck units are defined from these constants, so this is circular. - Five inputs: Planck mass, length, time, Planck charge, r_e [read: Spacetime p.3-4]. Of these, r_e and Planck charge carry the electron's information; the others carry G, h, c. So the electron charge and mass are inputs, not outputs, for the "dozen constants". - Alpha: 1/(4 pi^3 + pi^2 + pi) = 137.0363 vs 137.0360: a numerological form that fits to 7 significant figures but is excluded by the measurement at ~14,000 sigma (see Corrections). The paper's geometric story (sphere + cone, slant length l = pi r) is ad hoc: l = pi d is justified by the half-wave "pi cycle" in prose; the particular combination of areas is chosen so that it gives about 137. - Anomalous magnetic moment: alpha/(2 pi) only, see Corrections.

b. Experiments. Alpha is measured to 1.6e-10 [web: NIST above]; the formula fails. Anomalous moment is measured to ~1e-13 and the QED series beyond alpha/(2 pi) (negative second-order term of about -0.3285 (alpha/pi)^2, [memory: high], which accounts for the 1.76e-6 gap; I computed alpha/(2 pi) - 0.3285 (alpha/pi)^2 = 0.00115964, matching the measured 0.00115965 to about 6e-9 absolute) is needed. The model has nothing equivalent.

c. Tests. None needed beyond the comparison already above. The reproduction is a comparison with known values, not a prediction.

d. If true. One would need a derivation of alpha from geometry that matches to better than 1e-10; none exists. Wyler-type and other closed forms for 1/alpha have been proposed before and none survived the measurement precision [memory: medium].

Verdict. Split. Planck-unit "derivations" of c, h, G, k_e: Internally inconsistent (circular; they derive nothing). Alpha closed form: Refuted by calculation (2.2e-6 relative error, ~14,000 sigma). Anomalous moment: Refuted by calculation (leading term only; 0.15% off the measured value).

Claim 4. Planck's constant is the energy of one cycle of light at any frequency; the photon is an arbitrary one-second accumulation; MUA ("cycle" as base unit) removes quantum uncertainty

a. Mathematics. The unit manipulation is internally coherent as a bookkeeping convention: writing frequency as cycles per second and h as J per cycle is fine if the cycle is carried as a label [read: Planck V3 pp.1-3]. The error is the physical step. Take a monochromatic beam of power P and frequency f. The energy per cycle is P/f, which depends on P and f. Gardi's P = h_Delta f (eq. 8) is the power of a beam carrying exactly one "photon" per second (E = hf delivered in 1 s). For any other beam, energy per cycle differs. So "each oscillation carries the same energy h regardless of frequency" is true only for that special beam. Likewise the laser example [read: MUA Real Life Example] gives P/c by construction: the amplification factor A = P/(hf) multiplied by the single-photon force hf/c is algebraically P/c, so it cannot discriminate MUA from standard theory. Numerical sanity: h = 6.626e-34 J per cycle for a visible 5e14 Hz source implies one cycle carries about 4e-16 times less than one photon energy hf = 3.3e-19 J. The "removes uncertainty" argument (smallest detectable energy step 6.626e-34 J) does not follow; the uncertainty relation comes from the non-commutation of conjugate observables, not from a detector resolution step [memory: high].

Honest credit: treating cycle vs radian and the factor 2 pi between Hz and rad/s as a source of confusion is a genuine, long-standing metrology point (the radian is treated as dimensionless in the SI) [memory: high]. That is a bookkeeping issue and not a physical one.

b. Experiments. The photoelectric effect has a threshold frequency independent of intensity (Einstein 1905; Millikan 1916), where a single absorption event delivers hf. A per-cycle energy of 6.6e-34 J, 15 orders of magnitude below an eV, cannot do that however many cycles accumulate "in a second" without the quantised-absorption picture. Photon antibunching (Kimble, Dagenais, Mandel 1977) and heralded single-photon sources (Grangier et al 1986) show single-photon detection events with E = hf [memory: high; web check not done]. Blackbody spectrum and Compton scattering likewise.

c. Tests. Not tested by Gardi; the standard experiments above were not designed to test MUA but contradict it where it makes a physical claim. It was not dismissed without testing; it simply makes no distinct prediction in her papers.

d. If true. Photon counting statistics, the photoelectric threshold and the structure of quantum optics would need a different account.

Verdict. Split. MUA as a unit convention: Not testable as stated. "h is the energy of one cycle of any light" and "the photon is an arbitrary accumulation": Tested, not supported (contradicted by the photoelectric effect and single-photon experiments, which were not run against MUA specifically). "Uncertainty removed": Internally inconsistent (a detector step does not change the commutator). The laser force example: Not testable as stated (agrees by construction).

Claim 5. Michelson-Morley does not exclude an aether; GW150914 plus the Fermi transient 0.4 s later is an aether P-wave then S-wave

a. Mathematics. The paper only draws an analogy to seismic P and S waves [read: Medium paper p.4]. In seismology P and S speeds differ by a factor of about 1.7 [memory: high]. Here a 0.4 s lead over 1.3 billion light years means a fractional speed difference of 0.4 s / (1.3e9 yr x 3.156e7 s/yr) = 9.7e-18 [calculation], which is not a "P-wave versus S-wave" ratio. And the paper's own stated expectation, that the P-wave is only "slightly" faster, is the opposite of a seismic analogy. It also says the LIGO strain "morphologically looks like a seismic P-wave"; that is a visual comparison, not a fit. LIGO's signal is a transverse strain (h in two perpendicular arm directions) and is fitted by general-relativity inspiral-merger-ringdown templates [memory: high]. The claim that LIGO detected a longitudinal wave has no equation.

b. Experiments. The Fermi GBM transient: Connaughton et al. report a weak transient about 0.4 s after GW150914, lasting about 1 s, with false alarm probability 0.0022 (2.9 sigma), "not detected by any other instrument" [web: https://arxiv.org/abs/1602.03920]. So the premise ("it was followed by a gamma-ray signal") is a marginal 2.9 sigma coincidence that its own authors state could be chance. For comparison, GW170817 (a neutron-star merger) had GRB 170817A about 1.74 s after the gravitational wave, and the speed of gravity equals that of light to between -3e-15 and +7e-16 [web: https://arxiv.org/abs/1710.05834]. That event shows a gamma burst following a gravitational wave only for the source expected to produce one. The paper's own revisit criterion (all future events followed by gamma rays) was not met: later binary-black-hole events did not have confirmed gamma-ray counterparts [memory: medium; web check not completed because search budget ran out]. Stated plainly: I cannot give a source for the statement about later events beyond this memory; it should not be repeated without one.

c. Tests. The MM null has been repeated by many independent labs [memory: high]. The "gamma-ray follow-up of every LIGO event" test was effectively run by Fermi GBM and others over later observing runs, but I do not have a source for the results, so I classify this half on the strength of the GW170817 and Connaughton sources only.

d. If true. Gravitational waves would not be spin-2 tensor ripples of spacetime; GR's wave predictions (polarisation, speed) would be wrong, and the binary-pulsar orbital decay [memory: high] would need another cause.

Verdict. Split. "MM does not exclude every aether": true only for Lorentz-type ether theories, so as a claim for Gardi's medium, Not testable as stated; the "circular" argument itself is refuted by the logic above. LIGO P-wave / Fermi S-wave: Tested, not supported (the 2.9 sigma coincidence was marginal and singly observed; the one confirmed gamma-ray association with a GW event, GW170817, is a neutron-star merger with a 1.74 s delay and a speed bound of 1e-15). A fair caveat: nobody ran a neutral, targeted P-wave search for longitudinal gravitational polarisation in the aether sense, so the "longitudinal light" idea itself is Never properly tested in the narrow sense, but it has no quantitative statement to test.

Claim 6. Particles are standing-wave clusters of Planck-scale wave centres (neutrino fundamental); the Mandelbrot set models black holes (pear-shaped photon ring)

a. Mathematics. Particles: Step 1 and my reading of the abstract-level text report cluster counts K = 8, 20, 28, 50 for muon neutrino, tau neutrino, muon electron, tau electron, noted to coincide with nuclear magic numbers (2, 8, 20, 28, 50, 82, 126). The coincidence is with a handful of small integers chosen from a list that already contains many small integers, and nuclear magic numbers describe nucleon shells, not leptons; I did not see any calculation that outputs a mass, lifetime or branching ratio from the geometry. The equations are images that I did not read, so I cannot assess the algebra, only that the text claims no such outputs. Mandelbrot: the paper says "circumstantial and highly speculative" [read: PaperColour, abstract/introduction per Step 1]; z -> z^2 + c has no Einstein-equation content, so the mapping to Schwarzschild features is a visual analogy. The Hubble "pear" and the hand-aligned overlay are an alignment chosen by eye (Step 1 summary).

b. Experiments. Leptons are pointlike to below about 1e-18 m in scattering experiments [memory: medium], and lepton masses and lifetimes are measured; the lattice model gives no number to compare. For the black-hole prediction: the Event Horizon Telescope's M87 image shows "an asymmetric bright emission ring with a diameter of 42 +/- 3 micro-arcseconds, which is circular and encompasses a central depression", with the brightness asymmetry explained by relativistic beaming, and "consistent with ... the shadow of a Kerr black hole" [web: https://arxiv.org/abs/1906.11238]. That is the ring Gardi's prediction (oblate or pear-shaped) concerns, and the measured ring is circular.

c. Tests. Particle model: never tested because it makes no quantitative prediction (checked as far as my sampling goes). Pear-shaped ring: tested by EHT, not supported; the EHT was not designed to test the Mandelbrot idea but did measure the quantity it predicts. The paper's offset-of-M87 argument concerns the host galaxy emission region, a different object from the shadow.

d. If true. The Standard Model's point-particle leptons and quark confinement would be replaced by Planck-scale lattice clusters; neutrino oscillation and the Higgs mechanism would have to be reproduced.

Verdict. Split. Particles-as-clusters: Not testable as stated (no quantitative output seen; limited by my sampled reading, and the "magic numbers" are a Never properly tested coincidence rather than a result). Mandelbrot-black-hole analogy: Not testable as stated; the specific pear-shaped-ring prediction: Tested, not supported (EHT ring is circular).

Claim 7. Charge is obsolete; electric current is a diffusion wave of "electrigen"; E and H fields are exact duals (Steinmetz); free space is an LC circuit with impedance Z0/(2 pi)

a. Mathematics. The Steinmetz equations (flux = L i, Psi = C e, energies L i^2/2 and C e^2/2) are standard circuit theory [read: Steinmetz paper; memory: high]. Duality between L and C elements is genuine. "Exact duality" of E and B fields, however, fails in Maxwell's equations as measured because there is no magnetic charge (div B = 0) [memory: high]. Z0/(2 pi) = 59.958 ohm is arithmetic [calculation: 376.7303/(2 pi) = 59.9585]; the paper gets this by defining C and L per radian, so the "correct" impedance depends on the convention chosen for 2 pi. The measured ratio E/H of a plane wave in vacuum is 376.73 ohm and is unaffected. The "quantum of momentum" h/c and "quantum of flow" hbar c definitions [read: LC Circuit] are renamings.

b. Experiments. Electrigen: the carrier in metallic conduction has been identified as having the electron's charge-to-mass ratio (Tolman-Stewart 1916; Hall effect signs; Thomson) [memory: high]; shot noise in tunnel junctions shows current is carried in discrete charges of e [memory: high]. These directly conflict with "charge does not exist" and an unknown neutral element carrying current. A neutral element in Group 0 that conducts would also be seen in spectroscopy and mass spectra; none is reported [memory: medium].

c. Tests. Electrigen: no experiment designed to detect it is reported in the paper (no equations or quantitative predictions per Step 1). The Tolman-Stewart-type experiments tested the identity of the carrier and were not directed at electrigen specifically. Dielectric/magnetic analogy: textbook.

d. If true. Electron identity, band theory, the Hall effect and every vacuum-tube and semiconductor design would require a different carrier.

Verdict. Split. Steinmetz L-C analogy and the circuit equations: Established (as circuit theory; the claim of exact field duality is overstated because of magnetic monopoles). "Charge obsolete / electrigen as the current carrier": Tested, not supported (carrier identity measured; no direct test of electrigen exists). Z0/(2 pi) as "the correct impedance": Not testable as stated (a convention about where 2 pi goes).

Strongest honest case for the researcher

  1. Taking MM as proof that no medium of any kind exists overstates what it shows: Lorentz-ether theory is empirically equivalent, and a conventionalist reading of the one-way speed of light is respectable. That is philosophy of science, not support for her e+e- lattice.
  2. The Fermi GBM transient 0.4 s after GW150914 was real enough to be published by the Fermi team, and no other instrument confirmed it. It deserved scrutiny, and it received it; she did state a falsification criterion ("revisit if not followed"), which is more than most aether authors do.
  3. Treating the radian/cycle as a unit label is a legitimate metrology question, and the Gaussian-units point shows charge-as-mechanical-dimension is not absurd.
  4. Her model makes exact-number claims (alpha, anomalous moment) that can be and were checked; that falsifiability is a point in her favour even though the numbers fail.

Where the researcher falls short (same rigour as for critics)

  • Many "derivations" are tautologies presented as proof (c from e+e-; G from Planck units; h per cycle in the laser example).
  • Numerology is reported as a match: alpha misses by 2.2e-6 while experiment is 1.6e-10 precise; the paper rounds to 0.00729734 and calls it matching CODATA.
  • No quantitative prediction in the particle or electrigen papers distinguishes them from the standard model.
  • The LIGO argument rests on a visual resemblance and a 2.9 sigma coincidence reported by others.
  • Some work in the folder (Demystifying Complex Numbers) is ordinary correct pedagogy (C as 2x2 real matrices) and should not be lumped with the physics claims.