Gift-Stephan · Source catalogue

Physics Essays 14(4), Dec 2001, DOI 10.4006/1.3025500 [read]. Gift argues that Einstein's equivalence principle is false, so general relativity (GTR) fails, and proposes a replacement scalar-graviton gauge theory. Section 2: a rocket A hovering above a mass and a rocket B accelerating in empty space are said to differ dynamically (gravitational force plus floor thrust versus thrust only), and by Larmor radiation: a charge fixed in B radiates P = (1/4 pi eps0)(2/3) q^2 g^2/c^3 (eq 3), one fixed in A does not. Peierls's earlier resolution is dismissed as 'arbitrary' [read: p.4]. Section 3 lists complaints about GTR (Schwarzschild singularity, undetected gravitational waves, black holes, dark matter wrongly called a GTR prediction, causality violations, DI Herculis apsidal motion as 'most glaring failure'). Section 4.1 (PDF p.5-6, eqs 4-10): take the Dirac Lagrangian L = psi-bar(i gamma.d - m)psi (eq 4), make the phase vary with time only, psi -> exp(-i m alpha(t)) psi (eq 5), and restore invariance with a massless scalar field A: L = ... - m psi-bar gamma^0 psi A, with A -> A + d alpha/dt (eq 6). The term is admitted not to be Lorentz-boost invariant and is allowed by the postulate 'the gravitational interaction is not covariant under Lorentz boosts'. A kinetic term is 'added' (never written, ref 20 Roman) giving Box A = -m psi-bar gamma^0 psi (eq 7), relabelled Box A = -4 pi G rho_m (eq 8), whose static limit gives Poisson's equation and Newton's law (eqs 9-10). The coupling strength is the fermion's own mass, so only rest mass (m times number density) sources gravity; the paper never says which fermion psi_n is, and never discusses binding energy, composite bodies or antiparticles. Deep-read checks: the fermion part of eq 6 is correctly time-locally invariant (shift dalpha/dt, no 1/m); but the wave-equation kinetic term needed for eq 7 is not invariant unless alpha-double-dot = 0, and the only invariant kinetic term, -(1/2)(grad A)^2, gives instantaneous Poisson with no waves [calc, SymPy]. The source psi-bar gamma^0 psi counts particles minus antiparticles, so matter and antimatter would repel [calc/memory high; not stated in the paper]. Section 5: photons are given gravitational mass E/c^2 'heuristically' (eq 11), which borrows the equivalence idea the paper rejects; perihelion advance uses Aspden's law F=(1+3v^2/c^2)GMm/r^2 (eq 12, ref 22) not derived from eq 8, leading to eq 15 with the planet's mass m where the Sun's mass M belongs (printed typo/error), 43 arcsec per century; light bending (eq 17) 4GM/(c^2R) comes from dropping GM/h^2 by setting h^2 = infinity (section 5.2) [read: page image p.7]; redshift eqs 18-24 give delta nu/nu = GM/(Rc^2). Marmet and Couture (ref 25) are cited for large systematic errors in deflection measurements. Section 5.4: no black holes, no expanding universe, no dark matter, no Einsteinian gravity waves; binary-pulsar PSR 1913+16 decay is said to be accounted for qualitatively by traveling wave (8), asserted without calculation and in tension with the no-wave result above. Radar echo delay, orbit decay and DI Herculis are promised for later. Table I: GTR 10 nonlinear equations versus 1 linear equation. Table II (PDF p.9) lists gravity coupling 'Gm^2/(hbar c) = 3e-42' (matches the proton-electron product, not either mass squared [calc]), weak bosons 81.8/81.8/92.6 GeV, and strong force mediated by spin-0 pions with the claim that the same time-local gauge modification 'is exactly that used in the quantum generation of pi-mesons' (ref 29, Gift 1999 Speculat. Sci. Tech. 21, 143), unification of gravity and strong force and an explanation of why gravity is weak, none derived here. Nordenson is quoted calling relativity a 'serious logical incoherency'.

Key claims

• A charge fixed in a uniformly accelerating rocket radiates (Larmor power with a = g) while a charge fixed in a rocket hovering above a mass does not, so the two frames are distinguishable and the equivalence principle is false.
• Because the equivalence principle is false, general relativity is invalid.
• Making the Dirac phase vary in time only (time-local U(1)) and adding a scalar field A coupled as m psi-bar gamma^0 psi A gives a massless spin-0 graviton whose static field equation is Newton's law, if Lorentz boost covariance is dropped by postulate.
• The same coupling, with an added kinetic term, gives a wave equation Box A = -4 pi G rho_m that carries energy and so accounts for PSR 1913+16 orbit decay without Einsteinian gravity waves.
• The theory reproduces the gravitational redshift delta nu/nu = GM/(Rc^2), light bending 4GM/(c^2R) and perihelion advance 6 pi GM/(a c^2 (1-e^2)) (the 'three classical tests'), the latter two via a photon gravitational mass E/c^2 and Aspden's force law taken from other sources.
• There are no black holes, no expanding universe, no dark matter and no Einsteinian gravitational waves.
• The same time-local gauge modification, applied to isospin, generates pi-mesons and unifies gravity with the strong force (asserted, derived elsewhere).

Relevance: The gravity half of the project. Its time-local gauge trick is the template reused for magnetism in 2009, and the 2006 paper cites it as the 'new gauge theory of gravity' that Lorentz-covariance relaxation justifies.

Mainstream conflict: Contradicts GR and the weak equivalence principle, tested to about 1e-13 (torsion balances) and 1e-15 (MICROSCOPE) [memory high]. Composition dependence follows directly from the written coupling (source = rest mass of constituent fermions; nuclear binding energy, about 0.94 percent for iron-56 per nucleon, would not gravitate) [calc; binding energy memory high]. Gravitational waves (LIGO 2015) and black-hole observations are accepted; dark matter is an observational inference, not a GR prediction [memory high]. A purely scalar graviton is generally held to give wrong light bending [memory medium].

June 2018 preprint on ResearchGate (journal not stated) [read]. Epigraph from Buchwald and Thorne: non-zero aether drift would make relativity crumble. Section 2 (rotation): (1) the CCIR/ITU clock-synchronisation formula gives one-way travel time dt = l/c + l v/c^2 east and l/c - l v/c^2 west for stations at the same latitude (v = surface speed); dividing l by dt gives light speed c - v east, c + v west (eqs 1-8), said to match Marmet and Kelly's 414 ns extra eastward equatorial circumnavigation time and Sato [read]. (2) A modified Michelson-Morley layout with GPS clocks at P, M1, M2: east-west arm gives out-and-back difference 2 l w/c^2 (eq 17), north-south arm zero (eq 22). (3) The GPS range equation in the ECI frame for stations moving at v gives c - v and c + v (eqs 24-33), crediting Wang. (4) The ITU time-transfer formula to a geostationary satellite gives c - v cos(theta) (eqs 34-39). Section 3 (orbital): (5) Shtyrkov's aberration measurement, and (6) planetary radar and spacecraft ranging in the barycentric frame giving c +/- v at the antenna (eqs 41-55). Conclusion: light-speed variation is 'incontrovertible', a cosmic medium is 'confirmed', with multiple preferred frames carried by Earth and Sun; Hatch's clock-bias alternative 'requires further investigation'. It also reviews Michelson-Gale 1925, Bilger 1995 ring laser, Miller (via DeMeo 2014) with Shankland's temperature explanation, and resonator limits dc/c < 1e-17. Gift's own earlier GPS papers (refs 32, 34, 49 and a 2014 paper), Su's local-ether model and Allan/Ashby GPS relativity papers are cited. Deep-read facts on Shtyrkov (section 3.1, Fig. 5): Gift writes tan(alpha) = v/c for a star at the ecliptic pole, alpha = 20.49 arcsec gives 29.80 km/s (check: 29.78). He asserts aberration is independent of source motion and distance, so a geostationary satellite can replace a star, an extension to a co-moving source made without argument. Apparatus as reported: Intelsat 704 (66 deg E, inclination 0.02 deg), Earth-fixed radio telescope TAT-01B at Kazan, March 1998 to December 1999. Fig. 5 is a low-resolution raster copied from Shtyrkov [47]: y axis 'orbital ether drift velocity, km/s' 0-50, x axis 'time, days'; about 25-30 open circles dated 25 March 98 to 23 Dec 99, nearly all between about 28 and 31 km/s; no error bars, no fit curve, no uncertainty stated anywhere; reference lines at 29.4 km/s (apparently Shtyrkov's mean) and 'astron. 29.765 km/s'. The plot is a full-magnitude derived speed, not a residual, and is nearly flat; the true annual swing is only about 29.3 to 30.3 km/s [memory high], so seasonal variation is asserted, not shown. Gift never says how the angle was obtained, or what orbit determination, pointing, refraction or ephemeris model was applied. Gift's only conclusion is 'We await confirmation by others of this very interesting experiment' (printed p.18 = PDF p.19). Sources are an NPA conference paper (2005) and a 2011 Eur. J. Natural History paper. Key observation: in every derivation the starting assumption is that light moves at c in the ECI or barycentric frame and clocks are synchronised by that rule; the 'speed relative to the Earth' is distance divided by that clock time, so no synchronisation-independent quantity is measured [read; the paper itself states the GPS specification assumes constant c in ECI].

Key claims

• GPS-synchronised clocks measure a one-way light speed of c - v eastward and c + v westward relative to the rotating Earth's surface.
• A Michelson-Morley layout using GPS clocks gives an out-and-back time difference 2 l w/c^2 on an east-west arm and zero on a north-south arm, detecting ether drift from Earth's rotation.
• The GPS range equation applied to two stations on the rotating Earth gives light speeds c - v (east) and c + v (west).
• The ITU time-transfer formula to a geostationary satellite implies light speed c - v cos(theta) relative to a ground receiver.
• Shtyrkov's Kazan measurement with Intelsat 704 recovers the Earth's orbital speed (about 29.4 to 29.8 km/s) from an Earth-fixed source and receiver, implying a medium for light (reported second-hand; Fig. 5 shows no error bars and no extraction method).
• Planetary and spacecraft ranging equations in the barycentric frame imply light speeds c + v and c - v relative to the Earth's antenna, confirming a cosmic medium and multiple preferred frames.
• Light takes about 414 ns longer to circle the equator eastward than westward (Marmet and Kelly, quoted).

Relevance: Gift's latest and most developed ether claim and the only source offering a new anomalous experiment (Shtyrkov, Fig. 5). Its central move, reading a clock-synchronisation convention as a physical light speed, is the main thing to analyse.

Mainstream conflict: Conflicts with the invariance of c in special relativity. The one-way speed of light is generally treated as convention-dependent (Einstein synchronisation; Reichenbach, Mansouri-Sexl) and GPS Sagnac corrections are standard consequences of SR in a rotating frame [memory high]. Resonator experiments cited in the paper itself limit anisotropy to about 1e-17 [read]. For Shtyrkov: stellar aberration of a co-moving source such as a geostationary satellite has no first-order v/c effect in the standard account [memory medium; not web-checked]. Gift's 'aberration is independent of source motion' is the contested step.

Journal of Scientific Exploration 20(2), 201-214, 2006 [read]. Framework: the generalised Galilean transformation (GGT, after Gagnon et al. 1988): light moves at c in a preferred frame, velocities add Galilean-ly (eq 1.5), moving rods contract and clocks slow (FitzGerald-Lorentz-Larmor, FLL, eqs 1.1-1.4), apparent speed u'_a = (u - v)/(1 - v^2/c^2) (eq 1.6, credited to Levy). For u = c this gives c - v at low speed, 'what distinguishes the ether theory from special relativity'. Second-order experiments such as Michelson-Morley are said to be unsuitable (Ives). Roemer (eqs 2.1-2.12): Io's period T_o = lambda_o/c; receding Earth sees T_A = lambda_o/(c - v), approaching T_B = lambda_o/(c + v); with FLL factors the difference is DeltaT = 2 T_o v/c (eq 2.12), 'confirmed with great precision' (Hoyle) and equal to the textbook result. Deep-read check: with T_o = 1.769 d and v = 29.8 km/s, about 30 s per orbit [calc]; algebra 2.10 to 2.11 is correct; but eqs 2.3 and 2.7 shorten the measured period while eq 1.2 says a moving system's frequency is lowered, a second-order inconsistency that does not affect 2.12. Doppler (eqs 2.13-2.23): moving observer meets wavefronts at c +/- v, shift 2 f_E v/c (eq 2.23). The mixing of Jupiter's or a star's own motion is dismissed in one parenthesis as 'easily separated' (citing Zeilik and Smith), with no method or numbers; the Sun's motion through the ether is not separated at all. Gift concludes both tests demonstrate light-speed variation, 'invalidate special relativity', and that Mansouri and Sexl's equivalence result is 'incorrect' (an assertion, not a calculation). Discussion: Gift cites Sagnac 1913, Michelson-Gale 1925 and Allan, Weiss and Ashby 1985 (Science 228, 69) as rotational detections and claims 'for the first time' ether-drift detection for uniform motion (Earth's revolution). Maxwell's Io proposal (eq 3.1, dt = 2 t_o v/c, v the Sun's speed through the ether) is called identical to the Roemer relation; in fact t_o is the light time across Earth's orbit (about 1000 s) so it is a different, much smaller quantity [deep-read check]. The CMB (Penzias and Wilson; 2.7 K) is suggested to be blackbody radiation in equilibrium with the ether, a preferred frame; Conklin 1969 and Henry 1971 are cited for the dipole; Gift quotes 390 +/- 30 km/s toward Virgo (uncited; the usual figure is about 370 km/s [memory medium, not checked]). Weisskopf (Am. Scientist 71, 473, 1983, ref 43) is quoted: 'Nature has provided an absolute frame of reference' (quote not checked against the original). Marinov (ref 44, Gen. Rel. Grav. 12, 57, 1980) 'succeeded in measuring the absolute motion of the Earth relative to the ether by using an ingenious modification of the Sagnac experiment', obtaining 303 +/- 20 km/s, 'comparable' to the CMB value; no apparatus details, direction or replication are given (the two figures differ by about 2.4 sigma [calc]). The conclusion quotes Einstein ('no aether-drift') and answers 'Einstein was wrong!', and says light-speed variation 'justifies the relaxation of the Lorentz covariance requirement that led to a new gauge theory of gravity' (the 2001 paper).

Key claims

• The Roemer effect (fractional change in Io's eclipse period equal to v/c) shows light speed c - v relative to a receding Earth and c + v relative to an approaching one, so ether drift is detected.
• The first-order Doppler shift of starlight on the ecliptic (2 f v/c) shows light speed c + v toward and c - v away from the star, so ether drift is detected.
• Second-order experiments such as Michelson-Morley cannot detect ether drift because FLL contraction and clock slowing compensate; first-order experiments can.
• Light-speed variation in first-order tests contradicts light-speed invariance and invalidates special relativity, contrary to Mansouri and Sexl.
• The cosmic microwave background is blackbody radiation in equilibrium with the ether and gives a preferred frame; the Earth moves at 390 +/- 30 km/s relative to it.
• Marinov's modified Sagnac (coupled-mirror) experiment measured the Earth's absolute velocity as 303 +/- 20 km/s, comparable to the CMB value (reported second-hand; replication status unchecked).
• This is the first claimed detection of ether drift for uniform (orbital) motion, and the effect of Jupiter's and the Sun's motion can be 'easily separated' (asserted).

Relevance: First paper in the ether line. It sets out the GGT framework assumed in the 2018 GPS paper and contains the Roemer and Doppler reinterpretation, plus the only experimental anomaly beyond Shtyrkov (Marinov), and the link to the gravity paper.

Mainstream conflict: Conflicts with invariance of c. The Roemer and Doppler first-order results are the same in SR and in GGT-type ether theories (Mansouri-Sexl equivalence) [memory high], so they cannot discriminate; v there is Earth's velocity relative to the Sun and the source, not to any ether. The CMB is a cosmological rest frame, which SR accepts. Marinov's result is not recognised by mainstream physics; reception unchecked.

Progress in Physics 1, January 2009, from p.12 (submitted 18 July 2008, accepted 25 September 2008) [read]. Gift proposes 'quantum magneto-dynamics' (QMD): magnetism as a force separate from electromagnetism. Motivation: two parallel-spin electrons repel classically but attract through QED's magnetic component; magnets with zero net charge cannot exchange photons. Postulate 1: 'not all interactions are covariant under Lorentz boost transformations', 'the only assumption used'. Equations (deep read, page images pp.3-4): (3.1) 'magnetic charge' nu = div mu_m (divergence of a fermion's moment), 'physically unreal ... for the purposes of the development'; (3.2) free Dirac Lagrangian; (3.3) global phase exp(-i nu alpha); (3.4) 'conservation of magnetic charge' (Noether, actually conservation of particle number with a constant nu_i per particle); (3.5) phase made space-dependent only; (3.6) L adds -nu psi-bar gamma^k psi A_k with a 3-vector field A_k; (3.7) A_k -> A_k + d alpha/dx; (3.8) Box A_k = nu psi-bar gamma^k psi after a kinetic term that is never written; (3.9) modified Dirac equation; (3.10) Box A_k = mu J_k; (3.11) J_m = curl M quoted from a textbook; (3.12) static limit = Poisson equation of magnetostatics; (3.13) F = 3 mu mu_1 mu_2/(4 pi r^4) for parallel dipoles side by side, quoted from Roman. The quanta are 'magnatons': spin 1, mass 0, charge 0, odd parity. Deep-read findings: nu = div(mu_m) is zero or undefined for a single moment and the total pole strength of any finite magnet is zero, yet nu is later a nonzero constant per particle with no value given; the fermion-part invariance of eq 3.6-3.7 is correct (checked); no kinetic term is written and the Klein-Gordon form is not invariant for time-dependent sources, so eq 3.8 holds only in a fixed gauge for static sources; the coupling is a minimal (charge-type) coupling to the Dirac vector current, not to the measured moment (the worked example is a neutron, whose anomalous moment -1.913 nuclear magnetons minimal coupling cannot give); eq 3.11 is juxtaposed, not derived, from 3.10; eq 3.13 is repulsive only for the side-by-side geometry (coaxial parallel dipoles attract; angular average zero); no value for nu, coupling strength or energy appears, and Section 4 defers predictions. Second deep read (PDF pp.5-9, journal pp.15-19). Earnshaw: the whole treatment is one sentence in section 4.1, 'The existence of the magnetic interaction in the repulsive core mechanism is supported by Earnshaw's theorem [13]' (ref 13 = Gottfried and Weisskopf 1984); the paper does not state the theorem's hypotheses, apply it to an atom, or mention the uncertainty principle or antisymmetry, so the quantum-stability reply is the reviewers' own [memory high], not the paper's. Pauli/solidity: Gift adds the magnetic field of anti-parallel electron pairs to Gillespie's account, whose Pauli 'forces' Gillespie calls apparent, not real. Greenspan (section 4.2, refs 15, 16, Physics Essays 1992): a classical simulation of H2 with a Coulomb electron-electron force gives 2.20e14 Hz against 1.38e14 Hz as printed (the usual H2 value is about 1.32e14 Hz [memory high, not web-checked], so the printed figure is a discrepancy of about 4.5 percent); with a 'fully attractive' electron-electron force the correct frequency results. Molecules (page image p.6): ref 15 = H1H1, H2H2, H1H2, H1H3, Li7-Li7; ref 16 = Li7-Li7, B11-B11, C12-C12, N14-N14. The paper gives no bond diameters, force law, attraction magnitude or initial conditions, so 'correct diameters' is a bare assertion, and a classical simulation has no spin, so nothing shows the attractive force is the 1/r^3 spin-dependent magnatonic one. 'Extremely strong indication' is Gift's inference, not Greenspan's. Covalent, ionic and polar bonds, HCl (Fig. 1 is a drawing, not data), chemical reactivity (liquid oxygen 'clings to a magnet'), the 'rule of two' (Gillespie and Popelier) and Hund's rule are all qualitative narrative; Moore and Luder are cited only for wave mechanics being approximate, and Luder's book is an electron-repulsion theory, the opposite thesis to Greenspan's attraction, which the paper does not say. Barut (ref 9) is cited second-hand for sharp e+e- resonances 'possibly of magnetic origin', with no data. Quantitative content: none beyond the two H2 frequencies and the form 1/r^3 ('dominant at high energies or short distances', Conclusion, page image p.8); no coupling, nu value, energy or force; the ratio of magnetic dipole-dipole energy to Coulomb repulsion at about 0.74 angstrom is never computed (order 1e-4 eV against about 10 eV [memory high, not calculated]). Mainstream explanation of the singlet H2 bond (Heitler-London exchange) is never named. Conclusion: QMD 'establishes magnetism as a force entirely separate from electromagnetism' (digest wording: 'plausible explanations'); items deferred: renormalizability, quantitative Pauli/reactivity/bond application, molecular geometry, e+e- scattering (Barut), polarised pp (Krisch, ref 20), elastic e-n scattering (Halzen and Martin, ref 5). Future-work claims are hopes plus self-citation: a space-dependent isospin rotation U(x) should give massless rho-mesons that gain mass by symmetry breaking when unified with the magnatons; a time-dependent isospin rotation U(t) unified with spin-0 gravitons 'yielded pi-mesons' (ref 22, Gift 1998/99, not in the folder, no masses compared); the 2001 time-local scalar graviton (ref 21) is said to contain the Poisson equation 'and hence 300 years of Newtonian gravitation'. The paper 'supersedes' ref 23 (Hadronic Journal Supplement 2002) and thanks Dr F. Campayne for 'extensive assistance'.

Key claims

• Magnetism is a fundamental interaction separate from electromagnetism, mediated by massless spin-1 magnatons of a new 3-vector gauge field A_k.
• Relaxing Lorentz-boost covariance (Postulate 1) allows a gauge field to arise from a space-local, time-independent phase transformation of the fermion field.
• The magnaton field equation reduces to the Poisson equation of classical magnetostatics (eq 3.12), and A_k is the transmitting agent in the Aharonov-Bohm effect.
• A magnetic attraction between anti-parallel-spin electrons, mediated by magnatons and following a 1/r^3 potential, explains Pauli exclusion, chemical reactivity, covalent bonds and the solidity of matter.
• Greenspan's classical simulations (refs 15, 16: H2, D2, HD, HT, Li2, B2, C2, N2) give the correct vibrational frequency (H2: 1.38e14 Hz as printed versus 2.20e14 Hz with Coulomb repulsion) and 'correct molecular diameters' only if the bonding electrons attract, and this is an 'extremely strong indication' of the magnetic interaction (second-hand; no diameters or force law are given; Greenspan's papers not in the folder).
• Earnshaw's theorem [13] is cited in one sentence as support for the claim that the repulsive core of matter needs a magnetic component (the paper does not state the theorem or answer the quantum-stability reply).
• Covalent bonds, ionic and polar bonds (H2, HCl, NaCl) and the 'rule of two' arise from magnetic attraction between anti-parallel-spin electrons, though no force, energy or bond length is calculated.
• An isospin version of the construction (space-dependent U(x)) is hoped to give massive rho-mesons, and a time-dependent U(t) unified with spin-0 gravitons is said to have yielded pi-mesons (ref 22, self-citation, no numbers).

Relevance: Applies the 2001 method (relax Lorentz covariance, build a compensating gauge field, name a new particle) to magnetism, and is the project's non-ether, quantum-field strand.

Mainstream conflict: Conflicts with electricity and magnetism being one interaction (Maxwell, QED), with Pauli exclusion as a consequence of fermion antisymmetry and the spin-statistics theorem, and with Lorentz covariance in relativistic QFT. The Earnshaw argument is classical and static; quantum mechanics (uncertainty principle plus antisymmetry) gives stability [memory high]. No magnaton has been reported. Greenspan's standing is [memory low] and not checked.