Builder-G: Physicist review (PhD physicist, mainstream view)

Source gap (repeated from Step 1): The only document is Ether-And-Relativity-1957.pdf (G. Builder, Australian J. Physics 11, 279-297, 1958). It has no text layer. I re-read pp.2-9 and 10-18 from rendered page images [read: page images]. I did not read Dingle's 1957 Nature letter or Builder's earlier papers (Aust. J. Phys. 10:246, 10:424); statements about them rest on Builder's own paraphrase. Experimental statements below are tagged [web] and checked. Anything tagged [memory] is flagged as such.

Corrections to Step 1

1. Mach's hypothesis (claims.md item 4, facts key claim 5). The facts file says Builder rejects "Mach's hypothesis (instantaneous interaction)". Builder (p.291 footnote) explicitly separates Mach's hypothesis (instantaneous action at a distance, which he rejects) from Einstein's "Mach's principle" (distant masses determine the properties of space in the vicinity of bodies, i.e. a field-mediated version). He rejects only the first. His "no tenable alternative to the ether" conclusion therefore skips over the field-mediated version, which is the one GR-based work (frame dragging) actually uses. Not a Step 1 error as such, but claims.md compresses it.
2. Equations 6b, 7, 7a, 7b, 8 were not checked in Step 1; I checked them (see Claim 3). They are correct to O(v^2/c^2) for the simultaneous-position, non-radiating case.
3. Dingle citation. Builder cites Dingle 1957, Nature 179:866. A web search shows Dingle's piece is "The 'Clock Paradox' of Relativity", Nature 179, pp.865-866, April 1957 [web] https://www.nature.com/articles/1791071a0 . Step 1's description of the Dingle position (clock retardation is absolute, so relativity fails) is still only Builder's paraphrase here (p.1-2 of the paper), not verified from Dingle.
4. Scope of the Lorentz-ether "equivalence". facts.json says Lorentz ether theory (LET) is empirically equivalent to SR "so no conflict on predictions", tagged [memory: medium]. This is correct for LET in its full Lorentz-covariant form, and Builder explicitly builds it that way (p.294-295). Upgraded: the paper itself says the ether is not measurable by dynamical or electrodynamical experiments [read: p.286-287, 290]. Step 1 correct on this, just under-tagged.
5. Minor: the paper's pages carry journal pagination 280-296 (PDF pp.2-18), so the "pp." references in claims.md are PDF pages, not journal pages. Not an error, but worth stating.

Background: what a reader needs (brief)

Special relativity (SR) says physics takes the same form in every inertial frame (frames in uniform motion). A clock moving at speed v in frame S ticks slower by the factor sqrt(1 - v^2/c^2). Builder's eq (1)-(3) is exactly this: each clock's rate dt_a/dt = sqrt(1 - u^2/c^2), and the total difference after integrating between two meetings. Lorentz ether theory (LET, 1895-1904, Lorentz, Larmor, Poincare) says the same equations hold but with one frame (the ether frame) privileged and invisible. The two theories predict identical measurements. Builder argues the ether version is the more honest reading.

---

Claim 1: Clock retardation depends on each clock's individual speed, so it demands a causal role for velocity relative to a unique absolute inertial system (pp.2-3, 7-9)

a. Is the mathematics internally consistent? Eq (1)-(3) are correct SR: proper time of a worldline is tau = integral of sqrt(1 - v^2/c^2) dt [read: p.2 image]. Worked check: clock A at rest (u=0), clock B moving out and back at v = 0.6c for 10 years of S-time: tau_B = 10 x sqrt(1-0.36) = 8 years, so t_a - t_b = 2 years. Builder's statement that this is invariant for all observers is correct (proper time is the invariant length of a worldline).

The logic is where it fails. Builder says (p.3): the form of (1)-(3) is the same in every inertial frame, so the physical explanation "must be sought in the form of the equations rather than their numerical content"; the clocks are isolated from everything local; therefore they "must interact with something universal". The first premise (frame-independence) argues against an explanation involving any particular frame. The conclusion needs a further hidden step: that the retardation needs a dynamical cause rather than a geometric one. Compare an odometer: two paths between the same two points on a plane have different lengths. That requires no absolute direction or "universal something" acting on the odometers; path length is a property of the path. In SR the retardation is the Minkowski analogue (with a sign flip), a property of the worldline. The "causal role" demand is an added philosophical premise, not a consequence of the equations. Also the "causal" reading would give a measurable absolute speed if it were real, which Builder admits it does not (pp.4, 11). Verdict on (a): equations right; the inference "therefore an absolute system" is a non sequitur (needs the hidden premise that a worldline-length difference must have a dynamical cause).

b. Consistent with established experiments? Yes in the sense that the retardation is real and measured: Hafele-Keating (1971-72) flying cesium clocks: lost 59 +/- 10 ns eastward and gained 273 +/- 7 ns westward vs predicted 40 +/- 23 and 275 +/- 21 ns [web] https://www.science.org/doi/10.1126/science.177.4044.168 ; CERN muon storage ring: muons with gamma = 29.33 lived 64.419 us vs 2.197 us at rest, matching SR to 2x10^-3 [web] https://www.nature.com/articles/268301a0 . No experiment can distinguish "caused by absolute velocity" from "worldline geometry". Note also that the Hafele-Keating result includes gravitational terms from GR, which Builder (writing 1957) does not discuss.

c. Tested? The retardation: yes, many times, replicated (clocks, muons, Ives-Stilwell [memory: high], GPS [memory: high]). The "absolute velocity as cause" interpretation: not testable (Builder says so). Modern bounds on any preferred-frame effect on light speed reach ~10^-18 relative frequency change [web] https://www.nature.com/articles/ncomms9174 (Nagel et al. 2015: 9.2 +/- 10.7 x 10^-19, 95% CI).

d. What would have to change if true? Nothing in predictions. Interpretively, the geometric view of spacetime would be replaced by a privileged frame with an unobservable causal role; mainstream physics would have to accept a conspiracy of nature that always hides the preferred frame.

Verdict: split. The retardation itself: Established. "Therefore a causal role for absolute velocity": Not testable as stated (an interpretation with no measurable difference; Builder concedes this). Also a logical gap, so the argument as an argument for necessity is unsound; the claim of "compelled" is not supported.

---

Claim 2: An absolute system (universe as a whole, an ether) is compatible with and sufficient for SR, because absolute velocities cannot be measured by dynamical or electrodynamical experiments, so the relativity principle is not violated (pp.4, 11, 13-14)

a. Math. Consistent. If frame S0 is one of the inertial frames, SR's formulas hold in it (eq 4-5), and the observable consequences can be computed in any other frame (p.282). This is the Lorentz-Poincare reading and is mathematically equivalent to SR [read: pp.282, 293-294]. Builder is careful that Einstein's 1905 derivation is not affected if the postulate "stationary frame" is also the ether frame (p.293); that is correct.

b. Experiments. Consistent, because it predicts the same as SR. All null results of ether-drift searches (Michelson-Morley, Kennedy-Thorndike, Ives-Stilwell, modern cavity tests ~10^-18 [web] https://www.nature.com/articles/ncomms9174) are accommodated if the ether theory includes the compensating contractions and time dilation by construction. That is the "conspiracy" criticism: a theory whose unobservable ingredient is guaranteed unobservable.

c. Tested? No test can distinguish it from SR. Builder says this himself (p.295: "does not prove that it is the only possible causal explanation"). Test of the weaker statement (no preferred-frame effects) has been done to very high precision: nothing found.

d. What would change? Nothing predictive. Philosophically: Occam's razor; mainstream drops the ether as superfluous (Einstein's 1905 statement quoted by Builder p.292 is a methodological, not an experimental, point, as Builder rightly notes).

Verdict: Plausible, untested in the sense "consistent with known physics, no measurable difference." Strictly it is closer to Not testable as stated (interpretation with no measurable difference). I give: compatibility Established (as a mathematical fact); physical existence of an undetectable ether Not testable as stated.

---

Claim 3: Relativistic electrodynamics depends on individual (absolute) velocities; same relative velocity gives different forces; Newton's third law fails for moving charges (eq 6a, 6b, 7, 8) (pp.5-8)

a. Math (I checked the equations). Expand the field of a uniformly moving point charge at the present position r: E = q(1-beta^2) r-hat / [r^2 (1 - beta^2 sin^2 theta)^(3/2)] [memory: high; Griffiths Ch.12]. To O(v^2/c^2) the factor is 1 + (1/2) v^2/c^2 - (3/2)(v.r-hat)^2/c^2. That reproduces eq (7a) exactly [read: p.7 image]. Eq (7b): charge q1 moving in the static Coulomb field of a charge q2 at rest feels only q1q2 r-hat/r^2 (no B field), as printed. Eq (6a)/(6b): a moving monopole produces E = -(u x B)/c with B scaled by the same factor; setting u = -v gives the factor in (6b). Eq (8) follows from the same field plus the magnetic force v2 x (v1 x E)/c^2 with the sign for the opposite direction of r. All five are correct to the stated order, ignoring radiation/acceleration terms (Builder notes this, p.283). Eq (7) and (8) are asymmetric in 1 and 2, so F1 is not equal and opposite to F2: true (e.g. two charges moving at right angles, the standard Biot-Savart non-reciprocity; the missing momentum is in the field [memory: high; Griffiths Sec. 8.2]).

Where it goes wrong: the comparison cases (a) and (b) are different physical configurations as described in S, because in S one charge is at rest and the other moves vs vice versa. They are related by exchanging which particle moves in S, not by a Lorentz boost. Boost case (a) to the rest frame of the charge and you get something that looks like case (b) but with transformed forces and different simultaneity, so the two descriptions agree. So "same relative velocity, different force in S" is correct arithmetic; but "S" is the lab, and the lab frame is physically meaningful in a lab (it holds the equipment). Nothing here singles out S0; the dependence is on velocity in the observer's frame, as it is already in Newtonian mechanics (kinetic energy depends on velocity in the chosen frame). The statement "must necessarily be displayed in S0" (p.283) is a restatement of Lorentz covariance, not evidence of an absolute frame. Also fails: eq (7)-(8) are valid only in the limit of constant velocities, so Newton's third law fails for the two-particle system alone; momentum conservation is restored when field momentum is included, which Builder does not mention (he attributes it to potential retardation, p.284, which is equivalent).

b. Experiments. The formulas are standard classical electrodynamics and consistent with all of it (induction, Biot-Savart, magnetic forces between currents).

c. Tested? Yes, as ordinary electrodynamics (replicated every day). The interpretation that this shows absolute velocity is not tested and not testable.

d. What would change? Nothing; the mathematics is already mainstream. Only the inference changes: mainstream reads it as frame-dependent description; Builder reads it as evidence of the absolute system.

Verdict: split. The equations and the non-reciprocity of forces: Established. The inference that this "compels recognition of the causal significance of absolute velocities": Not testable as stated (and logically does not follow, since S is just the observer's frame; the same dependence exists in Galilean physics).

---

Claim 4: Absolute acceleration implies an absolute inertial system and hence absolute velocity; Mach's instantaneous-interaction hypothesis must be rejected, leaving the ether; absolute velocity may in principle be kinematically measurable, e.g. by Doppler asymmetry of starlight (pp.9-12)

a. Math / logic. The key step (p.287-288): "the acceleration of a body ... must, by definition, result in a change in its absolute velocity. Thus the existence of something corresponding to absolute acceleration entails the existence of something corresponding to absolute velocity." This is false as a matter of structure. Newtonian mechanics has absolute (frame-independent) acceleration but no absolute velocity: the Galilean group preserves acceleration while shifting velocity; the thing that is absolute is the inertial structure (which worldlines are straight), not a velocity. SR has the same: the Lorentz group shifts velocities, acceleration (proper acceleration) is invariant, and no frame is singled out. So the claim "absolute acceleration implies absolute velocity" is not valid (a standard point in the philosophy of spacetime [memory: high]). The "hence unique system" also assumes a causal mechanism; in GR, the local inertial frames are fixed by the metric field, which is dynamical, but defines no preferred velocity. The rejection of Mach: he rejects instantaneous action at a distance (reasonable), but treats the alternatives as binary. A field-mediated Mach principle (Einstein's own, which he notes in the footnote on p.291) is the version that GR incorporates in part (Lense-Thirring frame dragging). Frame dragging was measured: Gravity Probe B (2011) [memory: medium, not re-checked; see below]. This shows the "ether" role in GR is the metric field, which has no rest frame. Builder's "no tenable alternative" omits this. The Doppler claim: "if the absolute velocity of the solar system were an appreciable fraction of c, the Doppler asymmetry of starlight would show it" (p.290) is fine as a kinematic statement.

b. Experiments. The kinematic measurement Builder said "seems impractical" has in fact been done in a modified sense: the cosmic microwave background (CMB, discovered 1965, i.e. after this paper) has a dipole giving the Sun's velocity relative to the CMB rest frame of 369.82 +/- 0.11 km/s [web] https://arxiv.org/pdf/1807.06205 (Planck 2018) [web-search result, figure confirmed]. So Builder's "absolute velocity relative to the universe as a whole is a kinematic quantity" is vindicated in the weak form: there is a frame in which the cosmic radiation is isotropic and we can measure our velocity relative to it. This is not a violation of SR, because the CMB frame is a property of the cosmic state, not of the laws: no local experiment depends on it (cavity tests ~10^-18 [web] https://www.nature.com/articles/ncomms9174 ). Builder's strong claim that this frame "interacts with bodies and causes dynamical effects" is not supported: no such dependence is found.

c. Tested? Weak form (a cosmic matter/radiation rest frame exists and our velocity relative to it is measurable): measured many times by many groups (COBE, WMAP, Planck [memory: high for COBE/WMAP; Planck confirmed web]). Strong form (a dynamical ether acting on clocks and rods): tested via ether-drift and Lorentz-invariance tests, null at 10^-18 level [web]. Mach's principle (strong form): not demonstrated; frame dragging in GR is a related effect [memory: medium].

d. What would change if true? If absolute acceleration implied absolute velocity: the Galilean and Lorentz symmetry arguments would fail; GR's equivalence principle (no local detection of velocity) would need qualification. Mach's principle would be dispensed with in favour of an ether with matter-independent properties; Einstein's GR (metric as dynamical field) would have to be an ether theory with a preferred frame.

Verdict: split. (i) "Absolute acceleration entails absolute velocity": Internally inconsistent as a deduction (does not follow; counterexamples in Newtonian and Lorentz-invariant physics). (ii) "Mach's instantaneous action must be rejected": Established (reasonable; instantaneous action contradicts SR) but the leap to ether is Not testable as stated. (iii) "Absolute velocity measurable kinematically, e.g. relative to distant matter": Established (weak form), via CMB dipole, though neither Builder nor his era knew it; the claim that it has dynamical consequences: Tested, not supported by Lorentz-invariance tests [web].

---

Claim 5: A clock moving through the ether at v0 slows by sqrt(1 - v0^2/c^2) as a consequence of Lorentz-FitzGerald contraction, so one hypothesis (not two) accounts for Michelson-Morley, and the ether presentation is pedagogically superior (pp.14-16)

a. Math. The footnote (p.292 image, journal p.294): a light-clock built of a rod with mirrors at both ends, moving along its length at v0. Check: round trip time in the ether frame = L/(c-v) + L/(c+v) = 2L c/(c^2 - v^2). With contracted L = L0 sqrt(1 - v^2/c^2) this becomes 2 L0/c x 1/sqrt(1 - v^2/c^2), so the clock rate is lower by sqrt(1 - v^2/c^2). Correct for a longitudinal light-clock. Numerical check: v = 0.6c, L0 = 1 (c=1): round trip 2 x 0.8 / (1 - 0.36) = 2.5 = 2 x 1.25. Rest value 2. Ratio 0.8 = sqrt(1 - 0.36). Good. Gaps: (i) This ties the clock slowing to length contraction only for that one geometry. A transverse light clock needs no contraction and gives the same slowdown directly, so the two effects are not "one hypothesis" in general: Michelson-Morley-type null results fix only the longitudinal-to-transverse ratio (contraction); the slowing of clocks is separately measured by Ives-Stilwell (Doppler of moving atoms) and the Kennedy-Thorndike experiment (different arm lengths, test of velocity dependence of the contraction) [memory: high for both; not re-checked]. The Lorentz-Robertson-Mansouri-Sexl framework requires three independent parameters to be pinned (contraction, time dilation, and synchronisation convention), and Builder's single-hypothesis claim holds only if full Lorentz covariance (electromagnetic bonds, Lorentz 1904) is assumed [memory: medium; Mansouri-Sexl 1977]. Builder's reasoning ("the ether theory requires two independent hypotheses" is wrong, Broad 1923) is reasonable for a covariant ether theory and gets the footnote right; but it is a point about LET's internal economy, not a test. (ii) "Lorentz showed that if bonds are electromagnetic then all bodies contract" is only approximately true: Lorentz's argument covers electrostatically-bound systems; nuclear forces are not electromagnetic, which is why the contraction is built in through Lorentz covariance of all forces. [memory: high]. Hence claim 5a (clock slowdown follows from contraction in the light-clock geometry) correct; 5b (one hypothesis) correct only given Lorentz covariance of all interactions.

b. Experiments. Michelson-Morley null [memory: high] is consistent; Ives-Stilwell (time dilation of moving atoms) [memory: high]; muons [web] https://www.nature.com/articles/268301a0 ; clocks [web] https://www.science.org/doi/10.1126/science.177.4044.168 . All fit equally in SR and covariant LET.

c. Tested? Contraction and clock slowing are tested (the experiments above, many replicated). Ether interpretation: not distinguishable. The "pedagogical superiority" claim is an opinion; no controlled teaching study is cited and I know of none [memory: low]. The mainstream teaching of SR is geometric because the geometric version carries over to GR; a non-measurable ether adds a quantity that cannot be measured.

d. What would change? Nothing for predictions. For pedagogy and philosophy: a preferred frame would be reinstated, which GR (and quantum field theory with Lorentz-invariant vacuum) renders redundant.

Verdict: split. Clock slowing and length contraction as Lorentz-covariant effects: Established. "Caused by motion through an ether, so one hypothesis": Historically superseded but instructive (Lorentz-Larmor was a reasonable position, empirically equivalent to SR, and replaced by the postulate-based/geometric version because the ether adds no observable content). "Pedagogically superior": Not testable as stated (opinion; no data).

---

Overall assessment (physicist)

• The mathematics Builder uses is correct standard SR and classical electrodynamics (I verified eq 1-3, 6a/6b, 7, 7a/7b, 8 and the light-clock footnote). He does not make an arithmetic error.
• The paper makes no experimental claim, no data, no numerical example (as Step 1 says), and offers no distinguishing prediction. It is an interpretation argument, partly sound (LET is equivalent to SR; Einstein's 1905 dismissal of the ether was methodological), partly unsound (absolute acceleration does not entail absolute velocity; the "demand for causal explanation" of proper time is a hidden assumption).
• One piece of the paper that aged well: the notion that a preferred, kinematically measurable frame defined by the universe's matter exists. The CMB dipole (369.82 km/s [web] https://arxiv.org/pdf/1807.06205) is just that, with no dynamical effect on local physics. This is not a victory for the strong "ether causes clock slowing" claim.
• Caveat on my own verdict for Gravity Probe B: I mentioned frame dragging as [memory: medium] and did not re-check it on the web; it is used only as an illustration, not a verdict basis.
• Source gap reminder: Dingle's argument and Builder's earlier papers were not read; nothing here depends on them except the framing of the motivation.

Verdict summary