The Origin and Present Status of the Special Relativity Theory
| Scientific Paper | |
|---|---|
| Title | The Origin and Present Status of the Special Relativity Theory |
| Read in full | Link to paper |
| Author(s) | Herbert Dingle |
| Keywords | special relativity theory, Special Theory of Relativity, Ritz emission theory, Lorentz transformation, clock synchronization, crucial experiment |
| Published | 1960 |
| Journal | Science Progress |
| Volume | 48 |
| Number | 190 |
| No. of pages | 19 |
| Pages | 201-219 |
Read the full paper here
Abstract
The special theory of relativity is now so much an accepted part of physics that its origins and early history tend to be forgotten, and they are largely unknown to the younger theoretical physicists of today. In view of recent difficulties that have arisen in connection with the theory, and particularly in view of the fact that an alternative theory of Ritz's, which was thought to have been disproved, has now been shown to be a distinct possibility, the occasion seems opportune for a review of the circumstances in which the theory arose and developed, and for an appraisal of its present status. The purpose of this article is to supply these desiderata...
Overview
This 1960 review article in Science Progress is Herbert Dingle's attempt to separate two things he believed the physics community had allowed to merge: the internal logical consistency of the special theory of relativity, and its actual support by observation. His complaint is that "a theoretical demonstration that the theory contains no internal contradictions — that it could be right — has frequently been regarded as a proof that it is right", and that experiments cited as confirmations in fact establish only the compatibility of the theory with the Maxwell-Lorentz field equations rather than with the facts those equations are meant to represent.
Dingle's conclusion is deliberately narrow and unusually concrete. He does not claim in this paper that relativity is false — he explicitly sets aside the logical objection he had raised elsewhere, saying he does not wish to introduce controversial matters here. Instead he argues that the surviving live alternative is Walter Ritz's emission theory, that the 1913 de Sitter double-star argument long taken to have killed it does not do so, and that between Einstein and Ritz "at present there is not a scrap of experimental evidence even to make one appear more probable than the other". He proposes a crucial experiment. This is a considerably more restrained claim than the one Dingle became known for later, and it makes the paper a useful record of how his critique developed.
The historical reconstruction
Three theories in 1909
Dingle sets the scene with the failure of optical experiments to reveal the Earth's motion through a fixed aether. Either the electromagnetic field equations were wrong, or they had been wrongly applied. He classifies the responses:
- Lorentz — a stationary aether; Maxwell-Lorentz equations valid; relativity of motion not fundamental but apparent, because motion through the aether physically alters bodies; the Galilean transformation fundamentally true, the Lorentz transformation apparently operative.
- Einstein — no aether; Maxwell-Lorentz equations valid; relativity of motion fundamental; the Lorentz transformation universal and fundamentally true.
- Ritz — no aether; Maxwell-Lorentz equations false; relativity of motion fundamental; the Galilean transformation universal and fundamentally true, with light leaving its source at fixed speed relative to the source "just as though it were a material particle fired by an unvarying mechanism".
Ritz's account, Dingle notes, was in all respects the simplest, "but it had the one glaring disadvantage that it left electromagnetism without a general theory" — and Ritz died in 1909 before he could supply the replacement equations he knew were needed.
Why de Sitter did not refute Ritz
The standard disproof of emission theory is de Sitter's: if light from a component of a binary star reached the Earth at c ± v, the star's apparent motion would be a confusion of images rather than the simple Keplerian ellipse observed. Dingle's objection is precise. What de Sitter's argument excludes is light that leaves the star at c with respect to the star and thereafter maintains c ± v with respect to the Earth. It says nothing about light that leaves the star at c with respect to the star and thereafter maintains c with respect to the star. The latter, he argues, is the only version of Ritz's hypothesis acceptable as a law of nature, since "it would be fantastic to suppose that, having once been emitted, it must keep a constant velocity with respect to an arbitrary body like the Earth". He adds the historian's aside that the readiness with which so inconclusive an argument was taken as final suggests the Lorentz-Einstein view was already dominant for other reasons: "physicists were prepared to sacrifice almost anything rather than the electromagnetic equations".
Minkowski's contribution as psychology, not physics
Dingle treats the 1908 geometrization as "simply a piece of pure mathematics, contributing nothing at all to the physical problem". Its effect was psychological: it made the reversals of time order picturable, and it allowed physicists to keep belief in the objectivity of nature by putting "space-time" in the place vacated by space and time. That this apparatus is self-consistent is then repeatedly presented, he complains, as though it were proof that the theory is physically true.
The logical core: time is a definition
Dingle stresses "and the fact cannot be too strongly stressed" that Einstein's theory rests entirely on a freely chosen definition of distant simultaneity — the stipulation that the time light takes from A to B equals the time from B to A. He does not present this as a defect: "it is a necessary characteristic of every theory", and pre-relativity physics differed only in being unconsciously rather than consciously based on one.
His sharpening of the point is that the mention of light in the definition is "entirely gratuitous". Any messenger X will serve, provided synchronization by X is mutual — sound in still air would do equally well. What Einstein additionally had to assume, with no evidence then and, Dingle insists, none since, is that light in fact behaves like such an X: that beams emitted from relatively moving sources A and A' travel onward as a single beam. Define time in terms of an abstract X with those properties, and the Lorentz transformation follows as a matter of pure mathematics once the principle of relativity is granted; the theory then becomes "invulnerable to observation, and logically impeccable". The physical question — the only one Dingle is interested in — is whether the coordinate t so defined corresponds to readings of real clocks.
The sound analogy and the four cases
He works the whole scheme through with sound: two long rods ABC and A'B'C with clocks along them, one set in motion, the moving clocks resynchronized by sound signals. The resulting disagreement of the clocks is, he insists, "not a fact about nature… but simply a statement about their readings when we adjust them to accord with our definition of synchronization." Why then must the master clock B itself run slow? Only because of the principle of relativity: for sound the two descriptions of the same relative motion are genuinely distinguishable, since the air is at rest in one frame and not the other, whereas in Einstein's theory there is no aether to play that role, so equivalence can be preserved only by slowing the moving clock.
Two experiments therefore decide everything: (i) is the velocity of light independent of the motion of its source, and (ii) does a moving clock run slow. Dingle tabulates the four outcomes. Case 1a (single beam, clock slows) fully justifies Lorentz-Einstein. Case 1b (single beam, no slowing) violates the principle of relativity and makes motion with respect to an aether measurable; he judges it "scarcely conceivable". Case 2a (double beam, clock slows) he calls impossible, since the synchronization procedure would then force moving and stationary clocks to agree. Case 2b (double beam, no slowing) completely justifies Ritz.
Why the standard confirmations do not discriminate
The heart of the paper's polemic is that the observations cited for relativity all depend for their interpretation on the electromagnetic equations, which are already Lorentz-invariant by construction — so finding that their coordinates obey the Lorentz transformation "is simply confirming experimentally what the mathematics already requires."
- Mass increase with velocity. Deflection experiments show the momentum of a charged particle is mv/√(1 − v2/c2) — but v there is dx/dt with x, t the coordinates of the field equations, not rod-and-clock quantities. One may equally read the factor as attaching to the mass or to the velocity. "These experiments afford no evidence whatever as to which choice is to be made."
- Atomic clocks. An atom "is a clock only by virtue of its interpretation in terms of electromagnetic theory. We never observe an atom as a periodic system." The Ives-Stilwell experiment therefore "prove[s] nothing except the invariance of the electromagnetic equations to the Lorentz transformation".
- The Doppler effect. Returning to sound, an observer who inferred the frequency of moving sources from the sound he received would conclude the receding clocks had slowed — a fallacy transparent in the acoustic case, because what changed was the frequency of reception, not of the source.
Dingle also offers a constructive suggestion: if velocities w measured by rods and light-synchronized clocks relate to the coordinate velocities v by w = v/√(1 − v2/c2), then Minkowski's invariant ds is simply ordinary clock time, and Minkowski, "thinking he had discovered 'space-time', had actually disinterred pure time from its burial in this hybrid."
The proposed crucial experiment
He recommends the easier of the two tests: compare the laboratory velocity of light from relatively moving sources, taking care to avoid de Sitter's error by ensuring "the sources did not change their motion during the passage of the light". He urges that it be undertaken without delay.
Assessment
The paper's genuine strength is its insistence on the distinction between a convention and an empirical result. Dingle is right, and clear, that Einstein's synchronization rule is a stipulation rather than a discovery, that nothing in the rule requires the messenger to be light, and that a theory built on a definition cannot be confirmed by exhibiting its internal consistency. His warning against reading Minkowski's geometry as physical evidence is well aimed, and his historical point about de Sitter is technically correct: the 1913 argument as originally stated does not exclude every emission hypothesis, which is why the question was eventually closed by other means. As history of science, the reconstruction of the Lorentz/Einstein/Ritz three-way choice as it stood in 1909 is careful and fair — Dingle credits Einstein with "quite exceptional insight" and calls the time analysis "a permanent and fundamentally important contribution to physical theory".
The difficulties are equally clear. The claim that there is "not a scrap of experimental evidence" between Einstein and Ritz is asserted by a blanket argument — that every relevant measurement is interpreted through Lorentz-invariant field equations — rather than shown case by case, and the argument proves too much: on the same reasoning almost no measurement could ever bear on any theory whose apparatus is described by electromagnetism. The treatment of Ives-Stilwell is the weakest step. That experiment measures the transverse second-order shift, the term that the first-order Doppler effect cannot produce, and the acoustic analogy Dingle uses to dismiss it does not reach it. His stipulation that Ritz's light must maintain velocity c with respect to its source after emission is presented as the only reasonable form of the theory but is not derived from anything; it is chosen because it evades de Sitter, and Dingle does not ask what it means for light already in flight from a source that later accelerates.
Most decisively, the crucial experiment Dingle asks for has since been performed in the form he specified. Extinction-corrected measurements of the speed of gamma rays from fast-decaying neutral pions, and observations of X-ray and gamma-ray binaries where the emitting sources move at large speeds and the arrival times remain sharp, bound any source-velocity dependence of c to a very small fraction of the source speed. Independent muon-lifetime measurements and stored-particle experiments supply the moving-clock test of Dingle's second experiment directly. In his own tabulation this is Case 1a. The paper is therefore best read not as a standing challenge but as an unusually precise statement of what would have had to be true for emission theory to survive — one whose value lies in the sharpness of the question it posed rather than in the answer it expected.