I — The Foundations
Chapter 1: The Lie in the Textbook
Every introductory physics textbook in the English-speaking world contains a false claim about the most famous experiment in the history of the discipline. The claim is not ambiguous. It is not a pedagogical simplification that sacrifices detail while preserving the essential truth. It is a demonstrable logical fallacy, repeated across publishers, across editions, across decades, in textbook after textbook used to train hundreds of millions of physics students since the middle of the twentieth century.
The claim is this: that the Michelson-Morley experiment of 1887 disproved the existence of the luminiferous ether, and that Albert Einstein, recognising the implication, built a new physics without it. The claim arrives without fanfare, sandwiched between problem sets on kinematics and Doppler shifts. The language varies from publisher to publisher. The substance does not. The student encounters it in the first course, encounters it again in identical form in the intermediate course, the upper-division course, the graduate electrodynamics seminar. It is the origin myth of modern physics: the moment when science shed a medieval illusion and embraced the austere truth that space is empty.
The story is false. The experiment that supposedly disproved the ether cannot distinguish between a universe with an ether and a universe without one. Both theories -- Einstein's special relativity and Lorentz's ether theory -- predict the same null result. This is not a controversial claim among specialists. It is a mathematical fact, provable from the transformation equations that both theories share. Yet the false version persists in every undergraduate physics course in the English-speaking world, edition after edition, generation after generation.
The Basement in Cleveland
On a summer evening in 1887, in the basement of a dormitory at the Case School of Applied Science in Cleveland, Ohio, two physicists completed the most famous null result in the history of science.
Albert Abraham Michelson was thirty-four years old, a meticulous experimentalist of extraordinary skill, the first American to win the Nobel Prize in Physics. Edward Williams Morley was forty-nine, a chemist and precision measurer whose gift for careful laboratory work complemented Michelson's instinct for elegant experimental design. They had spent months constructing, calibrating, and operating an apparatus of unprecedented sensitivity -- an interferometer mounted on a massive sandstone slab, floated in a trough of mercury to permit smooth rotation, isolated from vibration by the building's basement foundation.
The apparatus was designed to answer a specific question. Not whether the ether exists -- that was taken as given by every working physicist in 1887 -- but how fast the Earth moves through it.
The logic was simple. Light propagates through the ether at speed c. The Earth orbits the Sun at approximately thirty kilometres per second. If the ether is stationary -- if it does not move with the Earth -- then the Earth ploughs through it like a ship through water, and the ether flows past the laboratory at thirty kilometres per second: an "ether wind." A beam of light sent into the wind and reflected back should take longer for the round trip than a beam sent perpendicular to the wind. The interferometer splits a beam of light into two perpendicular paths, sends each along an arm of equal length, reflects them back, and recombines them. Any difference in round-trip travel time produces a shift in the interference fringes -- a measurable displacement of the light-and-dark pattern where the beams recombine.
The expected fringe shift, calculated from the known orbital velocity of the Earth and the dimensions of the apparatus, was approximately 0.4 fringes. The sensitivity of the instrument was sufficient to detect a shift of 0.01 fringes. The margin was comfortable. The measurement should have been straightforward.
The interferometer was rotated slowly through 360 degrees, measurements taken at each orientation, the apparatus floated on its mercury bath to eliminate mechanical strain. The data were collected over multiple sessions, day and night, across several months.
The observed fringe shift was no more than approximately 0.01 fringes -- at or below the noise threshold of the instrument. Against a prediction of 0.4, the result was effectively zero.
The experimental record is straightforward. The question of what the authors concluded -- and what the textbooks subsequently attributed to them -- is more consequential than the measurement itself.
What the authors actually concluded
The paper was published in November 1887 in the American Journal of Science: Albert A. Michelson and Edward W. Morley, "On the Relative Motion of the Earth and the Luminiferous Ether," Series 3, Volume 34, Number 203, pages 333 to 345.
The final paragraph of that paper reads:
"It appears, from all that precedes, reasonably certain that if there be any relative motion between the earth and the luminiferous ether, it must be small; quite small enough entirely to refute Fresnel's explanation of aberration."
The phrasing warrants close attention. The authors concluded three things:
First, that if relative motion between the Earth and the ether exists, it must be very small -- far smaller than the thirty kilometres per second predicted by the stationary-ether model.
Second, that this result refutes Fresnel's specific explanation of stellar aberration, which required a completely stationary ether through which the Earth moves at full orbital velocity.
Third -- and this is the decisive point -- they did not conclude that the ether does not exist. The conditional phrasing is explicit: "if there be any relative motion." The authors left open the possibility that the ether exists but that the Earth's motion relative to it is negligible -- because the ether moves with the Earth, or because some other physical mechanism conceals the expected effect.
Michelson himself, in the years following the experiment, was explicit about what it had and had not shown. In his 1903 book Light Waves and Their Uses, published by the University of Chicago Press, he discussed the result with evident puzzlement, not triumph. He referred to it as a "failure" -- meaning a failure to detect the expected motion. The word choice is revealing. A man who believed his experiment had disproved the ether would not call it a failure. A man who expected to find something, and did not find it, would.
What was actually eliminated
The experiment tested one specific model: a rigid, stationary ether -- Fresnel's ether -- through which the Earth moves at its full orbital velocity without dragging any of the medium along. The predicted fringe shift of 0.4 fringes follows from this model and only from this model.
The experiment did not test:
A partially dragged ether, of the kind proposed by George Gabriel Stokes, in which the ether near massive bodies is carried along with them -- reducing or eliminating the expected relative motion.
A fully dragged ether, in which the local medium moves with the Earth entirely, producing zero relative velocity at the laboratory.
An ether in which objects moving through the medium undergo a physical contraction in the direction of motion -- the hypothesis proposed independently by George FitzGerald in 1889 and Hendrik Lorentz in 1892 -- which would cancel the expected fringe shift by shortening the interferometer arm aligned with the motion.
An ether with any number of other dynamical properties that could account for the null result while preserving the medium's existence.
To say that the Michelson-Morley experiment "disproved the ether" is to say that the elimination of one specific ether model eliminates all possible ether models. This is a logical error of the most elementary kind -- the fallacy of denying the antecedent applied to the relationship between a general concept and one of its specific instantiations.
The authors of the 1887 paper did not commit this error. The textbooks that describe their work do.
What the Textbooks Say
We now place the textbook account beside the original source. The comparison involves eight major textbooks used across the English-speaking world -- in the United States, the United Kingdom, Canada, Australia, and wherever English-language physics is taught. These are not obscure or fringe publications. They are the standard texts assigned to millions of undergraduate and graduate students over the past several decades. Together, they have shaped the understanding of an entire profession.
For each textbook, we document the claim it makes about the Michelson-Morley experiment, the specific logical error it commits, and the gap between its account and what the original paper actually says. The comparison that follows constitutes the central evidence establishing that the distortion is not incidental but systematic.
Textbook 1: Serway and Jewett
Raymond A. Serway and John W. Jewett Jr., Physics for Scientists and Engineers, various editions, published by Cengage Learning.
This is one of the most widely adopted introductory physics textbooks in North America, used in hundreds of universities and community colleges, taught to tens of thousands of students each year. In its chapter on special relativity, the Michelson-Morley experiment is presented as definitive proof that the ether does not exist. The language is unequivocal: the experiment showed that "the ether hypothesis was wrong."
No mention is made of Lorentz Ether Theory. No mention is made of the Lorentz-FitzGerald contraction as a prediction of ether physics. No mention is made of the fact that the null result is equally predicted by an ether theory with contraction. The student who reads this textbook is given a binary: the experiment was performed, the ether was disproved, Einstein built the correct theory. There are no other options. There is no room for nuance. The narrative is constructed so that the student cannot ask the question, "But what about Lorentz?" -- because the student does not know Lorentz exists.
The original 1887 paper said: "if there be any relative motion between the earth and the luminiferous ether, it must be small." Serway and Jewett say: the ether hypothesis was wrong. The gap between these two statements is not a pedagogical simplification. It is the distance between a careful conditional conclusion and a categorical declaration -- between what the experimenters found and what the textbook needs them to have found.
The logical error: the experiment tested one ether model (stationary, non-dragged). The textbook claims it tested the concept of ether as such. The authors of the 1887 paper concluded that relative motion, if it exists, must be small. The textbook concludes that the ether does not exist. The gap between the two is not a matter of emphasis or simplification. It is a distortion of what the experiment demonstrated.
Textbook 2: Halliday, Resnick, and Walker
David Halliday, Robert Resnick, and Jearl Walker, Fundamentals of Physics, various editions, published by Wiley.
This textbook -- one of the most widely used introductory physics texts in the world, translated into dozens of languages and adopted across six continents -- presents the Michelson-Morley experiment as establishing that "the speed of light is the same in all inertial reference frames." This is not what the experiment established. The constancy of the speed of light in all inertial frames is Einstein's second postulate, published in 1905 -- eighteen years after the experiment. It is a theoretical interpretation of the null result, not a direct conclusion of the measurement.
The experiment established that no fringe shift was observed at the level of sensitivity available. To leap from "no fringe shift" to "the speed of light is frame-independent" requires substantial theoretical work: the assumption that no physical mechanism (such as contraction) accounts for the null result, the assumption that the only explanation is frame-independence, and the assumption that no ether model can accommodate the observation. Each of these assumptions is either false or undemonstrated. Halliday, Resnick, and Walker present the conclusion as though it were self-evident.
The distortion here is particularly insidious because it is not crude. The textbook does not say, in blunt terms, "the ether was disproved." It says something subtler: that the experiment established a positive result -- the frame-independence of light speed. This reframing converts a null result (no fringe shift detected) into a positive discovery (light speed is constant in all frames). The conversion requires theoretical interpretation that the experiment does not supply. In Lorentz's ether theory, the null result arises not because the speed of light is frame-independent, but because the interferometer arm contracts by exactly the right amount to compensate for the difference in light travel times. The null result is consistent with both interpretations. Halliday, Resnick, and Walker present one interpretation as though it were the observation itself.
The logical error: the textbook attributes to the experiment a conclusion that belongs to a theory published nearly two decades later. The original experiment's authors drew no such conclusion. The student who reads this textbook is taught that an experiment performed in 1887 established a theoretical postulate published in 1905 -- a chronological impossibility that the textbook presents without comment.
Textbook 3: Young and Freedman
Hugh D. Young and Roger A. Freedman, University Physics with Modern Physics, various editions, published by Pearson.
This textbook -- the standard calculus-based text at many large research universities -- states that the experiment "failed to detect any ether" and presents this failure as a direct motivation for Einstein's special relativity. The narrative is clean and sequential: experiment disproves ether, Einstein builds new theory without it.
The language is important. "Failed to detect any ether" is ambiguous in a way that serves the textbook's purpose. It could mean the experiment failed to detect the ether wind predicted by the stationary model -- which is what happened. But in context, the phrasing functions as a statement that the experiment failed to find evidence for the ether's existence -- which is not what happened, because the experiment was not designed to test the ether's existence, only the Earth's motion relative to one particular model of it.
The ambiguity is not accidental. Precise language would require the textbook to say: "The experiment failed to detect relative motion between the Earth and a stationary ether." This formulation immediately raises the question: "What about a non-stationary ether?" -- a question the textbook cannot afford to provoke, because the answer ("A non-stationary ether, or an ether with contraction, predicts the same null result") undermines the entire narrative. The vague phrasing -- "failed to detect any ether" -- forecloses the question by implying that the experiment searched for the ether itself, not merely for one of its predicted effects.
The logical error: the textbook conflates the failure to detect one predicted effect (the ether wind of the stationary model) with the failure to find evidence for the ether as such. It then presents this conflation as motivation for abandoning the ether entirely -- a conclusion the original authors never drew.
Textbook 4: Tipler and Mosca
Paul A. Tipler and Gene Mosca, Physics for Scientists and Engineers, various editions, published by W.H. Freeman (now Macmillan Learning).
The pattern continues. Tipler and Mosca present the Michelson-Morley experiment as the death knell of the ether, leading directly to Einstein. The Lorentz-FitzGerald contraction, when mentioned at all, is characterised as an "ad hoc" device -- a desperate rescue attempt by physicists clinging to a discredited theory. The student encounters the contraction not as a physical prediction derived from electromagnetic theory, but as a historical curiosity, a wrong turn that was superseded by Einstein's correct insight.
No mention is made of the fact that the contraction was derived from first principles by Lorentz before Einstein's paper was published. No mention is made of the fact that FitzGerald reached the same conclusion independently, from the same physical reasoning. No mention is made of the mathematical identity between the two theories' predictions. Instead, Tipler and Mosca offer the student a morality tale: brave experimenters disprove a false theory, clever Einstein builds the correct one, and the adherents of the old theory engage in desperate rearguard actions (the "ad hoc" contraction) before being swept aside by the march of progress.
The narrative has the structure of a Whig history -- a story told from the perspective of the winners, in which every development leads inevitably toward the current orthodoxy. The losers are not merely wrong; they are foolish. Their ideas are not merely superseded; they are desperate. The student who reads this account has no reason to suspect that the "desperate" contraction was in fact a rigorous prediction derived from the electromagnetic properties of matter -- or that it remains, to this day, empirically indistinguishable from Einstein's formulation.
The logical error: the textbook presents the ether theory's response to the null result as ad hoc -- an invented patch -- when it was in fact a physical prediction derived from the electromagnetic nature of matter. The characterisation reverses the actual historical and logical situation.
Textbook 5: Krane
Kenneth S. Krane, Modern Physics, various editions, published by Wiley.
Krane's treatment, aimed at upper-division undergraduates -- students who have already completed introductory physics and are being trained to think more carefully about the foundations of the discipline -- discusses the Michelson-Morley experiment in the chapter on special relativity as motivating evidence for the theory. Lorentz Ether Theory is mentioned -- briefly -- but as a historical footnote rather than as a viable alternative interpretation. The student comes away understanding that the experiment motivated Einstein and that Einstein's theory superseded the ether. The possibility that Lorentz's theory remains empirically equivalent to Einstein's is not presented.
The logical error is one of omission rather than commission. Krane does not claim, in crude terms, that the experiment "disproved the ether." What the textbook does is present the ether theory as having been superseded -- implying that subsequent developments rendered it obsolete or empirically inferior -- without informing the student that no empirical distinction between the two theories has ever been demonstrated. The word "superseded" carries an implication of empirical defeat that the evidence does not support. A theory that has been superseded is one that made predictions which were falsified, or which could not account for new observations. Lorentz Ether Theory made no such failed predictions. It was not empirically defeated; it was socially displaced. Krane's treatment, by using the language of supersession, participates in the distortion even while avoiding the cruder forms.
For the upper-division student, the omission is arguably more damaging than the outright falsehoods in introductory texts. The introductory student may be forgiven for not knowing that alternatives exist. The upper-division student is being trained in professional judgement -- in how to evaluate competing theories, how to assess evidence, how to distinguish empirical from non-empirical arguments. Krane's omission deprives this student of the very example that would sharpen these skills: a case in which two empirically equivalent theories coexist, and the choice between them was made on philosophical, not empirical, grounds.
Textbook 6: Griffiths
David J. Griffiths, Introduction to Electrodynamics, fourth edition, published by Pearson.
Griffiths's electrodynamics textbook is used in virtually every upper-division undergraduate programme in the English-speaking world. It is the standard text for the course that teaches students how electromagnetic fields behave, how Maxwell's equations work, and how special relativity relates to electrodynamics. Its influence on the training of professional physicists is difficult to overstate. In its chapter on special relativity, the Michelson-Morley result is presented as evidence against the ether.
But Griffiths commits an additional distortion that is, in its way, more revealing than the others. He presents the Lorentz contraction as a consequence of special relativity -- as though the contraction were discovered by Einstein's theory, or at the very least as though it finds its natural home there. The historical reality is precisely reversed. The Lorentz contraction was derived by Lorentz, within ether theory, from the electromagnetic properties of matter, before Einstein's paper was published. Einstein's theory reproduces the contraction as a kinematic effect. The textbook attributes the discovery to the theory that borrowed it, not to the theory that produced it.
The irony is profound. Griffiths's textbook is devoted to electrodynamics -- the very subject from which Lorentz derived the contraction. The student who studies Griffiths learns Maxwell's equations, learns how electromagnetic fields behave under boosts, learns the full apparatus of relativistic electrodynamics. The student has all the tools needed to understand Lorentz's derivation. Yet Griffiths never tells the student that Lorentz used these very tools -- the tools the student is learning in this course -- to derive the contraction before Einstein. The student learns the effect as a consequence of Einstein's postulates, never suspecting that it was originally a consequence of the electrodynamics they have just mastered.
The logical error: the textbook reverses the historical priority. It presents a result of ether physics as a consequence of the theory that replaced it, erasing the originator from the narrative. A student reading Griffiths would have no reason to suspect that the Lorentz contraction was predicted by, and named after, a physicist working within an ether framework. The name "Lorentz contraction" becomes an empty label -- a historical fossil whose meaning has been stripped away.
Textbook 7: French
A.P. French, Special Relativity, published by W.W. Norton in 1968 as part of the MIT Introductory Physics Series.
French is somewhat more careful than the previous six. He acknowledges the historical development in greater detail. He discusses Lorentz's work and presents the contraction as a serious proposal, not merely as a desperate patch. His treatment of the Michelson-Morley experiment includes more historical context than most. But his framing ultimately arrives at the same destination: the experiment showed that "the ether concept had become untenable." The ether is presented as a stepping-stone to Einstein -- a useful scaffolding that was dismantled once the correct building was complete.
The logical error is subtle but consequential. French presents the ether as historically important but conceptually superseded. The word "untenable" does the work: it implies that the ether position can no longer be maintained, that the evidence has rendered it indefensible. But the evidence -- the null result -- is predicted equally by both theories. Nothing about the evidence makes the ether untenable. What makes the ether "untenable," in the context of French's narrative, is the community's preference for Einstein's interpretation -- a preference that French presents as though it were forced by experiment.
The sophistication of French's treatment makes its conclusion more misleading, not less. A student who reads a crude textbook knows they are reading a crude textbook. A student who reads French -- who encounters the careful historical discussion, the measured tone, the acknowledgement of Lorentz's contributions -- comes away with greater confidence in the conclusion. If even a careful, historically informed account at MIT concludes that the ether is untenable, the conclusion must be solid. The care of the treatment lends authority to the distortion. The student does not realise that the word "untenable" is doing work that the evidence cannot support -- that it imports a judgement of the community and presents it as a finding of the experiment.
Textbook 8: Feynman
Richard P. Feynman, Robert B. Leighton, and Matthew Sands, The Feynman Lectures on Physics, Volume I, Chapter 15-1.
Feynman requires separate treatment, because he was more careful than the others -- and his carefulness makes the sloppiness of the rest more conspicuous.
In his discussion of the Michelson-Morley experiment, Feynman presents the null result and Einstein's interpretation as the resolution. He does not, however, make the crude claim that the experiment "disproved the ether." He does not use the word "disproved." He does not claim the ether was shown not to exist. He presents the null result and then moves, without logical bridge, to Einstein's postulates, as though the elegance of the resolution were sufficient justification.
This is a different kind of error -- not a blunt distortion but an artful elision. Feynman's treatment is accurate about what the experiment showed. It simply omits what the experiment did not show: that Lorentz's ether theory predicts the same null result with equal precision, that the choice between the two interpretations was not forced by experiment, and that Einstein's postulates, however elegant, were a philosophical preference, not an empirical discovery.
Feynman was not a careless thinker. He was, by near-universal acknowledgement, one of the most penetrating and honest intellects of the twentieth century. If he avoided the crude distortions of the other textbooks, it was because his instinct for precision would not permit them. But the omission -- the failure to mention the empirical equivalence of the two theories -- is all the more significant because it comes from him. Feynman knew the physics. He knew Lorentz's theory. He knew the contraction was derived, not invented. His silence on these points was not ignorance. It was a choice -- a choice that reflects the depth of the consensus, the completeness of the paradigm's hold even on its most independent minds.
The significance of Feynman's treatment lies in the contrast. If the most careful and intellectually honest physicist of the twentieth century could not bring himself to mention the empirical equivalence of the two theories, the systematic nature of the omission is established. This is not one careless textbook author. This is the entire profession.
The pattern
Eight textbooks. Eight authors or author groups. Published across decades, across publishers, across levels from introductory to advanced. The specific language varies. The distortion does not. The pattern is consistent:
The experiment tested one model. The textbooks say it tested the concept.
The authors concluded that relative motion must be small. The textbooks conclude there is no ether.
Multiple ether theories accommodate the null result. The textbooks present no alternatives.
The null result is predicted identically by both Lorentz Ether Theory and special relativity. The textbooks present it as evidence for special relativity over the ether.
This is not a difference of emphasis. It is not a pedagogical simplification. It is the systematic misrepresentation of the logical content of a foundational experiment, repeated across the profession, across generations, for over a century. The number of physics students who have been taught this distortion is not thousands. It is not millions. It is hundreds of millions -- every physicist, every engineer, every science student who has taken a course in modern physics since the textbooks were standardised in the middle of the twentieth century.
Thomas Kuhn, the philosopher and historian of science, documented this process. In The Structure of Scientific Revolutions (1962), he described how textbooks are rewritten after a paradigm shift so that the new paradigm appears inevitable:
"The depreciation of historical fact is deeply, and probably functionally, ingrained in the ideology of the scientific profession."
Kuhn's word was "functionally" -- meaning the distortion serves a purpose. The textbook does not merely simplify history; it rewrites history so that the dominant theory appears to be the only theory that was ever viable. Alternative frameworks are not refuted; they are erased. The student does not learn that Lorentz's theory was rejected despite being empirically equivalent. The student learns that Lorentz's theory does not exist.
Kuhn used the word "Orwellian" in his later writings to describe this process. The textbook presents a Whig history -- a narrative in which every prior development was a step toward the current theory, in which every wrong turn is a "failed hypothesis" rather than a viable alternative that was displaced for non-empirical reasons. The student who reads the textbook believes they are learning history. They are learning mythology.
The Mathematical Identity
We now move from documentation to proof. The textbook distortion is not merely a historical error -- it is a logical impossibility. The Michelson-Morley experiment cannot distinguish between special relativity and Lorentz Ether Theory, because the two theories make identical predictions for every experiment that tests Lorentz invariance. This is not an approximation. It is not true only at low velocities, or only for electromagnetic phenomena, or only to first order. It is exact and complete.
The proof rests on a single fact: both theories use the same transformation equations.
The transformations
In 1904, Hendrik Lorentz published a paper in the Proceedings of the Royal Netherlands Academy of Arts and Sciences (Volume 6, pages 809 to 831) in which he derived a complete set of coordinate transformations relating measurements in a frame at rest in the ether to measurements in a frame moving through it. These transformations, now called the Lorentz transformations, are:
For a frame moving at velocity v in the positive x-direction:
The transformed spatial coordinate in the direction of motion equals the Lorentz factor multiplied by the quantity (the original coordinate minus velocity times time). That is: x' equals gamma times (x minus v times t).
The transformed coordinates perpendicular to the motion are unchanged: y' equals y, and z' equals z.
The transformed time coordinate equals the Lorentz factor multiplied by the quantity (the original time minus velocity times the original coordinate, divided by the square of the speed of light). That is: t' equals gamma times (t minus vx divided by c squared).
Here gamma -- the Lorentz factor -- equals one divided by the square root of (one minus v squared over c squared). These equations are presented in the companion monograph at etherphysics.org, Equations 2.25 through 2.28.
In 1905, Albert Einstein, in his paper "On the Electrodynamics of Moving Bodies" published in Annalen der Physik, derived the same transformations from two postulates: the principle of relativity (the laws of physics are the same in all inertial frames) and the constancy of the speed of light (light propagates at c in all inertial frames).
The mathematical content of the two derivations is identical. The same equations. The same gamma factor. The same transformation of coordinates. The same transformation of electromagnetic fields. The same velocity addition formula. The same predictions for length contraction, time dilation, the relativistic Doppler effect, and every other consequence of the transformations.
What differs is the physical interpretation:
In Lorentz's theory, the transformations describe real physical effects caused by motion through the ether. The contraction is a genuine compression of matter, caused by the modification of electromagnetic binding forces in the medium. The time dilation is a genuine slowing of physical processes, caused by the dynamics of matter-ether interaction. There is a preferred frame -- the ether rest frame -- in which the "true" coordinates are defined, but Poincare proved in 1905 that this frame is observationally inaccessible.
In Einstein's theory, the transformations describe the relationship between measurements made in different inertial frames, with no preferred frame. The contraction and time dilation are reciprocal -- each observer measures the other's rods as contracted and the other's clocks as slow. There is no underlying medium. The postulates are taken as brute facts about the structure of spacetime.
The interpretations differ profoundly. The predictions do not differ at all.
Theorem 1.1: Empirical Equivalence
The companion monograph states this as a theorem:
Theorem 1.1 (Empirical Equivalence). For all kinematic and electromagnetic phenomena expressible as functions of coordinates and field strengths, Lorentz Ether Theory and special relativity yield identical quantitative predictions, since both employ identical transformation equations applied to identical dynamical laws.
The proof is immediate. Both theories compute observable quantities -- the position of a fringe, the reading of a clock, the deflection of a particle, the frequency of a photon -- using the same mathematical operations on the same equations. The Lorentz transformations are the same in both theories. Maxwell's equations, which govern electromagnetic phenomena, are the same in both theories. The dynamical laws are the same. The calculations are the same. The numbers that emerge are the same.
This equivalence is not an accident or a coincidence. It is a structural identity. As the philosopher of physics Pablo Acuna documented in a detailed 2014 analysis, "On the Empirical Equivalence Between Special Relativity and Lorentz's Ether Theory" (Studies in History and Philosophy of Modern Physics, Volume 46B, pages 283 to 302), the two theories are empirically equivalent in the strongest sense: no possible observation can distinguish between them, because they agree on all observable consequences.
What this means for the textbook claim
The implication for the Michelson-Morley experiment is devastating.
The null result -- the absence of a fringe shift -- is predicted by special relativity. It is also predicted, with identical precision, by Lorentz Ether Theory. In Lorentz's framework, the null result arises because the interferometer arm aligned with the direction of motion through the ether contracts by exactly the factor required to compensate for the difference in light travel times. The contraction is not assumed; it is derived from the electromagnetic properties of the material composing the arm.
If both theories predict the same result, the observation of that result cannot be evidence for one theory over the other. This is a principle of elementary logic. If hypothesis A predicts observation X, and hypothesis B also predicts observation X, then the observation of X does not distinguish between A and B. To present X as evidence for A requires the silent assumption that B has been eliminated on other grounds -- an assumption that must be stated and defended, not smuggled in.
The textbooks do not state this assumption. They do not defend it. They do not mention it. They present the null result as evidence for special relativity and against the ether, as though the empirical equivalence did not exist.
The Contraction That Was Not Ad Hoc
There is a word that appears in textbook after textbook when the Lorentz-FitzGerald contraction is mentioned: "ad hoc." The contraction, the student is told, was an ad hoc hypothesis -- a patch, an invention, a desperate manoeuvre to save a failing theory. This characterisation has become so standard that it functions as a synonym for the contraction itself. Students learn it as: "Lorentz proposed an ad hoc contraction to explain away the null result."
This characterisation is false. It is not merely misleading or overstated. It is the reverse of the truth. The contraction was derived from first principles. It was predicted by the electromagnetic theory of matter. And two physicists reached the same conclusion independently, from the same physics -- which is the hallmark of a natural physical consequence, not a desperate patch.
Lorentz's derivation
Hendrik Antoon Lorentz first proposed the contraction in 1892, in a paper titled "De relative beweging van de aarde en den aether" ("The Relative Motion of the Earth and the Aether"), published in Zittingsverslagen Akad. v. Wet., Volume 1, pages 74 to 79. (An English version appears in his Collected Papers, Volume IV, pages 219 to 223.)
Lorentz's reasoning was not ad hoc. It was physical, specific, and derived from the established properties of electromagnetism. The derivation proceeds from four premises to an inescapable conclusion, and the general reader deserves to see the logic laid bare -- because it is the logic that the textbooks bury.
Premise 1: Matter is held together by electromagnetic forces. The atoms that compose a solid body -- a measuring rod, an interferometer arm, any material object -- are held in their positions by electromagnetic interactions. The electric fields between electrons and nuclei, the electromagnetic bonds between atoms, the forces that maintain the equilibrium spacing of a crystal lattice: all are electromagnetic. This was understood in the 1890s. It is established beyond question today.
Premise 2: Electromagnetic fields transform when their source moves through the ether. This is a consequence of Maxwell's equations, derivable by direct calculation. When a charged particle moves through the ether at velocity v, the electric field it produces becomes anisotropic. In the direction of motion, the field is unchanged. In the direction perpendicular to the motion, the field is enhanced by the Lorentz factor gamma. The magnetic field that appears is related to the electric field by the velocity. These are not assumptions -- they are mathematical consequences of the field equations that every electrodynamics student derives in their third-year course. The companion monograph gives the explicit expressions in Equations 2.21 and 2.32-2.33.
Premise 3: If the binding forces change, the equilibrium shape changes. This is elementary physics. If the forces that hold a crystal together in one configuration are altered -- if the field between every pair of atoms is modified by motion through the medium -- then the equilibrium configuration must change. The atoms will rearrange themselves to find a new equilibrium under the new force law. The shape of the body will be different from its shape at rest.
Conclusion: Objects in motion through the ether contract in the direction of motion. The anisotropy of the field -- enhanced perpendicular to the motion, unchanged parallel to it -- produces a new equilibrium in which the body is compressed along the direction of motion. The factor of contraction is precisely one divided by gamma, which equals the square root of (one minus v squared over c squared). This is not an adjustable parameter. It is not chosen to fit the data. It follows from the transformation properties of the electromagnetic field -- the same transformation properties that every upper-division physics student learns in their electrodynamics course.
This deserves attention. The contraction is derived from Maxwell's equations. It is a consequence of the physics that every textbook teaches. The tools are in every electrodynamics course. The derivation is in the companion monograph, Section 2.4. It is not abstruse. It is not speculative. It is a straightforward application of electromagnetic theory to the structure of matter. And it was published in 1892 -- thirteen years before Einstein's paper.
In 1904, Lorentz published the landmark paper in which he assembled the contraction, time dilation, and the full coordinate transformations into a complete theory: "Electromagnetic phenomena in a system moving with any velocity smaller than that of light" (Proceedings of the Royal Netherlands Academy of Arts and Sciences, Volume 6, pages 809 to 831). The mathematical content of this paper is identical to what Einstein published the following year. The physical interpretation is different: for Lorentz, these are real effects caused by motion through a physical medium. But the equations are the same.
FitzGerald's independent derivation
George Francis FitzGerald proposed the same contraction independently in 1889 -- three years before Lorentz's paper and sixteen years before Einstein's. His proposal appeared in a brief letter published in Science, Volume 13, page 390. It is one of the shortest significant publications in the history of physics: a few lines stating the essential idea.
FitzGerald's reasoning was similar to Lorentz's but more general. If the forces that hold matter together are electromagnetic in nature -- or, more broadly, if they propagate at the speed of light -- then a body moving through the ether would be expected to contract. The interaction between the body and the medium modifies the equilibrium configuration. The contraction follows.
Oliver Lodge discussed FitzGerald's proposal at length in his 1893 paper "Aberration Problems," published in Philosophical Transactions of the Royal Society A, Volume 184, pages 727 to 804, establishing that FitzGerald's reasoning was taken seriously by his contemporaries. FitzGerald was not a fringe figure. He held the Erasmus Smith Chair of Natural and Experimental Philosophy at Trinity College, Dublin. He was a respected physicist, a member of the Royal Society, a contemporary of Lodge, Thomson, and Heaviside. His proposal was discussed in the major physics journals of the day.
The significance of the independent derivation cannot be overstated. When two physicists, working independently from the same physical principles, reach the same quantitative conclusion, the conclusion is not ad hoc. It is a natural consequence of the physics. An ad hoc hypothesis is one invented to explain away an unwelcome result, with no independent theoretical motivation. The Lorentz-FitzGerald contraction had independent theoretical motivation: the electromagnetic nature of matter and its interaction with the medium. Two physicists saw the same implication in the same physics. This is not desperation. It is derivation.
An analogy clarifies the point. If two engineers, working independently, predict that a bridge will deflect under load -- not because they measured the deflection and then invented a theory to match it, but because they applied the known properties of steel to the known forces on the structure -- no one would call their prediction "ad hoc." The prediction follows from the physics. The same is true of the Lorentz-FitzGerald contraction. It follows from the electromagnetic nature of matter, just as the bridge deflection follows from the elastic properties of steel.
The "ad hoc" slander: origins and function
Where, then, does the characterisation of the contraction as "ad hoc" come from? This question matters, because the label did not arise from a careful analysis of Lorentz's reasoning. It arose from the narrative needs of a paradigm.
The label appears to have its origins in the generation of physicists who adopted Einstein's approach and needed to justify their choice against a rival that was, embarrassingly, empirically identical. The word "ad hoc" -- Latin for "for this purpose," meaning improvised for a specific situation without broader justification -- was applied to Lorentz's contraction as a way of marking it as inferior. The contraction was said to have been invented solely to explain away the Michelson-Morley null result, with no independent motivation.
This is historically false, as we have just demonstrated. But the label served a crucial function. If the two theories make the same predictions, the choice between them must be justified on non-empirical grounds. The most common such justification is theoretical virtue: one theory is said to be more elegant, more parsimonious, less ad hoc than the other. If Lorentz's contraction is ad hoc, then his theory is inferior to Einstein's on grounds of theoretical economy -- it introduces an arbitrary assumption that Einstein's theory does not require. The choice is justified. The narrative works.
If, however, Lorentz's contraction is derived from first principles -- as it was -- then the theoretical virtue argument collapses. The contraction is not an arbitrary assumption; it is a prediction. It adds no free parameters. It follows from the same Maxwell's equations that Einstein's theory also uses. The "economy" of Einstein's approach is not that it avoids an arbitrary assumption, but that it provides no mechanism for a phenomenon that Lorentz's approach explains. This is not parsimony. This is the absence of explanation masquerading as elegance.
Harvey Brown makes this point with characteristic precision. If anything is ad hoc, Brown argues, it is Einstein's elevation of the light-speed postulate to the status of an axiom -- a brute fact to be accepted without explanation. Why does light travel at the same speed in all inertial frames? Einstein's answer: it just does. It is a postulate. Lorentz's answer: because the measuring instruments -- the rods and clocks used to determine speed -- are made of matter held together by electromagnetic forces, and those forces transform in a way that produces exactly this result. Lorentz explains the postulate. Einstein assumes it.
Which of these deserves the label "ad hoc"?
The textbooks needed Lorentz's contraction to be ad hoc. Without that characterisation, the story they tell does not work. The experiment cannot have disproved the ether if the ether theory predicted the null result from first principles. The "ad hoc" label is not a description of Lorentz's reasoning. It is a requirement of the textbook narrative. It was constructed after the fact, propagated through generations of authors who did not read the original papers, and reinforced by every repetition until it acquired the appearance of established truth. The label is itself evidence -- not of bad physics, but of narrative maintenance.
The Experiment That Found Something
The Michelson-Morley experiment of 1887 is not the end of the interferometer story. It is not even the most extensive chapter. In the decades that followed, larger and more sensitive interferometers were built, operated at different altitudes, in different locations, across different seasons. The textbooks present these as a series of null results confirming the 1887 finding. The textbooks are wrong. One of the most extensive interferometer programmes in the history of physics obtained persistent positive results -- and was explained away rather than engaged.
Dayton Miller at Mount Wilson
Dayton Clarence Miller was Professor of Physics at the Case School of Applied Science in Cleveland -- the same institution where Michelson and Morley had performed their experiment. Miller was not a fringe figure. He was a meticulous experimentalist, a member of the National Academy of Sciences, and president of the American Physical Society in 1925-1926. He had collaborated with Morley himself on interferometer experiments in the early 1900s. His credentials were impeccable, his laboratory skills beyond question.
Between 1921 and 1926, Miller conducted an extensive series of ether-drift experiments at the Mount Wilson Observatory in California, at an altitude of approximately 1,750 metres. His reasoning was that altitude might matter: if the ether is partially dragged by the Earth, the dragging effect might be weaker at higher altitudes, farther from the Earth's surface. A basement laboratory in Cleveland, surrounded by heavy masonry, might be in a region of stronger ether drag than an exposed observatory on a mountain top.
Miller's interferometer was larger than Michelson and Morley's original, with an effective arm length of approximately 32 metres (compared to 11 metres in 1887). He performed over 200,000 individual observations across multiple years, in different seasons, at different times of day and night. The sheer volume of data was unprecedented.
The results were not null. Miller consistently observed a small but persistent fringe shift, corresponding to an apparent ether-drift velocity of approximately 8 to 10 kilometres per second. This was smaller than the 30 kilometres per second predicted by the stationary ether model (which had already been eliminated by the original Michelson-Morley experiment), but it was larger than zero. The direction of the apparent drift showed a sidereal pattern -- it rotated with the stars rather than with the Sun -- which Miller interpreted as evidence for the Earth's motion through a cosmic ether.
Miller published his results extensively, culminating in a comprehensive review: Dayton C. Miller, "The Ether-Drift Experiment and the Determination of the Absolute Motion of the Earth," Reviews of Modern Physics, Volume 5, Number 3 (July 1933), pages 203 to 242. This was not a letter to an obscure journal. Reviews of Modern Physics is one of the most prestigious physics journals in the world, published by the American Physical Society. The paper was peer-reviewed and accepted for publication by the editorial standards of the time.
The contested dismissal
Miller's results posed a problem for the emerging consensus. If there was a genuine ether drift, even a small one, the entire narrative -- Michelson-Morley disproved the ether, Einstein built the correct theory without it -- would require revision. The results were not ignored. They were addressed. But the manner in which they were addressed is itself revealing.
In 1955, twenty-two years after Miller's Reviews of Modern Physics paper and two years after Miller's death, Robert S. Shankland and three colleagues at Case Institute of Technology published a reanalysis: R.S. Shankland, S.W. McCuskey, F.C. Leone, and G. Kuerti, "New Analysis of the Interferometer Observations of Dayton C. Miller," Reviews of Modern Physics, Volume 27, Number 2 (April 1955), pages 167 to 178.
Shankland and colleagues reanalysed Miller's data and attributed the positive results to systematic errors, primarily temperature gradients in the optical path. Their conclusion was that the apparent ether drift was an artefact of thermal effects, not a genuine signal.
The reanalysis has been accepted by the physics community as the definitive assessment. Textbooks that mention Miller at all cite Shankland as having disposed of his results. The matter is considered settled.
But the settlement is not as clean as it appears. Several aspects of the situation deserve attention.
First, Miller himself had been aware of temperature effects and had taken extensive precautions against them, including insulating his apparatus and performing measurements at different times of day when temperature conditions varied. He argued that the systematic pattern of his results -- the sidereal dependence, the consistency across years and seasons -- was not consistent with a simple temperature artefact.
Second, the Shankland reanalysis was conducted after Miller's death, when he could not respond to its arguments. This is not, by itself, evidence of bad faith -- scientific reanalysis is legitimate and necessary. But the timing meant that the debate was one-sided. Miller's extensive knowledge of his own apparatus, his awareness of potential systematic errors, and his arguments for the genuineness of his results died with him.
Third, and most significant for our argument: the institutional irony. Miller worked at the Case School of Applied Science -- the same institution where the Michelson-Morley experiment was performed. He was Michelson's colleague. His results, had they been taken seriously, would have complicated the most famous achievement associated with his own institution. The institution that produced the null result that "disproved" the ether also produced the positive result that challenged that narrative. The positive result was explained away. The null result became the origin myth of modern physics.
Whether Miller's results were genuine or artefactual is a question that remains open to investigation. The documented facts are these: a meticulous experimentalist, president of the American Physical Society, using a larger interferometer than the original Michelson-Morley apparatus, obtained persistent positive results over five years of observation, published them in the most prestigious review journal in physics, and had those results attributed to temperature effects by a reanalysis conducted after his death. The strength of this dismissal is for history to judge.
What is not in dispute is this: the textbooks do not mention Miller. They do not mention the positive results. They do not mention the Shankland reanalysis. They present the Michelson-Morley experiment and its successors as a uniform series of null results. This is factually incorrect. At minimum, intellectual honesty requires acknowledging that one major programme obtained positive results that were contested -- not that the entire experimental record is uniformly null. The omission of Miller from the textbook record is itself a piece of evidence in the pattern this chapter documents.
The Subsequent Experiments
The Michelson-Morley experiment was not the last word. In the century that followed, experiments of increasing precision were performed, each one testing Lorentz invariance with greater sensitivity. The textbooks present these as a cumulative confirmation of special relativity -- experiment after experiment proving Einstein right. The textbooks are committing the same logical error each time. Every one of these experiments is predicted identically by both theories. The "evidence" for special relativity over Lorentz Ether Theory, as presented in the textbooks, does not exist.
Kennedy-Thorndike (1932)
Roy J. Kennedy and Edward M. Thorndike performed an experiment in 1932, published in Physical Review (Volume 42, pages 400 to 418), that modified the Michelson-Morley design in a significant way. Where Michelson and Morley used interferometer arms of equal length, Kennedy and Thorndike used arms of deliberately unequal length. The purpose was to test a different prediction: if there is an ether drift, and if the Earth's velocity through the ether changes over the course of months (as it must, since the Earth's orbital velocity changes direction), then an interferometer with unequal arms should show a variation in the interference pattern over time, even if no instantaneous fringe shift is observed.
The experiment found no variation over months of observation. The null result is predicted by special relativity, which holds that no preferred frame exists and therefore no time-dependent drift should appear.
The null result is also predicted by Lorentz Ether Theory. In Lorentz's framework, the combination of length contraction and time dilation -- both derived from the electromagnetic nature of matter and its interaction with the ether -- cancels the expected variation. The contraction adjusts the arm lengths; the time dilation adjusts the comparison clock. The two effects conspire to produce a null result. This is not a coincidence -- it is a mathematical consequence of the Lorentz transformations, which both theories employ.
The textbooks present Kennedy-Thorndike as additional evidence for special relativity. It is additional evidence for the Lorentz transformations, which is not the same thing.
Ives-Stilwell (1938)
Herbert E. Ives and G.R. Stilwell performed an experiment in 1938, published in the Journal of the Optical Society of America (Volume 28, pages 215 to 226), that directly measured the transverse Doppler effect -- the frequency shift of light emitted by an atom moving perpendicular to the line of sight. This effect is a direct consequence of time dilation: a moving clock runs slow, and a moving atom emits light at a lower frequency. The experiment confirmed the predicted frequency shift to high precision.
The result is predicted identically by both theories. In special relativity, the moving atom's clock runs slow because of the structure of Minkowski spacetime. In Lorentz Ether Theory, the moving atom's clock runs slow because electromagnetic processes in the atom are modified by motion through the ether. The predicted frequency shift is the same in both cases -- it depends only on the Lorentz factor gamma, which is the same in both theories.
But the Ives-Stilwell experiment contains a historical detail that the textbooks never mention, a detail that is devastating to the standard narrative. Herbert Ives himself was an ether theorist. He did not interpret his own experiment as confirming Einstein's theory. He interpreted it as confirming Lorentz's theory. The man who performed the experiment, who understood its design, its systematic errors, and its implications better than anyone, believed it supported the ether.
Ives published extensively on this interpretation. In a 1951 paper, "Revisions of the Lorentz Transformations" (Proceedings of the American Philosophical Society, Volume 95, Number 2, pages 125 to 131), he argued explicitly for the Lorentz interpretation over the Einstein interpretation. He was not a crank. He was the head of electrooptical research at Bell Telephone Laboratories, a distinguished experimentalist, the recipient of the Frederic Ives Medal of the Optical Society of America (a medal named after his father). His interpretation of his own experiment was ignored by the profession that canonised the experiment as evidence for Einstein.
The irony is layered. The experiment is named after Ives. It is taught in textbooks as confirming special relativity. The man whose name it bears believed it confirmed the ether. The textbooks do not mention this. The student who learns the Ives-Stilwell experiment as evidence for special relativity has no way of knowing that Ives himself would have rejected that interpretation.
Modern optical cavity tests
In the early twenty-first century, experiments testing Lorentz invariance achieved sensitivities undreamt of by Michelson and Morley. Experiments by Herrmann and colleagues (2009, Physical Review D, Volume 80, paper 105011) and by Eisele and colleagues (2009, Physical Review Letters, Volume 103, paper 090401) used ultra-stable optical resonators -- cavities in which light bounces back and forth between mirrors of extraordinary quality -- to search for anisotropies in the speed of light at the level of one part in ten to the seventeenth power. This is a sensitivity approximately twelve orders of magnitude beyond what Michelson and Morley could achieve.
All yield null results. All are cited in the literature as "confirming special relativity" or "testing the foundations of special relativity." None mention Lorentz Ether Theory. None acknowledge that the null results are predicted by both theories.
The mathematical situation is exactly as it was in 1887, only more precise. The Lorentz transformations -- which both theories employ -- guarantee the null result by mathematical construction. No experiment that tests Lorentz invariance can produce a result that distinguishes between the two theories, because both theories are constructed to be Lorentz invariant. To test Lorentz invariance at one part in ten to the seventeenth is to confirm the Lorentz transformations at one part in ten to the seventeenth. Both theories pass. Neither is favoured. The "evidence" for special relativity, at any level of precision, is not evidence for special relativity over Lorentz Ether Theory.
The pattern is absolute
Every experiment that has ever been performed to test Lorentz invariance -- from the basement in Cleveland to the Mount Wilson Observatory to the most precise optical cavities in modern laboratories -- is predicted identically by both frameworks. Not approximately. Not to first order. Identically. The "evidence" for special relativity over Lorentz Ether Theory, as presented in the textbooks, does not exist. It has never existed. The choice between the two theories was made on other grounds entirely.
Harvey Brown and the Explanatory Inversion
The most thorough modern analysis of the relationship between special relativity and Lorentz Ether Theory is provided by Harvey Brown, professor of philosophy of physics at the University of Oxford, in his 2005 book Physical Relativity: Space-time Structure from a Dynamical Perspective, published by Oxford University Press.
Brown's argument is technical, rigorous, and devastating. It does not merely establish that the two theories are empirically equivalent -- that much was already known. It establishes something deeper and more unsettling: that the standard narrative has the explanatory situation precisely backwards. The theory that was adopted does not explain more than the theory it replaced. It explains less. The community chose the shallower theory and called it progress.
Constructive versus principle theories
Brown's argument rests on a distinction that Einstein himself articulated: the distinction between constructive theories and principle theories. Einstein drew this distinction in a 1919 essay published in the Times of London, and it is one of his most illuminating contributions to the philosophy of physics -- illuminating because it applies, with uncomfortable precision, to his own theory.
A constructive theory explains phenomena from the bottom up. It posits a physical mechanism -- particles, forces, a medium -- and derives the observed phenomena from the behaviour of that mechanism. The kinetic theory of gases is a constructive theory: it explains temperature, pressure, and the gas laws by positing that gases are composed of molecules in random motion. The explanation is mechanical. You understand why the gas behaves as it does. You can ask: what happens if we change the molecular mass? What happens at higher density? What happens near a phase transition? The theory answers these questions because it has a mechanism.
A principle theory constrains phenomena from the top down. It posits general principles -- laws that must be obeyed -- and derives consequences from those principles, without specifying the underlying mechanism. Thermodynamics is a principle theory: it tells you that energy is conserved and entropy increases, without telling you what energy or entropy are at the microscopic level. The principles are powerful and general, but they do not explain. They constrain. They tell you what must happen, not why it happens.
Einstein acknowledged, in his own writings, that special relativity is a principle theory. He elevated the constancy of the speed of light and the relativity principle to the status of postulates -- axioms to be accepted, not explained. The Lorentz transformations follow from these postulates. Length contraction and time dilation follow from the transformations. But the theory provides no mechanism. It does not say why rods contract. It does not say what physical process causes clocks to slow. The postulates describe. They do not explain.
Einstein also acknowledged -- and this is the point the textbooks suppress -- that constructive theories provide deeper understanding. In the same 1919 essay, he wrote that constructive theories, which build up the phenomena from underlying mechanisms, offer "the most complete kind of understanding." Principle theories, by contrast, offer logical rigour and economy but leave the underlying mechanism unspecified. Einstein preferred his principle-theory approach for its elegance. He did not claim it provided deeper understanding.
The Minkowski illusion
Lorentz's theory was a constructive theory. It specified the mechanism: electromagnetic binding forces in matter, propagating through the ether, modified by motion. The contraction follows from the mechanism. The time dilation follows from the mechanism. You understand why rods contract -- because the forces holding them together are altered by motion through the medium. You understand why clocks slow -- because the electromagnetic processes that constitute the clock's operation are modified by the medium.
In 1908, Hermann Minkowski reformulated special relativity as geometry in four-dimensional spacetime, and this reformulation has dominated physics ever since. Minkowski's famous declaration -- "Henceforth space by itself, and time by itself, are doomed to fade away into mere shadows, and only a kind of union of the two will preserve an independent reality" -- established the framework in which special relativity is taught today. Length contraction and time dilation are presented as consequences of the geometry of Minkowski spacetime. The student learns that rods contract "because" spacetime has a certain structure, that clocks slow "because" of the geometry of the four-dimensional manifold.
Brown's central argument is this: Minkowski spacetime geometry does not explain why rods contract and clocks slow. It describes the fact that they do. The geometry is a mathematical codification of the behaviour of rods and clocks, not a physical explanation of that behaviour. To say "rods contract because spacetime has Minkowski geometry" is to say "rods contract because rods contract" -- the geometry is constructed from the behaviour of physical objects (rods and clocks are the physical standards by which spacetime intervals are operationally defined), so it cannot serve as an independent explanation of that behaviour. The reasoning is circular.
Brown puts the point precisely: the Minkowski metric encodes the Lorentz covariance of the laws governing the microstructure of matter. If the laws governing molecular binding, electromagnetic interaction, and atomic structure are Lorentz covariant, then rods and clocks built from this matter will exhibit Lorentz contraction and time dilation. The geometry describes this fact. It does not explain it. The explanation must come from the dynamics -- from the actual physics of how matter is constituted and how it behaves under boosts. The geometry is a summary of the dynamical facts, written in a particularly elegant mathematical language. But a summary is not an explanation, any more than a map is the territory it represents.
The explanatory arrow
The implications are profound. If Brown is correct -- and his argument has been extensively discussed in the philosophy of physics community, including substantive responses by Michel Janssen ("Drawing the Line Between Kinematics and Dynamics in Special Relativity," Studies in History and Philosophy of Modern Physics, Volume 40, 2009, pages 26 to 52) and by Oliver Pooley, with no successful refutation of the core point -- then the explanatory arrow in special relativity points in the opposite direction from what the textbooks claim.
The textbooks say: spacetime has Minkowski geometry, and this explains why rods contract and clocks slow. Brown says: the dynamics of matter and fields are Lorentz covariant, and this explains why spacetime has Minkowski geometry. The arrow runs from dynamics to geometry, not from geometry to dynamics.
Lorentz and FitzGerald had the right explanatory instinct. They sought a dynamical, constructive account of the phenomena -- an explanation rooted in the physical constitution of matter and its interaction with the medium. Einstein's and Minkowski's approaches, by elevating the phenomena to postulates or geometric features, traded explanation for description. The theory that replaced the ether did not provide a deeper account of why things happen. It provided a more elegant description of what happens. The constructive theory was discarded in favour of the principle theory. The explanation was discarded in favour of the description.
The century of describing without explaining
Brown's analysis implies that physics has spent a century describing length contraction and time dilation without explaining them. The Minkowski framework tells us that these effects occur. It tells us how to calculate them. It provides a beautiful geometric picture. It does not tell us why a rod made of atoms held together by electromagnetic forces contracts when it moves. It cannot tell us, because its postulates are taken as brute facts. The question "why does the speed of light appear the same in all inertial frames?" is answered with: "It is a postulate." The question "why do rods contract?" is answered with: "It is a consequence of the postulates." The underlying dynamics -- the actual physics of matter-medium interaction that would constitute a genuine explanation -- are not in the theory, because the theory was constructed to avoid needing them.
This is the deepest layer of the textbook distortion. The textbooks do not merely misrepresent the history. They do not merely commit logical errors in presenting the Michelson-Morley experiment. They present a description as an explanation, a codification as a theory, a summary as a discovery. The student who learns special relativity from a textbook comes away believing that the geometry of spacetime explains the behaviour of matter. Brown has shown that the relationship runs the other way: the behaviour of matter explains the geometry of spacetime.
If Brown is correct, then the irony is complete. The theory that explains -- Lorentz's constructive theory, with its physical mechanism and its derivation from electromagnetic principles -- was abandoned in favour of the theory that describes -- Einstein's principle theory, with its unexplained postulates and its geometric codification of phenomena. The theory with the deeper understanding was discarded. The theory with the shallower understanding was adopted. And the textbooks, to justify the adoption, labelled the deeper theory "ad hoc."
Poincare and the Erasure
The textbook distortion extends beyond the Michelson-Morley experiment and beyond the treatment of Lorentz. There is a third figure whose contributions to the physics of relativity have been systematically diminished in the standard narrative -- a figure whose work is so closely intertwined with what we now call "special relativity" that a major historian of science titled his chapter on the subject "The Relativity Theory of Poincare and Lorentz."
Henri Poincare submitted two papers in 1905 on the electrodynamics of moving bodies. The short paper, "Sur la dynamique de l'electron," was submitted to the Comptes Rendus de l'Academie des Sciences on 5 June 1905. The long paper, with the same title, was submitted to the Rendiconti del Circolo Matematico di Palermo on 23 July 1905. Einstein's paper, "Zur Elektrodynamik bewegter Korper," was received by Annalen der Physik on 30 June 1905.
The chronology is important. Poincare's short paper was submitted before Einstein's paper was received. The long paper was submitted shortly after. The mathematical content of Poincare's papers overlaps substantially with Einstein's -- and in certain respects goes further.
Poincare's contributions in these papers include: the demonstration that the Lorentz transformations form a group (now called the Lorentz group or, with translations, the Poincare group) -- a fundamental mathematical insight that Einstein did not provide; the proof that Maxwell's equations are fully covariant under the Lorentz transformations; the explicit statement that no experiment can detect absolute motion -- a statement equivalent to Einstein's first postulate; the discussion of how gravitation would need to be modified in a Lorentz-covariant framework, anticipating aspects of the problem Einstein would address over the next decade; and the introduction of a four-dimensional formalism (using ict as the fourth coordinate) that foreshadowed Minkowski's 1908 spacetime.
Crucially, Poincare did all of this within an ether framework. He did not abandon the ether. He showed that the ether could be retained while deriving all of the mathematical content of what we now call special relativity. His ether was undetectable -- he proved as much -- but Poincare did not see undetectability as a reason to deny existence. The ether remained, for Poincare, the physical substrate that explained why electromagnetic waves propagate.
Edmund Taylor Whittaker, the distinguished mathematical physicist and historian of science, assessed this situation in 1953. In the second volume of his A History of the Theories of Aether and Electricity, published by Thomas Nelson and Sons, Whittaker titled his chapter on special relativity "The Relativity Theory of Poincare and Lorentz." He treated Einstein's 1905 contribution as one among several -- important, but not the singular breakthrough that the standard narrative claims. This provoked considerable controversy, most notably from Max Born, who defended Einstein's priority. The debate has been thoroughly documented in the historical literature, including works by Arthur I. Miller (Albert Einstein's Special Theory of Relativity, Addison-Wesley, 1981), Olivier Darrigol ("The Mystery of the Einstein-Poincare Connection," Isis, Volume 95, 2004, pages 614 to 626), and Peter Galison (Einstein's Clocks, Poincare's Maps, W.W. Norton, 2003).
We introduce the Poincare dimension here not to resolve the priority question -- that is the subject of the next chapter -- but to add a layer to the textbook forensics. Every textbook examined in this chapter attributes the theory of special relativity to Einstein alone. Not one of them gives Poincare's 1905 papers substantive treatment. Not one mentions that Poincare's short paper was submitted before Einstein's. Not one notes that Poincare derived the group structure of the Lorentz transformations, which Einstein did not. Not one acknowledges that Poincare accomplished his work within an ether framework.
The textbooks that misrepresent the Michelson-Morley experiment also misrepresent the history of the theory that supposedly replaced the ether. The distortion is not confined to one experiment or one historical episode. It is a comprehensive revision of the early twentieth century, in which one physicist is elevated to sole authorship, his collaborators and competitors are diminished or erased, and the framework they worked within is declared dead -- all in the service of a narrative in which experiment disproves the ether, genius builds the replacement, and the replacement is uniquely correct.
The textbooks that omit Poincare commit historical distortion atop logical distortion. The logical distortion is the misrepresentation of the Michelson-Morley experiment. The historical distortion is the erasure of a figure whose contributions are, by any honest reckoning, foundational. The two distortions reinforce each other: if Poincare's work is invisible, the student cannot discover that relativity was developed within an ether framework. If the ether framework is declared dead, the student has no reason to look for Poincare's contributions to it. The student is left with Einstein -- alone, unprecedented, working from pure thought to produce a theory that experiment had already demanded. The story is clean. It is compelling. It is false.
The Implications
The evidence presented in this chapter establishes the following.
The Michelson-Morley experiment tested one specific ether model -- a rigid, stationary, non-dragged ether -- and found that model inconsistent with the data. The original authors said so, in precisely those terms. They did not conclude that the ether does not exist.
Eight major textbooks -- Serway, Halliday, Young, Tipler, Krane, Griffiths, French, and the Feynman Lectures -- systematically misrepresent this result, presenting it as evidence against the ether as such, or as evidence for special relativity over ether theory, or as motivation for abandoning the ether entirely. The distortion varies in its crudeness but not in its direction. Every textbook pushes the same way. Each commits a specific logical error -- from the crude (Serway: "the ether hypothesis was wrong") to the subtle (Feynman: an artful elision that omits the empirical equivalence) -- but the cumulative effect is identical: the student learns that the ether was disproved by experiment.
The mathematical proof of empirical equivalence -- Theorem 1.1 of the companion monograph -- establishes that no experiment testing Lorentz invariance can distinguish between special relativity and Lorentz Ether Theory, because both theories use the same transformation equations. The "evidence" cited by the textbooks is not evidence. It is a logical fallacy: the presentation of a result predicted by both theories as evidence for one over the other.
The Lorentz-FitzGerald contraction, far from being an "ad hoc" patch, was derived from first principles by two independent physicists before Einstein's paper was published. The derivation proceeds from established electromagnetic theory -- the same theory that every textbook teaches -- through a chain of reasoning that is straightforward, physical, and motivated by the known properties of matter. The "ad hoc" label was applied after the fact, by the victors of a paradigm dispute, and has been propagated without scrutiny through generations of textbooks. Harvey Brown's analysis demonstrates that if anything deserves the label "ad hoc," it is the unexplained postulates of the theory that replaced Lorentz's.
The subsequent experiments -- Kennedy-Thorndike, Ives-Stilwell, modern optical cavity tests -- are each predicted identically by both theories. Each is presented in the textbooks as confirming special relativity. Each equally confirms Lorentz Ether Theory. The Ives-Stilwell experiment was performed by an ether theorist who interpreted his own results as confirming Lorentz. The textbooks that bear his name in their narrative do not mention his interpretation.
Dayton Miller's extensive interferometer programme at Mount Wilson obtained persistent positive results over five years. These results were explained away by a reanalysis conducted after Miller's death. The textbooks do not mention Miller. They do not mention the positive results. They present the experimental record as uniformly null, which it is not.
Harvey Brown's analysis of the constructive-principle distinction demonstrates that the standard narrative has the explanatory situation backwards. The theory that was adopted (Einstein's principle theory) describes the phenomena without explaining them. The theory that was abandoned (Lorentz's constructive theory) explains the phenomena from first principles. The community chose description over explanation and called it progress.
Poincare's 1905 contributions -- submitted before Einstein's paper, accomplishing much of the same mathematical work within an ether framework -- are invisible in the textbooks. The historical distortion reinforces the logical distortion: by erasing the figure who developed relativity within an ether framework, the textbooks make it impossible for the student to discover that such a development occurred.
The question that the textbooks never raise -- and that the evidence documented in this chapter forces into the open -- concerns the scope of the distortion.
The scale of the distortion
The pattern documented above is not confined to one textbook or one careless author. It spans eight textbooks -- the dominant texts across the English-speaking world -- published by Cengage, Wiley, Pearson, Macmillan, Norton, and Addison-Wesley, spanning decades, spanning levels from introductory to advanced. Each was reviewed by editors, vetted by peer review, adopted by departments, purchased by students, taught by lecturers who learned from the same textbooks a generation earlier. The same distortion appears in every one.
The scale can be quantified. In the United States alone, approximately 700,000 students per year take introductory physics courses. Over fifty years, that is thirty-five million students in the United States alone. Extended to the United Kingdom, Canada, Australia, India, and every other country where English-language physics textbooks are used, the number reaches the hundreds of millions. Every one of them was taught that the Michelson-Morley experiment disproved the ether.
Incompetence or paradigm maintenance
Two explanations are possible, and the evidence is consistent with either.
The first is incompetence. Textbook authors, themselves trained on earlier textbooks that contained the same distortion, simply repeated what they were taught without consulting the original source. The error propagated from generation to generation, each cohort of authors assuming the previous generation's account was accurate. This is plausible. Textbook authors write about dozens of topics. Not every author reads every original paper. The narrative was established early, and inertia carried it forward.
The second explanation is what Kuhn described: the functional depreciation of historical fact. The distortion serves the paradigm. If students understood that the Michelson-Morley experiment does not distinguish between the two theories -- that the choice was made on philosophical grounds, not empirical ones -- they might ask questions. They might investigate. The textbook distortion forecloses this possibility. It tells the student that the question is settled, that the evidence is in, that the ether is dead. It closes a door.
The two explanations are not mutually exclusive. Incompetence and paradigm maintenance can operate simultaneously, each reinforcing the other. Authors who do not know the original source are less likely to challenge the narrative. A narrative that has never been challenged is less likely to be checked against the original source. The result is a self-reinforcing system in which the distortion persists not because anyone actively defends it, but because no one in the chain -- author, editor, reviewer, lecturer -- has the incentive to correct it.
The Duhem-Quine thesis, as established by Pierre Duhem in La Theorie physique (1906) and elaborated by W.V.O. Quine in "Two Dogmas of Empiricism" (The Philosophical Review, Volume 60, 1951, pages 20 to 43), provides the logical framework for understanding why the textbook distortion works. Duhem and Quine showed that individual hypotheses cannot be tested in isolation -- only whole theoretical systems face the tribunal of experience. The Michelson-Morley experiment does not refute "the ether." It refutes a specific cluster of assumptions: the ether + complete stationarity + no contraction + full orbital velocity of the Earth relative to the medium. Change any one of these assumptions -- as Lorentz did, by introducing the contraction derived from electromagnetic theory -- and the cluster survives. The textbooks present the experiment as refuting a single hypothesis (the ether). Logic requires it to be assessed as refuting a cluster. The textbooks ignore this distinction because acknowledging it would undermine the narrative.
The broader implications
The documented distortion of the Michelson-Morley experiment is not an isolated error. It is an instance of a general pattern whose scope can be assessed. If the foundational claim of modern physics education -- the claim every student encounters in the first week -- is a demonstrable logical fallacy sustained for over a century across publishers, countries, and generations, then the same mechanisms of narrative maintenance are available to sustain other distortions. The Lorentz-FitzGerald contraction, derived from first principles by two independent physicists, has been labelled "ad hoc" in every textbook without challenge for decades. The empirical equivalence of two major physical theories has been omitted from every textbook -- not refuted, not argued against, simply omitted. Poincaré's 1905 contributions, submitted before Einstein's, have been rendered invisible. A major experimental programme -- Miller's five years of positive results at Mount Wilson -- has been omitted from the textbook record entirely.
The evidence documented in this chapter does not demonstrate that every aspect of the standard narrative is wrong. It establishes something more precise: that the mechanisms by which a demonstrable falsehood is sustained in physics education are systematic, self-reinforcing, and capable of operating across the entire profession for more than a century. The questions this raises about what else may have been sustained by the same mechanisms are not speculative. They are the logical consequence of the documented facts.
The four non-empirical reasons that determined the 1905 choice -- parsimony, geometric elegance, the path to general relativity, and positivist philosophy -- are legitimate grounds for preference. Reasonable physicists can prefer Einstein's framework for these reasons. What is not legitimate is presenting the choice as though it were forced by experiment. What is not legitimate is teaching students that the ether was "disproved" when it was displaced. What is not legitimate is erasing the alternative from the curriculum so that no student can discover that it exists.
The textbook distortion documented in this chapter established the foundation for everything that followed. It told every physics student, in the first week of studying modern physics, that the ether had been experimentally disproved. It foreclosed investigation before it could begin. It closed a door that was never locked -- because the experiment that supposedly locked it does not distinguish between the two theories it purported to test.
A distortion that persists for over a century, however, requires more than textbook error. It requires that no one of consequence noticed -- that the discipline's most penetrating minds never identified the logical fallacy, never stated publicly that the textbooks were wrong.
This is not what happened. Many noticed. The most distinguished physicists of the twentieth century noticed. Einstein himself, fifteen years after declaring the ether superfluous, said publicly at the University of Leiden in 1920 that space without ether is "unthinkable." Dirac wrote about it in Nature. Bell endorsed a preferred-frame interpretation. Laughlin and Wilczek -- both Nobel laureates -- said so in published books.
The distortion persisted not because no one saw through it, but because seeing through it made no institutional difference. The voices that maintained the ether tradition, and the mechanisms by which they were marginalised, are the subject of the next chapter.