I taste a liqueur finest brewed —  
From French mountains remote —
Not all the world’s Botanicals
Can hit those tasting notes! 

Inebriate of herbs — am I —
And Debauchee of Spleen —
Reeling — thro’ endless summer days — 
From rivers yellow-green — 

When the Pope turns the drunken monks
Out of the alpine flowers — 
When Carthusians — renounce their drams —
I shall but drink empowered! 

Till Rivals find the recipe — 
And Barkeeps — to cocktails run — 
You’ll see this little Tippler
Tasting the pure green — Sun!

With thanks to Emily.

Miller was not some dogmatic or irrational opponent of relativity, but a renowned and well-respected scientist. In fact, he collaborated with Edward Morley, one of the physicists behind the so-called ‘crucial experiment’ supporting relativity theory: the Michelson-Morley experiment. This was an attempt to measure the motion of the Earth relative to the luminiferous aether, a postulated medium for the propagation of light waves. It was thought at the time that light waves must travel through a medium, just as sound waves travels through air. If this were so, then the direction of the Earth’s motion should affect the speed at which the light travels. The experiment’s null result (that is, no relative motion was detected, the speed of light was the same in every direction) was taken to be evidence against aether theories of light. So when Einstein’s special relativity theory came along, where the speed of light is a constant and no medium needed to be postulated to explain its travel, the experiment was often taken as support for his theory.

In the period between 1900-1933, Miller personally and collaboratively conducted hundreds of thousands of tests to try to confirm the reliability of this result. He improved on the apparatus used to measure the relative speed of light, called an interferometer, such that it was the most sensitive instrument of its kind in the world. He tried it out at various sites in different conditions, including atop a mountain peak at Mount Wilson Observatory. Within that time, he also published a manual for physics students on experimental practice which became the standard for years - literally writing the book on proper scientific method for this field of physics.

His findings were hard to swallow. With better instruments, more variables controlled, and more data, Miller consistently detected a small positive effect of the Earth’s motion on the speed of light, which he called the ‘aether drift’. This aether drift was roughly 8-10 km/s. He saw this as an extension of the initial Michelson-Morley result, rather than its contradiction. He claimed that the original ‘crucial experiment’ discounted a small positive effect because of the assumption of a static aether - Miller considered the possibility that the medium of aether was itself in dynamic motion, which would predict the small effects measured.

These findings were taken seriously at first. In a letter, Einstein wrote, “I believe that I have really found the relationship between gravitation and electricity, assuming that the Miller experiments are based on a fundamental error. Otherwise, the whole relativity theory collapses like a house of cards.”1 Luckily for Einstein, when others attempted to repeat Miller’s experiments, the results failed to replicate.2 Many assumed there must be some unknown systematic error afflicting Miller’s results. But this was just an assumption, belied by Miller’s reputation as a careful experimentalist. The English physicist Charles G. Darwin summed up the situation at the time as follows: “We cannot see any reason to think that this work would be inferior to Michelson’s, as he had at his disposal not only all the experience of Michelson’s work, but also the very great technical development of the intervening period, but in fact he failed to verify the exact vanishing of the aether drift. What happened? Nobody doubted relativity. There must therefore be some unknown source of error which had upset Miller’s work.”3

For a while, the scientific community offered explanations for that unknown source of error. The problem is, those explanations were not very good. When Einstein wrote to Miller suggesting that altitude or temperature variation may have been the source of the anomaly, Miller scoffed: “The trouble with Professor Einstein is that he knows nothing about my results (…) He ought to give me credit for knowing that temperature differences would affect the results. He wrote to me in November suggesting this. I am not so simple as to make no allowance for temperature.”4 However, the mainstream consensus is that Miller was mistaken in his analysis of the effect of temperature. In the 1950s, the physicist Robert Shankland declared in an influential report that temperature variation (in addition to random statistical fluctuations) introduced a systematic error in Miller’s data. Although this report was not conclusive, it argued that this was a more plausible explanation of the positive result than aether drift. The scientific community largely accepted Shankland’s judgement, and found justification for why they rejected Miller’s results in the past.5

Polanyi’s argument in Personal Knowledge implies that Shankland’s post-mortem report, whether correct or not, was largely irrelevant to the acceptance of relativity theory. Most physicists were convinced of relativity even in the face of this unexplainable anomaly. Recall the Darwin quote above on the effect of Miller’s results: “Nobody doubted relativity.” This may seem fine, especially since relativity theory turned out to be vindicated, but it should make us pause and think of why this happened and whether it was justified. Is this how science should be done? Shouldn’t doubt, especially doubt prompted by empirical evidence, be a crucial part of the scientific method?

A popular account of the scientific method, and the one most of us learn in school, is roughly the following: form a hypothesis, test with experiment, analyze results, confirm/refute hypothesis. A more philosophically respectable version of that account, like Karl Popper’s, would say that science proceeds through conjecture and refutation - an iterative process of eliminating false hypotheses and theories through empirical testing. These accounts tend to emphasize experimental results above all, and say that every scientist should be ready to reject a theory (or at least, suspend judgement) in the face of experiment. To a large extent, public trust in science as an objective source of knowledge is a result of these accounts seeping into public consciousness. We like to think of scientists as detached, impartial observers of the phenomena, and these theories of scientific method reinforce that impression.

Miller could not have conceived of himself as such, but he was a dedicated Popperian. He remained a defender of aether theories until his death, because he trusted the empirical results that his own experiments showed him. According to our popular accounts of scientific method, he did everything right. His results were a significant anomaly for relativity theory, but one which most physicists hastily explained away, without seriously reckoning with the evidence.

To this day, there remains a minority group of dissident scientists who believe Miller’s results falsify relativity.6 The mainstream of physics simply assumes his results are invalid along the lines of Shankland’s report, but, as mentioned before, that report was not conclusive.7 Naturally, all of this is fodder for conspiracy theories and those seeking to tear down the cult of Einstein for political reasons. Miller is the perfect hero of this story, a victim of Einstein’s domination. And he is legitimately a tragic figure; his experiments were the best of their kind, while worse versions were heralded as evidence for relativity theory.

What I want to suggest in this short post is that our predominant conceptions of scientific method make it difficult to argue against those dissident scientists and conspiracy theorists. In order to do so effectively, we need a different conception of science. Following Polanyi, I would argue that it is a mistake to tie scientific method and objectivity to solely formal and empirical methods. The truth is, science hardly ever proceeds through strict falsification criteria or immediate deference to empirical results. As Polanyi argued, “The experience of D.C. Miller demonstrates quite plainly the hollowness of the assertion that science is simply based on experiments which anybody can repeat at will.”8 This is not to deny the important role of experiment, but to acknowledge that scientific practice often can and should depart from merely considering experiments.

This point can be shown by considering how Einstein actually arrived at his special theory of relativity. In textbook histories of relativity theory, the Michelson-Morley experiment’s null result is often presented as the origin of Einstein’s discovery, the ‘crucial experiment’ which prompted him to develop the theory. This gels with a strongly empirical or Popperian view of science as driven by experimental conjectures and refutations. But it’s no wonder, then, that the plight of Miller should be seen as biased or conspiratorial suppression! For Miller’s experiments were better than the Michelson-Morley one, and it was even arguably consistent with an aether drift. And yet, all of his work was disregarded and subsequently forgotten.

But this may all be a red herring. In truth, there is good evidence that Einstein barely considered the experiment in his discovery of relativity theory. Instead, he was primarily motivated by theoretical and aesthetic values. Famously, he relied on thought experiments, like imagining what an observer would see of a light beam if they were also traveling at the speed of light. He was also influenced by Ernst Mach’s philosophical critique of absolute space and time. Finally, perhaps the greatest impetus of his discovery was the theoretical asymmetry in Maxwell’s electrodynamics, which framed the seminal 1905 paper introducing the special theory. In the genesis of the theory, there was little to no discussion of the aether experiments, or any experiments at all.9

These theoretical, rather than experimental, arguments were not just convincing to Einstein. Relativity theory was beautiful and coherent, and this was enough to convert many of the old guard to his side, even in the face of anomalies. Polanyi acknowledges that, now, relativity theory is among the most experimentally confirmed theories in history. But this misses the point: “such verifications of relativity are but confirmations of the original judgment of Einstein and his followers, who committed themselves to the theory long before these verifications.”10 He makes the controversial claim that Einstein and his followers were right to reject Miller’s results, for theoretical and aesthetic values sometimes trump empirical ones - and this deserves to be called legitimate, objective scientific method. Regardless of whether you’re willing to go as far as Polanyi in this argument, it is undeniable that this case shows scientific method is much messier in practice than orthodox views would have us believe. His conclusion still strikes home:11

“The triumph of the Michelson-Morley experiment, despite its giving the wrong result, the tragic sacrifice of D. C. Miller’s professional life to the pursuit of purely empirical tests of a great theoretical vision, are sardonic comments on the supposed supremacy of experiment over theory.”

So what is the best way to respond to the dissident scientists and conspiracy theorists? We should admit that Miller falsified relativity theory, according to the standards of the time. But this will require us to rethink our characterizations of scientific confirmation and explanation, of scientific method in general. We have to acknowledge that experiment isn’t everything. Other values like coherence, simplicity, fruitfulness, explanatory power, perhaps even intuition and aesthetics, all bear on theory confirmation and acceptance, and are all part of a well-functioning, rational scientific method.