MYTH 18

THAT THE MICHELSON-MORLEY EXPERIMENT PAVED THE WAY FOR THE SPECIAL THEORY OF RELATIVITY

Theodore Arabatzis and Kostas Gavroglu

Michelson and Morley found that the speed of the earth through space made no difference in the speed of light relative to them. The inference is clear … that all observers must find that their motion through space makes no difference in the speed of light relative to them. The above inference was clear, at least to Einstein, who … made it a cornerstone of his special theory of relativity.

—James Richards et al., Modern University Physics (1960)

What exactly do we mean by “myths in science”? Often we mean the propagation of stories that are at odds with the historical record—be it because their protagonists have specific views on how science has (or ought to have) developed or because teachers and textbook writers find them educationally expedient. When encountering such stories, historians of science have an obligation to dispel them by setting the record straight. In this chapter, however, we are interested in less blatant myths, the main characteristic of which is not the distortion of historical evidence. Rather, they slowly establish themselves by an appropriation of various aspects of the historical record, simplifying and transforming them for pedagogical, ideological, or philosophical aims. Within such a framework, we would like to examine the character of a rather idiosyncratic myth: the close—almost causal—connection that is stated in many textbooks, monographs, and popular writings between the null result of the Michelson-Morley (M-M) experiment and the emergence of the special theory of relativity (STR).

In the late nineteenth and early twentieth centuries, many well-known physicists were facing a rather awkward situation. A highly sensitive experiment, performed by Albert A. Michelson (1852–1931) and Edward W. Morley (1838–1923) in 1887 and designed to measure the relative velocity of the earth with respect to the ether, gave consistently null results. The ether was the necessary substratum of James Clerk Maxwell’s (1831–1879) electrodynamics, since its prediction of the existence of electromagnetic waves required a medium through which they would travel. With the benefit of hindsight, the null result could have meant that the ether was not needed for the propagation of electromagnetic waves, but this was not something that could be entertained within the framework of nineteenth-century physics. All the explanations given for the null result were problematic or artificial: either the ether would have to possess contradictory properties or it would have unheard-of effects on the forces holding matter together, causing the contraction of material bodies along the direction of their motion.

The connection between the M-M experiment and the emergence of STR has been exposed as a myth, most notably by Gerald Holton (b. 1922), who made much of the fact that in Albert Einstein’s (1879–1955) original 1905 paper on STR there is no explicit reference to the M-M experiment. Holton proceeded to a systematic examination of the origins of STR, concluding that the “role of the Michelson experiment in the genesis of Einstein’s theory appears to have been so small and indirect that one may speculate that it would have made no difference to Einstein’s work if the experiment had never been made at all.”1 Holton’s assessment remained the same after new material surfaced in later years: “the influence of the famous experiment was neither direct and crucial nor completely absent, but small and indirect.”2 His influential and elegant debunking of that myth has stood the test of time and has been further refined by others who also examined additional archival material.³

Holton characterized the myth as a textbook account of the birth of relativity theory, heavily influenced by the positivist philosophy of science, which inflated the role of the M-M experiment as Einstein’s main motivation. It is, indeed, the case that textbook writers have taken some liberties with Einstein’s 1905 paper—and, as will be discussed further, they did the same thing with Max Planck’s (1858–1947) 1900 paper on blackbody radiation as well as Niels Bohr’s (1885–1962) 1913 papers on atomic structure. However, Einstein was not all that innocent in the development of the myth. Even though in his 1905 paper he remained silent about the M-M experiment, beginning in 1907, in a review article appearing that year, he was quite vocal and sometimes contradictory about the significance of that experiment.⁴ Historians have been obliged to discuss this problem within the quagmire brought about by Einstein himself, not because of his silence about the role of this experiment in the formulation of the STR but because of his ambivalence about it for the rest of his life.

This qualifies as an idiosyncratic myth for two reasons: first, even though in the original paper of 1905 there is no mention of an explicit connection between the M-M experiment and the formulation of the STR, Einstein in subsequent papers, speeches, books, and interviews mentioned the M-M experiment and its positive role in the emergence and acceptance of the STR. Second, although historians of science do not altogether deny this connection, they do not consider the role of the null result to be as important as many writers of textbooks, monographs, and popular texts make it out to be, thus relegating the myth to something that is only partially false. Remarkably, there is virtual unanimity among Einstein scholars about the role of the M-M experiment in the creation of the STR: if it did play a role, it was by no means significant.5 Our task, then, is less to set the record straight than to examine the character and implications of the M-M myth.

The history of the relation between the M-M experiment and the STR has an aspect not sufficiently emphasized by Holton.⁶ In 1907, Einstein, in the aforementioned review article on “The Relativity Principle and the Conclusions Drawn from It,” made repeated references to the M-M experiment and emphasized that it suggested “the simplest possible assumption,” that is, the principle of relativity.⁷ Holton mentions this review article only in one footnote and refers to Einstein’s allusions to the M-M experiment with the cryptic phrase that “here again we find a sequence of sentences which can be considered implicit history.”⁸ It is also worth noting that in 1911, Max von Laue (1879–1960)—who had visited Bern to meet Einstein while he was still in the patent office and who wrote the first monograph on relativity—claimed that the M-M experiment was “the fundamental test for the theory of relativity,” thus giving a green light to subsequent textbook writers for underlining the importance of the M-M experiment in the formulation of the STR.⁹ In view of these statements, which may have set the tone for subsequent textbook writing, the “textbook myth” of the origins of relativity starts to look justifiable, especially within the framework of the pedagogical aims of textbook writers.

It appears, then, that the authors of scientific texts and popular writings “accused” of creating and propagating a myth have very little to be apologetic about. They could have a perfectly legitimate answer: we have just been following what the master himself said and wrote on various occasions, such as in his 1907 review article and in his excursions into the history of physics.¹⁰ In fact, the situation is even more complicated. Einstein was, throughout his life, of two minds regarding the role of the M-M experiment in the origins of STR. He appeared undecided as to what the true story was, at times denying that he knew about it and at times giving it a decisive role; thus, he created an ideal framework for the perpetuation of the myth.

Might it be the case, then, that it is the historians of science who are off the mark in debunking a nonexistent myth? Might it be the case that they are, at least partly, responsible for creating another myth in response to the myth they attribute to textbook writers? The answer is yes and no. To the extent that textbook authors and popularizers have overblown the significance of the M-M experiment for the genesis of the STR, the answer is yes, these authors have created a myth. To the extent that they have tried to reconstruct, in a pedagogically plausible way, the origins of STR, focusing on most of what Einstein wrote about it after 1905, the answer is no.

This is basically the situation regarding the myth involving the M-M experiment and the origins of the STR. However, most textbook authors have contributed decisively to the creation of another myth, which has shaped rather strongly the historical consciousness of many generations of physicists: in many textbooks, the chapter on relativity is part of the treatment of mechanics and rarely covers topics from electromagnetism. The STR is presented as the “correct” mechanics, which reduces to Newtonian mechanics when the speed of light approaches infinity. This is what physics students have been taught, and this is the “feeling” of the overwhelming majority of physicists. This, then, is a “real” myth, having very little to do with Einstein and the history of the STR. Einstein formulated the STR in response to problems in electrodynamics, and one of his main aims was to get rid of “asymmetries” in electromagnetic theory. Furthermore, when he chose to speak of the M-M experiment, Einstein emphasized, time and again, its significance for supporting the principle of relativity—not the constancy of the speed of light. But in many textbooks one finds a different inference: that the M-M experiment demonstrated the constancy of the speed of light regardless of the motion of its source.

Take, for instance, two textbooks that have helped educate hundreds of thousands of university students and have provided a basic knowledge and the overall framework of modern physics. They have been repeatedly revised, reprinted, translated, and used around the world. In both of them, the STR is introduced toward the end of the section on mechanics, with no reference to electrodynamics, and the null result of the M-M experiment is related to Einstein’s postulate about the speed of light:

The null result of the Michelson-Morley experiment to detect the drift of the earth through an ether … can only be understood by making a revolutionary change in our thinking; the new principle we need is simple and clear: The speed of light is independent of the motion of the light source or receiver.11

The result of the Michelson Morley experiment … shows that the velocity of light is the same in all directions on the moving earth.¹²

Here, then, are the ingredients of another myth that has been systematically cultivated by textbook writers: (1) that the M-M experiment proved the constancy of the velocity of light, regardless of the velocity of its source; (2) that the M-M experiment refuted the ether hypothesis; (3) that special relativity was developed in the context of mechanics rather than electrodynamics; and (4) that special relativity was a generalization of Newtonian mechanics.

In such physics textbooks, the M-M experiment is presented in the context of “pedagogical reconstructions” of relativity theory, the main aim of which is to facilitate the comprehension of the theory and to present a plausible and straightforward account of its origins (or its evidential base). Of course, these reconstructions are not historiographically innocent. To serve their purpose, they inevitably oversimplify and often distort the historical record. Furthermore, these reconstructions are framed in terms of an intuitively plausible “schema” of how new theories are born and come to be accepted, which consists in “putting the phenomena first”: new theories are motivated by puzzling phenomena that cannot be explained by those theories’ predecessors.¹³ Besides accounting for puzzling phenomena, new theories always include their predecessors as limiting cases, which hold within a well-circumscribed domain. Thus, the continuity and progress of scientific change is secured.

Such a schema often captures the way all of modern physics is presented in textbooks. It expresses the collective aim of textbook writers to articulate a “canonical” textbook historiography. Consider, for instance, Planck’s 1900 paper, in which he is preoccupied with the second law of thermodynamics and entropy. He discusses the experimental measurements for blackbody radiation and the difficulties in providing a satisfactory theoretical explanation for a wide range of the emitted frequencies; he then proceeds to derive Wien’s law, announced in 1893 by the German physicist Wilhelm Wien (1864–1928), who proposed it in an empirical manner. Nowhere in the paper is there any mention of the Rayleigh-Jeans law for low frequencies. However, textbook accounts of how Planck introduced his theory suggest a different approach: They first present blackbody radiation and discuss Wien’s law and its success in accounting for high frequencies. They then discuss the Rayleigh-Jeans formula for low frequencies. And they finally present Planck’s work as providing a unified account of the whole spectrum of blackbody radiation. Besides the fact that the Rayleigh-Jeans law is not mentioned in Planck’s paper, James Jeans’s (1877–1946) improved version of Lord Rayleigh’s (1842–1919) derivation did not appear until 1905.14

Similar problems beset textbook accounts of Bohr’s model of the atom. Bohr, in the first part of his 1913 trilogy “On the Constitution of Atoms and Molecules,” starts with the paradox that within the framework of classical electrodynamics, Ernest Rutherford’s (1871–1937) model of the atom cannot account for the stability of matter. This is what Bohr tries to alleviate, and only at the end of his paper does he show that his model, which guarantees the stability of atoms, also provides a very satisfactory account of the hydrogen spectrum. Textbooks, however, provide a different picture. Almost invariably they start with the empirical Paschen-Balmer formulas for the hydrogen spectrum, and present Bohr as aiming directly at their explanation.

Ideas explicitly expressed in Einstein’s, Planck’s, and Bohr’s original papers are “tampered with” in order to fit into the textbook culture of pedagogical expediency. The works of Einstein, Planck, and Bohr are presented as being motivated by their wish to explain experimental results that had been either unaccounted for or explained in an unsatisfactory manner. In textbook accounts, Einstein appears dissatisfied with the Lorentz-Fitzgerald explanation of the M-M experiment; Planck not satisfied with the lack of explanation for the Wien and the Rayleigh-Jeans laws; and Bohr not content with the hitherto unexplained Paschen-Balmer series. It appears that textbooks have followed a coherent historiographical viewpoint, which, while differing in its details, has an incredibly strong grip on both writers and readers: the heroes of modern physics are those who provided ingenious explanations of recalcitrant experimental results.