In a morning digest, I came across a story I initially took for one of those “everything was stolen from a poor genius” yarns. It was about a 1927 paper: a Belgian abbé had concluded that the Universe was expanding, calculated the rate of that expansion, and then, two years later, Hubble published the same thing—and the whole world named the discovery after him. The 1931 English translation, which ought to have corrected the injustice, came out shorter than the original: precisely the paragraphs calculating the expansion rate had vanished. For eighty years, historians of science searched for whoever had struck out those lines.
But there was a second half to the story, and that was what really snagged me. I already knew about the text—I had a breakdown of the “missing paragraphs” published in the neighboring column just days ago. It turned out the question was far more interesting: why was the relationship derived in 1927 called Hubble’s law for nine decades, while the second man’s name appeared in it only in 2018—and even then by a vote of astronomers, not by force of habit? As it turns out, a law of physics gets its name not where it is derived, but where someone votes for it.
Before we get into the marriage of labels, it is worth understanding what exactly the disputants were measuring. In the 1920s, astronomers saw the same oddity in the spectra of distant galaxies: the lines were shifted toward the red, and the fainter and more distant the object appeared, the stronger the shift. There was no classical explanation for it. The redshift might have been explained by the galaxies’ own motion—like the whistle of a receding train, only stretched to astronomical scale—but the resulting velocities were implausibly enormous.
The spectral shifts of these objects were collected by the American astronomer Vesto Slipher, working at Lowell Observatory. By then, theorists had noticed that in de Sitter’s empty Universe—a world with no matter whatsoever, where distant bodies nevertheless appear to be fleeing because of the curvature of space—such a picture emerged on its own. It was elegant and nearly meaningless: an empty Universe fit poorly with the fact that there were, after all, galaxies all around us.
Then came the Belgian priest and cosmologist Georges Lemaître. He trained under Eddington in Cambridge, attended Shapley’s lectures at Harvard, and at MIT studied everything that would soon become observational cosmology. In 1927, he took the empirical data and did what no one had thought to do: he read redshift not as the actual motion of galaxies, but as the stretching of the very fabric of space between the moment light was emitted and the moment it was caught. This was the expanding Universe—not as a metaphor, but as a property of the metric, following directly from Einstein’s equations.
Lemaître arrived at specific numbers. He relied on a list of 42 radial velocities, compiled by the Swedish astronomer Gustaf Strömberg at that same Mount Wilson Observatory—and on the empirical “distance–absolute magnitude” formula formulated before Hubble by his colleague Hopmann. Depending on how the data were weighted, the coefficient of proportionality in his formula came to 625 or 575 kilometers per second per megaparsec (without weighting—up to 670). Lemaître did not hesitate to name the radii of the Universe as well: an initial 2.7×10⁸ parsecs and a present-day 6×10⁹ parsecs.
Two years later, Edwin Hubble did the same thing with data for 24 galaxies and nine groups of galaxies, arriving at 465±50 and 513±60 km/s per megaparsec. In effect, it was the same law with the same constant. The difference was not in the mathematics, but in presentation: Lemaître’s paper appeared in French in the low-circulation Annals of the Scientific Society of Brussels, while Hubble published in the Proceedings of the National Academy of Sciences—and worked with the largest telescope in the world.
Then comes the part that turned scientific history into a detective story. In 1931, the Monthly Notices of the Royal Astronomical Society published an English translation of Lemaître’s paper—the very translation that should have brought the work out of the shadows. But the translation was shorter than the original: the table of observational data was missing, and, crucially, so were the calculations of the expansion coefficient. To historians, it looked almost like an established fact—someone had been bothered by the fact that the constant named after Hubble had been calculated earlier, by somebody else.
For four decades, the editors of MNRAS were under suspicion. Others pointed toward Hubble himself, a man with a reputation for fiercely guarding priority—he wrote plainly to de Sitter that he considered the velocity-distance relation a contribution specifically of Mount Wilson and was deeply interested in having it recognized as such. Then people dug through papers for another twenty years, and in 2011 the astrophysicist Mario Livio of the Space Telescope Science Institute combed through hundreds of letters from the Royal Astronomical Society, minutes of its meetings, and the Lemaître archive—and found the answer.
The lines had been struck out by Lemaître himself. His letter to the editors preserves the sentence: “I did not find it advisable to reprint the provisional discussion of radial velocities which is clearly of no actual interest, nor the geometrical note, which could be replaced by a short bibliography of ancient and new works on the subject.” In other words, the man personally removed the numerical proof of his own priority from the translation—because by 1931 Hubble had more precise data, and Lemaître saw no point in repeating his earlier, provisional conclusions.
This is perhaps the most unexpected part of the whole story. It has no villain. It has a man for whom the question of “whose law” meant precisely nothing.
A separate thread is how those we are accustomed to regarding as the fathers of cosmology reacted to expansion. Lemaître met Einstein in 1927 at the Solvay Congress in Brussels. He himself recalled the scene: as Einstein walked along the avenues of Leopold Park, he told him that, from the point of view of physics, the idea seemed absolutely abominable to him. In the taxi on the way to the university, the conversation about nebular velocities only reinforced Lemaître’s impression that Einstein did not understand astronomical phenomena especially well—he was a genius of gravitation, not an observer of the sky.
Einstein had his own drama behind him. In 1917, he introduced the cosmological constant into the equations—an extra term to keep the Universe from collapsing in on itself, so that the world would remain eternal and motionless. When Friedmann showed in 1922 that the same equations had non-stationary solutions, Einstein first wrote that Friedmann’s results seemed “suspicious” to him, and a year later admitted an error in his calculations—but immediately added that it was “hardly possible to give a physical meaning” to dynamic solutions. Stubbornness worthy of an essay in its own right.
Lemaître, however, had no better luck elsewhere. He said that when he spoke to Einstein about the primeval atom—the idea that the Universe had a beginning—Einstein stopped him: “No, this is no good, this suggests creation too much.” A few years later, in Pasadena, their conversations continued, always circling that same lambda which Einstein had “cleverly introduced into the equations, but with which he was never satisfied and which he tried to withdraw.” Journalists, overhearing these walks, turned the “little lambda” into a “little lamb,” supposedly trailing after them everywhere.
The historical anecdote about “Einstein’s greatest blunder”—that he supposedly crossed out lambda after learning of expansion—turns out, on inspection, to be heavily tidied up. Lambda did not vanish; it returned at the end of the 20th century, when it became clear that expansion was accelerating. So the man who was publicly wrong twice in this story ultimately turned out to be partly right in precisely the thing for which he was mocked.
Formally, Lemaître’s priority had been obvious since 1927, but no one rushed to rename anything. Science went on as usual: in 1930 Eddington rediscovered the model of the expanding Universe for himself, found that Einstein’s static world was as unstable as a pencil balanced on its point, and publicly apologized to Lemaître for not having read his work in time. It was Eddington who sponsored the English translation of the 1927 paper in Monthly Notices. The term “Eddington–Lemaître model” entered circulation, but the constant in the equation remained Hubble’s all the same.
This went on for nearly nine decades. Then the International Astronomical Union suddenly decided that the name in the law needed correcting. Resolution B4, adopted in 2018, proposed renaming Hubble’s law the Hubble–Lemaître law. The vote was electronic—one code for each individual member of the union. A total of 11,072 codes were issued; 4060 people voted.
The result—78% in favor, 20% against, 2% abstaining. Turnout: 36.7%. It was announced on 29 October 2018.
This is where the most interesting thing is hiding. First, note the wording of the resolution itself: “on a suggested renaming of the Hubble Law”—suggested renaming. This was not an order to astronomers, but a recommendation. No IAU resolution can force a textbook author to write a double name if they do not want to. Second, they were not voting on who first derived the relationship—that fact was common knowledge by 2018 and disputed by no one. They were voting on whether the law should bear a second name, and 798 astronomers said no.
I think that is the real anomaly, the reason it was worth digging. Science is built like a machine for producing verifiable statements: data must be reproducible, conclusions must follow from calculation. But this machine has a layer that follows from no calculation whatsoever and yet decides what we will remember. Names, priorities, labels on constants—all of it exists in another mode: bulletins, reputation, habit, whose telescope is bigger.
Eighty years between two numbers is not a span during which science “failed to notice” an error. It is a span during which the machine could formally have assigned the name correctly several times and each time chose not to. In 1930, Eddington apologized, but did not rename the constant. In 1984, historians noticed the missing paragraphs, but convinced no one. In 2011, Livio proved there had been no saboteur—and the question of the name only grew sharper. Finally, in 2018, an assembly decided that the law would now have two names.
My personal irony is that the decision was adopted by a vote of a minority of a minority: at the time of the vote, four thousand people out of eleven thousand eligible voters spoke in favor of a law describing the entire observable Universe. Thirty-seven percent. This is not a reproach—it is a figure that says more about the nature of the procedure itself than any article could. No one compared Lemaître’s and Hubble’s data in the vote: they compared reputation and fairness.
Incidentally, about the data. If the chronology is to be believed, Hubble never read Lemaître’s paper. He explained redshift as a purely Doppler effect—that is, actual motion of galaxies—and remained skeptical of the relativistic interpretation to the end of his life. And the myth that Hubble discovered expansion was first set in motion by his colleague Humason as early as 1931. The quality of a good label is this: it sticks to the person who talked loudest about it, not the one who calculated it first.
So the next time you read “the Hubble–Lemaître law,” it is useful to remember that its double name is not the result of someone finally studying the primary sources. It is the result of a ballot in which seventy-eight percent of thirty-seven percent of those assembled decided that a law could be renamed if one asked long enough and officially enough. The calculations and drafts, meanwhile, had been studied forty years earlier—and all that time the constant had calmly carried a single name.
The ending of the story, if you look for a moral in it, sounds dry: fairness in naming scientific laws does not follow from the fairness of the discoveries themselves. Lemaître’s idea was more correct and earlier—but its label was issued only after people voted for it. What we call knowledge rests on measurements. What we call memory rests on procedure. They are different things, and to confuse them is to expect from a ballot what only a telescope can provide.