In 1805, the French mathematician Adrien-Marie Legendre published a book on the orbits of comets. Buried in an appendix was a mathematical technique he called the “method of least squares,” a procedure for fitting a curve to a set of data points by minimizing the sum of the squared differences between the observed values and the predicted values. The method was elegant, practical, and immediately useful. Astronomers, surveyors, and physicists adopted it within years.
Four years later, in 1809, Carl Friedrich Gauss published his own treatise on celestial mechanics, Theoria Motus Corporum Coelestium. In it, he presented the method of least squares and claimed that he had been using it since 1795, a full decade before Legendre’s publication. Gauss offered no published proof of this claim. He simply stated it as fact.
Legendre was furious. What followed was one of the most bitter priority disputes in the history of mathematics, a conflict that poisoned the relationship between two of the greatest mathematical minds of the 19th century and raised questions about scientific credit that remain relevant today.
What Is the Method of Least Squares?
The method of least squares is one of the most widely used techniques in all of science. Its purpose is simple: given a set of observations that contain measurement errors, find the mathematical curve or line that best fits the data.
Suppose you are an astronomer tracking the position of a comet across the sky. Each observation contains small errors due to atmospheric distortion, instrument limitations, and human imprecision. If you have more observations than the minimum needed to determine the orbit (which you should, to reduce the effect of errors), the observations will not perfectly agree. The method of least squares finds the orbit that minimizes the total squared error across all observations.
Why squared errors? Because squaring ensures that positive and negative errors do not cancel each other out. A measurement that is two units too high and one that is two units too low would average to zero error, hiding the fact that both measurements were wrong. Squaring both errors (giving four in each case) reveals the true magnitude of the discrepancy.
The method is the foundation of regression analysis, which is itself the foundation of modern statistics, econometrics, machine learning, and virtually every field that fits models to data. When a scientist draws a “best fit line” through a scatter plot, they are almost certainly using least squares. When a machine learning algorithm trains on data, it is often minimizing a squared error function that descends directly from the technique that Legendre published and Gauss claimed.
Legendre’s Publication (1805)
Legendre presented the method of least squares in his 1805 book Nouvelles méthodes pour la détermination des orbites des comètes (New Methods for the Determination of the Orbits of Comets). The relevant section appears in an appendix titled “Sur la Méthode des moindres quarrés” (On the Method of Least Squares).
Legendre’s presentation was clear, practical, and complete. He described the principle, derived the key equations, and demonstrated the method’s application to astronomical data. He did not claim a long history of private use. He presented it as a new technique, gave it a name, and published it for the world to use.
The mathematical community received the method enthusiastically. It solved a genuine problem that astronomers and geodesists had been struggling with for decades: how to extract the best possible estimate from a collection of imperfect measurements. Within a few years, the method of least squares was being used across Europe.
Gauss’s Claim (1809)
When Gauss published his Theoria Motus in 1809, he included a treatment of least squares that was mathematically deeper than Legendre’s. Gauss connected the method to probability theory, showing that if measurement errors follow a normal (bell curve) distribution, then the method of least squares gives the most probable values for the unknown parameters. This was a significant theoretical advance.
But Gauss also claimed priority. He wrote that he had been using the method since 1795 and that his successful prediction of the orbit of the asteroid Ceres in 1801 (which had made him famous across Europe) had relied on it. He offered no published evidence for this claim, only his word and the testimony of a few colleagues who said they had seen him use the technique in private correspondence.
The claim was plausible. Gauss was an extraordinary prodigy who had made major discoveries as a teenager, and his prediction of Ceres’s orbit was so accurate that it would be difficult to explain without a sophisticated fitting method. Several of Gauss’s private letters and notebooks do contain references to least squares from before 1805. But none of this material was published or publicly available at the time.
Legendre’s Fury
Legendre took Gauss’s claim as a direct assault on his priority. In a letter to Gauss, Legendre wrote bitterly about the injustice of having his published discovery claimed by someone who had never made the method public. “I do not dispute that you had the same idea before me,” Legendre wrote, “but the one who publishes first acquires the right.”
Legendre had a point. Scientific priority is traditionally assigned to the first person to publish, not the first person to think of an idea. Private notebooks, letters to friends, and unpublished manuscripts do not establish priority in the formal sense. Gauss’s habit of withholding results until they were “ripe” (his word) was well known and had caused priority conflicts before. He had done the same with non-Euclidean geometry, the fast Fourier transform, and several results in number theory, in each case developing ideas privately and then watching others publish them first.
The dispute was never formally resolved. Gauss never apologized or withdrew his claim. Legendre continued to assert his priority in subsequent publications. The two mathematicians, who had previously maintained a cordial professional relationship, became permanently estranged.
Who Deserves Credit?
Modern historians generally agree on the following. Gauss almost certainly did use the method of least squares before Legendre published it. The evidence from his notebooks and correspondence is convincing, and his prediction of Ceres’s orbit is difficult to explain otherwise. However, Legendre published first, named the method, and made it available to the scientific community. Without Legendre’s publication, the method might have remained in Gauss’s private notebooks for years or decades longer.
The fairest assessment is that both mathematicians deserve credit: Gauss for the earlier discovery and the deeper theoretical foundation, Legendre for the publication and dissemination. Science depends on both invention and communication. An idea that remains private, however brilliant, does not contribute to the collective enterprise of knowledge until it is shared.
The Pattern of Priority Disputes
The Gauss-Legendre conflict is not unique. Priority disputes have been a recurring feature of scientific history, often arising when two brilliant minds independently converge on the same idea.
- Newton vs Leibniz: the most famous priority dispute in all of mathematics, over the invention of calculus. Newton developed his method of “fluxions” earlier, but Leibniz published first. The resulting controversy consumed decades and damaged the mathematical communities of both England and continental Europe.
- Darwin vs Wallace: both Charles Darwin and Alfred Russel Wallace independently developed the theory of evolution by natural selection. Their joint paper was presented to the Linnean Society in 1858, but Darwin’s more comprehensive On the Origin of Species (published in 1859) received most of the credit.
- Hilbert vs Einstein: David Hilbert and Albert Einstein both worked on the field equations of general relativity in November 1915. The question of who reached the final equations first has been debated ever since.
These disputes reveal something important about the nature of scientific discovery. Major advances are often “in the air” at a particular moment, when the necessary tools, data, and theoretical frameworks have reached a point where the next step becomes almost inevitable. Independent discovery is not a coincidence; it is a sign that the time is ripe.
Gauss’s Private World
Gauss’s role in the least squares dispute reflects a broader pattern in his career. He was famously secretive about his work, publishing only a fraction of what he discovered. His private notebooks, which were not fully examined until decades after his death, contained results that would have earned any other mathematician lasting fame.
Kronecker Wallis’s edition of Gauss’s Selected Visual Notebooks offers a direct window into this private mathematical world. The collection presents five of Gauss’s manuscripts, vectorized line by line from the originals at the University of Göttingen, including the famous diary entry marked EUREKA that records his proof about triangular numbers. Bound in Japanese stab binding with waxed linen thread, these notebooks reveal the working methods of a mind that operated far ahead of its contemporaries but chose to keep much of its output hidden.
The mathematical tradition that both Gauss and Legendre inherited begins with Euclid’s Elements, the text that established the standard of rigorous mathematical proof that every mathematician since has followed. The method of least squares, like all of modern mathematics, rests on the axiomatic foundations that Euclid laid down over two thousand years ago.
An Unresolved Question
The Gauss-Legendre dispute remains instructive because the question it raises has no easy answer. Should scientific credit go to the person who has the idea first, or to the person who shares it first? Gauss believed that ideas belonged to their creators. Legendre believed that ideas enter the public domain through publication. Both positions have merit, and the tension between them continues to shape debates about scientific credit, intellectual property, and the ethics of discovery.
What is not in dispute is the importance of the method itself. The method of least squares is used billions of times every day, in every field from astronomy to economics to artificial intelligence. It is one of those rare mathematical inventions that is both profoundly simple and endlessly useful. Whether you credit Gauss, Legendre, or both, the method they gave the world remains one of the most consequential tools in the history of science.
No products in the cart. 