In 1905 a 26 year old clerk with no academic post published four papers that changed physics. Albert Einstein was then working as a technical expert at the Swiss patent office in Bern, judging inventions by day and thinking about physics in his spare hours. One of those papers, on special relativity, is famous. Another, on light, was the one Einstein himself told a friend was truly revolutionary, and it took the physics world nearly twenty years to accept it.

A heuristic point of view

The paper, published in March 1905, carried the cautious title "On a Heuristic Point of View Concerning the Production and Transformation of Light." A heuristic is a guide to discovery rather than a finished theory. Einstein did not start from Planck's formula. He studied the entropy of light at high frequencies and found that dilute radiation's entropy behaved like that of a gas of separate particles. From this he drew a bold conclusion. In an English translation of the paper, he proposed that the energy of a light ray "consists of a finite number of energy quanta which are localized at points in space, move without dividing, and can be absorbed or generated only as a whole."

This went far beyond Planck. Planck had quantized the way matter gives off and takes in energy. Einstein suggested that light itself traveled in lumps.

The photoelectric puzzle

Einstein used his idea to explain several odd experiments, the best known being the photoelectric effect, in which light knocks electrons out of a metal. Philipp Lenard's experiments of 1902 had shown something strange: the energy of each freed electron depended on the color of the light, not on its brightness. Brighter light freed more electrons, but not faster ones. Waves could not easily explain this. Lumps of light could: each quantum hands its whole energy to one electron, and the energy of a quantum depends on frequency. Einstein's account made a sharp prediction about how electron energy should rise with frequency.

A fusion of wave and particle

Einstein knew the problem with his idea. Interference and diffraction, the patterns light makes passing through slits, were explained beautifully by waves. Particles seemed unable to produce them. In a talk at Salzburg in 1909, he predicted that the next phase of theoretical physics would bring a theory of light fusing the wave and particle views, "a kind of fusion of the wave and the emission theory," as he put it. It was a remarkably accurate forecast, though he would later dislike the form the fusion took.

Nearly everyone said no

For years almost no leading physicist accepted light quanta. In 1913, when Planck and others recommended Einstein for the Prussian Academy of Sciences, they praised him warmly but asked colleagues to forgive him for overshooting in some speculations, naming light quanta as the example.

The American Robert Millikan spent years testing the photoelectric prediction, expecting to refute it. His results, published in 1916, did the opposite. Millikan confirmed Einstein's equation while still doubting the physical picture behind it, calling it a "bold, not to say reckless" hypothesis. Einstein received the Nobel Prize in Physics for 1921, cited for the photoelectric law, while the light quantum itself remained controversial.

Compton tips the balance

The turn came in 1923. Arthur Compton, at Washington University in St. Louis, fired X-rays at carbon and measured the X-rays that scattered off. They came out with longer wavelengths, having lost energy. Compton showed that the numbers matched a collision in which X-rays bounce off electrons like particles carrying momentum, just as one billiard ball strikes another. Peter Debye reached the same analysis independently.

Some physicists still resisted. In 1924 Niels Bohr, Hendrik Kramers and John Slater proposed a way to keep light as waves, at the price of giving up strict conservation of energy in each single event, keeping it only on average. Experiments by Walther Bothe and Hans Geiger, and by Compton and Alfred Simon, tested individual collisions in 1925 and found energy conserved every time. The proposal collapsed. The word itself arrived a year later: Gilbert Lewis coined the name "photon" in 1926, although he meant something rather different from what the word came to mean.

Worked example: how a minority view won

Trace the light quantum from 1905 to 1925 as a case study in how evidence persuades. In 1905 it rested on an argument from entropy and a few loose experiments, and the experts reasonably held back, since waves explained so much. In 1916 Millikan's careful data confirmed its prediction, yet the experimenter himself still rejected the picture, showing that a confirmed equation does not force anyone to accept the story behind it. In 1923 Compton's scattering gave a direct, quantitative sign of particle momentum. In 1925 the last serious rival, which required giving up exact energy conservation, failed a clean test.

No single experiment did the work. What changed minds was a growing set of independent results, each hard to explain any other way, and the failure of the best alternative. Those who resisted were not fools. They were weighing a radical idea against a theory that had worked brilliantly for a century. The lesson applies well beyond physics: a new claim earns acceptance by surviving tests that its rivals fail.