The Question That Was Answered, and the One That Stayed — Bohr and Einstein at the Solvay Conferences, 1927 and 1930

Last time we followed electrons through solids (Part 4 is here). Quantum mechanics had by then explained the chemical bond, the atomic nucleus, and the difference between a metal and an insulator. Part 5 turns back to the tool itself. It works whenever it is used. But does it say everything there is to say about nature? At the Solvay conferences of 1927 and 1930 in Brussels, Einstein put questions to Bohr, and Bohr answered. In the usual telling, Bohr won twice. The records of the time suggest something else: Einstein had two questions, and only one of them was answered.

A conference on "Electrons and Photons", and Bohr's complementarity

The fifth Solvay conference met in Brussels from 24 to 29 October 1927. Its theme was "Electrons and Photons". As we saw in Part 5 of the earlier series, Heisenberg had published his uncertainty relation that year: the more sharply you pin down where a particle is, the more blurred its momentum (roughly, how hard it would hit) becomes, and the other way round.

The month before, at Como in Italy, Bohr had set out his own view in a lecture (it appeared in Nature in April 1928). He called it complementarity. An electron, or light, behaves like a particle in some situations and like a wave in others. Which one shows up depends on the apparatus you use to look. The two pictures cannot be used at the same time, yet neither is enough alone. So, Bohr said, accept both, as pictures that complete each other.

What Bohr stressed was that the thing being examined (the electron) and the thing doing the examining (the apparatus) cannot be cleanly separated. In the quantum world the disturbance caused by a measurement cannot be made as small as you like.

Einstein had not been at Como. The two met in Brussels.

The question on the record — far apart, at the same instant

The published proceedings hold a surprise. In the formal sessions, neither man said very much. Bohr did not even give one of the reports; what the proceedings print under his name is the text of his Como lecture.

Einstein did speak up once, clearly, in the general discussion. His example went like this. Send electrons at a small hole in a screen. Beyond the hole the electron's wave spreads out in a hemisphere and reaches a photographic film. Yet what the film records is always a mark at one single spot.

If the wave only describes how a great many electrons are spread out, there is no difficulty. But suppose the wave is the complete description of this one electron. Then, the instant the mark appears at one spot, the chance of a mark must vanish everywhere else on the film at once. That, Einstein said, would require a very peculiar mechanism of action at a distance linking the separate places. And it sat badly with relativity, which says that no influence travels faster than light.

So the question on the record is not whether the uncertainty relation can be broken. It is whether the theory says everything about each individual event; and, if it does, what is going on between places that are far apart.

Chess at the breakfast table — the movable screen

The famous exchanges did not happen in the sessions. They happened over breakfast and dinner. We know them mainly from two sources. One is the account Bohr himself wrote in 1949, twenty-two years later. The other is a letter written a few days after the meeting, on 3 November, by Paul Ehrenfest, a friend of both men who was there, to his former students. He wrote that the conversations were like a game of chess. Einstein popped up each morning like a jack-in-the-box with a fresh example, and Bohr took the examples apart one by one.

Bohr's account includes a well-known moment. Einstein asked, teasingly, whether they could really believe that God plays dice. Bohr says he answered by pointing out that ancient thinkers had already urged great caution in describing Providence in everyday language. That is a good deal milder than the retort usually quoted, in which Bohr tells Einstein to stop giving God instructions.

Here is one of Einstein's examples. A first screen has a single slit; a second screen beyond it has two. Send particles through one at a time, and stripes slowly build up on a plate at the back. The stripes are the sign that a wave went through both slits and overlapped with itself. Which slit did each particle go through? Try to find out, and the stripes disappear. That was complementarity's claim.

Einstein's twist was to leave the first screen free to move. A particle heading for the upper slit gives that screen a slightly different kick from one heading for the lower slit. Measure the screen's motion and you know the path. And the stripes are still there on the plate. You would then know both.

Bohr's reply was that the screen obeys quantum mechanics too. To know its momentum well enough to tell the two kicks apart, you must, by the uncertainty relation, accept that its position is blurred. Work it out, and the blur is just as large as the spacing of the stripes. The stripes are smeared away. You can trace the path, or you can see the stripes. Not both.

1930: a box of light, a balance and a clock

Three years later, at the sixth Solvay conference in 1930 (its theme was magnetism), they met again. In Bohr's account the discussion now took a dramatic turn.

This time the target was energy and time. The more sharply an energy is fixed, the more blurred is the moment it belongs to. It is the partner of the relation between position and momentum.

Einstein imagined a box filled with light. There is a hole in its wall, and a shutter driven by a clock inside the box opens it for an instant at a chosen time, letting out a single particle of light. The clock tells you when it left. Then weigh the box before and after. Energy has weight (the famous result of relativity), so the loss of weight gives the energy of the light that escaped. Both the time and the energy, it seemed, could be known as exactly as one pleased.

Bohr wrote that this was a serious challenge. The key to his answer lay in what weighing actually involves. Hang the box on a spring and read its height against a scale. The more exactly you read the height, the more blurred the box's up-and-down momentum becomes. Until the weighing is finished, it stays unclear how far the box has moved up or down in the pull of gravity. And by Einstein's own theory, the rate of a clock depends on its height in a gravitational field (the effect known from the prediction that sunlight should be shifted slightly toward the red). If the height is blurred, so is the rate of the clock. Work it out, and the more exactly the energy is weighed, the less you know about when the shutter opened. The uncertainty relation is exactly preserved.

It was the same kind of answer as in 1927: the measuring instruments cannot escape the rules of quantum mechanics either. Bohr added that Einstein himself contributed a great deal to working this out. Later studies have pointed out that the argument needs only the assumption that energy has weight, not the whole theory of gravity.

The box had been built for something else

All this is usually summed up as "Einstein tried twice to break the uncertainty relation, and lost twice". But that story rests mainly on Bohr's account of 1949. One scholar has described it as written from a winner's perspective.

A letter from the time says something a little different. On 9 July 1931 Ehrenfest wrote to Bohr to pass on what Einstein had told him. Einstein had not doubted the uncertainty relations at all for a long time. And he had by no means invented the box in order to defeat them. He had invented it for a completely different purpose.

The purpose was this. Let the particle of light fly from the box to a mirror far away, say half a light-year. About the time it reaches the mirror, choose one of two things to do with the box at home. Weigh the box again, and you learn the energy of the light. Read the clock, and you learn when the light will come back. The choice is yours. But whatever you choose at home, the light itself, far away, can hardly be changed by it. Yet quantum mechanics gives that same light a different description depending on your choice. Then perhaps the quantum description does not say everything about the real state of the light.

This has the same shape as the question on the record from 1927. The issue is not whether the uncertainty relation is right. It is whether the theory is complete, and what holds between places far apart.

In February of that same year, 1931, Einstein had published a short paper with Tolman and Podolsky. Using an imagined experiment, it argued that in quantum mechanics the past, and not only the future, is uncertain. This is not a paper that tries to knock the uncertainty relation down. It accepts the relation and thinks onward from it.

What exactly Einstein was aiming at in the room in 1930 cannot be settled from the surviving sources alone (some scholars read his aim as different from the very start). What is certain is that by the following year, at the latest, his interest had moved away from whether the relation could be broken. Bohr himself noted in his account that Ehrenfest had told him Einstein was far from satisfied, and he described the new use of the box, with its choice between weighing again and reading the clock.

What could be done at this point, and what could not

First, what was achieved. The uncertainty relation was tested against the hardest counter-examples anyone could think of, and it held. A lesson came out of this clearly. The measuring instruments obey quantum mechanics as well. The screen, the balance and the clock are no exceptions. Treat the instruments alone by the old physics, and the account stops adding up.

"The path or the stripes" has since been checked in the laboratory. In 2025 a group at the University of Science and Technology of China used a single atom, held in mid-air by laser light, in the role of the movable screen, and sent particles of light at it one at a time. As they loosened the grip on the atom, the stripes faded. It was the result Bohr's answer called for.

What was not achieved is just as clear. Einstein's other question got no answer. Does the theory say everything about each individual event? What is going on between two places that are far apart? The scholars who retranslated and studied the 1927 proceedings conclude that the question of interpretation was not settled at that conference, and that no consensus was reached.

This series has returned several times to the difference between being able to produce an answer and feeling that one understands. Einstein may have been the person who put that difference most plainly. He accepted that quantum mechanics gives the right answers. Having accepted it, he kept asking whether that was all. Complementarity tells you what you will see once the apparatus is fixed. It does not tell you what is there before the apparatus is fixed. Bohr held that giving up that question was itself the lesson of the new physics. Einstein did not give it up.

Ehrenfest, who watched it all, wrote in his 1927 letter that he was almost entirely on Bohr's side. Yet, by Bohr's account, he also said that his mind would not be at rest until he and Einstein agreed. Even for those in the room, it seems, the matter was not neatly closed.

Next time is the final part. In 1935 Einstein put this question into a paper with Podolsky and Rosen, and Schrödinger brought out his cat. The link between two separated things that they found there would later become the raw material of the quantum computer.