Why quantum mechanics says the past isn’t real

The following is an excerpt from our Lost in Space-Time newsletter. Every month, we dive into fascinating ideas from around the world. You can sign up for Lost in Space-Time here.

Adolf Hitler died on April 30, 1945. At least that’s what the official story says. But a handful of historians have disputed this evidence and insisted that the Führer fled war-torn Berlin and was living in hiding somewhere. Although the latter account is today widely dismissed as a baseless conspiracy theory, no rational historian would doubt that, whatever the disputed evidence, there was at least one “proven fact”. Either Hitler died that day or he didn’t. It would make no sense to say that Hitler was both alive and dead on May 2, 1945. Yet replace Adolf Hitler with Erwin Schrödinger’s famous cat, and the historical “facts” become seriously obscure.

Schrödinger was one of the founders of quantum mechanics, the most successful scientific theory in history. It is the basis of all chemistry, particle physics, materials science, molecular biology and much of astronomy, and has given us dazzling technological marvels, from lasers to smartphones. The problem is that, for all its triumphs, quantum mechanics ultimately seems to make no sense.

In everyday life, we assume that there is a real world “out there” in which objects such as tables and chairs have well-defined properties, such as having a position and orientation, regardless of whether someone is looking. When we observe an object in the macroscopic world, we are simply discovering a pre-existing reality. But quantum mechanics deals with the microworld of atoms and subatomic particles, where reality evaporates into uncertainty and vagueness.

Quantum uncertainty implies that the future is not entirely determined by the present. For example, if an electron is fired at a known speed over a thin barrier, it may bounce or pass through the barrier and fly to the far side. Or if an atom is put into an excited state, then a microsecond later it may still be excited, or it may have decayed and emitted a photon. In either case, we cannot predict with certainty what will be the case; only betting odds can be given.

And most people have no problem accepting that the future is somewhat open. But quantum fuzziness also implies that the pass it’s not a done deal either. Look at a fine enough scale and the story dissolves into an amalgam of alternate realities, technically called a superposition.

The blur of the quantum microworld comes into focus when a measurement is made. For example, you can perform a position measurement on an electron and find that it has a specific location. But according to quantum mechanics, this does not mean that the electron was already there before measurement, observation simply revealing where exactly. Rather, the measurement projects like an electron to a location from a prior state of positionlessness.

If so, how should we think about the electron before it is observed? Imagine a plethora of half-real “ghost electrons” distributed in space, each representing a potential reality, hovering in a state of blur. Sometimes this is described by saying that the electron is in several places at once. So – bam! – a measurement is carried out which serves to transform a specific “phantom of the winner” into concrete reality, annihilating the competitors.

Does the experimenter have a choice about the outcome? Not when it comes to choosing the winning ghost – it’s a matter of chance. But there is nevertheless an element of choice at play, and it is crucial to understanding quantum reality. If, instead of making a position measurement, the experimenter chooses to measure the speed of the electron, then the previous fuzzy state becomes a net result again – but this time creating not an electron at a location, but an electron with a speed. And an electron with speed behaves like a wave. It is not the same entity as an electron in a place, which is a particle. Obviously, electrons are somehow both waves and particles; which aspect they manifest depends on how someone chooses to question them.

Ultimately: What happens to the electron – whether it behaves as a wave or a forward particle – depends on the type of measurement the experimenter decides to make to observe it. It’s strange, certainly, but that’s where it gets really weird: it also happens that what has arrived at the atom Before the measurement depends on the experimenter’s decision! That is to say the nature of the electron in the past – wave or particle – is determined by this choice. It seems like something is going back in time and affecting the way the world “out there” wasbefore measurement.

Is this time travel? Retrocausality? Telepathy? All of these words come up in popular articles on quantum physics, but the most apt description was given by John Wheeler, the physicist who coined the term black hole: “The past doesn’t exist unless it’s recorded in the present,” he said.

Wheeler’s description sounds profound as a saying, but is there an actual experiment to prove it? This is indeed the case, as I first learned from Wheeler himself when we met for breakfast at the Hilton Hotel in Baltimore in 1980. The meal began with a cryptic question, typical of the man: “How do you hold on to the ghost of a photon?” he asked. Seeing my puzzlement, Wheeler went on to explain a new twist he had imagined for a classical quantum experiment. This is easiest to do with light, although it can just as easily be done with electrons or even whole atoms.

The experiment, first performed by English mathematician Thomas Young in 1801, was an attempt to demonstrate the wave nature of light. Young installed a screen pierced with two narrow slits close together and illuminated it with a point of light. The light passes through the slits and falls on a second screen a little further from the light source. What did Young see? These are not two blurry bands of light, as you might imagine, but a series of bright and dark bands, called interference fringes. They appear because the light waves passing through each slit spread out, and where they arrive in rhythm – peak to peak, trough to trough – they reinforce to form a bright spot, and where they arrive out of sync, they cancel out and produce a dark spot.

Quantum mechanics arose when physicists were debating whether light was made of waves or particles, called photons. We now know that, just like electrons, the answer is both. And thanks to modern technology, you can carry out Young’s experiment, one photon at a time. Each photon forms a small dot on the second screen, and over time, many dots form a speckled pattern to display the distinctive stripes discovered by Young. This seems confusing: if a photon is a tiny particle, surely it must pass through either a slot Or the other. But both slits are needed to create the interference pattern.

So what happens if a cunning experimenter decides to see which slit a given photon passes through? This can easily be achieved by placing a detector near the slots. Once this is done, the interference pattern disappears. The interference detection effectively induced the photon to manifest as a particle, thereby eliminating its wave nature. You can do exactly the same thing with electrons: find out which slit they passed through and find no striped pattern, or leave each electron’s path ambiguous and observe the stripes (once many electrons have built the pattern). So the experimenter must decide, photon by photon or electron by electron, whether it behaves like a wave or a particle when it hits the image screen.

Now we come to Wheeler’s twist. This decision – to watch or not to watch – should not be made in advance. In fact, it can be left until the photon (or electron) has passed through the slit system and is well on its way to the image screen. Indeed, the experimenter can choose to look back and see from which slit the photon emanates or not. This configuration, which naturally bears the name of delayed choice experiment, was carried out and, of course, the results were as expected. When the experimenter decides to take a look, the photons do not collectively form bands; when they go unnoticed, they do it. The conclusion? The reality that was – whether the light behaves as a wave passing through both slits or as a particle passing through one of them – is determined by the experimenter’s subsequent choice. I should mention that, in the real experiment, the “choice” is automated and randomized to avoid bias that could skew the results, and because everything happens faster than human reaction times.

The experience of deferred choice does not change the past. On the contrary, in the absence of experience, there are many pasts – multiple intermingled realities. When a choice is made about what to measure, some of these histories are selected. The effect of this choice is to reduce some of the past quantum vagueness and, if not to determine a single story, at least to reduce the number of contenders. This is why it is sometimes called the quantum eraser experiment.

In the actual experiment, the lookback time is only about a nanosecond, but in principle it could go all the way back to the origin of the universe. And indeed, this was the meaning of Wheeler’s enigmatic question about the retention of the photon ghost. He envisioned a distant cosmic light source gravitationally lensed from our perspective through an intervening black hole, with twin light paths curving around opposite sides of the black hole before converging on Earth, much like the cosmic-scale twin-slit experiment. A ghost of the photon might arrive by one route, while another ghost taking the other, perhaps longer, route might not arrive here for a month. To perform such a cosmic interference experiment, one would have to somehow store, or “hold,” the first ghost to wait for the second to arrive before merging them, so that the waves overlap at the same time, as is the case in Young’s original experiment.

Einstein once wrote that the past, present and future are illusions. In this he was wrong. The error lies in the word “the”. A The past exists today in historical records, but it consists of a vast multiplicity of jumbled “phantom pasts,” grouped together to form a single narrative on a macroscopic scale. At the quantum level, however, this blends into an amalgam of fuzzy partial realities that are beyond human experience.

Paul Davies is a theoretical physicist, cosmologist, astrobiologist and bestselling author. His book, Quantum 2.0, published in November 2025 by Penguin.