Wheeler's Delayed-Choice Experiment: When Future Decisions Change the Past

A Photon That Has Already Passed Through an Experiment Can Retroactively Appear to Have Taken a Different Path

In 1978, physicist John Archibald Wheeler proposed a thought experiment that seemed to violate one of science's most basic assumptions: that the past is fixed. In Wheeler's delayed-choice experiment, a photon is sent through an optical apparatus — a beam splitter, an interferometer — and the experimenter's decision about what to measure is made after the photon has already entered the device. Depending on that late measurement choice, the photon appears to have behaved either as a wave (traveling both paths simultaneously) or as a particle (traveling only one path) — as if the photon "knew in advance" what would be measured. This is not a violation of causality; it is something arguably stranger: evidence that quantum systems do not have definite properties until measured, regardless of when the measurement is made.

The Experimental Setup

The basic apparatus involves a Mach-Zehnder interferometer. A photon enters a first beam splitter (BS1), which sends it into a superposition of two paths (upper and lower). The paths recombine at a second beam splitter (BS2) and arrive at two detectors (D1 and D2). The key variable is whether BS2 is present or absent.

• BS2 present (closed configuration): The two paths recombine and interfere. If the path length difference is tuned correctly, all photons arrive at D1 — constructive interference at D1, destructive at D2. This demonstrates wave behavior: the photon must have traveled both paths to produce interference.
• BS2 absent (open configuration): Each photon arrives at either D1 or D2 with 50/50 probability, and detecting at one tells you which path the photon took. This demonstrates particle behavior: the photon traveled only one path.

Wheeler's twist: make the decision to insert or remove BS2 after the photon has already passed through BS1 — using a random switch triggered while the photon is in flight between BS1 and the detection region.

The Experimental Results

Wheeler's thought experiment became a real experiment in 2007, when Alain Aspect's group at the Institut d'Optique in Paris implemented it using a fast electro-optic switch that could change the interferometer configuration in nanoseconds — well after the photon had passed the first beam splitter. The results were unambiguous:

• When BS2 was in place, interference fringes appeared — consistent with wave behavior on both paths
• When BS2 was removed, which-path information was available and interference vanished — consistent with particle behavior on a single path
• The switch decision made after the photon entered the device perfectly predicted the photon's behavior, as if the photon retroactively "chose" to be a wave or a particle

A further experiment by the same group in 2015 used a quantum random number generator ensuring the choice was truly random and spacelike-separated from the photon's entry into the interferometer — ruling out any classical hidden-variable explanation.

The Cosmic Version: Quasars as Beam Splitters

Wheeler extended the thought experiment to cosmological scales. Light from a distant quasar is gravitationally lensed by an intervening galaxy — the galaxy acts as a cosmic beam splitter, sending quasar light along two different curved paths around the galaxy, both of which arrive at a telescope on Earth. An astronomer can choose to measure either which path the light came from (particle behavior) or set up an interference measurement (wave behavior). The photons making this choice started their journey billions of years ago — before the solar system existed. Wheeler's point was ontological: the photon does not have a definite mode of travel (wave or particle) until the measurement is made, regardless of how far in the "past" the photon's journey began.

What the Experiment Does and Does Not Show

Wheeler's Participatory Universe

John Wheeler drew a radical philosophical conclusion from his thought experiment — one that most physicists find either profound or overreaching, depending on their philosophical commitments. Wheeler proposed that the universe is fundamentally participatory: the act of observation does not merely reveal a pre-existing reality but partly constitutes it. He coined the aphorism "no phenomenon is a phenomenon until it is an observed phenomenon." In his view, quantum mechanics implies that a universe of pure potentiality collapses into actuality through the acts of observation of conscious beings — a view called the participatory anthropic principle.

Most physicists separate Wheeler's technical contribution (which is verified and important) from his participatory universe philosophy (which requires much stronger assumptions). The experiment definitively shows that quantum systems do not have definite mode-of-travel properties until the relevant measurement is made — it does not definitively show that consciousness plays a special role. Decoherence theory handles the appearance of classical reality without requiring a conscious observer.

Connection to the Quantum Eraser

The delayed-choice experiment connects directly to the quantum eraser (a separate experiment), in which which-path information is first recorded, destroying interference, then the recorded information is erased — and interference is restored. Together, these experiments establish that what determines wave-like or particle-like behavior is not the physical trajectory of the photon but whether which-path information exists anywhere in the universe — accessible or not. The universe keeps track of information in ways that shape quantum behavior, and the timing of information acquisition or erasure can be decoupled from the timing of the photon's physical journey through an apparatus.