The ‘Negative Time’ Photon Experiment Passed Peer Review—It Still Isn’t Time Travel
The photon experiment that attracted “negative time” headlines in 2024 is no longer merely a preprint: its current record lists publication in Physical Review Letters in 2026. The result survived peer review, but its meaning remains narrower than the time-travel language suggests.
What the researchers measured was a negative weak value for atomic excitation time among selected photons transmitted through an atom cloud. That is a real laboratory result involving light–matter interaction—not evidence that a photon followed an ordinary trajectory into the past, reversed causality or enabled messages to arrive before they were sent.
What changed after the original 2024 report
The most important update is the study’s publication status. The researchers’ current paper record identifies the work by Daniela Angulo, Kyle Thompson, Vida-Michelle Nixon, Andy Jiao, Howard M. Wiseman and Aephraim M. Steinberg as Physical Review Letters 136, 153601 (2026), after its initial submission on September 5, 2024.
Peer review strengthens the result as an experimental contribution, but it does not turn “negative time” into a general discovery about time itself. The published claim concerns a particular operationally defined quantity: the time integral of a weak probe signal, normalized as an atomic excitation time and averaged only over runs in which the signal photon was detected after transmission.
The distinction matters because an earlier headline-friendly description can sound as if a detector watched a photon leave the cloud before that same photon visibly entered. The experiment instead compares statistical timing and weak-measurement results accumulated over many repetitions. Neither measurement supplies a classical timestamped itinerary for an individual photon.
How the experiment asked the atoms what happened
The setup used photon pulses resonant with a cloud of rubidium atoms. Resonance permits the photon’s energy to be stored temporarily as a collective atomic excitation, while spontaneous emission can scatter light away from the forward path. The researchers focused on the less likely cases in which the signal photon was ultimately transmitted through the cloud.
A direct, precise check of whether the atoms were excited at every instant would strongly disturb the interaction being studied. The team therefore used the cross-Kerr effect: a separate, weak, off-resonant probe beam passed through the atoms, and tiny changes in its optical phase served as a minimally disturbing indicator of excitation. Individual readings were imprecise, so the result emerged from calibrated averages over many runs.
This is also why “the photon spent minus some amount of time inside” should not be read like a stopwatch result. The experiment measured a weak value, a conditional quantum statistic obtained after selecting the transmitted outcomes. Weak values can fall outside the ordinary range of values that a strong, single-shot measurement would return.
The published measurements nevertheless contain a specific surprise. Across different pulse durations and optical depths, the inferred excitation time agreed with the group delay of the transmitted light. It ranged from −0.82 ± 0.31 times a reference excitation time for the narrowest-band pulse to +0.54 ± 0.28 times that reference for the broadest-band pulse; the reference equals the scattering probability multiplied by the atoms’ spontaneous lifetime.
Why a negative result can emerge without backward motion
Group delay describes the displacement of a pulse’s characteristic timing—commonly its peak or average arrival time—after it passes through a medium. Near an atomic resonance, absorption and coherent re-emission can reshape the long photon pulse so that the transmitted distribution is shifted forward. Extrapolating that shift can assign a negative group delay even though no usable signal has physically retraced time.
Pulse reshaping was already a conventional way to understand such early arrival statistics. The new contribution is that the atoms gave a corresponding answer: the independently inferred weak excitation time matched the group delay, including when the value was negative. An explanation written by co-author Howard Wiseman says that averaging the weak probe over many runs produced an accurate dwell-time estimate that could not be reduced simply to the front of the transmitted pulse surviving.
That does not establish that an atom literally remained excited for a negative duration. It establishes that the weakly measured, postselected effect carries a minus sign and tracks a timing quantity that physicists had often treated mainly as a feature of pulse propagation. The experiment therefore gives negative group delay more operational significance without converting it into ordinary negative elapsed time.
Why this is not a time machine
The researchers have been unusually direct about this boundary. A University of Toronto clarification says the experiment does not claim that photons travel backward in time and identifies the subject as the long-known phenomenon of negative group delay.
No controllable information is shown leaving before it is introduced, and the experiment is described by standard quantum optics. Postselection is essential: researchers keep the subset of trials in which a photon reaches the forward detector and then calculate a conditional average. The completed detection record is needed before that subset—and therefore its weak value—can even be identified.
The durable lesson is subtler than time travel. A quantum duration defined through a weak interaction and a later selection need not behave like time read from a clock attached to one particle. The 2026 publication confirms that the negative quantity corresponds to a measurable response of the atom cloud, while leaving chronological order and causal communication intact.
Also read:
Subscribe to our newsletter
Get the latest Web3, AI, and crypto news delivered straight to your inbox.