Physicists Measured “Negative Time”—But the Real Quantum Experiment Is Stranger Than Time Travel
A photon enters a cloud of atoms. The atoms interact with the light. The photon emerges on the other side.
Ask how long that interaction lasted and common sense demands one thing: the answer cannot be less than zero.
Yet physicists at the University of Toronto measured exactly that.
In experiments led by Daniela Angulo and Aephraim Steinberg's research group, the measured average atomic excitation time associated with transmitted photons became negative under particular quantum conditions. For narrowband light near atomic resonance, the team reported a value of −0.82 ± 0.31 times a reference excitation time. The result was first posted as an arXiv preprint in September 2024 and, after peer review, was published in Physical Review Letters on April 13, 2026.
That sounds like a photon completed part of its journey before it began.
It did not.
The experiment does not demonstrate backward time travel, faster-than-light communication, reversed causality, or a loophole in Einstein's special relativity. The University of Toronto researchers themselves have repeatedly warned against those interpretations. What they measured was a negative weak value of atomic excitation time, a subtle quantum quantity associated with photons selected because they successfully passed through an ultracold cloud of rubidium atoms.
The real discovery is less sensational than a time machine—but considerably more interesting scientifically.
For decades, physicists often treated negative optical delays mainly as consequences of wave reshaping and interference. The Toronto experiment indicates that the same negative quantity also predicts the sign and magnitude of a measurable physical effect on the atoms and a separate probe beam.
In other words, “negative time” may not mean what the headline suggests.
But it may mean more than physicists previously thought.
What Did the University of Toronto Physicists Actually Discover?
The experiment asks a deceptively simple question:
When a photon passes through a cloud of atoms without being scattered away, how much time does it spend in the form of an atomic excitation?
Light traveling through matter does not behave as though atoms are simply transparent obstacles.
The electromagnetic field interacts with the atoms.
Near resonance, energy can temporarily occupy a collective state involving both the light and the atomic medium. Physicists can describe part of that interaction as the atoms being excited while the photon propagates through the material.
The Toronto researchers wanted to measure that excitation time experimentally.
Their result was peculiar.
Depending on pulse bandwidth and properties of the atomic cloud, the inferred excitation time for photons that were successfully transmitted changed from positive to negative. The peer-reviewed paper reports measured values ranging from −0.82 ± 0.31 τ₀ for the narrowest-band pulse to +0.54 ± 0.28 τ₀ for the broadest-band pulse.
Here, τ₀ is not an ordinary stopwatch reading.
It represents the average, non-post-selected atomic excitation time, calculated from the probability that an incident photon will scatter multiplied by the atoms' spontaneous lifetime.
The negative result appears only after considering a particular subset of events:
Photons that eventually make it through the cloud.
That detail—known as post-selection—is central to understanding why negative values become possible.
The Paper Is No Longer Just an arXiv Preprint
An important update has occurred since the original story began circulating.
The DOI often associated with the research,
is the DataCite DOI for the arXiv version of the study.
The original preprint, titled Experimental evidence that a photon can spend a negative amount of time in an atom cloud, was submitted on September 5, 2024.
The experiment subsequently passed peer review.
A revised version was published in Physical Review Letters, one of the major peer-reviewed journals in physics, on April 13, 2026 under the title:
Experimental Observation of Negative Weak Values for the Time Atoms Spend in the Excited State as a Photon Is Transmitted.
The final journal citation is Physical Review Letters 136, 153601, and its journal DOI is 10.1103/gjfq-k9dv. The paper lists Daniela Angulo, Kyle Thompson, Vida-Michelle Nixon, Andy Jiao, Howard M. Wiseman and Aephraim M. Steinberg as authors.
That distinction matters because some coverage still describes the finding as an unpublished or merely preliminary 2024 result.
As of 2026, that is outdated.
The experiment has undergone peer review and been formally published.
Why Light Can Have a “Negative Delay”
To understand the experiment, first forget photons for a moment and imagine a pulse of light as a wave packet.
A pulse has a front.
It rises toward a maximum.
Then it falls.
When the pulse passes through a material, different frequencies within it may experience different absorption and phase shifts.
The output pulse can therefore change shape.
Its peak may emerge earlier than the position at which the original pulse peak would have arrived if the pulse had traveled through empty space.
Physicists describe this using a quantity called group delay.
Normally, a positive group delay means the transmitted pulse appears delayed.
A negative group delay means its center or peak appears advanced.
This phenomenon has been understood in optics for decades and does not require an object or information to literally move backward through time.
The theoretical work behind the Toronto experiment explains that, near resonance, the trailing portion of a light pulse can be attenuated differently from the leading portion because atoms respond over a finite time. That reshapes the surviving pulse and shifts its center forward.
Imagine a long line of runners passing behind a wall.
If the slow runners toward the back are removed while the faster runners toward the front remain, the average position of the surviving group suddenly moves forward.
No runner traveled backward in time.
The composition of the group changed.
Negative group delay can behave similarly.
But that familiar explanation created a much harder question.
Was Negative Group Delay Merely a Mathematical Illusion?
For years, one common interpretation of negative group delays was essentially:
Do not take the negative time too literally.
The pulse was reshaped.
The output peak moved.
Nothing actually “spent negative time” inside the material.
Steinberg's group wanted to test whether that dismissal went too far.
Instead of measuring only when a transmitted pulse appeared at the output, they tried to ask the atoms themselves how much interaction had occurred.
The key question became:
If the negative group delay is merely an artifact of pulse reshaping, would an independent measurement of the atoms' excitation still produce the same negative quantity?
The experiment says yes.
The group found that the atomic excitation time inferred from the probe tracks the group delay—including when the group delay is negative.
That is the scientifically interesting result.
It suggests negative group delay is not merely a meaningless artifact attached to the position of a distorted pulse.
The same number describes another measurable physical effect.
The Experiment Used Ultracold Rubidium Atoms
The researchers created a cloud of rubidium-85 atoms cooled to approximately 60 to 70 microkelvin.
That is only a tiny fraction of a degree above absolute zero.
The atoms were confined using a magneto-optical trap.
The cloud was roughly one millimeter long, while the interacting beams were focused to a waist of about 25 micrometers. The atomic transition studied had a spontaneous lifetime of approximately 26 nanoseconds.
Two light beams passed through the cloud.
They had very different jobs.
The signal beam
The signal was a pulsed beam tuned to resonance with an atomic transition.
It was the light whose interaction with the atoms the researchers wanted to investigate.
The probe beam
A second, continuous-wave laser was deliberately tuned away from resonance.
Its purpose was not to undergo the same interaction.
Instead, it acted as a sensitive monitor of the atoms.
The two beams traveled in opposite directions through the same atomic cloud.
This arrangement allowed the scientists to detect an extraordinarily small physical signature left by the signal photons.
How Do You Ask an Atom How Long It Was Excited?
Obviously, an atom does not contain a tiny clock.
The researchers therefore needed an indirect measurement.
They used the cross-Kerr effect.
When the signal pulse slightly excites or saturates the atoms, the optical properties experienced by the probe beam change.
That change alters the probe's phase.
The instantaneous phase shift therefore contains information about how much atomic excitation exists at a particular moment.
Integrate that phase shift over time and the result provides a measure related to the total excitation time.
The principle is remarkably elegant:
Signal photon interacts with atoms.
Atoms slightly alter probe beam.
Probe phase records the interaction.
The paper describes the instantaneous probe phase shift as proportional to the amount of atomic excitation present in the medium.
The effect, however, was extraordinarily small.
The Signal Was Tiny Compared With the Noise
The expected phase shift associated with a transmitted photon was on the order of only 10 to 20 microradians.
Meanwhile, the experiment's phase noise was approximately 120 milliradians.
The signal researchers wanted was therefore buried beneath noise vastly larger than the effect itself.
They could not solve that problem by taking one beautiful measurement.
They solved it statistically.
The experiment required tens of millions of atomic measurement cycles for a single set of parameters.
According to the researchers' description of the experiment, one parameter set could require roughly 10 hours of data collection.
That is an important detail when headlines say physicists “watched a photon spend negative time.”
They did not observe one photon and see a clock run below zero.
The result emerged statistically from enormous ensembles of carefully selected quantum measurements.
They Did Not Simply Fire One Photon at a Time
Another subtle point is often lost in simplified accounts.
The researchers did not rely exclusively on ideal single-photon input states.
Instead, they used coherent light pulses containing an average of approximately 100 photons before later attenuation and detection.
Their method exploits a previously developed post-selection technique.
When a single-photon detector registers transmission, researchers compare the probe phase associated with detector-click events against the probe phase when there is no click.
That difference isolates the inferred contribution associated with an additional transmitted photon. The authors explain that this technique should reproduce the result expected from a genuine single-photon experiment while offering much higher experimental data rates.
This does not invalidate the result.
It simply makes the phrase “they fired individual photons through atoms” an oversimplification of the actual apparatus.
Post-Selection Is Where Quantum Mechanics Gets Strange
Suppose researchers perform an experiment a million times.
They do not necessarily average every result together.
Instead, they can ask a conditional question.
For example:
Among all the photons sent toward the atoms, what was the measured excitation associated specifically with the photons that eventually appeared in the transmitted detector?
That is post-selection.
You prepare a quantum system.
You perform a very weak measurement while it evolves.
Then you keep only trials with a particular final outcome.
In this case, the final condition is transmission.
Quantum mechanics allows the resulting conditional quantity—the weak value—to behave very differently from an ordinary expectation value.
What Is a Weak Value?
Weak values originated in a famous 1988 proposal by Yakir Aharonov, David Albert and Lev Vaidman.
The basic idea is to measure a quantum system so gently that the measurement extracts only a tiny amount of information and causes correspondingly little disturbance.
One weak measurement tells you almost nothing.
Repeat the experiment many times and statistical patterns emerge.
Then post-select the data according to the system's final state.
The resulting weak value can lie outside the ordinary range of possible values associated with the measured observable.
The Toronto paper explicitly emphasizes this feature: unlike conventional expectation values, weak values do not have to remain inside the eigenvalue range of the operator being measured.
That means an observable normally associated with values between zero and one can produce a conditional weak value that is:
- Larger than one
- Smaller than zero
- Or even complex
This does not mean an ordinary measurement would reveal an impossible physical state.
It means the weak measurement plus post-selection is probing interference between quantum alternatives in a different way.
A Negative Weak Value Is Not a Negative Stopwatch Reading
This distinction is perhaps the single most important point in the entire story.
When the experiment reports a negative excitation time, it does not mean:
An atom became excited at 12:00:00.000000020 and somehow stopped being excited at 12:00:00.000000010.
No microscopic clock was observed running backward.
Instead, the weak interaction with the probe creates a pointer shift.
In ordinary conditions, excitation pushes that pointer in one direction.
For the post-selected transmitted photons in the negative-delay regime, the statistical pointer shift reverses direction.
When that phase shift is converted into the experimentally defined excitation-time quantity, the answer is negative.
Steinberg explained the distinction in a University of Toronto clarification, noting that saying an atom literally spends “minus” an ordinary duration doing something would make little sense. What is physically significant is that the measurable effect used as the time probe acquires the opposite sign.
That is much stranger than a simple measurement error.
But it is not time travel.
Why Quantum Interference Can Produce the Negative Result
The theoretical explanation relies heavily on quantum interference.
A transmitted photon cannot always be pictured as following one classical trajectory through the cloud.
Instead, amplitudes associated with different histories interfere.
One possible history resembles the photon propagating through without becoming an atomic excitation.
Another involves the excitation temporarily residing collectively in the atoms before being coherently returned to the optical field.
Additional higher-order histories can occur in denser media.
Quantum amplitudes associated with these alternatives combine.
They can reinforce one another.
They can also cancel.
The theoretical analysis published in APL Quantum shows how interference among such pathways can generate a negative post-selected dwell time.
The important word is post-selected.
If researchers ignore the photon's eventual fate and average over everything, the total excitation time remains an ordinary nonnegative quantity.
The anomalous negative value appears when asking about the conditional history of the subset that was successfully transmitted.
A Simple Analogy for Negative Weak Values
No classical analogy perfectly reproduces quantum interference, but consider a financial portfolio.
Suppose three investments generate:
- +$100
- +$100
- −$150
The total is +$50.
Nothing requires every contribution to be positive simply because the final total is positive.
Quantum amplitudes are much stranger than financial transactions because they interfere before probabilities are calculated.
Still, the analogy captures one useful idea:
A conditional component of a larger calculation can carry a negative contribution even though the total physical quantity behaves normally.
In the Toronto experiment, the negative weak value does not mean negative time exists as an ordinary substance.
It means the transmitted subset contributes to the probe measurement in a way mathematically and experimentally described by a negative excitation-time value.
Why the Result Is More Than Pulse Reshaping
This is the strongest point made by the researchers.
Before this experiment, someone could say:
Negative group delay only tells us where the center of a reshaped output pulse appears.
Do not interpret it as a physical duration.
The Toronto team measured something else entirely.
They monitored a second beam interacting with the atoms.
Yet the measured phase response still tracked the same group-delay quantity.
The peer-reviewed paper concludes that the weak-valued atomic excitation time for transmitted photons equals the optical group delay across the conditions tested.
The original experimental paper goes further, arguing that group delay predicts not only the position of a transmitted pulse but also the magnitude and sign of the effect that transmitted photons produce on another system.
That is why the result attracted genuine scientific attention.
The experiment is not interesting because somebody attached the word “negative” to a mathematical expression.
It is interesting because the negative expression corresponds to an independent measurable effect.
Does a Photon Really Leave Before It Enters?
Not in the ordinary causal sense.
This phrase comes from the behavior of negative group delay.
Imagine comparing two pulses:
One propagates through free space.
The other enters the atomic cloud.
Under certain conditions, the center of the transmitted pulse can emerge earlier than the center of the reference pulse would be expected to emerge.
If the group delay is sufficiently negative, extrapolating the pulse centers can create language that sounds like the output happens before the input.
But the output pulse is formed from the leading portions of an extended incoming wave packet.
The medium does not receive a completely unknown message and reproduce that information before the message arrives.
No usable information travels into the past.
No causal signal outruns light in vacuum.
The University of Toronto's own clarification specifically stresses that the experiment does not demonstrate photons traveling backward through time.
Does This Violate Einstein’s Special Relativity?
No.
Special relativity forbids information or causal influence from propagating faster than the speed of light in vacuum in a way that would permit violations of causality.
A negative group velocity or group delay is not automatically the same thing as the velocity at which new information propagates.
This distinction has been understood in wave physics for a long time.
A pulse has a shape containing many frequencies.
A dispersive medium can alter that shape.
The location of its maximum therefore does not necessarily behave like a little object following a trajectory.
The Toronto experiment does not report superluminal information transfer, and the researchers explicitly reject interpretations involving backward time travel or broken relativity.
Einstein survives.
Did the Experiment Reverse Cause and Effect?
No.
A detector click still occurs after the physical experiment that produced it.
The researchers cannot know in advance which photons will successfully transmit.
Post-selection happens after the outcome is recorded.
When scientists condition earlier weak measurements on that later result, the calculated weak value can be negative.
This does not allow someone to use a future detector decision to change something that already happened.
There is no message sent to the past.
There is no ability to choose the future outcome and thereby rewrite the earlier state.
The counterintuitive result belongs to the statistical structure of quantum measurement and interference.
Causality remains intact.
Is This Evidence That Time Itself Is Quantum?
Not directly.
The experiment concerns a quantum-mechanical definition of dwell or excitation time.
It does not demonstrate that spacetime itself is quantized.
That distinction is important because “quantum time” is sometimes used loosely in popular discussions.
A full quantum theory of gravity would need to explain the relationship between quantum mechanics and dynamical spacetime.
This experiment addresses a much more focused problem:
How should physicists define and measure the duration associated with a photon interacting with an atomic medium when they condition the measurement on a particular quantum outcome?
That is already a deep foundational question.
It does not require claiming that the experiment discovered the fundamental nature of time.
Why Did the Scientific Community Debate the Terminology?
The phrase “negative time” is irresistible.
It is also dangerous.
One side of the interpretive issue emphasizes that negative group delays are already understood through wave interference and pulse reshaping.
From that viewpoint, saying a photon “spent a negative amount of time” inside atoms risks turning an unusual conditional measurement into an unjustified literal story about particle trajectories.
The Toronto researchers agree that the number should not be interpreted as an ordinary clock duration.
Where they push further is in arguing that it should not be dismissed as physically meaningless either.
Their experiment showed that the negative value predicts the sign of an independently measured phase response.
That leaves a fascinating middle ground:
The negative time is not a literal classical duration, but neither is it merely an arbitrary mathematical artifact.
That is where much of the scientific interest lies.
The Experiment Builds on Earlier Work
The 2024–2026 experiment did not appear from nowhere.
Steinberg's group had already investigated how much atomic excitation is associated with photons that pass through an absorbing medium.
A 2022 experiment found that a transmitted photon could be associated with substantial atomic excitation even though it was not ultimately scattered out of the transmitted mode.
That result challenged a tempting classical intuition:
Perhaps photons that scatter are absorbed and excite the atoms, while photons that pass through do essentially nothing.
The experiment showed the situation was more complicated.
A later theoretical study by Kyle Thompson, Howard Wiseman, Steinberg and collaborators predicted something even stranger: the weak-valued excitation time for transmitted photons should equal their group delay—including when that delay becomes negative. The theory was published in APL Quantum in 2025.
The newest experiment tested that prediction.
How Strong Was the Experimental Result?
The most dramatic reported value was:
−0.82 ± 0.31 τ₀.
That is clearly suggestive of a negative quantity, although the experiment is not a situation where every parameter point perfectly matches theory.
The authors openly discuss discrepancies.
For certain pulse durations, measured excitation times were more negative than predicted. They identify possible spurious correlations between probe phase and signal transmission as an important source of systematic error and describe multiple checks designed to investigate those effects.
This is worth emphasizing because responsible science writing should not portray the experiment as flawless confirmation of every theoretical curve.
The broader result is that measurements across multiple pulse durations and optical depths were consistent overall with the predicted relationship between weak-valued excitation time and group delay.
Peer review does not turn an experiment into unquestionable truth.
It means the work survived expert scrutiny sufficiently to enter the formal scientific literature.
Replication and further experiments remain valuable.
What Happens to the Photons That Are Scattered?
This question may provide one of the most interesting next tests.
The theory distinguishes between photons that are transmitted through the cloud and photons that scatter away from the original direction.
For transmitted photons, the predicted atomic excitation time equals the group delay.
For scattered photons, theory predicts an additional positive contribution known as the Wigner time delay.
The theoretical paper predicts that the excitation time associated with scattered photons should equal the group-delay contribution plus this elastic-scattering delay.
Testing that second half experimentally would provide another stringent check of the framework.
It could help determine how broadly weak-value descriptions capture the temporal history of light-matter interactions.
What Is Wigner Time Delay?
Wigner time delay is a familiar concept in scattering theory.
When a wave or quantum particle interacts with a scattering potential, the interaction changes its phase.
That phase change can be translated into an effective delay relative to free propagation.
In the atom-photon problem, scattered light carries information about both propagation through the medium and the act of scattering itself.
The theory therefore predicts a different temporal quantity for scattered photons than for photons that remain in the transmitted mode.
This matters because it offers a useful consistency test.
If transmitted and scattered outcomes can be measured independently and their weighted behavior recombines into the ordinary nonnegative excitation time for the entire ensemble, researchers gain a more complete picture of where the anomalous negative contribution fits into standard quantum mechanics.
Could “Negative Time” Be Used for Faster Computers or Communication?
There is currently no demonstrated technology that uses this effect to send information backward, transmit data faster than light, or make a processor perform operations before they begin.
The research is foundational.
Its immediate value lies in understanding quantum measurement, light-matter interactions, interference and the physical meaning of quantities such as group delay and dwell time.
That does not mean applications will never emerge.
Research on apparently abstract quantum phenomena has historically contributed to technologies such as:
- Atomic clocks
- Lasers
- Quantum sensors
- Quantum cryptography
- Quantum computing
- Precision interferometry
- Quantum communication
- Optical information processing
But claiming a direct technological breakthrough from this experiment would be premature.
The researchers themselves have emphasized interpretation and fundamental physics rather than a near-term device.
Could Negative Weak Values Be Useful Anyway?
Possibly.
Weak-value techniques have already been investigated as ways of detecting extremely small physical effects.
Because post-selection can generate anomalously large or unusual pointer shifts, weak measurements can sometimes amplify particular signatures.
The Toronto experiment demonstrates another feature:
The sign of a weak value itself can encode physically meaningful information about interference and conditional dynamics.
Understanding such effects could improve theoretical and experimental tools used in:
- Quantum metrology
- Quantum optics
- Precision measurement
- Light-matter interfaces
- Quantum information experiments
The important distinction is between potential relevance and demonstrated application.
This experiment establishes the former much more clearly than the latter.
Does a Photon Have a Definite History Inside the Cloud?
This question gets close to the philosophical heart of quantum mechanics.
Classically, we imagine an object following a continuous trajectory.
It was here.
Then here.
Then here.
Therefore it spent a specific amount of time in each place.
Quantum mechanics does not always permit such a simple narrative.
When alternatives remain coherent, assigning one definite history can destroy exactly the interference responsible for the observed phenomenon.
Weak measurements offer a way to ask limited questions about what happened between preparation and detection without forcing the system into the same type of definite classical history.
The resulting answers can therefore look bizarre.
A weak value may suggest a particle spent:
- More time than seems available
- Zero time
- Or a negative amount of time
Those numbers are not necessarily durations belonging to an invisible classical trajectory.
They are experimentally accessible properties of quantum histories conditioned on preparation and final outcome.
That distinction is subtle.
It is also why the experiment is scientifically valuable.
Does Quantum Mechanics Say the Future Affects the Past?
The mathematics of weak values often treats initial preparation and final post-selection symmetrically.
That can make the formalism look retrocausal.
The measurement at an intermediate time depends statistically on which later outcomes are selected.
But correlation with a future measurement is not automatically physical influence traveling backward through time.
Imagine sorting photographs today according to something you discover tomorrow.
Tomorrow's information changes how you classify yesterday's photographs.
It does not physically change what the camera recorded yesterday.
Quantum post-selection is more subtle because interference is involved, but the logical warning remains useful.
A later measurement can determine which subset of earlier weak-measurement data researchers analyze without allowing controllable messages into the past.
The Toronto experiment offers no mechanism for retrocausal communication.
What Did the Researchers Mean by “Physical Significance”?
This may be the most important phrase in the original paper.
The authors do not claim:
“Time itself became negative.”
Their argument is more precise.
The quantity called group delay can become negative.
Physicists have long known that.
But the same quantity also correctly predicts the phase shift written onto the probe beam by transmitted photons.
Therefore, negative group delay appears to govern a real experimentally measurable interaction—not merely the arbitrary location of a distorted pulse peak.
That is what gives the negative quantity greater physical significance.
It is a predictor of measurable behavior.
Whether one should describe that behavior verbally as a photon “spending negative time” in an atomic excitation remains partly a question of interpretation and terminology.
Why the Headline “Physicists Discover Negative Time” Is Misleading
It contains a piece of truth wrapped in an easy misunderstanding.
The researchers really obtained a negative time-valued quantity.
The quantity really came from experimental measurements.
It really corresponds to a physical phase shift.
And the result really survived peer review.
But they did not discover:
- Time flowing backward
- Photons returning to the past
- Effects preceding causes
- Faster-than-light messaging
- A violation of relativity
- A time machine
- A new dimension of negative time
A better headline would be:
Physicists measured a negative weak value for how long transmitted photons excite an atomic medium.
It is less dramatic.
It is also more accurate.
Ironically, once properly understood, the actual experiment is arguably more intellectually interesting than the sensational version.
Why the Result Matters for Quantum Foundations
Quantum mechanics is spectacularly successful at predicting experimental outcomes.
Yet questions about what happens between preparation and measurement remain conceptually difficult.
How long does tunneling take?
Where was a photon before detection?
How much time did energy reside in an atom?
What does a conditional quantum history mean?
Can quantities outside an observable's normal spectrum correspond to measurable physical effects?
These are not merely philosophical distractions.
They determine how physicists connect mathematical quantum descriptions to real laboratory procedures.
Steinberg's research program has spent decades probing exactly these uncomfortable questions.
The negative-excitation-time experiment contributes another piece of evidence that quantities once dismissed as mathematical oddities can sometimes encode experimentally meaningful information.
The Bigger Lesson: Quantum Mechanics Does Not Owe Us Classical Stories
Human intuition evolved to understand falling rocks, moving animals and objects traveling at ordinary speeds.
It did not evolve to interpret weak measurements of single-photon interactions with coherent atomic ensembles.
When quantum mechanics produces a negative weak value, our instinct is to force it into a familiar story.
The photon went backward.
The atom was excited before the photon arrived.
Time reversed.
Cause followed effect.
Those interpretations are seductive because they turn difficult mathematics into a movie scene.
But nature is not obligated to provide a classical story behind every quantum statistic.
Sometimes the experimentally correct statement is simply stranger:
A weak probe of atomic excitation, conditioned on successful photon transmission, produces an integrated phase shift whose sign corresponds to a negative group delay.
That sentence is harder to visualize.
It is also closer to what the laboratory actually found.
So Did Physicists Really Measure Negative Time?
Yes—with a major qualification.
They experimentally measured a quantity interpreted within weak-measurement theory as a negative weak-valued atomic excitation time.
The value was not an ordinary duration recorded by a clock.
It arose from a very weak measurement of atomic excitation combined with post-selection on photons that were successfully transmitted through the atomic cloud.
The most negative reported mean value was approximately −0.82 ± 0.31 τ₀, and the results were broadly consistent with a theoretical prediction that the excitation time for transmitted photons equals their group delay.
The finding therefore deserves neither dismissal nor science-fiction exaggeration.
The experiment does not break relativity.
It does not permit time travel.
It does not allow information to reach the past.
But it does show something genuinely strange:
A quantity that looks mathematically like a negative duration can correctly describe the sign and magnitude of a real physical interaction between photons, atoms and another beam of light.
That forces physicists to confront a deeper question.
Perhaps the mistake was never allowing time to become negative in the equations.
Perhaps the mistake was assuming every measurable “time” in quantum mechanics must behave like the time displayed on a clock.
Frequently Asked Questions About the Negative Time Quantum Experiment
Did scientists really discover negative time?
Scientists measured a negative weak value for the atomic excitation time associated with transmitted photons.
That is a real experimental result, but it does not mean ordinary time flowed backward. The quantity describes a conditional quantum measurement obtained after post-selecting photons that successfully passed through the atomic cloud.
Who performed the negative time experiment?
The peer-reviewed study was authored by Daniela Angulo, Kyle Thompson, Vida-Michelle Nixon, Andy Jiao, Howard M. Wiseman and Aephraim M. Steinberg.
Most of the experimental team was based at the University of Toronto's Department of Physics and Centre for Quantum Information and Quantum Control, while Wiseman is affiliated with Griffith University in Australia.
When was the negative time study published?
The original preprint was posted to arXiv on September 5, 2024.
The peer-reviewed paper was published in Physical Review Letters on April 13, 2026.
What is the DOI of the negative time study?
The arXiv version uses the DataCite DOI 10.48550/arXiv.2409.03680.
The final peer-reviewed Physical Review Letters article has the DOI 10.1103/gjfq-k9dv.
Was the negative time experiment peer reviewed?
Yes.
Although the research first attracted widespread attention while available as a 2024 arXiv preprint, it was later accepted and published in Physical Review Letters in April 2026.
What exactly was negative?
The weak-valued average atomic excitation time associated specifically with photons that were transmitted through the rubidium cloud became negative under narrowband resonant conditions.
The most negative reported value was −0.82 ± 0.31 τ₀.
What does τ₀ mean in the experiment?
τ₀ is the non-post-selected average atomic excitation time.
It is related to the probability that an incident photon scatters from the atoms multiplied by the atomic spontaneous lifetime.
Did a photon literally spend minus time inside an atom?
Not in the ordinary stopwatch sense.
The negative quantity is a weak value inferred from an integrated phase shift on a separate probe beam after conditioning the data on successful photon transmission.
The researchers themselves caution against treating it as an ordinary negative clock duration.
Did photons travel backward in time?
No.
The University of Toronto team explicitly states that the experiment does not show photons traveling backward through time.
Did the photons travel faster than light?
The experiment does not demonstrate faster-than-light information transmission.
Negative group delays can make a pulse's peak appear advanced because the medium reshapes the pulse, but that does not allow a new causal signal to propagate faster than light in vacuum.
Does the experiment violate Einstein’s theory of relativity?
No.
Nothing in the experiment enables superluminal transfer of usable information or a violation of causality.
The researchers explicitly reject interpretations suggesting that relativity or standard physics has been broken.
What is group delay?
Group delay describes how the center or envelope of a wave packet is shifted in time as it passes through a system.
In dispersive media, group delay can become negative because different parts of the pulse are absorbed and phase-shifted differently.
How can group delay be negative?
Near atomic resonance, the medium can attenuate and reshape different portions of the incoming pulse.
The trailing portion may be suppressed relative to the leading portion, shifting the center of the transmitted pulse forward in time and producing a negative group delay.
What is new if negative group delays were already known?
The novelty is not simply observing negative group delay.
The Toronto team independently probed atomic excitation using another laser beam and found that the measured weak-valued excitation time followed the same group delay, including its negative sign.
That suggests negative group delay has measurable physical significance beyond merely describing the position of a reshaped pulse.
What atoms were used in the experiment?
The experiment used a cloud of ultracold rubidium-85 atoms.
The atoms were cooled to roughly 60 to 70 microkelvin and confined in a magneto-optical trap.
How long is the atomic lifetime in the experiment?
The relevant rubidium transition has a spontaneous lifetime of approximately 26 nanoseconds.
What is the cross-Kerr effect?
The cross-Kerr effect allows one light field to influence the optical phase experienced by another.
In the Toronto experiment, excitation caused by the resonant signal beam altered the phase of a separate off-resonant probe beam, allowing researchers to weakly monitor the atomic excitation.
What is a weak measurement?
A weak measurement interacts only slightly with a quantum system, extracting very little information from any single trial while also producing relatively little disturbance.
By repeating the experiment many times and combining the measurements statistically, researchers can obtain information about the system's behavior.
What is post-selection?
Post-selection means retaining only experimental runs with a particular final outcome.
In this experiment, researchers focused on cases in which the signal photon was eventually transmitted through the atomic cloud.
What is a weak value?
A weak value is the conditional result obtained from a weak measurement when a quantum system is prepared in one state and later post-selected in another.
Unlike ordinary expectation values, weak values can fall outside the usual range associated with an observable, including becoming negative.
Why can a weak value be negative?
Quantum amplitudes corresponding to different possible histories interfere.
When researchers condition their analysis on a particular final outcome, destructive and constructive interference can produce a weak value outside the normal range, including a negative value.
The theoretical model associated with this experiment attributes negative excitation times to interference between different transmission pathways.
Did researchers measure individual photons directly?
The experiment used coherent signal pulses rather than relying exclusively on single-photon input states.
The team used photon detection and post-selection to isolate the inferred effect associated with an additional transmitted photon, a technique the researchers argue reproduces the corresponding single-photon result.
How many photons were in the signal pulses?
The incident coherent signal pulses contained an average of approximately 100 photons before the downstream attenuation and detection arrangement used for post-selection.
How difficult was the measurement?
Extremely difficult.
The expected phase signal from a transmitted photon was only around 10 to 20 microradians, while experimental phase noise was much larger. Individual parameter sets required tens of millions of atomic cycles and roughly 10 hours of data acquisition.
Is the result statistically perfect?
No experiment is.
The authors report overall agreement with theoretical predictions but also discuss specific parameter points that departed from predicted values and possible systematic effects involving unwanted correlations between the signal and probe.
Can negative time be used to build a time machine?
No.
The experiment provides no mechanism for sending matter, people or information into the past.
Could negative time enable faster-than-light communication?
No demonstrated method follows from this experiment.
Negative group delay does not allow controllable new information to propagate faster than light in vacuum.
Could the discovery have technological applications?
There are no immediate applications demonstrated by the experiment.
The research is primarily foundational, although improved understanding of weak measurements and light-matter interactions could eventually inform quantum optics, sensing, metrology and quantum-information research.
What happens to photons that are scattered instead of transmitted?
Theory predicts a different excitation-time relationship for scattered photons.
Their average atomic excitation time should include both a group-delay contribution and an additional positive Wigner scattering delay.
Has that prediction been experimentally tested?
The published experiment focused on transmitted photons.
Further measurements of scattered photons would provide another important test of the theoretical framework.
Does this prove that time is not real?
No.
The experiment says nothing that directly establishes whether time is fundamental, emergent or illusory.
It examines how a particular quantum-mechanical time quantity behaves during photon-atom interactions.
Does this prove quantum mechanics is wrong?
Quite the opposite.
The unusual result is described using ordinary quantum mechanics, particularly weak measurements, post-selection, quantum interference and quantum trajectory theory.
The experiment tests predictions derived from that framework rather than demonstrating its failure.
Why are physicists interested in the result?
The experiment connects a counterintuitive mathematical quantity—a negative group delay—to an independent, physically measurable effect on another optical field.
That may deepen understanding of how time, conditional histories and interference should be interpreted in quantum mechanics.
What is the biggest misconception about the negative time experiment?
The biggest misconception is that researchers watched photons move backward through time.
They actually measured a negative weak value associated with atomic excitation after post-selecting successfully transmitted photons.
It is an anomalous quantum measurement, not a time machine.
What is the simplest way to understand the result?
A transmitted light pulse can be reshaped so that its average timing shifts forward, producing a negative group delay.
The surprising discovery is that when researchers independently measured how that transmitted light affected the atoms, the measured effect carried the same negative sign.
So the negative time is not an ordinary duration.
But it is also not merely an empty mathematical illusion.
That is precisely what makes the experiment important.
