Time Travel Paradoxes Explained: Can Physics Let Us Change the Past?

If You Could Change Yesterday, Would Tomorrow Still Exist?

Time travel paradoxes, Einstein’s universe, and the science of going back.

Imagine receiving a letter in your own handwriting. It is dated 2056, thirty years from now. Inside is a warning: “Whatever you do, do not open the blue door tomorrow.” Naturally, the first thing you want to know is what is behind that door. The second is how your future self knew you would be tempted.

Now imagine you ignore the warning. Behind the door is a working time machine. You use it to travel to yesterday and make sure your younger self never opens that door. But if you never find the machine, you never make the trip. So who came back to stop you? One simple decision has turned into a contradiction. The difficulty is not the technology. It is asking the same event both to happen and not to happen.

This is why time travel has a place in serious physics, even without a single working time machine. Einstein showed that time does not tick at the same rate for everyone. Some solutions to his equations even contain routes into the past. But a possible universe on paper is not necessarily a universe nature will allow. The real questions are sharper than science fiction usually admits: Can such a route exist? Could you change what already happened? Or would the laws of physics stop you before the journey began?

A traveler faces a glowing time portal connecting a historical city with a futuristic city, illustrating the mysteries of time travel.
What would happen if we could travel into the past? Time travel paradoxes challenge our understanding of causality, history, and the laws of physics.

The first surprise: traveling into the future is already real

The easiest kind of time travel is also the least cinematic: you take a journey, return, and discover that more time has passed for everyone else than for you. This is a consequence of special relativity, not a loophole invented for science fiction. Two people following different paths through spacetime can accumulate different amounts of time between their meetings.

Think of two identical twins. One remains on Earth. The other boards an extraordinarily fast spacecraft, turns around, and comes home. Each has lived every second of their own life normally; neither notices their wristwatch acting strangely. Yet the traveling twin can arrive younger than the one who stayed behind. The difference is not a trick of appearance. The clocks—and the biological processes they measure—have genuinely accumulated different elapsed times.

Here is a striking number. At 99% of light speed, five years measured aboard a spacecraft correspond to about 35 years in Earth's frame, in a simplified calculation that ignores acceleration, turning around and gravity. Your family could age by decades while you age by five years. No crewed spacecraft can approach that speed today. But the difference between those clocks follows from physics that has passed countless experimental tests.

Gravity changes clock rates, too. Near a massive body, where gravity is stronger, a clock generally ticks more slowly than a comparable clock farther away. Researchers have measured the resulting gravitational redshift across a millimeter-scale sample of atoms. And your phone depends on the same basic physics: GPS satellite clocks lose about 7 microseconds per day because of their speed but gain about 45 because gravity is weaker in orbit. The net difference is roughly 38 microseconds a day. Engineers must account for it to keep navigation accurate.

The important distinction is that this is one-way time travel into somebody else’s future. You cannot choose to return to Earth on the morning before launch. Speed and gravity let clocks disagree about how much time passed. They have not been shown to reverse the order of cause and effect. For a different science-fiction way to skip the years between stars, our guide to interstellar hibernation examines the biology of sleeping through a journey rather than changing the physics of time.

Why traveling into the past is a different problem

In ordinary life, we think of time as a one-way street. First you drop a glass; then it shatters. You can film the pieces and play the video backward, but the reversed sequence is not what happens spontaneously on your kitchen floor. More importantly, yesterday’s events help determine today’s world. Our memories, records, and very existence depend on that history.

A journey into the past would be more radical than arriving in 2100 without aging much. It would put an agent or an object where it could influence events that, from its own perspective, have already happened. This creates a conflict between local action—what a traveler seems able to do at a particular moment—and the complete history those actions must fit into.

The key word is causality: the rules determining which events can influence which others. A time paradox appears when a journey seems to make those rules contradict themselves. But a strange circle of causes is not always a logical impossibility. Some loops are perfectly consistent, even if they leave us asking where the story began. That difference matters.

The grandfather paradox: can you erase your own existence?

The classic example is brutally simple. You travel to a year before your grandfather meets your grandmother and prevent their relationship. Your parent is never born. Consequently, neither are you. But if you never existed, who took the trip to interfere?

You do not need violence—or even a family—to make the paradox work. Suppose you travel back five minutes and disable the time machine just before your younger self activates it. If the machine is disabled, you cannot have made the trip that disabled it. The same contradiction appears in miniature.

It is tempting to say that the universe would simply rewrite itself. You might imagine a ripple traveling forward, removing photographs, records, memories, and eventually the traveler. Movies can show that process with wonderfully unsettling visuals. But physics does not provide a tested mechanism for a history that is first one way, then is globally edited into another. To speak of the entire timeline changing raises an awkward question: changing when, according to what additional clock outside time?

And what about the butterfly effect? Chaos theory shows that tiny differences in initial conditions can sometimes grow into large differences later. Miss a train by ten seconds, and you might meet a different person or miss a life-changing conversation. In a fictional journey to the past, small actions could have sprawling consequences. Yet the butterfly effect says nothing about whether the journey is possible. Nor can it explain how you could prevent the very trip that brought you there.

A traveler from 2026 enters a glowing portal leading to a 1960s street, with arrows representing a paradoxical loop of cause and effect.
The grandfather paradox asks what happens when a time traveler changes the past in a way that prevents their own journey from ever taking place.

The bootstrap paradox: a book that nobody wrote

The grandfather paradox threatens to make something both happen and not happen. The bootstrap paradox is stranger because its events might all fit together without contradiction.

Picture a physicist in 2050 receiving a notebook from a visitor from 2080. It contains the equations needed to build a time machine. She publishes them and becomes famous. Thirty years later, she gives her old notebook to a younger visitor, who carries it back to her in 2050. The equations have traveled around a closed loop. At every step, someone obtained the notebook from someone else. But who discovered the equations in the first place?

Or imagine a songwriter carrying a famous tune back to a time before it was composed. A young musician hears it, releases it, and becomes the celebrated author whose recording inspired the traveler. The song seems to exist without a first composer. Information has a history, but no obvious beginning within that history.

This is the bootstrap paradox, sometimes called an ontological paradox. Notice that nobody needs to disappear, and no event must both occur and fail to occur. The loop might be consistent, yet the information has no obvious inventor. There is another complication: a physical notebook would age and wear out. An endlessly circulating object is harder to imagine than a set of equations copied onto fresh paper at each turn. A story can ignore that detail; a physical model cannot simply assume it away.

The predestination paradox gives the loop a more personal twist. A traveler tries to prevent a disaster because they remember it happening; their intervention turns out to be one of the causes of the very disaster they hoped to avert. They did not change history. They helped make it happen. Stories such as Dark and Predestination explore related ideas; theoretical physics asks what, if anything, could make them consistent.

Einstein’s equations contain some very strange roads

How did time travel move from fiction to serious physics? The answer lies in general relativity. Einstein replaced the idea of a rigid stage called space and time with spacetime, a geometry shaped by matter and energy. Objects and light follow paths through that geometry. In everyday conditions the paths behave predictably, and the future remains separate from the past.

Some mathematically valid spacetime geometries, however, contain a closed timelike curve, usually shortened to CTC. Imagine a traveler following a perfectly continuous path forward according to their own clock. The path bends through the geometry of spacetime and eventually reaches an event in the traveler’s own past. From the traveler’s local point of view, they never drove faster than light or watched their clock tick backward. Yet the route itself curves back to an earlier event.

Picture walking around Earth's equator. You keep moving forward, yet eventually return to the same place. A closed timelike curve is far stranger: the route loops through spacetime, not just space. If you could follow it, you might return to an event you have already lived through, even while your own watch continues to move forward. The equator is only an analogy, not a recipe for bending time.

In 1949, mathematician Kurt Gödel found a solution to Einstein’s equations describing an idealized rotating universe with closed timelike curves. This did not establish that our universe rotates that way, or that time machines can be built. It established something subtler: the equations of general relativity, by themselves, do not exclude every geometry that permits a return to the past.

Other theoretical proposals involve extreme rotating objects, engineered warp geometries, and wormholes. Many require idealizations or physical conditions we have no evidence can be realized. A result can be mathematically legitimate and still describe a universe that nature refuses to create.

Wormholes: the time machine hidden inside a shortcut

A wormhole is a proposed connection between two regions of spacetime. Think of a tunnel joining far-apart places, though the real geometry is vastly more complicated than a folded-sheet-of-paper drawing. No traversable wormhole has ever been observed, and we do not know how to create or stabilize one.

In 1988, Michael Morris, Kip Thorne, and Ulvi Yurtsever explored a remarkable conditional idea. If an advanced civilization could maintain a traversable wormhole, then suitably changing the motion or gravitational environment of one entrance might put its two ends out of sync in time. Passing through could let a traveler emerge at an earlier external date than the date on which they entered.

Imagine that two connected wormhole mouths begin with synchronized clocks. One mouth takes a high-speed trip and returns, so its clock has accumulated less time than the stationary mouth’s clock. If the wormhole’s internal connection maintains the relevant time relationship assumed by the model, the mouths now link different external times. Under suitable arrangements, a journey through the wormhole plus an ordinary trip outside it could form a loop into the past.

That is a fascinating mathematical proposal, not an available machine. Traversable wormholes run into severe requirements involving exotic stress-energy and restrictions from quantum field theory. Ordinary matter does not simply hold open a useful tunnel. Quantum effects that permit small negative-energy densities under specific circumstances do not imply that a human-sized wormhole can be supported. For a more detailed separation of spacetime shortcuts from actual human transport, see our science-first guide to teleportation and wormholes.

There is a popular misconception worth fixing: even an engineered wormhole time machine would not necessarily be a ticket to Ancient Rome. In typical schemes where the temporal offset is created by manipulating a pair of mouths, the reachable past is constrained by the wormhole’s history and the time-machine setup. “Visit any date ever” does not follow from the model.

A theoretical wormhole connects two regions of curved spacetime, with different clock readings illustrating a possible time difference between its entrances.
Some theoretical models suggest that wormholes with differently synchronized entrances could connect distant moments in time. No traversable wormhole has ever been observed.

Solution one: the past is self-consistent and cannot be changed

One influential response is associated with physicist Igor Novikov and collaborators. It is commonly called the self-consistency principle. The idea is not that every action is easy to predict. It is that the only physical histories permitted in a spacetime containing a time loop are histories that fit together without contradiction.

Suppose you travel to yesterday to stop yourself from discovering the machine. You pull a cable, thinking it will shut everything down. But the noise draws your younger self to the hidden room; that is how they discover the machine in the first place. You have not changed the history you remember. You have helped produce it. Under this idea, there is no original version of yesterday that existed before your arrival. You were always there.

Does that mean fate is secretly controlling your hands? Not in the mathematical argument. The idea is that only complete histories without contradictions count as possible solutions. It does not tell us why a particular cable breaks or guarantee that every apparently paradoxical plan has a solution. It asks whether the whole chain of events can fit together.

Physicists have tested the logic with surprisingly ordinary objects. In a 1991 study, Fernando Echeverria, Gunnar Klinkhammer, and Kip Thorne investigated a billiard ball entering a wormhole, traveling backward in time, and colliding with its younger self. At first the setup looks paradoxical: if the older ball knocks the younger one away from the entrance, how does the older ball ever come out? Yet certain trajectories are self-consistent. A glancing collision can nudge the younger ball into the path that eventually produces the older ball.

These are mathematical models, not observed billiard balls arriving from the future. They also do not prove that every possible initial condition has a unique consistent outcome. Their importance is conceptual: a situation that looks contradictory at first glance may have a consistent solution when the full loop is considered.

In 2020, Germain Tobar and Fabio Costa further examined mathematical frameworks allowing nontrivial closed timelike curves while preserving reversible dynamics and avoiding inconsistencies. This work did not show that real people can change the past; it explored what a consistent theory could look like if such curves existed.

Solution two: changing the past creates another history

A second possibility lets the traveler make a genuinely different choice without erasing their own departure. Imagine going back and preventing your grandparents from meeting. The branch you enter might contain no version of you born to that family, while the history you departed from still exists. The two outcomes no longer contradict each other because they happen in different histories.

This is the branching-timeline solution beloved by fiction. It is also the subject of actual mathematical investigation. In a 2020 Physical Review D paper, Jacob Hauser and Barak Shoshany developed models with multiple histories in which paradoxes can be treated using branching spacetime or covering-space constructions. Their work asks how many histories are needed, whether some can reconnect, and how the models could differ in principle.

There is an important catch: a set of possible histories in a mathematical model is not evidence that our universe creates parallel timelines when somebody moves a chair in 1850. The many-worlds interpretation of quantum mechanics is another separate idea. It provides a way to understand quantum outcomes, but it does not by itself supply an engine that takes a person to the past or prove that altered histories are accessible.

There is also a distinct quantum approach. In 1991, physicist David Deutsch developed a model in which a quantum system interacting with a closed timelike curve must reach a self-consistent state: in simplified terms, what comes out of the loop must fit what goes in. The mathematics can avoid classical contradictions, but it makes unusual assumptions about how quantum systems behave. It is not the same thing as proving that parallel universes exist—or that a quantum computer can create a time portal.

There is a haunting consequence. You might save someone in a new history without undoing their loss in the world you left. The paradox would be gone, but the original tragedy might remain. Fiction loves that possibility for a reason: it solves a logical puzzle while creating a human one.

Scientific diagram comparing three hypothetical time travel scenarios: a self-consistent time loop, a branching timeline, and a blocked closed timelike curve.
Could history remain unchanged, split into alternative timelines, or prevent backward time travel entirely? These three scenarios illustrate different theoretical approaches to time travel paradoxes.

Solution three: Hawking’s universe protects causality

Stephen Hawking proposed a less accommodating possibility: perhaps the laws of nature prevent closed timelike curves from forming in the first place. His 1992 chronology protection conjecture suggested that quantum effects and the back-reaction of matter and energy on spacetime could spoil a would-be time machine before it becomes operational.

One possible mechanism is almost like acoustic feedback. Place a microphone too close to a loudspeaker and the sound can build into a piercing squeal. In certain theoretical spacetimes, radiation circulating around a nearly completed time loop could produce large quantum-energy effects. The resulting gravitational back-reaction might spoil the geometry before the loop becomes usable. This is an analogy to particular calculations, not a demonstrated universal safety system built into the cosmos.

Hawking’s proposal is powerful partly because it does not require a cosmic rulebook saying, “You may travel, but you must never change anything.” It says that the dangerous route may never exist in a physically realizable universe. Subsequent work, including studies by Matt Visser, has examined ways wormholes and time machines might destabilize themselves.

The honest status is unresolved. Some strong arguments support chronology protection under particular conditions. There is still no generally accepted proof that all possible backward-time machines are forbidden. That uncertainty is not a promise that one can be built. It reflects a deeper problem: we do not yet have a complete experimentally confirmed theory uniting gravity with quantum physics in the regimes where these questions matter most. Our guide to string theory and quantum gravity explores one ambitious, still unconfirmed attempt to bridge that gap.

What scientists actually tested in 2025 and 2026

Recent headlines sometimes make it sound as though researchers have already sent a message into yesterday. A closer reading reveals a more careful and more interesting story: physicists are studying the information-processing rules of hypothetical time loops and experimentally simulating some of those rules in ordinary laboratories.

A 2025 paper in Physical Review D by Achintya Sajeendran and Timothy Ralph described closed timelike paths in a modified rotating warp-drive spacetime, allowing researchers to analyze how a freely moving particle might interact with a loop for a finite interval. The authors explicitly cautioned that the physical realizability of the geometry was questionable. It was a theoretical model, not a warp-drive demonstration.

In April 2026, Yi-Te Huang and colleagues described an experiment on Quantinuum and IBM quantum processors. They used postselection: after an experiment, they kept only the outcomes that met a specified condition. The selected results can be interpreted mathematically as if scrambled information had been decoded through a consistent time loop. That is an unusual way to study quantum information. It does not let anyone read tomorrow's message today.

In June 2026, Kaiyuan Ji, Seth Lloyd and Mark Wilde studied a related question on paper: if a mathematical channel could carry information backward in time, how much information could survive noise? They calculated limits within that model. The result is about what the equations permit under their assumptions, not a communications link anyone has built.

Most recently, on October 19, 2026, Ran Yang and colleagues published an experiment with photons. They used a framework called an erasure-based closed timelike curve to help characterize how quantum processes work. Behind the futuristic name is a measurement method involving quantum correlations, an auxiliary system and carefully chosen projections. The experiment demonstrated a way to describe and probe quantum operations in a laboratory—not a physical tunnel connecting two dates.

These studies matter because theoretical time loops raise serious questions about consistency, quantum information, and how future and past conditions constrain a calculation. They do not mean a quantum computer can send tomorrow’s lottery numbers to its operator today. There is a gulf between simulating a model that resembles time travel and physically making information arrive before it was sent.

A scientist adjusts optical equipment in a quantum photonics laboratory, with a glowing infinity-shaped graphic symbolizing theoretical time loops.
Quantum researchers investigate mathematical models of closed timelike curves using advanced simulations and photonic experiments. These studies do not demonstrate actual travel into the past.

Does the arrow of time forbid a return to yesterday?

There is one last puzzle: why does ordinary time feel so one-directional? You remember yesterday, not tomorrow. Coffee cools; eggs break; you never watch a scrambled egg leap back into its shell. Thermodynamics connects this familiar arrow of time to entropy. Very roughly, entropy reflects how many microscopic arrangements are compatible with what we see, and it tends to increase for isolated systems.

The thermodynamic arrow is one reason ordinary life unfolds in a direction we recognize. But it is not, by itself, a proof that closed timelike curves are mathematically impossible. The equations of many microscopic processes work similarly when time is reversed; a cosmic time machine would be a question about global spacetime geometry and the consistency of physical states, not simply the reversal of a movie clip.

Could a traveler experience events backward, becoming younger second by second? That is not what a CTC predicts. A traveler following such a path still experiences their own time moving forward: meals are digested, memories form, their watch advances. The oddity is where the path carries them in the larger history. Likewise, observing a galaxy as it looked millions of years ago is not visiting that galaxy’s past; it is receiving light that took millions of years to arrive.

We should also be careful with quantum jargon. Entanglement does not provide a faster-than-light messaging service, and experiments involving delayed choices are not proof that people can change decisions already made. Genuine quantum mysteries are interesting enough without borrowing powers that the experiments have not shown.

So what would happen if you stepped out in 1926?

Let us grant the impossible-looking device for a moment. You open a door in 2026 and walk out into a city exactly a century earlier. Three broad endings correspond to the major ideas above.

A fixed history: You can walk the streets, speak to people, and leave footprints. But everything you do was always part of 1926. Any apparent attempt to change what you know about history will turn out consistent with what actually happened. This is the Novikov-style possibility.

A different history: Your arrival initiates or accesses a separate set of events. Your decisions genuinely matter in that world, but they do not rewrite the world you left. This is the multiple-histories possibility, not an experimentally established fact.

No arrival: The underlying spacetime geometry cannot be created or maintained. The machine fails, the tunnel collapses, or the required matter distribution proves impossible. This is the broad spirit of chronology protection, though no single failure mechanism has been universally established.

Notice what is missing: a scientifically established rule saying the universe will erase you in a flash if you move one object, or that every tiny change must instantly dissolve the future. Those are narrative choices, not predictions of confirmed physics. We simply do not know what a real backward-time journey would entail because no such journey has been demonstrated.

The question that makes time travel worth studying

A time machine would be an astonishing invention, but the deeper mystery is older than any machine: why can the universe contain mathematical descriptions in which a cause loops back toward itself? Why do our actual experiences come with such a firm division between memory and expectation? And are the equations of general relativity telling us something profound about time—or showing us the edge of a theory that still needs quantum gravity?

A convincing reason why time loops can never form would teach us something profound about the universe. So would evidence that one can form without contradiction. The question is useful precisely because it forces relativity, quantum theory and our everyday idea of cause and effect into the same room.

So return to that letter on the table. If the past is fixed, the warning may be exactly what sends you to the blue door. If histories branch, opening it might start a different story without erasing your own. If Hawking's conjecture is right, no working door to yesterday will ever appear.

What can we say with confidence today? Different clocks really can experience different amounts of time. Some solutions of Einstein's equations contain paths into the past. No experiment has shown a person, an object or a usable message traveling backward through time. The space between those three facts is where some of the most fascinating questions in science still live.

FAQ: Time travel and temporal paradoxes

Is time travel scientifically possible?

Travel into the future through time dilation is consistent with well-tested physics. Travel into the past appears in certain mathematical models of general relativity, but it has not been observed or demonstrated to be physically possible.

What is the grandfather paradox?

It is the contradiction created when a traveler changes the past in a way that prevents the trip that produced the change. A simple version is going back five minutes to stop yourself from activating the machine you used.

What is the bootstrap paradox?

It is a closed causal loop in which information or an object appears to have no original source—for example, a future scientist giving their younger self the only copy of a formula they will later publish.

Can the past be changed without a paradox?

Some mathematical models consider self-consistent histories; others explore multiple histories in which a changed past belongs to a different branch. Neither has been experimentally verified as a mechanism for backward time travel.

Did quantum computers prove that time travel is real?

No. Experiments can simulate features of theoretical closed timelike curves using quantum circuits or photonics. They do not establish that objects or controllable messages traveled into the real past.

Could a time machine visit the dinosaurs?

No verified time machine exists. Furthermore, proposed wormhole schemes generally restrict how far backward a traveler could go according to the wormhole’s own history; they do not automatically offer access to arbitrary prehistoric dates.

Does Hawking’s chronology protection conjecture prove time travel is impossible?

No. It is a serious proposal that quantum effects may prevent time machines from forming, supported in certain theoretical settings. It is not a universally proven prohibition covering every conceivable spacetime.

Next Horizon Guides

Explore Topics

Explore More Topics

Go deeper with curated guides that connect our best reporting around a single subject.

Dive into our in-depth guides and discover more stories around the ideas shaping our world.

Comments