In What Time Is It? we untangled the human mess of time zones, calendars, and daylight saving. In How Clocks Work we traced the physics of the second from quartz crystals to optical lattice clocks that won’t lose a tick in the lifetime of the universe. Both of those stories treated time as something that flows at the same rate everywhere – a backdrop against which clocks are merely more or less accurate. That assumption is wrong. Time itself bends.
Einstein enters the chat
So far we’ve been treating time as though it flows at the same rate everywhere. It doesn’t.
Einstein’s special theory of relativity, published in 1905, showed that time passes more slowly for objects moving at high speeds relative to an observer. This isn’t a theoretical curiosity – it’s measurable. In 1971, Hafele and Keating flew caesium clocks on commercial airliners around the world and compared them to reference clocks on the ground. The flying clocks disagreed with the ground clocks by exactly the amount relativity predicted (Hafele & Keating, 1972, Science).
The speed of light as a universal speed limit. Nothing with mass can reach the speed of light. As you approach it, time dilation increases without bound. At the speed of light, time stops entirely – from a photon’s frame of reference (to the extent that’s meaningful), no time passes at all. A photon emitted from a star ten billion light-years away has, from its own perspective, arrived at your eye instantaneously.
Muon decay provides one of the cleanest experimental demonstrations. Cosmic ray muons are created in the upper atmosphere and should decay in roughly 2.2 microseconds, which at near-light speed would let them travel only about 660 metres. But we detect them at sea level, roughly 15 kilometres below where they were created. How? At 0.99c, their time is dilated by a factor of roughly seven. They “live” long enough to reach us. Rossi and Hall first confirmed this in 1941 (Physical Review), and it remains one of the most intuitive demonstrations of special relativity.
The twin paradox
Special relativity produces a result so counterintuitive that it has its own name. Imagine two twins: one stays on Earth, the other takes a round trip to a distant star at near-light speed. When the travelling twin returns, less time has passed for them. They are younger than their sibling. This is not an illusion or an accounting trick – it’s a real, physical difference in elapsed time.
The “paradox” label is misleading. There’s no logical contradiction. The resolution is that the two twins are not in symmetric situations: one of them accelerated (turned around), and that breaks the symmetry. The twin who stayed home followed an inertial (non-accelerating) path through spacetime, which turns out to always be the path of maximum elapsed time between two events. Any deviation from an inertial path – any acceleration – reduces the elapsed time. This is why the travelling twin ages less.
The effect doesn’t require a spaceship. The International Space Station orbits at about 7.7 kilometres per second. Astronauts on the ISS age very slightly slower than people on the ground – roughly 0.01 seconds less per year. Scott Kelly, who spent 340 days aboard the ISS in 2015-2016 while his identical twin Mark stayed on Earth, returned about 5 milliseconds younger than he would have been had he stayed home. Not enough to matter biologically. Enough to prove the physics is real.
Gravity bends time
Einstein’s general theory of relativity, from 1915, added another twist: time passes more slowly in stronger gravitational fields. The closer you are to a massive object, the slower your clock ticks relative to someone further away. A clock on the floor of your house runs very slightly slower than a clock on your roof. The difference is about 10 nanoseconds per year per metre of altitude, which doesn’t affect your morning routine but absolutely matters for GPS.
GPS satellites orbit at about 20,200 km above the Earth. Their clocks tick faster than ground clocks by about 45 microseconds per day due to weaker gravity up there. They tick slower by about 7 microseconds per day due to their orbital speed. The net effect is that satellite clocks gain roughly 38 microseconds per day relative to the ground. If this weren’t corrected, GPS positions would drift by about 10 kilometres per day. Every GPS satellite has its clock rate deliberately adjusted before launch to compensate.
This means that when you use your phone to navigate to a restaurant, you are relying on corrections derived from general relativity. Einstein helps you find pizza.
The gravitational effect has been measured with astonishing precision. In 2010, optical clocks at NIST detected the difference in time flow between two clocks separated by just 33 centimetres of altitude (Chou et al., 2010, Science). Time really does run at different speeds depending on where you are in a gravitational field. There is no single “correct” rate at which time passes. It’s always relative to something.
This has practical consequences beyond GPS. The definition of UTC itself requires a choice: the clocks that contribute to UTC are at different altitudes and latitudes, so they tick at slightly different rates due to gravity. The BIPM corrects all contributing clocks to the rate they would tick at the “geoid” – the mean sea-level gravitational potential of the Earth. A clock in Boulder, Colorado (1,655 metres above sea level) ticks faster than one in London (near sea level) by roughly 15 microseconds per year. Without the geoid correction, the ensemble average would be meaningless – you’d be averaging clocks that are physically keeping different times. The concept of “a second” on Earth is, in a gravitational sense, a political decision about which altitude to use.
The young heart of the Earth
We don’t need black holes to see gravitational time dilation at work on a grand scale. The core of the Earth, being under more gravitational stress than the surface, has experienced less elapsed time since the planet formed. The centre of the Earth is roughly 2.5 years younger than the surface – not metaphorically, but in actual measured atomic clock ticks. Time has passed more slowly down there for 4.5 billion years, and it adds up. Feynman mentioned a version of this calculation; it was rigorously computed by Uggerhoj et al. (2016, European Journal of Physics).
This is not a thought experiment. It’s a straightforward consequence of general relativity applied to the known density and gravitational profile of the Earth. If you could somehow place a clock at the centre of the planet when it formed and retrieve it today, it would show a date 2.5 years behind a clock that had spent its life on the surface. The rock beneath your feet is, in a physically meaningful sense, younger than the rock you’re standing on.
Black holes and the edge of time
Near a black hole, gravitational time dilation becomes extreme. At the event horizon – the boundary beyond which nothing, not even light, can escape – time, from an outside observer’s perspective, stops entirely. An object falling toward a black hole appears to slow down asymptotically, growing dimmer and redder, never quite crossing the horizon from the viewpoint of someone watching from a safe distance. The object falling in experiences time perfectly normally from its own point of view. Neither observer is wrong. Time is doing something different in each location.
The mathematics are well-established. The Schwarzschild solution to Einstein’s field equations, published in 1916 – just months after general relativity itself – describes the geometry of spacetime around a non-rotating, uncharged mass. It predicts that at the event horizon, the gravitational time dilation factor goes to infinity. Time, as experienced by a distant observer, literally ceases to advance for anything at the horizon.
Inside the horizon, things get stranger still. The roles of space and time effectively swap – the radial direction becomes timelike, meaning that falling inward is no longer a choice but an inevitability, in the same way that moving forward in time is an inevitability for the rest of us. The singularity at the centre isn’t a place in space. It’s a moment in time – the future that everything inside the horizon is headed toward.
Supersonic time travel
Concorde, that beautiful, impractical supersonic airliner, offered a surreal temporal experience. You could leave London at 10:30 AM and arrive in New York at 9:30 AM the same day – arriving before you departed, by clock time. The crossing took about three and a half hours, but the five-hour time difference meant you gained more than you spent.
This wasn’t relativity – it was time zones. But the special relativistic effect was real too, if tiny. Concorde flew at roughly Mach 2, about 600 metres per second. At that speed, the time dilation factor is approximately 1 + 2 x 10^-12, which means passengers aged about 0.000000002% less than people on the ground per flight. Over a career of flying Concorde, a pilot might have “saved” a few hundred nanoseconds of biological time. Not enough to notice. Enough to measure.
The more interesting effect was the experience itself. Westbound on Concorde, the sun appeared to move backwards in the sky. You were flying faster than the Earth rotates at that latitude. For the duration of the flight, you were outrunning the planet’s spin. It’s the closest any commercial passengers ever came to the intuitive experience of time running in an unusual direction.
Beautiful ideas nobody uses
Every now and then someone looks at the mess of time zones and leap seconds and local conventions and says: surely we can do better.
TAI64 is one such attempt. Proposed by Daniel J. Bernstein (the same person behind qmail and djbdns – and the plaintiff in Bernstein v. United States, the case that established code as protected speech under the First Amendment, which comes up in the story of how TLS works), TAI64 is a 64-bit representation of TAI – a simple count of seconds from a fixed epoch, with no leap seconds, no time zones, no daylight saving. It’s monotonically increasing, which means it’s ideal for log timestamps and event ordering. Every TAI64 label refers to exactly one second of real time, and the labels never go backwards or repeat. The extended form, TAI64N, adds nanosecond precision.
It’s elegant. It solves almost every practical problem with timestamps in one clean design. Almost nobody uses it.
Swatch Internet Time took a completely different approach. In 1998, the Swatch watch company proposed dividing the day into 1,000 “.beats”, with no time zones at all. The whole world would share a single time – @500 would mean the same moment for someone in Tokyo as in Toronto. The meridian was set at Biel, Switzerland (Swatch’s headquarters, naturally). One .beat is 86.4 seconds.
It was a lovely idea. Time zones exist because of the sun, but in an increasingly connected digital world, coordinating across zones creates constant friction. A universal internet time would eliminate “my 3 PM or your 3 PM?” forever. The notation was fun, the concept was sound, and it was backed by a major brand.
Nobody used it. The sun is still there. People still wake when it rises and sleep when it sets, more or less, and local time still reflects that biological reality. Swatch Internet Time lives on as a curious footnote and the occasional novelty watch face.
Both TAI64 and Swatch Internet Time failed for the same fundamental reason: they solved a technical problem while ignoring the human one. We don’t just use time to coordinate machines. We use it to coordinate lives, and lives are lived in places where the sun rises and sets at particular local times. Any scheme that ignores this is swimming against a very strong current.
It’s the same pattern I wrote about in A Gentle Guide to Typography – the technically “correct” solution (a universal encoding, a universal timescale) only wins when it also solves the human problem. UTF-8 succeeded where other encodings failed because it was backwards-compatible with ASCII. A universal time system would need to be backwards-compatible with the sun.
The deep weirdness
Step back far enough and time gets stranger still.
The arrow of time. At the quantum level, the equations of physics are mostly time-symmetric – they work just as well running backwards. Maxwell’s equations, the Schrodinger equation, even the equations of general relativity: none of them distinguish past from future. Run the film backwards and the physics still works. Yet we experience time as having a clear direction. Eggs break but don’t unbreak. You remember yesterday but not tomorrow. What gives time its arrow?
The standard answer is the second law of thermodynamics: entropy – roughly, the disorder of a system – tends to increase over time. A broken egg is more disordered than an intact one, and the path from order to disorder is overwhelmingly more probable than the reverse. But this just pushes the question back a step: why does entropy increase? The second law is statistical, not fundamental. It says that higher-entropy states are more probable, so systems tend to evolve toward them. But that only works if the universe started in a low-entropy state – a highly ordered initial condition. Why did it? This is one of the deepest unsolved problems in physics, and it sits at the intersection of cosmology, thermodynamics, and the foundations of quantum mechanics. Roger Penrose has devoted much of his career to it; he estimates the probability of the universe’s initial low-entropy state arising by chance at roughly 1 in 10^(10^123) – a number so absurdly large that writing it out would require more paper than exists in the observable universe.
The block universe. At the cosmic level, time is inseparable from space. General relativity describes them as a single four-dimensional fabric – spacetime – that can be curved, stretched, and warped by mass and energy. The notion of “now” is surprisingly hard to define across large distances. In special relativity, simultaneity is relative: two events that happen at the same time for one observer happen at different times for another moving at a different speed. There is no universal “now”. There is only your now, defined by your position and velocity, and it disagrees with everyone else’s.
This leads some physicists to the “block universe” interpretation: the idea that past, present, and future all exist equally and simultaneously. The four-dimensional spacetime block simply is – complete and unchanging. What we experience as the flow of time is an artefact of our consciousness moving through this block. In this view, the future is as real as the past; we just haven’t encountered it yet. It’s a view that Einstein himself appears to have held. After the death of his lifelong friend Michele Besso in 1955, Einstein wrote to Besso’s family: “For those of us who believe in physics, the distinction between past, present, and future is only a stubbornly persistent illusion.”
Does time exist at all? Some physicists have gone further. Julian Barbour, in The End of Time, argued that time doesn’t fundamentally exist. What we call time is just the way we experience the relationships between configurations of matter. The universe doesn’t evolve through time; it simply is a collection of states, and our brains string them into a narrative. Carlo Rovelli, in The Order of Time, takes a related but more nuanced position: time as we experience it – flowing, universal, directed – is an emergent property that arises from our limited perspective as macroscopic beings who interact with the world thermodynamically. At the most fundamental level of quantum gravity, the equations may contain no time variable at all. The Wheeler-DeWitt equation, which attempts to describe quantum gravity, is notably time-independent – it describes a static universe in which nothing happens. How a timeless equation produces our very much time-full experience is, to put it mildly, an open question.
This is philosophy as much as physics, and it’s nowhere near settled experimentally. But it illustrates how deep the rabbit hole goes. We started with a simple question – “what time is it?” – and ended up questioning whether time exists at all.
So what time is it?
After all of this – the human history of sundials and railways and political time zones, the physics of caesium atoms and clock ensembles, the relativity that bends time near massive objects and at high speeds – the answer is: it depends.
It depends on where you are in a gravitational field. It depends on how fast you’re moving. It depends on which timescale you’ve chosen and why. It depends on whether you care about the sun’s position, or the purity of atomic seconds, or the agreement between your timestamp and everyone else’s.
The phone in your pocket hides all of this heroically. It receives signals from GPS satellites that have been corrected for both special and general relativistic effects. It knows your time zone from your location. It knows about DST transitions from a regularly updated database. It adjusts for leap seconds, or at least it tries to. It presents you with a number that looks simple and authoritative, and you glance at it and get on with your day.
Underneath, it’s leaning on millennia of astronomy, centuries of mechanical engineering, decades of atomic physics, and Einstein. It’s a tower of clever hacks and hard-won compromises, and it’s a miracle it works at all.
What time is it? It’s always, and only, approximately now.