16 August,2026 08:52 AM IST | Mumbai | Nishant Sahdev
Speed tells you time. And note which way it runs — the faster the car, the more recently it left. The universe is the same. We can see roughly how far apart things are and how fast they’re separating. So a faster-growing universe is a younger one. Pics/iStock
Somewhere above your roof tonight there are stars that appear to be older than the universe they sit in. Which is impossible, obviously. Nothing can be
older than everything. But that is what the numbers have been saying for about ten years now, and no one has managed to make the problem go away.
Scientists have been hoping somebody made a mistake. A mistake would be the easy ending. Someone's instrument was slightly off, we find the error, we fix it, and everybody goes back to work. That's how these arguments normally finish.
This one isn't finishing that way, and I've come to think that's the most interesting thing happening in science right now.
Two answers, a billion years apart
There are two ways to work out how old the universe is. One method says about 13.8 billion years. The other says something closer to 12.8 billion. A billion years apart. Both groups are careful. Both have checked their work for a decade. Neither answer will move.
Neither method measures age directly - you can't ask the universe for a birth certificate. What both do is measure how fast it's growing, and that turns out to be the same question. Picture a car driving away from you. Say it's now sixty miles down the road. If it's traveling sixty miles an hour, it left an hour ago. If it's crawling at thirty, it must have set off two hours ago to have gotten that far.
Speed tells you time. And note which way it runs: the faster the car, the more recently it left. The universe is the same. We can see roughly how far apart things are and how fast they're separating. So a faster-growing universe is a younger one.
Astronomers call this growth rate the Hubble constant. One group measures it at about 73, the other at about 67. Ignore the units. Just remember: 73 means a younger universe, 67 an older one. One of these numbers was never measured. The two methods could hardly be more different. The first group measures, in steps, the way you'd measure a long road with a short tape measure. Start with stars close enough to judge directly. Use those to work out the true brightness of a kind of star that pulses like a lighthouse. Use those to reach exploding stars visible almost anywhere. Slow, unglamorous work. It gives you 73.
The second group doesn't measure the growth rate at all. They predict it. There's a faint glow filling the whole sky, left over from when the universe was 3,80,000 years old - in effect, a baby photograph. You hand that photo to our best theory of how the universe works, let the theory grow the baby up, and ask what the growth rate should be today. The theory answers 67.
So this was never two rulers disagreeing. One side made a prediction, and the sky declined to match it. The thing on trial isn't anyone's telescope. It's the theory. Everyone assumed somebody was being sloppy.
We have been here before, and last time it was embarrassing. In the 1980s the field was split between two famous astronomers. Allan Sandage was certain the number was around 50. Gérard de Vaucouleurs was equally certain it was 100. They did not like each other. The true answer eventually turned up in the middle, roughly where a cynic would have guessed.
So everyone expected this round to end the same way. Two proud teams, two dusty telescopes, one boring average. That hope mostly died this April. A large group redid the measurement using several kinds of star at once, so no single step could be blamed, across three of the world's best telescopes.
One was the James Webb, which sees in infrared - meaning it looked straight through the space dust that had been everybody's favourite suspect.
The answer came back at 73.5, with an error of less than one percent. It hadn't moved.
Then a different group went looking for an independent opinion. They took the oldest stars they could find - ones burning slowly enough to have survived since nearly the beginning - and dated them one by one. More than 155,000 of them. The oldest came in at about 13.7 billion years. Allow a couple hundred million years at the start for the first stars to switch on, and you get a universe of roughly 13.8 billion years.
That's the older answer - the one supposedly in trouble. And here's the trap. Some of the leading ideas for fixing all this work by adding new physics to the universe's earliest moments. They speed the universe up, which makes it younger - around 12.9 billion years.
In that universe, the stars we just spent years carefully dating could not exist. They would be older than the thing they're sitting inside. To fix today, you have to change the beginning. The last twist: any repair probably has to be made to the distant past. When the universe was very young it was a hot fog, and ripples ran through it like sound through air. As it cooled and cleared, those ripples froze in place, leaving a pattern in the sky of one particular size. Astronomers use that frozen pattern as a ruler. Almost everything about the 67 answer is measured against it.
To make today's number bigger, you generally have to make that ruler smaller. Which means rewriting what happened in the universe's first few hundred thousand years - long before there was a single star, long before there was anything to point a telescope at?
And nothing comes free. Nearly every fix that solves this problem creates another one elsewhere.
What I keep coming back to isn't the physics. It's how people behaved. For ten years, both sides have spent most of their energy attacking their own work. Adam Riess's team hunted its own measurements for errors with the enthusiasm of a hostile auditor. Wendy Freedman argued in public, repeatedly, that the whole problem might be an illusion created by her own field's methods. Nobody fudged a number. Nobody let a result quietly drift toward the popular answer to make the discomfort stop.
The gap got sharper precisely because everyone tried honestly to make it disappear and failed. I'm not sure there's a better picture of how finding things out is supposed to work - or of how rarely we manage it.
The likeliest ending is not that somebody turns out to be wrong. It's that everyone is right, and something is missing from the theory that connects them. The last time this happened, it ended with the discovery of dark energy and a Nobel Prize.
In the meantime, there are stars over your house tonight that looks older than time itself. Somebody's arithmetic is going to have to give.
Nishant Sahdev is a physicist at the University of North Carolina at Chapel Hill, US. He chases ideas that refuse to stay in laboratories, for your favourite Sunday mid-day.
