Some stars are older than the universe. That’s the good news

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 beolder 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 awayScientists have been hoping somebody made a mistake A mistake would be the easy ending Someonersquos instrument was slightly off we find the error we fix it and everybody goes back to work Thatrsquos how these arguments normally finishThis one isnrsquot finishing that way and Irsquove come to think thatrsquos the most interesting thing happening in science right nowTwo answers a billion years apartThere are two ways to work out how old the universe is One method says about 138 billion years The other says something closer to 128 billion A billion years apart Both groups are careful Both have checked their work for a decade Neither answer will moveNeither method measures age directly mdash you canrsquot ask the universe for a birth certificate What both do is measure how fast itrsquos growing and that turns out to be the same question Picture a car driving away from you Say itrsquos now sixty miles down the road If itrsquos traveling sixty miles an hour it left an hour ago If itrsquos crawling at thirty it must have set off two hours ago to have gotten that farSpeed 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 theyrsquore separating So a fastergrowing universe is a younger oneAstronomers 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 yoursquod 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 73The second group doesnrsquot measure the growth rate at all They predict it Therersquos a faint glow filling the whole sky left over from when the universe was 380000 years old mdash 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 67So this was never two rulers disagreeing One side made a prediction and the sky declined to match it The thing on trial isnrsquot anyonersquos telescope Itrsquos the theory Everyone assumed somebody was being sloppyWe 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 Geacuterard 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 guessedSo 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 worldrsquos best telescopesOne was the James Webb which sees in infrared mdash meaning it looked straight through the space dust that had been everybodyrsquos favourite suspectThe answer came back at 735 with an error of less than one percent It hadnrsquot movedThen the stars spoke upThen a different group went looking for an independent opinion They took the oldest stars they could find mdash ones burning slowly enough to have survived since nearly the beginning mdash and dated them one by one More than 155000 of them The oldest came in at about 137 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 138 billion yearsThatrsquos the older answer mdash the one supposedly in trouble And herersquos the trap Some of the leading ideas for fixing all this work by adding new physics to the universersquos earliest moments They speed the universe up which makes it younger mdash around 129 billion yearsIn that universe the stars we just spent years carefully dating could not exist They would be older than the thing theyrsquore 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 itTo make todayrsquos number bigger you generally have to make that ruler smaller Which means rewriting what happened in the universersquos first few hundred thousand years mdash long before there was a single star long before there was anything to point a telescope atAnd nothing comes free Nearly every fix that solves this problem creates another one elsewhereNobody cheatedWhat I keep coming back to isnrsquot the physics Itrsquos how people behaved For ten years both sides have spent most of their energy attacking their own work Adam Riessrsquos 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 fieldrsquos methods Nobody fudged a number Nobody let a result quietly drift toward the popular answer to make the discomfort stopThe gap got sharper precisely because everyone tried honestly to make it disappear and failed Irsquom not sure therersquos a better picture of how finding things out is supposed to work mdash or of how rarely we manage itThe likeliest ending is not that somebody turns out to be wrong Itrsquos 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 PrizeIn the meantime there are stars over your house tonight that looks older than time itself Somebodyrsquos arithmetic is going to have to giveNishant 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 midday