IT WAS fun while it lasted. Last March, cosmologists celebrated what seemed like evidence that space-time had shaken violently during the big bang. The discovery of the apparent gravitational waves was hailed as the 鈥渟moking gun鈥� for a theory that the infant universe experienced an epic growth spurt known as inflation. Physicists in elation and dreamed of a Nobel prize.
But 11 months later, this smoking gun has itself gone up in smoke, and researchers are nursing a hangover. 鈥淲e are pretty much back to where we were before,鈥� says of the Massachusetts Institute of Technology, who proposed the theory of inflation in 1981.
It all started on 17 March, when astronomers using a telescope called at the South Pole reported seeing telltale signs of gravitational waves in a tiny patch of sky viewed at a particular wavelength in the microwave range. That was exciting because although inflation should have produced gravitational waves, there was no guarantee they would be strong enough to detect. There are hundreds of models of inflation, each with its own prediction about how fast the universe expanded 鈥� and therefore how powerful the resulting gravitational waves would be.
Advertisement
So it was even more astounding that the strength of BICEP2鈥檚 waves fitted the simplest version. In this model, inflation proceeded more or less like a ball rolling down the inside of a U-shaped bowl, with expansion starting fast then slowing down. 鈥淭he simplest theory yields predicted gravitational waves right where BICEP2 seemed to see them,鈥� says at the University of Edinburgh in the UK.
But the excitement was short-lived. A series of studies soon suggested that dust within our galaxy may have muddied the picture. Observations at other wavelengths were needed to clear up the confusion, as dust shines more brightly at certain wavelengths than at others.
In September, researchers used Europe鈥檚 to show that BICEP2鈥檚 entire signal could be due to dust (, ). The final nail in the coffin came last week, when a study combining BICEP2 and Planck data showed that the dust observed with Planck lined up with the signal the BICEP2 team had attributed to gravitational waves. This all but rules out the simplest model: if they are out there, any gravitational waves from inflation must have been no more than about half as strong as those seen with BICEP2, in line with the Planck team鈥檚 earlier estimates.
鈥淧lanck showed BICEP2鈥檚 entire signal could be due to dust. The final nail in the coffin came last week鈥�
Upcoming observations will put more models to the test by improving measurements of possible contaminants. That means studying the sky with great sensitivity at a range of places and microwave wavelengths. 鈥淭o claim a detection of a primordial signal, one has to exclude the possibility, to the fullest extent possible, that something else hasn鈥檛 generated the signal,鈥� says William Jones of Princeton University, who leads a balloon-borne mission called SPIDER that is expected to release its observations in a year or so.
Gravitational waves are not inflation鈥檚 only prediction. For example, ultra-fast expansion can explain how the universe, which could have started out with any curvature, came to appear so flat. But a rival theory, which says the universe cycles between periods of expansion and contraction, can also account for those mysteries, says of Princeton, a pioneer of both inflation and its cyclic competitor.
The cyclic model predicts that we should not see any gravitational waves from the early universe, so BICEP2鈥檚 initial announcement seemed to deal it a fatal blow. Now, without a discovery of gravitational waves, the theory of inflation has lost its most powerful line of evidence. 鈥淭he current non-detection certainly does not rule inflation out, but equally, without a detection, many, including me, would not consider the theory to be proved true,鈥� says Peacock.
Steinhardt fears that inflation is so flexible it cannot be proved false. Once started, inflation is hard to stop, and should have spawned a zoo of universes, each with different properties. 鈥淎ny result can fit somewhere in the multiverse,鈥� Steinhardt says. Inflation鈥檚 simplest model barely fits the observations now, leaving only more complicated models still alive, he adds. 鈥淪houldn鈥檛 that give some people pause?鈥�
鈥淚t worries me,鈥� admits of the Institute for Advanced Study in Princeton, who helped figure out how to spot signs of gravitational waves nearly 20 years ago. But he adds: 鈥淣ature is how it wants to be. It doesn鈥檛 follow from any logic that the simpler thing is the true one.鈥�
This article appeared in print under the headline 鈥淏ig bang discovery crumbles to dust鈥�
