“Some people take [the observed lopsidedness] and calculate how probable it was that we would see something like that,” Bennett says, and come up with the answer that “it was astonishingly not probable and therefore there’s something wrong with the theory. So what’s wrong with that analysis? Here’s the analogy I use: Suppose you’re walking along on the sidewalk, and a leaf falls off a tree and hits you on the head at the precise moment you’re walking by it, and I come along and say, ‘That was astonishing! What were the chances that that particular leaf on that particular tree should fall just at the moment that you’re walking by?’ Well, if I calculated the probability of that, it would be tiny. But it’s a ridiculous thing to calculate. The problem is that lots of improbable things happen all the time because so many things happen.
“Once you identify something unusual in the data that you didn’t predict, you’ve already started from the answer and worked backward, which is the same problem with the tree and the leaf. You can’t say it’s improbable after you’ve observed it; you would have to ask the question up front. What are the chances of any leaf falling from any tree hitting you on the head at the moment you walk by? And you’d find that that’s not so improbable.”
Bennett concludes with the opinion expressed by many Planck scientists: “The lopsidedness may have some significance. You can’t say that it absolutely doesn’t. But you can’t really say that it does.”
The Planck data includes other intriguing observations that cosmologists are currently debating. The data indicates that hundreds of galaxy clusters are streaming together toward the constellation Centaurus, as if tugged by something beyond the visible universe in what has been termed “dark flow.” Most Planck scientists concluded again that the data was not statistically significant, but one, Fernando Atrio-Barandela of Spain’s University of Salamanca, found the opposite, and withheld his name from the official Planck results. He suggests that inflation could have lasted long enough to magnify some features beyond what we can see, and that they pull on parts of the universe we can. This thinking “does not contradict the inflation paradigm; it complements it,” Atrio-Barandela says.
WMAP had spotted traces of these “anomalies,” but scientists largely dismissed them as likely a matter of chance. Planck’s precision makes that more difficult. It also exposes the unease of cosmologists without a satisfying theory to explain what they see. They talk a little like acrobats without a net, unsure whether the new observations will push them off the standard theories that have been guiding them for decades.
One team published a study in 2006 predicting that Planck would find a “smoking gun” of oddities in the CMB—such as the cold spot and lopsidedness—that reveal other universes beyond our view are pulling on or have collided with ours. This theory of a frothy “multiverse” could also help explain dark flow. But others argue that if we indeed do live in a multiverse, Planck would have revealed it. “If there were really a multiverse, it would be much clearer,” says Paul Steinhardt, a theoretical cosmologist and the Albert Einstein Professor in Science at Princeton.
Steinhardt then turns the tables. He says the lack of evidence for a multiverse, plus other complications, undercuts the idea of inflation itself. His argument would be easier to discount if Steinhardt himself had not helped fashion inflation back in the 1980s. The problem, he says now, is that the same quantum fluctuations credited with creating galaxies also make it impossible for inflation to ever completely stop; while it may stop here or there, the random fluctuations keep it going elsewhere. Eventually it produces an infinite number of island-like universes, with an infinite number of physical properties. So where are they?
“The Planck results reveal a remarkably simple and uniform universe,” says Steinhardt. “And that is the problem. Inflationary theory predicts a multiverse, and the multiverse should be evident.”
“Unless a convincing theoretical explanation can be found, the anomalies are likely to remain statistical curiosities,” says Hiranya Peiris, a cosmologist at the University College London. Her team searched WMAP data for bubble universe collisions and found several candidates, but none were definitive. They are doing the same with Planck’s data while also running computer simulations of such collisions, using the equations of general relativity to better understand exactly what the collisions should look like. “The statistical significance of these ‘anomalies’ is not at a level where you would say the evidence is overwhelming,” Peiris adds. “So in this sense, the ball is in the theorist’s court.”
Last October, with its fuel depleted, Planck was turned off, but team members are still analyzing the data it collected. In a few months, they will release a second set of results that could offer the strongest evidence yet for inflation. “One of the biggest single [pieces of evidence] that would tell us about inflation has to do with the pattern of polarization on the sky at large angular scales,” says Charles Lawrence. “If we could observe this—if there were a significant background of gravitational waves in that 10-35 seconds after the Big Bang—it would leave an imprint that we could detect, and it would tell us what the energy scale of inflation was.” Inflation would have produced gravity waves, and, says Lawrence with satisfaction, “the pattern of polarization produced by gravitational waves isn’t produced by anything else.” Planck was not optimized to detect this pattern of polarization; it was optimized to detect temperature differences. Charles Bennett and other astronomers are working on ground-based instruments—some in the Atacama desert in Chile, others in Antarctica, the regions where there is the least atmospheric interference—to look for the polarization patterns that may not show up in the Planck data.