[Note: October 5 update and clarification added at the bottom]

Neutrinos that go beyond light speed? Not so fast, say two theoretical physicists.

In a terse, peremptory-sounding paper posted online on September 29, Andrew Cohen and Sheldon Glashow of Boston University calculate that any neutrinos traveling faster than light would radiate energy away, leaving a wake of slower particles analogous to the sonic boom of a supersonic fighter jet. Their findings cast doubt on the veracity of measurements recently announced at CERN (and posted online here) that clocked neutrinos going a sliver faster than light.

For someone who may have just helped to save the edifice of modern physics (if it was ever really at risk of crumbling down), Cohen is not especially upbeat or relieved. “On the contrary, I am saddened and disappointed,” he says. After all, a lot physicists would love the shocking measurement to be correct. For the experimentalists who made it, it could mean that they had made the discovery of the century. For theorists, it could be the start of an exciting period of creative upheaval. “It gets boring if [nature] always works the same way you expected,” Cohen says.

The result announced at CERN on September 23 (although the news had leaked out ahead of time) was certainly unexpected. By now, if you haven’t heard of it, you must have been a straggler from the Imperial Japanese Army coming out of Iwo Jima’s tunnels. Anyway, to recap, the war in the Pacific is over, and a team of physicists has released data on neutrinos they beamed through the Earth’s crust, from Geneva to the Gran Sasso Massif, near Rome, in an experiment known as OPERA. According to the physicists’ estimates, the neutrinos arrived at destination around 60 nanoseconds too fast, violating the cosmic speed limit set by Albert Einstein’s theory of relativity.

Experts urged caution, especially because another measurement of neutrino velocity—one done in 1987 by detecting particles from a supernova that had gone off in the Magellanic Cloud, just outside our Milky Way—indicated to high precision and accuracy that neutrinos do respect the cosmic speed limit.

The neutrinos coming from that supernova, however, were relatively weak; by comparison, those shot from from CERN have more than 1,000 times the energy. What if supercharged neutrinos could be superluminal even as less-energetic ones were confined to our boring, relativistic world?

So, Cohen and Glashow (the latter a Nobel Prize winner) looked at precisely the high-energy kind of neutrinos that are detected at Gran Sasso. From basic principles such as the conservation of energy and momentum, they deduced that if superluminal particles indeed existed, they could decay into other particles that are bound to a lower speed limit. “When all particles have the same maximal attainable velocity, it is not possible for one particle to lose energy by emitting another,” Cohen explains. “But if the maximal velocities of the particles involved are not all the same,” then it can happen.

An effect of this type is well-known in the case in which electrons have the higher speed limit (light speed) and light itself has the lower one. This can happen because when light propagates in a medium, such as water, air or glass, its speed gets substantially reduced--a change that's at the basis of the familiar refraction effect in which a pencil half-dipped in water looks as if it’s broken in two. (The universal speed limit of relativity is, to be precise, the speed of light in the vacuum.)

Electrons then, can move in a medium at a speed higher than the maximums speed of photons in that medium, and lose energy by emitting photons. This process is called Cherenkov radiation, and it makes the reactor pools of nuclear power stations (such as the one pictured here) glow with a bluish light. It is also used to detect electrons that shower down on Earth after a high-energy cosmic ray crashes on the upper atmosphere.

The possibility of a transfer of energy between particles with different speed limits was well known, Cohen says, and is a fact that he often gives to his undergraduate physics students as a homework problem. But in their paper, he and Glashow go further. They discuss the exact mechanisms by which such a conversion can happen, and make precise quantitative estimates of how often the neutrinos would decay into each type of particle.

The emission that is most likely is, by far, that of an electron paired with its antimatter twin, a positron, the authors conclude. (The high-energy neutrino would create them by interacting with one of the “virtual particles” that incessantly and fleetingly froth out of the vacuum—in this case, a Z boson, one of the carriers of the weak nuclear force; it was by understanding precisely that type of interaction that Glashow shared a Nobel Prize in Physics in 1979.)

Crucially, the rate of production of these electron-positron pairs is such that a typical superluminal neutrino emitted at CERN would lose most of its energy before reaching Gran Sasso. “The beam sent from CERN would be significantly depleted” of high-energy neutrinos, Cohen says. The neutrinos picked up at the Italian lab, however, do not seem to have lost any of their energy.

But then, perhaps they were not superluminal to begin with.

"I think this seals the case," says Lawrence Krauss, a theoretical physicist and the director of the Origins Projet at Arizona State University. "It is a very good paper." Krauss has been among the most critical of the OPERA team's decision to go public with their findings, as he has written in an op-ed for the Los Angeles Times and told my former colleague John Matson.

The OPERA collaboration did not respond to a request for comments.

Carlo Rovelli, a theoretical physicist a the University of the Mediterranean in Marseille, says that Glashow and Cohen’s results are “plausible,” and did not seem particularly surprised. “It seems that the majority of physicists, including myself, strongly suspects that there is some mistake in OPERA’s measurements.”

Some physicists have suggested that neutrinos could be finding shortcuts in spacetime--for example, by moving in extra dimensions of space--that would allow them to get there faster while still respecting the speed limit. Such a possibility may not be ruled out by neutrinos' Cherenkov radiation, but may begin to look increasingly contrived.

“Let’s put it this way: physicists who work on string theory for more than 20 years have assumed that there are additional dimensions, and yet none of them had ever consider the possibility that a particle could find shortcuts in other dimensions, and go faster than light,” Rovelli says.

As to what may or may not have gone wrong with the experiment, Cohen does not want to speculate (though others have: see for example the blog of theoretical physicist Lubos Motl). “I am not the right person to say what happened,” a task that he says is best left to other experimentalists.

Was Einstein right after all? Einstein's relativity superseded Isaac Newton's physics, and probably something else will some day supersede it. And physicists will still continue to use either one, when appropriate. "All of our scientific results have some domain of validity," says Cohen; no scientific theory is "right" or "wrong" in an absolute sense--each just is in more or less accurate agreement with experiment. Meanwhile, others will no doubt keep trying to find glitches in Einstein's theories. “We never stop testing our ideas," says Cohen. "Even those that have been established well.”

October 5 Update and Clarification:

Since the post went up, there has been a very interesting exchange of viewpoints in the comments section: thanks to everyone who has been contributing.

First, the clarification: andrewgdotcom pointed out that was sloppy when I described the production of electron/positron pairs as a decay. Indeed, the neutrinos wouldn't decay; they would proceed on their superluminal journey, but with less energy.

I disagree with the interpretation that some have given below of Cohen and Glashow's result as an attempt at using theory to disprove facts; to me the point seems to be whether (something that for the authors amounts to) a back-of-the-envelope calculation can show that a certain experimental finding is in contradiction with scores of other experimental findings which also were done with high precision and accuracy.

But couldn't there be could assumptions in the paper (such as basic facts in quantum field theory) that one could imagine failing to hold true here? After all, the paper deals wtih the weird and hypothetical realm of superluminal particles: who knows how much of the "old" theory we would have to throw away.

Rovelli says there do not seem to be any "hidden" assumptions in the paper that would invalidate it in a superluminal world.

"There are many things we don't know," Krauss says. "But there are many things we know. And one of these things involves the interactions of neutrinos. The process presented by Glashow and Cohen must occur, given all the existing measurements of neutrinos."

In any event, it will be exciting to see what OPERA and other neutrino experiments will do in coming months and whether their incredible findings will hold up.

I will be doing my own experiment, too. Gran Sasso is just two hours away from where I live; I will go hiking just to the south of it to see if I can catch some of the neutrinos that bypass the OPERA detector and emerge overground. And I will keep you posted.

Paper: "New Constraints on Neutrino Velocities," by Andrew G. Cohen and Sheldon L. Glashow.

Added October 11: Physicist Matt Strassler blogged about the Cohen-Glashow paper too.

Images courtesy of Istituto Nazionale di Fisica Nucleare, Australian Nuclear Science and Technology Organization and Wikimedia Commons