In a paper published in the Oct. 16 issue of the research journal "Nature," the three researchers -- Daniel Gauthier and Michael Stenner of Duke University and Mark Neifeld of the University of Arizona - reached their findings by applying information theory to experiments with lasers. Information theory is a statistical theory dealing with the limits and efficiency of information processing

In the process, the researchers recorded experimental conditions in which the posted light speed limit appeared to be vastly exceeded ñ- until they subtracted the time it took them to reliably detect the results.

Albert Einstein proposed in his 1905 special theory of relativity that nothing can exceed the speed of light in a vacuum -- about 186,000 miles a second. However some theoreticians immediately challenged that light speed limit, said Gauthier, a Duke associate professor of physics.

Those theoreticians' calculations suggested that some components in a pulse of light could move faster while passing through particular types of transparent media, Gauthier said. These calculations posed a problem to the relativity theory itself, which stipulates that anything breaking that light speed limit would arrive at its destination sooner than it left, thus violating principles of cause-and-effect.

To resolve the problem, the early 20th century scientists agreed to a "reformulation" of Einstein's theory by stipulating that it was the velocity of "information" carried on a light beam that cannot exceed that speed, wrote Gauthier, Stenner and Neifeld in their paper.

The general idea was that only the earliest arriving part of the light pulse that carried information had to conform to relativity's speed limit guidelines and thus preserve the principle that no effect can happen before its cause.

"The early theorists thought of information as the stuff that travels between a cause and a resulting effect," said Stenner, Gauthier's graduate student and the first author of the Nature paper.

The light speed question has been reintroduced in recent years by experimentalists now equipped with lasers and advanced detectors. Their experiments involved measuring light pulses moving within what the Nature paper called special "'fast light' media" consisting of gases of laser-energized atoms.

Those experiments raised anew the possibility of faster-than-light speeds such that "the peak of a light pulse may exit the optical material before it passes through the entrance face," the authors wrote.

Scientists have long known that light is made up of different frequencies, Gauthier explained. As the Nature paper noted, in a given pulse of light each component frequency can be envisioned as a separate wavelet. And each wavelet can be said to move through a light-transmitting medium at a slightly different velocity.

Usually when light passes through such media -- an optical lens or window glass, for example -- the speeds of all those component frequencies proportionally decrease. But in a few materials there are cases of "anomalous dispersions" in which a limited number of light pulse components speed up instead of slowing down.

It was in those special situations that some early 20th century theoreticians saw the possibility of Einstein's light speed limit being violated when the fastest frequency components leaped ahead of the pack.

The issue has been resurrected in recent years by contemporary theoreticians who have reconsidered the dynamics of light pulses, as well as by experimentalists, Gauthier said.

Gauthier's group followed up on an experiment published in 2000 in Nature that created a special kind of anomalous dispersion by beaming pulses from two lasers into a gas of atoms. This created a fast light pulse that was both boosted in energy and tuned to a narrow range of frequencies.

As a result of these manipulations, there was evidence that the bulk of the light pulse, rather than just a few wavelets, appeared to move faster than light speed. But the design of that experiment made it difficult for the Duke team to determine whether information in the fast light could also exceed the speed limit, Gauthier said.

So Gauthier and Stenner set up their own two-laser experiment that made information speed much easier to distinguish by beaming each laser into a different vapor cell filled with potassium atoms.

That new experimental design permitted them to coordinate and tune the frequencies of both laser beams to create a pulse that was significantly advanced. They could also re-tune the lasers in a way that made the light behave as though it were in a vacuum.

When the lasers were tuned to advance the pulse the results appeared to be "highly superluminal," or faster than light speed, Gauthier said. As a result, the apparent speed differences of fast light and "normal" light were easier to compare.

To determine whether special relativity's rules of cause and effect were really being violated, Gauthier -- who is also a member of the Fitzpatrick Center for Photonics and Communications headquartered in Duke's Pratt School of Engineering -- and Stenner collaborated with Neifeld, an engineering expert in information science and theory at the University of Arizona's Optical Sciences Center.

Basically, the researchers encoded information into the light pulses and compared the speed of delivery of that information in both the advanced pulse and the "vacuum."

The comparison revealed that information in the "fast light" did not really travel faster than light speed limit. In fact, "it came out a bit slower" than the exact posted speed limit for light that was recorded in the vacuum, Gauthier said.

That means that, even though the peak of the light pulse was superluminal, the speed of its information delivery definitely wasn't.

As the authors concluded in their Nature paper, "our observations are consistent with relativistic causality and help to resolve the controversies surrounding superluminal pulse propagation."