# Information not light

1. Dec 29, 2003

### wolram

im not sure if this is some sort of scientific double talk,
but to circumvent problems with speed >C, it is now more
popular to say that "information" cannot travel >C, it seems
experimentaly part of the spectrum, "frequency" of light can
travel >C but any encoded information traveles <C or =C, so
irispective of "information", light can travel >C.
---------------------------------------------------------------------
http://www.spaceref.com/news/viewpr.html?pid=12797

That means that, even though the peak of the light pulse was superluminal, the speed of its information delivery definitely wasn't.
----------------------------------------------------------------------
the above is just one of many sites that seem to say the same.

so how does this fit in with LQG and other theories?
why can only the peak of the light pulse travel >C ?

2. Jan 1, 2004

### DW

It has always been superluminal "information" transfer that would not be allowed by the combination of special relativistic physics and the principle of causality. Consider a superluminal pulse sent from an emitter to a receiver. "If" the pulse width could be considered negligable then the emition and reception constitute two spacelike seperated events having each occured at a specific location and time. Due to relative simultaneity a simple velocity boost for a frame transformation can be done in order to reverse order the spacelike seperated events. Then according to the new frame, the effect of reception would preceed the cause of emmition. This is a violation of the principle of causality according to which effects never preceed their causes. Where the controversy comes in is that there are anomolous disspersive media for which the group velocity of light is greater than the Lorentz speed c. Experiments have been done sending laser pulses through such media with group speeds over 300c. Now for the catch. In the above scenario the pulse width had to be considered negligable. In reality quantum mechanics puts limits on a pulse width or period really for a given laser frequency in accordance with an energy time uncertainty principle. In order for the pulse to travel faster than c the laser has to be within a very narrow frequency range given by the properties of the media. In order for the pulse not to be lost to disspersion in the transmition it has to be gain assisted even over a short distance of transmition. The end result is that in the actual experiment the pulse width is even wider than the distance of transmition. They may be able to change that at some point, but there will always be a limit on how narrow the period of the pulse is. Now the group velocity is the speed at which the pulse peak travels, but once a given small segment of the pulse arrives at the receiver the information has still not been completely read. The information is not completely received until after the complete period of pulse reception after the front of the pulse arrived. One could then say that the information speed is not the group speed when this period can't be negleted but is instead given by distance of transmition divided by the sum of the group's peak transmition time and the period of the pulse itself. This information speed turns out to be $$5.0x10^{-5}c$$ for the 310c group speed experiment. Thats not so impressive. One could hypothetically assymtotically increase the information speed to the group speed in such an experiment by far expanding the distance of transmittion. The problem is that the information will ultimately be lost to the disspersion as the information speed approaches c.

3. Jan 1, 2004

### Hurkyl

Staff Emeritus
Maybe this thought experiment can explain the difference between something like a peak travelling superluminally and information travelling superluminally:

I have a large number of LCDs which I have arranged at rest (in some reference frame of my choosing) into a line a light-year long. Attached to each LCD is a power source, a microprocessor, and an atomic clock. The microprocessors for adjacent LCDs are also connected to each other. All of the atomic clocks are synchronized in the same rest frame.

Scenario 1: Information travelling FTL:

Each microprocessor is programmed to, when it receives a signal from its left, flash its light on and off then send a signal to its right.

I send a signal to the leftmost microprocessor. Half a year later, the rightmost processor flashes its light.

Scenario 2: A "peak" travelling FTL:

Someone comes along and programs into each microprocessor a time at which it will blink its light, and the times are chosen in such a way that the light that is on will travel down the line, and there will be half a year between the leftmost light blinking and the rightmost light blinking.

The former is what is forbidden by relativity. Something like the latter is what's happening in FTL experiments.

In laymen's terms, scenario 1 is forbidden because the information about the starting event propagates faster than light, but scenario 2 is allowed because all the relevant information was distributed in a slower than light fashion (even though the visible result is identical to scenario 1)

4. Jan 1, 2004

### wolram

i am sorry, but this is hard to fully understand, if something
the "light peak", can be judged to have traveled superluminally
and detected ,then the experiment has already provided information
that exceeds the speed of light.

5. Jan 1, 2004

Staff Emeritus
No because the last device didn't receive a signal from the first, it received instructions from its own local setting. The light peak was no more an information carrier thn is the spot of light you produce on a fence by swinging a flashlight.

6. Jan 2, 2004

### wolram

this is all very confusing, thanks for trying to explain, im
none the wiser though, i cant see the point in publishing a
paper that claimes superluminal speed ,when in reality nothing has exceeded C, ok the wave peak has ,can travel 300 times C but
as nothing useful can be done with it, its not breaking any
"limits"?

7. Jan 2, 2004

### jby

Forgive me but how does this be true? By energy-time uncertainty pcp, delta E*delta t >= hbar/2

E = hf,

delta f*delta t >= (1/h)*(hbar/2)

......

Did I overlook anything?