Arguments against FTL communication are usually based on causuality arguments, which I tend to agree with. But it seems to me that the 'backwards in time' problem seems to make an assumption that might not always apply, ie that FTL velocity would be relative to C for all observers. Here's another (most likely incorrect) idea for sending FTL signals ... I have a powerful laser, and a fancy optical system. I can use this to create small traveling pulses containing negative energy relative to vacuum, (ie 'squeezed light'). This in itself has already been done (to a small extent), so is not totally unrealistic. Assuming that I can create these pulses as the result of interference in the beam, I could possibly make these 'bubbles' travel at very high speed, (ie >C phase velocity - the group velocity of the beam is of course just C). Assume that I can 'inject' an extra photon into some of these bubbles, and since there is 'less space' that the photon experiences relative to free space, the photon can 'ride' the bubble and move FTL. The above might be *totally* impossible in practice, but is meant mostly as thought experiment. To me it demonstrates an interesting twist on the paradox problem, because as the transmitter/receiver speed up, the velocity that they can transmit becomes slower, ie to another observer, the 'FTL' transmission can be much slower than C, since the speed of the interference pattern is related to the transmitter's clock, not C itself. Could such an arrangement be used to send a message 'back in time'? It seems at first glance to me that the answer is no, which would at least not rule out this method of FTL transmission based on causuality problems. I would much appreciate a comment from someone in the know.