alt Hacker NewsIIRC the group velocity in a medium can also exceed c when you send pulses through certain nonlinear media. Basically, the "lump" of the pulse can appear to "exit before it entered," but what really happens is that the tiny tip of the pulse has already gone through at <= c, and starts growing into the body at the exit before it has completely shrank at the entrance because of the nonlinearity. So it is actually the information/"signal" velocity (the tip of the signal) that cannot exceed c.
AFAIK there are three scenarios;
1) Non-Dispersive medium (eg. vacuum) : v(g) = v(p) = c
2) Normal Dispersion medium (eg. in water) : v(g) < v(p)
3) Anomalous Dispersion medium (eg. absorbing materials) : v(g) > v(p)
where (simplified);
v(g) : group velocity which can be thought of as a function of the overall wave packet (i.e. pulse) and hence contains the actual energy/information/signal.
v(p): phase velocity which can be thought of as a function of a single wave and hence the crest of that wave.