What is faster -- light or gravity?

General relativity lacks the notion of the speed of gravity. This is inconvenient and the present paper is aimed at filling this gap up. To that end I introduce the concept of the "alternative" and argue that its variety called the "superluminal alternative" describes exactly what one understands by the "superluminal gravitational signal". Another, closely related, object called the "semi-superluminal alternative" corresponds to the situation in which a massive (and therefore gravitating) body reaches its destination sooner than a photon \emph{would}, be the latter sent \emph{instead} of the body. I prove that in general relativity constrained by the condition that only globally hyperbolic spacetimes are allowed 1) semi-superluminal alternatives are absent and 2) under some natural conditions and conventions admissible superluminal alternative are absent too.