UMLR — Ambiguity in description of TransitionKind

The spec seems to be contradictory regarding local transitions: in p. 378 is says that a local transition "[...] will not exit the composite (source) State, but it will exit and re-enter any state within the composite state that is in the current state configuration". Since it says "re-enter" it is implying that the current configuration doesn't change.
However on p.328 it states that "for local Transitions the target Vertex must be different than the source Vertex". The only way I can reconcile the two is if by "source Vertex" and "target Vertex" they mean an exit point and entry point of the same composite state, i.e., being a self-transition.

In summary, the formal definition in p.378 suggests that local transitions must be self-loops, while the description of p.328 suggests that local transitions cannot be self-loops. Both cases seem to be unnecessary restrictions, and there are no constraints in Transition enforcing those restrictions