I wonder if it's an all-moving vertical surface. If it has to rotate as the wing sweeps anyway, why not move the whole tail for flight control as well?
OK, I looked at the rest of that paper, and there's just enough detail in one image to see what's happening. Basically, the tail is attached to a small horizontal structure that seems to be pivoted at the leading edge. As the wing moves forward, the back part of this structure slides into a slot in the wingtip. So yes, it is a bit tricky, though not much worse than the normal swing mechanism.
It seems to be the same basic shape that Boeing offered for one of the pre-ATF studies: