@TimCampion Zig-zags? But you can approximate spacelike paths that way. :/

@TimCampion Yeah, I should drop the local triviality condition. (I really should have guessed that something like this should happen, with the discussion of curvature above.) Not sure what the correct condition is.