Aversive control: A separate domain?