A model of rotated mirror/normal letter discriminations