Later visual areas can adapt to adapted input from earlier visual areas.