The contribution of distinct component configurations to object recognition across changes of view