in their research, griffin and bock (2000) found that participants who are asked to describe a visual scene featuring more than one character or component tend to