Proceedings of the Annual Meeting of the Cognitive Science Society

Studies suggest that people use the least possible effort to gen-erate natural language descriptions of sets of objects. Thismeans that they base descriptions on what is perceptually avail-able to them. For instance, people can subitize, i.e., rapidlyassess the exact quantity of small numbers of objects, so whenthe quantity of objects in the visual scene is beneath this thresh-old, they give numeric descriptions; when the quantity is abovethis threshold, they generate non-numeric descriptions. How-ever, no research examines how people describe visual scenesof items in groups. As such, it is unclear how people will formdescriptions of scenes that contain a large total number of itemsin groups. We report on a novel experiment designed to in-vestigate how people produce quantified descriptions of scenescomposed of salient visual groups. The results corroborate theleast effort hypothesis, and suggest that people’s incrementalperception of quantity drives their descriptions.