The form of a kind concept and its role in human thought

Human beings are the only species to develop such a vast repertoire of concepts for kinds of things—concepts like table, tiger, artist, universe, democracy, and hexagon. The kind concepts we possess have a profound impact on the way we think. They frame our interpretation of the external world, and we combine them together to produce complex, intricate thoughts using natural language.

A proposal about these representations which traces back to Aristotle, but which remains under-explored empirically, is that each one possesses a central form—a structure that represents a subset of the information we store about the kind in a privileged manner (e.g., representing serving hot drinks, for the concept cafe). The central form changes how we conceive of the information that is represented as part of it (e.g., assigning norms—cafes should serve hot drinks—as well as assigning a suite of other expectations).

In the first part of this talk, I will sketch the background to this idea and present empirical research which suggests that from early in conceptual development, each new kind concept we acquire possesses a central form. Following this, I will present two lines of inquiry which explore the computational role of this structure. The first explores the role of a concept’s central form when concepts are combined to create more complex structures using natural language. The second investigates whether a concept’s central form underpins the human tendency to reason about people using stereotypes, as well as the human tendency to reason about rules in, at times, an oversimplified manner.