ABSTRACT
This paper reviews evidence that formal explanation is intrinsic to an important class of generic knowledge and to the mechanisms proposed for representing this type of knowledge. Formal explanations are preferentially generated and rated as better when the characterizing property is thought to be an aspect of being the kind of thing in question compared to cases where it is not, but is merely prevalent among things of this kind, or causally connected to it. These and other data suggest that formal explanations unify two deep characteristics of human cognition: (i) the pervasive tendency to think and talk about things as instances of kinds, and (ii) the drive to understand and explain. In identifying something as an instance of a kind, we understand some of its properties as being due to its being the kind of thing it is. The cognitive mechanism that provides the basis for thinking of things as instances of kinds also provides the means for thinking about a kind and the properties that the kind has by virtue of its being the kind of thing it is. This mechanism represents generic knowledge that involves formal explanation in a non-propositional, non-predicational manner via co-indexation. Furthermore, the connection between the kind and the property is cognitively represented as immediate (i.e. there is nothing that mediates between being a dog and barking), and explanatory. The final section speculates that the mechanisms that represent generic knowledge that involves formal explanations may shed light on the representation and acquisition of Aristotelian first principles.
