Abstract

Nelson Goodman's Calculus of Individuals and Lesniewski's Mereology are the best-known examples of what Peter Simons calls Classical Extensional Mereology. They are classical because they are equivalent and because they are representative of many attempts at mereologies ; they are extensional because they incorporate the principle that individuals which share all proper parts are identical. I synthesize these two systems into Generic Mereology; the result is a model for discussing problems with extensional mereologies in general. In the course of the synthesis, it is found that some relation--unacknowledged by Goodman or Lesniewski--holds between parts, a relation that is crucial to a whole's structure. ;There are philosophical problems with reflexivity and transitivity in Generic Mereology, and with the extensionality principle already stated. Even more objectionable is the principle of arbitrary sums: for any x and any y there exists a z such that z is the sum of x and y. Simons avoids these problems by limiting the application of Classical Extensional Mereology to spatiotemporal masses . My project is to develop an alternative extensional mereology which resolves these problems without losing general applicability to all kinds of wholes. ;George Lakoff offers arguments for the view that gestalt wholes, as preconceptual structures, are fundamental to knowledge and perception, and he constructs an informal logic in which whole is implicitly primitive. I show that this account is confused, and that another term he introduces, configuration, refers to a concept more fundamental than either part or whole. ;A critical survey of different approaches to the structure of wholes by philosophers from Plato to David Lewis concludes that Lakoff's configuration refers to the relation holding between the parts of a whole. Aristotle gives the best account of the nature of this relation. ;New Mereology takes configuration as primitive, and defines part, whole, and the other terms of standard mereologies in terms of it. The resulting calculus resolves the problems with Generic Mereology and captures our intuitive notions of parthood