BFO

We are trying to make sure that appropriate connections are made between independently developed sources of scientific information. To connect there must be compatibility of express - things and classes must be comparable or relatable in some way, e.g. classes in one source being subclasses of classes on another source. Therefore any method for designing ontologies (systems of classes, relations, and individuals) that increases the chance that two people doing similar or related designs will do them in compatible ways should be embraced - regardless of its philosophy. The methodology of naive realism seems to work because in our experience it helps build consensus - people can discuss alternative designs rationally, and evaluate them according to a shared set of criteria.

Naive realism is captured by BFO, an "upper ontology."

-

(let's write a cheat sheet. following is an incomplete sketch based on the iFOMIS report)

BFO has two parts: SNAP and SPAN.
 * SNAP starts with continuants: things that can endure through change (without temporal parts!?)
 * SPAN starts with occurrents: e.g. states of affairs (bound to a time)

Independent continuants:
 * substance
 * boundary of substance
 * fiat part
 * aggregate

Dependent continuants:
 * quality
 * role
 * function
 * disposition

Temporally extended:
 * portion of spacetime
 * process (fiat part, aggregate, ...)