BFO (Basic Formal Ontology)

The foundational partition of BFO. Continuants are entities that persist through time and maintain their identity while undergoing change (e.g., a heart, a person, a city). They have no temporal parts. Occurrents are entities that unfold in time and have temporal parts (e.g., a heartbeat, a life, a storm). This distinction prevents category errors in data integration.

A critical sub-partition of continuants. Independent continuants are entities that can exist without being borne by another entity (e.g., a molecule, a cell, an organism). Specifically dependent continuants require a bearer to exist. These include:

• Qualities: The redness of an apple, the temperature of a liquid.
• Realizable entities: Dispositions (fragility), functions (to pump blood), and roles (being a surgeon).

BFO is a realist ontology, meaning it is built to represent reality as it is understood by empirical science, not concepts or data models. It distinguishes between:

• Universals: Types or kinds that are instantiated by many particulars (e.g., the kind Electron).
• Particulars: Individual entities in space and time (e.g., a specific electron in a specific atom). This commitment to realism makes BFO suitable for integrating scientific data across disciplines.

BFO handles the problem of scale through a theory of granular partitions. It distinguishes between bona fide boundaries (physical discontinuities like a cell membrane) and fiat boundaries (abstract demarcations imposed by human convention, like a county line). This allows the ontology to represent objects at different levels of granularity without logical contradiction, crucial for bridging molecular biology and population health data.

BFO provides a rigorous mereological (part-whole) framework for occurrents. Processes have temporal parts that are themselves processes. A temporal region is the dimension in which occurrents unfold. This allows for precise reasoning about the sequence, overlap, and containment of events, which is essential for modeling experimental protocols, clinical pathways, and manufacturing workflows.

BFO is designed to be combined with domain ontologies through a hub-and-spoke architecture. Domain ontologies (like the Gene Ontology or Protein Ontology) are built by specializing BFO's high-level classes. This ensures that entities from different domains are classified under a common, philosophically coherent backbone, enabling cross-domain querying and inference that would be impossible with siloed taxonomies.