This is a summary of the book titled “Building Ontologies with Basic Formal Ontology” written by Robert Arp, Andrew Spear and Barry Smith and published by MIT Press, 2015. Modern scientific research is producing data at a pace and scale that far exceed the capacities of traditional analytical methods. This transformation is especially visible in the life sciences, where advances such as high‑throughput gene sequencing and multidimensional imaging generate vast amounts of information every day. As researchers confront this deluge of data, the question of how to store, integrate, interpret, and share it efficiently has become increasingly urgent. Robert Arp, Andrew Spear, and Barry Smith address this challenge, presenting ontology as a powerful solution for achieving interoperability, accessibility, and coherence across scientific domains. 

Ontologies, as the authors explain, emerge from philosophy’s long tradition of studying what exists and how different entities relate to one another. In contemporary scientific and computational contexts, an ontology functions as a representational structure—essentially, a taxonomy—designed to categorize and relate types of entities according to their defining characteristics. A classic example is the familiar biological hierarchy that starts with broad categories such as “vertebrate animals” and branches into more specific groups such as mammals, reptiles, primates, and snakes. Such structured classification enables scientists to clarify how individual items fit within broader categories, enhancing clarity and communication. 

This philosophical grounding underlies the ontology’s central purpose: representing reality as faithfully as possible. Ontological realism—the idea that the categories and relations described in an ontology correspond to entities in the real world—plays an important role here. For instance, the classification “mammal” is not a linguistic convenience but a label for a genuine biological class of organisms. Ontologies used in applied fields such as biomedical informatics depend on this realism, enabling researchers to use consistent terminology and shared conceptual frameworks across diverse technological platforms. 

The authors distinguish among different kinds of ontologies, showing how they operate at varying levels of specificity. A general ontology might describe broad types of organisms, while a domain ontology focuses on particular systems or phenomena—such as the human heart, with its chambers, valves, and functions. Domain ontologies are indispensable for specialized research areas, but they also risk creating isolated conceptual systems that do not integrate well with each other. To avoid this fragmentation, the authors emphasize the importance of beginning every ontology with universal, top‑level categories that provide a common foundation for more specific structures. This top‑down approach improves interoperability and supports scientific collaboration across disciplines. 

Designing an effective ontology also requires adherence to several foundational principles. Ontologists must assume the existence of real-world entities, acknowledge the complexity of systems, recognize the limitations of scientific theories, and strive to represent reality as accurately as possible given current knowledge. They must also design ontologies so that entities at various levels of granularity—from broad categories to fine distinctions—can be represented. Because science evolves, ontologies must remain flexible, open to revision as new discoveries emerge. 

Basic Formal Ontology (BFO) framework distinguishes between continuants and occurrents. Continuants are entities that persist over time while retaining their identity—like a human being or a piece of fruit—even though their parts may change. Occurrents, by contrast, are processes or events unfolding in time, such as infections or biological functions. These two types of entities require different representational strategies, and BFO provides the conceptual tools to integrate both within a single coherent ontology. 

The relationships among entities are equally crucial. Ontologies go beyond hierarchical classification, incorporating relationships among universals, between universals and particulars, and among individual entities. These relational structures reflect the complexity of scientific reality—for example, the shared atomic composition of different organisms or the dependence of certain qualities on larger structures. 

An ontology must become a practical tool—not just a conceptual model but a computer‑implementable artifact. Using tools such as the Protégé ontology editor and the Web Ontology Language (OWL), ontologists translate conceptual structures into software systems capable of supporting large‑scale data analysis and knowledge integration. These digital ontologies already underpin major scientific efforts in fields ranging from cell biology to mental health research. 

Through their systematic exposition, Arp, Spear, and Smith demonstrate that ontologies, when properly constructed, serve as vital infrastructure for modern science. They provide the shared language and structure necessary to manage overwhelming volumes of data, bridge disciplinary divides, and ensure that scientific knowledge remains coherent, accessible, and continually adaptable.