The following is a comprehensive book review of "Applied
Ontology: An Introduction," edited by Katherine Munn and Barry
Smith. This volume presents a powerful argument for the systematic
application of philosophical ontology to the practical challenges of modern
information management, particularly within the domains of medicine and
biology.
An Overview of Themes, Arguments, and Structure
"Applied Ontology: An Introduction" is a
collaborative volume that asserts the necessity of integrating rigorous
philosophical methods, particularly those derived from ontology, into the
development of automated information systems. The book's central argument is
that current information systems (such as electronic databases, terminologies,
and taxonomies) are often designed idiosyncratically, making them unable to
share data (interoperable) or serve as reliable inputs for automated reasoning.
The volume introduces the core theme that ontology is the
philosophical discipline which aims to understand how things in the world are
divided into categories and how these categories are related together. The
goal of the contributors is to show how philosophical insights can inform the
creation of structured, automated representations, also called 'ontologies'.
The key arguments revolve around advocating for the realist
orientation as an antidote to prevailing deficiencies. This orientation
holds that sound knowledge representation must aim to represent the world
itself, not merely convenient conceptualizations of it (a methodology
critiqued as "pragmatist conceptualism"). The book emphasizes the
doctrine of realist fallibilism, which combines a "healthy
intellectual humility" with the conviction that humans can take measures
to procure true beliefs about the world.
The book is organized into an introduction and thirteen
chapters contributed by various philosophers and information scientists. The
chapters systematically explore the theoretical foundation (formal ontology and
realism), essential structural tools (granular partitions and classification
theory), and deep ontological distinctions (continuants vs. occurrents,
universals vs. particulars, and relational logic). Although the methods are
intended to be broadly applicable, the book limits its focus primarily to the domains
of biology and medicine.
Analysis of Style, Tone, and Perspective
The overall tone of "Applied Ontology" is urgent,
expert, and reformative. The contributors stress the "urgency of applying
rigorous philosophical methods" because doing so has "vast
potential" for making information systems interoperable and efficient.
The perspective is consistently one of realist
ontological perspectivalism, which argues that we can gain knowledge of
reality via a multiplicity of veridical granular partitions. This stance is
consciously pitted against conceptualism, a view widespread in
information science where the definition of 'concept' is often loosely applied,
leading to confusion about the representation target.
The writing style is designed to bridge the gap between
abstract philosophy and practical engineering. Katherine Munn’s introduction is
explicitly non-technical, using accessible metaphors, such as imagining a
brilliant scientist scribbling a theory on a beer mat. However, the volume
quickly moves toward rigorous, formalized treatments, including discussions of
first-order logic (FOL) and description logics (DLs), making the style
necessarily dense in later chapters. The clarity is enhanced by drawing on a deep
tradition, particularly Aristotelian methodology concerning categories
and hierarchical taxonomy, which is shown to have new relevance for modern
bioinformatics.
Evaluation of Strengths and Weaknesses
Strengths: A major strength is the book’s unwavering
commitment to foundational rigor. The authors provide crucial clarity on
distinctions often blurred in information science, such as the difference
between universals (kinds) and particulars (instances) or the
critical difference between sets and classes.
The critique of existing biomedical terminologies (the
"concept orientation") is a strong point, yielding critical insights:
- The
volume addresses why systems like the UMLS Metathesaurus fail to provide
true integration, sometimes treating non-interchangeable entities (like
'aspirin' and 'Aspergum') as if they referred to the same concept in an
"artificial world".
- It
exposes the flaws in systems like the National Cancer Institute Thesaurus
(NCIT), demonstrating how it violates fundamental criteria for good
classifications, such as disjointness or uniformity. For
example, classifying 'cancer patient' and 'outpatient' as disjoint
subtypes of 'patient' is problematic because many cancer patients are
treated as outpatients.
- The
argument that type-level relations (such as $part_of$ between universals)
must be rigorously defined by reverting to consideration of their
instances is foundational. The precise delineation of four distinct $is_a$
relations—genus-subsumption, determinable-subsumption, specification,
and specialization—resolves ambiguities that plague information
systems.
Weaknesses: The authors acknowledge that the proposed
transition to a realist orientation would be arduous, necessitating a
global paradigm shift. Furthermore, the complexity of the domain—for example,
the challenge of defining biological kinds given the "sheer number"
of animal kinds (estimated at 30 million) and the existence of borderline
cases—means that no classification system can be perfectly exhaustive. The core
ideas, being novel in application, are still "being developed".
Connections to Broader Academic and Social Contexts
The book grounds its methodology in a broad sweep of
philosophical history. It draws heavily on Aristotle's Categories to
build a reliable top-level ontology, recommending the adoption of the Basic
Formal Ontology (BFO) structure, influenced by Aristotelian ideas. The
discussion of the four-category ontology (the ontological square, later
expanded to a sextet) highlights the deep metaphysical concerns underlying
effective information structure.
Academically, the volume directly combats influential
non-realist positions. Ingvar Johansson's chapter explicitly clarifies and
rejects alternative philosophies of science that discourage the pursuit of
objective reality, specifically critiquing Myrdal’s Biasism and Vaihinger’s
Fictionalism, in favor of Popper’s Epistemological Realism and its
central notion of truthlikeness.
In a social and practical context, the book highlights
immediate applications in health management. The lack of interoperability
caused by non-standardized systems is a critical barrier. The proposed realist
approach, particularly using Referent Tracking (assigning an instance
unique identifier, or IUI, to every clinically relevant real-world entity), is
presented as essential for linking information in Electronic Health Records
(EHRs) to general scientific knowledge, thereby supporting diagnostics, scientific
discovery, and public health surveillance.
Overall Contribution and Significance
"Applied Ontology: An Introduction" constitutes a
seminal contribution to the emerging discipline of applied ontology. Its
significance lies in its uncompromising stance that the intellectual rigor of
philosophy is not a luxury but a necessary condition for overcoming the
limitations of current, non-interoperable information systems. The book
provides a theoretical blueprint, formal methodology, and clear practical
guidelines for building reference ontologies that mirror reality.
The volume successfully demonstrates that foundational
philosophical problems—such as what constitutes a natural kind or how processes
unfold in time (occurrents)—have tangible practical consequences when building
computerized systems. As the complexity and volume of scientific data continue
to increase, the book’s argument that "the means by which knowledge is
conveyed are every bit as important as that knowledge itself" serves as a
powerful call to action.
Conclusion
This book is indispensable reading for information
scientists, knowledge engineers, and bioinformaticians who are involved in
developing, curating, or implementing controlled vocabularies and ontologies,
particularly those within the biomedical domain.
It is also highly recommended for philosophers—especially
those working in metaphysics, philosophy of science, and logic—as it showcases
the profound and urgent practical relevance of their disciplines in the 21st
century. The explicit intention of the volume is to improve systems for everyone
whose life may be affected by scientific research, making its core
principles vital for anyone concerned with the quality and reliability of
digitally stored scientific knowledge.